From cfe3b0005f6e6e2b5384851a153ee831bad56d44 Mon Sep 17 00:00:00 2001
From: Craig Topper <craig.topper@sifive.com>
Date: Sat, 1 May 2021 13:18:21 -0700
Subject: [PATCH] [RISCV] Reorder masked builtin operands. Use
 clang_builtin_alias for all overloaded vector builtins.

This patch makes the builtin operand order match the C operand order
for all intrinsics. With this we can use clang_builtin_alias for
all overloaded intrinsics.

This should further reduce the test time for vector intrinsics.

Differential Revision: https://reviews.llvm.org/D101700
---
 clang/include/clang/Basic/riscv_vector.td          |  84 ++-
 .../RISCV/rvv-intrinsics-overloaded/vaadd.c        | 176 ++---
 .../CodeGen/RISCV/rvv-intrinsics-overloaded/vadd.c | 176 ++---
 .../RISCV/rvv-intrinsics-overloaded/vamoadd.c      | 140 ++--
 .../RISCV/rvv-intrinsics-overloaded/vamoand.c      | 140 ++--
 .../RISCV/rvv-intrinsics-overloaded/vamomax.c      | 140 ++--
 .../RISCV/rvv-intrinsics-overloaded/vamomin.c      | 140 ++--
 .../RISCV/rvv-intrinsics-overloaded/vamoor.c       | 140 ++--
 .../RISCV/rvv-intrinsics-overloaded/vamoswap.c     | 210 +++---
 .../RISCV/rvv-intrinsics-overloaded/vamoxor.c      | 140 ++--
 .../CodeGen/RISCV/rvv-intrinsics-overloaded/vand.c | 176 ++---
 .../RISCV/rvv-intrinsics-overloaded/vasub.c        | 176 ++---
 .../RISCV/rvv-intrinsics-overloaded/vcompress.c    | 106 +--
 .../CodeGen/RISCV/rvv-intrinsics-overloaded/vdiv.c | 176 ++---
 .../RISCV/rvv-intrinsics-overloaded/vfabs.c        |  18 +-
 .../RISCV/rvv-intrinsics-overloaded/vfadd.c        |  36 +-
 .../RISCV/rvv-intrinsics-overloaded/vfclass.c      |  18 +-
 .../RISCV/rvv-intrinsics-overloaded/vfcvt.c        | 108 +--
 .../RISCV/rvv-intrinsics-overloaded/vfdiv.c        |  36 +-
 .../RISCV/rvv-intrinsics-overloaded/vfirst.c       |  14 +-
 .../RISCV/rvv-intrinsics-overloaded/vfmacc.c       |  36 +-
 .../RISCV/rvv-intrinsics-overloaded/vfmadd.c       |  36 +-
 .../RISCV/rvv-intrinsics-overloaded/vfmax.c        |  36 +-
 .../RISCV/rvv-intrinsics-overloaded/vfmerge.c      |  18 +-
 .../RISCV/rvv-intrinsics-overloaded/vfmin.c        |  36 +-
 .../RISCV/rvv-intrinsics-overloaded/vfmsac.c       |  36 +-
 .../RISCV/rvv-intrinsics-overloaded/vfmsub.c       |  36 +-
 .../RISCV/rvv-intrinsics-overloaded/vfmul.c        |  36 +-
 .../RISCV/rvv-intrinsics-overloaded/vfncvt.c       | 104 +--
 .../RISCV/rvv-intrinsics-overloaded/vfneg.c        |  18 +-
 .../RISCV/rvv-intrinsics-overloaded/vfnmacc.c      |  36 +-
 .../RISCV/rvv-intrinsics-overloaded/vfnmadd.c      |  36 +-
 .../RISCV/rvv-intrinsics-overloaded/vfnmsac.c      |  36 +-
 .../RISCV/rvv-intrinsics-overloaded/vfnmsub.c      |  36 +-
 .../RISCV/rvv-intrinsics-overloaded/vfrdiv.c       |  18 +-
 .../RISCV/rvv-intrinsics-overloaded/vfrec7.c       |  18 +-
 .../RISCV/rvv-intrinsics-overloaded/vfredmax.c     |  18 +-
 .../RISCV/rvv-intrinsics-overloaded/vfredmin.c     |  18 +-
 .../RISCV/rvv-intrinsics-overloaded/vfredsum.c     |  36 +-
 .../RISCV/rvv-intrinsics-overloaded/vfrsqrt7.c     |  18 +-
 .../RISCV/rvv-intrinsics-overloaded/vfrsub.c       |  18 +-
 .../RISCV/rvv-intrinsics-overloaded/vfsgnj.c       | 108 +--
 .../RISCV/rvv-intrinsics-overloaded/vfslide1down.c |  18 +-
 .../RISCV/rvv-intrinsics-overloaded/vfslide1up.c   |  18 +-
 .../RISCV/rvv-intrinsics-overloaded/vfsqrt.c       |  18 +-
 .../RISCV/rvv-intrinsics-overloaded/vfsub.c        |  36 +-
 .../RISCV/rvv-intrinsics-overloaded/vfwadd.c       |  32 +-
 .../RISCV/rvv-intrinsics-overloaded/vfwcvt.c       |  76 +-
 .../RISCV/rvv-intrinsics-overloaded/vfwmacc.c      |  16 +-
 .../RISCV/rvv-intrinsics-overloaded/vfwmsac.c      |  16 +-
 .../RISCV/rvv-intrinsics-overloaded/vfwmul.c       |  16 +-
 .../RISCV/rvv-intrinsics-overloaded/vfwnmacc.c     |  16 +-
 .../RISCV/rvv-intrinsics-overloaded/vfwnmsac.c     |  16 +-
 .../RISCV/rvv-intrinsics-overloaded/vfwredsum.c    |  20 +-
 .../RISCV/rvv-intrinsics-overloaded/vfwsub.c       |  32 +-
 .../CodeGen/RISCV/rvv-intrinsics-overloaded/vid.c  |  44 +-
 .../RISCV/rvv-intrinsics-overloaded/viota.c        |  44 +-
 .../CodeGen/RISCV/rvv-intrinsics-overloaded/vle.c  | 106 +--
 .../RISCV/rvv-intrinsics-overloaded/vloxei.c       | 382 +++++------
 .../CodeGen/RISCV/rvv-intrinsics-overloaded/vlse.c | 106 +--
 .../RISCV/rvv-intrinsics-overloaded/vluxei.c       | 382 +++++------
 .../RISCV/rvv-intrinsics-overloaded/vmacc.c        | 176 ++---
 .../RISCV/rvv-intrinsics-overloaded/vmadd.c        | 176 ++---
 .../CodeGen/RISCV/rvv-intrinsics-overloaded/vmax.c | 176 ++---
 .../RISCV/rvv-intrinsics-overloaded/vmerge.c       | 194 +++---
 .../RISCV/rvv-intrinsics-overloaded/vmfeq.c        |  36 +-
 .../RISCV/rvv-intrinsics-overloaded/vmfge.c        |  36 +-
 .../RISCV/rvv-intrinsics-overloaded/vmfgt.c        |  36 +-
 .../RISCV/rvv-intrinsics-overloaded/vmfle.c        |  36 +-
 .../RISCV/rvv-intrinsics-overloaded/vmflt.c        |  36 +-
 .../RISCV/rvv-intrinsics-overloaded/vmfne.c        |  36 +-
 .../CodeGen/RISCV/rvv-intrinsics-overloaded/vmin.c | 176 ++---
 .../RISCV/rvv-intrinsics-overloaded/vmsbf.c        |  14 +-
 .../RISCV/rvv-intrinsics-overloaded/vmseq.c        | 176 ++---
 .../RISCV/rvv-intrinsics-overloaded/vmsge.c        | 176 ++---
 .../RISCV/rvv-intrinsics-overloaded/vmsgt.c        | 176 ++---
 .../RISCV/rvv-intrinsics-overloaded/vmsif.c        |  14 +-
 .../RISCV/rvv-intrinsics-overloaded/vmsle.c        | 176 ++---
 .../RISCV/rvv-intrinsics-overloaded/vmslt.c        | 176 ++---
 .../RISCV/rvv-intrinsics-overloaded/vmsne.c        | 176 ++---
 .../RISCV/rvv-intrinsics-overloaded/vmsof.c        |  14 +-
 .../CodeGen/RISCV/rvv-intrinsics-overloaded/vmul.c | 440 ++++++------
 .../RISCV/rvv-intrinsics-overloaded/vnclip.c       | 120 ++--
 .../RISCV/rvv-intrinsics-overloaded/vncvt.c        |  60 +-
 .../CodeGen/RISCV/rvv-intrinsics-overloaded/vneg.c |  44 +-
 .../RISCV/rvv-intrinsics-overloaded/vnmsac.c       | 176 ++---
 .../RISCV/rvv-intrinsics-overloaded/vnmsub.c       | 176 ++---
 .../CodeGen/RISCV/rvv-intrinsics-overloaded/vnot.c |  88 +--
 .../RISCV/rvv-intrinsics-overloaded/vnsra.c        |  60 +-
 .../RISCV/rvv-intrinsics-overloaded/vnsrl.c        |  60 +-
 .../CodeGen/RISCV/rvv-intrinsics-overloaded/vor.c  | 176 ++---
 .../RISCV/rvv-intrinsics-overloaded/vpopc.c        |  14 +-
 .../RISCV/rvv-intrinsics-overloaded/vredand.c      |  88 +--
 .../RISCV/rvv-intrinsics-overloaded/vredmax.c      |  88 +--
 .../RISCV/rvv-intrinsics-overloaded/vredmin.c      |  88 +--
 .../RISCV/rvv-intrinsics-overloaded/vredor.c       |  88 +--
 .../RISCV/rvv-intrinsics-overloaded/vredsum.c      |  88 +--
 .../RISCV/rvv-intrinsics-overloaded/vredxor.c      |  88 +--
 .../CodeGen/RISCV/rvv-intrinsics-overloaded/vrem.c | 176 ++---
 .../RISCV/rvv-intrinsics-overloaded/vrgather.c     | 314 ++++-----
 .../RISCV/rvv-intrinsics-overloaded/vrsub.c        |  88 +--
 .../RISCV/rvv-intrinsics-overloaded/vsadd.c        | 176 ++---
 .../CodeGen/RISCV/rvv-intrinsics-overloaded/vse.c  | 226 +++---
 .../RISCV/rvv-intrinsics-overloaded/vsext.c        |  56 +-
 .../RISCV/rvv-intrinsics-overloaded/vslide1down.c  |  88 +--
 .../RISCV/rvv-intrinsics-overloaded/vslide1up.c    |  88 +--
 .../RISCV/rvv-intrinsics-overloaded/vslidedown.c   | 106 +--
 .../RISCV/rvv-intrinsics-overloaded/vslideup.c     | 106 +--
 .../CodeGen/RISCV/rvv-intrinsics-overloaded/vsll.c | 176 ++---
 .../RISCV/rvv-intrinsics-overloaded/vsmul.c        |  88 +--
 .../RISCV/rvv-intrinsics-overloaded/vsoxei.c       | 764 ++++++++++-----------
 .../CodeGen/RISCV/rvv-intrinsics-overloaded/vsra.c |  88 +--
 .../CodeGen/RISCV/rvv-intrinsics-overloaded/vsrl.c |  88 +--
 .../CodeGen/RISCV/rvv-intrinsics-overloaded/vsse.c | 212 +++---
 .../RISCV/rvv-intrinsics-overloaded/vssra.c        |  88 +--
 .../RISCV/rvv-intrinsics-overloaded/vssrl.c        |  88 +--
 .../RISCV/rvv-intrinsics-overloaded/vssub.c        | 176 ++---
 .../CodeGen/RISCV/rvv-intrinsics-overloaded/vsub.c | 176 ++---
 .../RISCV/rvv-intrinsics-overloaded/vsuxei.c       | 764 ++++++++++-----------
 .../RISCV/rvv-intrinsics-overloaded/vwadd.c        | 240 +++----
 .../RISCV/rvv-intrinsics-overloaded/vwcvt.c        |  60 +-
 .../RISCV/rvv-intrinsics-overloaded/vwmacc.c       | 210 +++---
 .../RISCV/rvv-intrinsics-overloaded/vwmul.c        | 180 ++---
 .../RISCV/rvv-intrinsics-overloaded/vwredsum.c     |  72 +-
 .../RISCV/rvv-intrinsics-overloaded/vwsub.c        | 240 +++----
 .../CodeGen/RISCV/rvv-intrinsics-overloaded/vxor.c | 176 ++---
 .../RISCV/rvv-intrinsics-overloaded/vzext.c        |  56 +-
 clang/utils/TableGen/RISCVVEmitter.cpp             | 106 +--
 128 files changed, 7068 insertions(+), 7090 deletions(-)

diff --git a/clang/include/clang/Basic/riscv_vector.td b/clang/include/clang/Basic/riscv_vector.td
index 0502441..46f1a31 100644
--- a/clang/include/clang/Basic/riscv_vector.td
+++ b/clang/include/clang/Basic/riscv_vector.td
@@ -189,12 +189,6 @@ class RVVBuiltin<string suffix, string prototype, string type_range,
   // parameter of the unmasked version. k can't be the mask operand's position.
   list<int> IntrinsicTypes = [];
 
-  // When the order of the parameters of clang builtin do not match the order of
-  // C/C++ api, we use permutation index to mapping the operand from clang
-  // builtin to C/C++. It is parameter of the unmasked version without VL
-  // operand. If empty, the default permutation is [0, 1, 2, ...].
-  list<int> PermuteOperands = [];
-
   // If these names are not empty, this is the ID of the LLVM intrinsic
   // we want to lower to.
   string IRName = NAME;
@@ -562,6 +556,8 @@ let HasNoMaskedOverloaded = false,
       Ops[0] = Builder.CreateBitCast(Ops[0], ResultType->getPointerTo());
     }],
     ManualCodegenMask= [{
+      // Move mask to right before vl.
+      std::rotate(Ops.begin(), Ops.begin() + 1, Ops.end() - 1);
       IntrinsicTypes = {ResultType, Ops[3]->getType()};
       Ops[1] = Builder.CreateBitCast(Ops[1], ResultType->getPointerTo());
     }] in {
@@ -608,6 +604,8 @@ multiclass RVVVLEFFBuiltin<list<string> types> {
       }],
       ManualCodegenMask = [{
       {
+        // Move mask to right before vl.
+        std::rotate(Ops.begin(), Ops.begin() + 1, Ops.end() - 1);
         IntrinsicTypes = {ResultType, Ops[4]->getType()};
         Ops[1] = Builder.CreateBitCast(Ops[1], ResultType->getPointerTo());
         Value *NewVL = Ops[2];
@@ -643,6 +641,8 @@ multiclass RVVVLSEBuiltin<list<string> types> {
         Ops[0] = Builder.CreateBitCast(Ops[0], ResultType->getPointerTo());
       }],
       ManualCodegenMask= [{
+        // Move mask to right before vl.
+        std::rotate(Ops.begin(), Ops.begin() + 1, Ops.end() - 1);
         IntrinsicTypes = {ResultType, Ops[4]->getType()};
         Ops[1] = Builder.CreateBitCast(Ops[1], ResultType->getPointerTo());
       }] in {
@@ -661,6 +661,8 @@ multiclass RVVIndexedLoad<string op> {
         Ops[0] = Builder.CreateBitCast(Ops[0], ResultType->getPointerTo());
       }],
       ManualCodegenMask = [{
+        // Move mask to right before vl.
+        std::rotate(Ops.begin(), Ops.begin() + 1, Ops.end() - 1);
         IntrinsicTypes = {ResultType, Ops[2]->getType(), Ops[4]->getType()};
         Ops[1] = Builder.CreateBitCast(Ops[1], ResultType->getPointerTo());
       }] in {
@@ -680,16 +682,19 @@ multiclass RVVIndexedLoad<string op> {
 }
 
 let HasMaskedOffOperand = false,
-    PermuteOperands = [1, 0], // C/C++ Operand: (ptr, value, vl). Builtin: (value, ptr, vl)
     ManualCodegen = [{
+      // Builtin: (ptr, value, vl). Intrinsic: (value, ptr, vl)
+      std::swap(Ops[0], Ops[1]);
       Ops[1] = Builder.CreateBitCast(Ops[1], Ops[0]->getType()->getPointerTo());
       IntrinsicTypes = {Ops[0]->getType(), Ops[2]->getType()};
     }],
     ManualCodegenMask= [{
+      // Builtin: (mask, ptr, value, vl). Intrinsic: (value, ptr, mask, vl)
+      std::swap(Ops[0], Ops[2]);
       Ops[1] = Builder.CreateBitCast(Ops[1], Ops[0]->getType()->getPointerTo());
       IntrinsicTypes = {Ops[0]->getType(), Ops[3]->getType()};
     }] in {
-  class RVVVSEMaskBuiltin : RVVBuiltin<"m", "0mPUe", "c"> {
+  class RVVVSEMaskBuiltin : RVVBuiltin<"m", "0PUem", "c"> {
     let Name = "vse1_v";
     let IRName = "vse1";
     let HasMask = false;
@@ -699,9 +704,9 @@ let HasMaskedOffOperand = false,
         IRName = "vse",
         IRNameMask = "vse_mask" in {
       foreach type = types in {
-        def : RVVBuiltin<"v", "0vPe", type>;
+        def : RVVBuiltin<"v", "0Pev", type>;
         if !not(IsFloat<type>.val) then {
-          def : RVVBuiltin<"Uv", "0UvPUe", type>;
+          def : RVVBuiltin<"Uv", "0PUeUv", type>;
         }
       }
     }
@@ -713,19 +718,22 @@ multiclass RVVVSSEBuiltin<list<string> types> {
       IRName = "vsse",
       IRNameMask = "vsse_mask",
       HasMaskedOffOperand = false,
-      PermuteOperands = [1, 2, 0], // C/C++ Operand: (ptr, stride, value, vl). Builtin: (value, ptr, stride, vl)
       ManualCodegen = [{
+        // Builtin: (ptr, stride, value, vl). Intrinsic: (value, ptr, stride, vl)
+        std::rotate(Ops.begin(), Ops.begin() + 2, Ops.begin() + 3);
         Ops[1] = Builder.CreateBitCast(Ops[1], Ops[0]->getType()->getPointerTo());
         IntrinsicTypes = {Ops[0]->getType(), Ops[3]->getType()};
       }],
       ManualCodegenMask= [{
+        // Builtin: (mask, ptr, stride, value, vl). Intrinsic: (value, ptr, stride, mask, vl)
+        std::swap(Ops[0], Ops[3]);
         Ops[1] = Builder.CreateBitCast(Ops[1], Ops[0]->getType()->getPointerTo());
         IntrinsicTypes = {Ops[0]->getType(), Ops[4]->getType()};
       }] in {
     foreach type = types in {
-      def : RVVBuiltin<"v", "0vPet", type>;
+      def : RVVBuiltin<"v", "0Petv", type>;
       if !not(IsFloat<type>.val) then {
-        def : RVVBuiltin<"Uv", "0UvPUet", type>;
+        def : RVVBuiltin<"Uv", "0PUetUv", type>;
       }
     }
   }
@@ -733,12 +741,15 @@ multiclass RVVVSSEBuiltin<list<string> types> {
 
 multiclass RVVIndexedStore<string op> {
   let HasMaskedOffOperand = false,
-      PermuteOperands = [1, 2, 0], // C/C++ Operand: (ptr, index, value, vl). Builtin: (value, ptr, index, vl)
       ManualCodegen = [{
+        // Builtin: (ptr, index, value, vl). Intrinsic: (value, ptr, index, vl)
+        std::rotate(Ops.begin(), Ops.begin() + 2, Ops.begin() + 3);
         Ops[1] = Builder.CreateBitCast(Ops[1],Ops[0]->getType()->getPointerTo());
         IntrinsicTypes = {Ops[0]->getType(), Ops[2]->getType(), Ops[3]->getType()};
       }],
       ManualCodegenMask= [{
+        // Builtin: (mask, ptr, index, value, vl). Intrinsic: (value, ptr, index, mask, vl)
+        std::swap(Ops[0], Ops[3]);
         Ops[1] = Builder.CreateBitCast(Ops[1], Ops[0]->getType()->getPointerTo());
         IntrinsicTypes = {Ops[0]->getType(), Ops[2]->getType(), Ops[4]->getType()};
       }] in {
@@ -747,9 +758,9 @@ multiclass RVVIndexedStore<string op> {
           defvar eew = eew_list[0];
           defvar eew_type = eew_list[1];
           let Name = op # eew  # "_v", IRName = op, IRNameMask = op # "_mask" in  {
-            def : RVVBuiltin<"v", "0vPe" # eew_type # "Uv", type>;
+            def : RVVBuiltin<"v", "0Pe" # eew_type # "Uvv", type>;
             if !not(IsFloat<type>.val) then {
-              def : RVVBuiltin<"Uv", "0UvPUe" # eew_type # "Uv", type>;
+              def : RVVBuiltin<"Uv", "0PUe" # eew_type # "UvUv", type>;
             }
           }
         }
@@ -774,6 +785,7 @@ multiclass RVVAMOBuiltinSet<bit has_signed = false, bit has_unsigned = false,
             Ops[0] = Builder.CreateBitCast(Ops[0], ResultType->getPointerTo());
           }],
           ManualCodegenMask = [{
+            std::rotate(Ops.begin(), Ops.begin() + 1, Ops.end() - 1);
             IntrinsicTypes = {ResultType, Ops[1]->getType(), Ops[4]->getType()};
             Ops[0] = Builder.CreateBitCast(Ops[0], ResultType->getPointerTo());
           }] in {
@@ -801,6 +813,7 @@ multiclass RVVPseudoUnaryBuiltin<string IR, string type_range> {
       }],
       ManualCodegenMask = [{
       {
+        std::rotate(Ops.begin(), Ops.begin() + 1, Ops.end() - 1);
         // maskedoff, op1, mask, vl
         IntrinsicTypes = {ResultType,
                           cast<llvm::VectorType>(ResultType)->getElementType(),
@@ -830,6 +843,7 @@ multiclass RVVPseudoVNotBuiltin<string IR, string type_range> {
       }],
       ManualCodegenMask = [{
       {
+        std::rotate(Ops.begin(), Ops.begin() + 1, Ops.end() - 1);
         // maskedoff, op1, mask, vl
         IntrinsicTypes = {ResultType,
                           cast<llvm::VectorType>(ResultType)->getElementType(),
@@ -876,6 +890,7 @@ multiclass RVVPseudoVFUnaryBuiltin<string IR, string type_range> {
       }],
       ManualCodegenMask = [{
       {
+        std::rotate(Ops.begin(), Ops.begin() + 1, Ops.end() - 1);
         // maskedoff, op1, mask, vl
         IntrinsicTypes = {ResultType,
                           Ops[1]->getType(),
@@ -907,6 +922,7 @@ multiclass RVVPseudoVWCVTBuiltin<string IR, string MName, string type_range,
       }],
       ManualCodegenMask = [{
       {
+        std::rotate(Ops.begin(), Ops.begin() + 1, Ops.end() - 1);
         // maskedoff, op1, mask, vl
         IntrinsicTypes = {ResultType,
                           Ops[1]->getType(),
@@ -941,6 +957,7 @@ multiclass RVVPseudoVNCVTBuiltin<string IR, string MName, string type_range,
       }],
       ManualCodegenMask = [{
       {
+        std::rotate(Ops.begin(), Ops.begin() + 1, Ops.end() - 1);
         // maskedoff, op1, mask, vl
         IntrinsicTypes = {ResultType,
                           Ops[1]->getType(),
@@ -1216,9 +1233,17 @@ defm vwmaccus : RVVOutOp1Op2BuiltinSet<"vwmaccus", "csi",
 }
 
 // 12.15. Vector Integer Merge Instructions
-// C/C++ Operand: (mask, op1, op2, vl), Builtin: (op1, op2, mask, vl)
-let HasMask = false, PermuteOperands = [2, 0, 1] in {
-  defm vmerge : RVVCarryinBuiltinSet;
+// C/C++ Operand: (mask, op1, op2, vl), Intrinsic: (op1, op2, mask, vl)
+let HasMask = false,
+    ManualCodegen = [{
+      std::rotate(Ops.begin(), Ops.begin() + 1, Ops.begin() + 3);
+      IntrinsicTypes = {ResultType, Ops[1]->getType(), Ops[3]->getType()};
+    }] in {
+  defm vmerge : RVVOutOp1BuiltinSet<"vmerge", "csil",
+                                    [["vvm", "v", "vmvv"],
+                                     ["vxm", "v", "vmve"],
+                                     ["vvm", "Uv", "UvmUvUv"],
+                                     ["vxm", "Uv", "UvmUvUe"]]>;
 }
 
 // 12.16. Vector Integer Move Instructions
@@ -1335,11 +1360,15 @@ let Name = "vfclass_v" in
 
 // 14.15. Vector Floating-Point Merge Instructio
 // C/C++ Operand: (mask, op1, op2, vl), Builtin: (op1, op2, mask, vl)
-let HasMask = false, PermuteOperands = [2, 0, 1] in {
+let HasMask = false,
+    ManualCodegen = [{
+      std::rotate(Ops.begin(), Ops.begin() + 1, Ops.begin() + 3);
+      IntrinsicTypes = {ResultType, Ops[1]->getType(), Ops[3]->getType()};
+    }] in {
   defm vmerge : RVVOutOp1BuiltinSet<"vfmerge", "fd",
-                                    [["vvm", "v", "vvvm"]]>;
+                                    [["vvm", "v", "vmvv"]]>;
   defm vfmerge : RVVOutOp1BuiltinSet<"vfmerge", "fd",
-                                    [["vfm", "v", "vvem"]]>;
+                                    [["vfm", "v", "vmve"]]>;
 }
 
 // 14.16. Vector Floating-Point Move Instruction
@@ -1503,13 +1532,18 @@ defm vrgatherei16 : RVVOutBuiltinSet<"vrgatherei16_vv", "csil",
                                      [["vv", "Uv", "UvUv(Log2EEW:4)Uv"]]>;
 
 // 17.5. Vector Compress Instruction
-let HasMask = false, PermuteOperands = [2, 0, 1] in {
+let HasMask = false,
+    ManualCodegen = [{
+      std::rotate(Ops.begin(), Ops.begin() + 1, Ops.begin() + 3);
+      ID = Intrinsic::riscv_vcompress;
+       IntrinsicTypes = {ResultType, Ops[3]->getType()};
+    }] in {
   // signed and floating type
   defm vcompress : RVVOutBuiltinSet<"vcompress", "csilfd",
-                                    [["vm", "v", "vvvm"]]>;
+                                    [["vm", "v", "vmvv"]]>;
   // unsigned type
   defm vcompress : RVVOutBuiltinSet<"vcompress", "csil",
-                                    [["vm", "Uv", "UvUvUvm"]]>;
+                                    [["vm", "Uv", "UvmUvUv"]]>;
 }
 
 // Miscellaneous
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vaadd.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vaadd.c
index cc3a9e3..1f8361a 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vaadd.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vaadd.c
@@ -890,7 +890,7 @@ vuint64m8_t test_vaaddu_vx_u64m8(vuint64m8_t op1, uint64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vaadd.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vaadd.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vaadd_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
@@ -901,7 +901,7 @@ vint8mf8_t test_vaadd_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vaadd.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vaadd.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vaadd_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
@@ -912,7 +912,7 @@ vint8mf8_t test_vaadd_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vaadd.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vaadd.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vaadd_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
@@ -923,7 +923,7 @@ vint8mf4_t test_vaadd_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vaadd.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vaadd.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vaadd_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
@@ -934,7 +934,7 @@ vint8mf4_t test_vaadd_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vaadd.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vaadd.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vaadd_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
@@ -945,7 +945,7 @@ vint8mf2_t test_vaadd_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vaadd.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vaadd.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vaadd_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
@@ -956,7 +956,7 @@ vint8mf2_t test_vaadd_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vaadd.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vaadd.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vaadd_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
@@ -967,7 +967,7 @@ vint8m1_t test_vaadd_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vaadd.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vaadd.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vaadd_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
@@ -978,7 +978,7 @@ vint8m1_t test_vaadd_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vaadd.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vaadd.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vaadd_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
@@ -989,7 +989,7 @@ vint8m2_t test_vaadd_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vaadd.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vaadd.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vaadd_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
@@ -1000,7 +1000,7 @@ vint8m2_t test_vaadd_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vaadd.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vaadd.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vaadd_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
@@ -1011,7 +1011,7 @@ vint8m4_t test_vaadd_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vaadd.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vaadd.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vaadd_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
@@ -1022,7 +1022,7 @@ vint8m4_t test_vaadd_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vaadd.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vaadd.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vaadd_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
@@ -1033,7 +1033,7 @@ vint8m8_t test_vaadd_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vaadd.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vaadd.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vaadd_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
@@ -1044,7 +1044,7 @@ vint8m8_t test_vaadd_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vaadd.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vaadd.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vaadd_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -1056,7 +1056,7 @@ vint16mf4_t test_vaadd_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vaadd.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vaadd.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vaadd_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -1067,7 +1067,7 @@ vint16mf4_t test_vaadd_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vaadd.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vaadd.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vaadd_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -1079,7 +1079,7 @@ vint16mf2_t test_vaadd_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vaadd.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vaadd.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vaadd_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -1090,7 +1090,7 @@ vint16mf2_t test_vaadd_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vaadd.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vaadd.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vaadd_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -1101,7 +1101,7 @@ vint16m1_t test_vaadd_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vaadd.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vaadd.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vaadd_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -1112,7 +1112,7 @@ vint16m1_t test_vaadd_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vaadd.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vaadd.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vaadd_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -1123,7 +1123,7 @@ vint16m2_t test_vaadd_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vaadd.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vaadd.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vaadd_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -1134,7 +1134,7 @@ vint16m2_t test_vaadd_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vaadd.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vaadd.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vaadd_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -1145,7 +1145,7 @@ vint16m4_t test_vaadd_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vaadd.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vaadd.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vaadd_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -1156,7 +1156,7 @@ vint16m4_t test_vaadd_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vaadd.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vaadd.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vaadd_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -1167,7 +1167,7 @@ vint16m8_t test_vaadd_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vaadd.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vaadd.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vaadd_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -1178,7 +1178,7 @@ vint16m8_t test_vaadd_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vaadd.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vaadd.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vaadd_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -1190,7 +1190,7 @@ vint32mf2_t test_vaadd_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vaadd.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vaadd.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vaadd_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -1201,7 +1201,7 @@ vint32mf2_t test_vaadd_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vaadd.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vaadd.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vaadd_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -1212,7 +1212,7 @@ vint32m1_t test_vaadd_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vaadd.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vaadd.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vaadd_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -1223,7 +1223,7 @@ vint32m1_t test_vaadd_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vaadd.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vaadd.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vaadd_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -1234,7 +1234,7 @@ vint32m2_t test_vaadd_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vaadd.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vaadd.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vaadd_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -1245,7 +1245,7 @@ vint32m2_t test_vaadd_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vaadd.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vaadd.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vaadd_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -1256,7 +1256,7 @@ vint32m4_t test_vaadd_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vaadd.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vaadd.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vaadd_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -1267,7 +1267,7 @@ vint32m4_t test_vaadd_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vaadd.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vaadd.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vaadd_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -1278,7 +1278,7 @@ vint32m8_t test_vaadd_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vaadd.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vaadd.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vaadd_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -1289,7 +1289,7 @@ vint32m8_t test_vaadd_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vaadd.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vaadd.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vaadd_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -1300,7 +1300,7 @@ vint64m1_t test_vaadd_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vaadd.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vaadd.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vaadd_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -1311,7 +1311,7 @@ vint64m1_t test_vaadd_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vaadd.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vaadd.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vaadd_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -1322,7 +1322,7 @@ vint64m2_t test_vaadd_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vaadd.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vaadd.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vaadd_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -1333,7 +1333,7 @@ vint64m2_t test_vaadd_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vaadd.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vaadd.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vaadd_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -1344,7 +1344,7 @@ vint64m4_t test_vaadd_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vaadd.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vaadd.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vaadd_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -1355,7 +1355,7 @@ vint64m4_t test_vaadd_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vaadd.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vaadd.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vaadd_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -1366,7 +1366,7 @@ vint64m8_t test_vaadd_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaadd_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vaadd.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vaadd.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vaadd_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -1377,7 +1377,7 @@ vint64m8_t test_vaadd_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vaaddu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vaaddu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vaaddu_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
@@ -1389,7 +1389,7 @@ vuint8mf8_t test_vaaddu_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vaaddu.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vaaddu.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vaaddu_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
@@ -1400,7 +1400,7 @@ vuint8mf8_t test_vaaddu_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vaaddu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vaaddu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vaaddu_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
@@ -1412,7 +1412,7 @@ vuint8mf4_t test_vaaddu_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vaaddu.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vaaddu.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vaaddu_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
@@ -1423,7 +1423,7 @@ vuint8mf4_t test_vaaddu_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vaaddu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vaaddu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vaaddu_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
@@ -1435,7 +1435,7 @@ vuint8mf2_t test_vaaddu_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vaaddu.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vaaddu.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vaaddu_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
@@ -1446,7 +1446,7 @@ vuint8mf2_t test_vaaddu_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vaaddu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vaaddu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vaaddu_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
@@ -1457,7 +1457,7 @@ vuint8m1_t test_vaaddu_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vaaddu.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vaaddu.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vaaddu_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
@@ -1468,7 +1468,7 @@ vuint8m1_t test_vaaddu_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vaaddu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vaaddu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vaaddu_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
@@ -1479,7 +1479,7 @@ vuint8m2_t test_vaaddu_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vaaddu.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vaaddu.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vaaddu_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
@@ -1490,7 +1490,7 @@ vuint8m2_t test_vaaddu_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vaaddu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vaaddu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vaaddu_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
@@ -1501,7 +1501,7 @@ vuint8m4_t test_vaaddu_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vaaddu.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vaaddu.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vaaddu_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
@@ -1512,7 +1512,7 @@ vuint8m4_t test_vaaddu_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vaaddu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vaaddu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vaaddu_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
@@ -1523,7 +1523,7 @@ vuint8m8_t test_vaaddu_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vaaddu.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vaaddu.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vaaddu_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
@@ -1534,7 +1534,7 @@ vuint8m8_t test_vaaddu_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vaaddu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vaaddu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vaaddu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -1546,7 +1546,7 @@ vuint16mf4_t test_vaaddu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vaaddu.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vaaddu.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vaaddu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -1558,7 +1558,7 @@ vuint16mf4_t test_vaaddu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vaaddu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vaaddu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vaaddu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -1570,7 +1570,7 @@ vuint16mf2_t test_vaaddu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vaaddu.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vaaddu.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vaaddu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -1582,7 +1582,7 @@ vuint16mf2_t test_vaaddu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vaaddu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vaaddu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vaaddu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -1594,7 +1594,7 @@ vuint16m1_t test_vaaddu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vaaddu.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vaaddu.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vaaddu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -1605,7 +1605,7 @@ vuint16m1_t test_vaaddu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vaaddu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vaaddu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vaaddu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -1617,7 +1617,7 @@ vuint16m2_t test_vaaddu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vaaddu.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vaaddu.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vaaddu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -1628,7 +1628,7 @@ vuint16m2_t test_vaaddu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vaaddu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vaaddu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vaaddu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -1640,7 +1640,7 @@ vuint16m4_t test_vaaddu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vaaddu.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vaaddu.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vaaddu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -1651,7 +1651,7 @@ vuint16m4_t test_vaaddu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vaaddu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vaaddu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vaaddu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -1663,7 +1663,7 @@ vuint16m8_t test_vaaddu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vaaddu.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vaaddu.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vaaddu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -1674,7 +1674,7 @@ vuint16m8_t test_vaaddu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vaaddu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vaaddu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vaaddu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -1686,7 +1686,7 @@ vuint32mf2_t test_vaaddu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vaaddu.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vaaddu.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vaaddu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -1698,7 +1698,7 @@ vuint32mf2_t test_vaaddu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vaaddu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vaaddu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vaaddu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -1710,7 +1710,7 @@ vuint32m1_t test_vaaddu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vaaddu.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vaaddu.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vaaddu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -1721,7 +1721,7 @@ vuint32m1_t test_vaaddu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vaaddu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vaaddu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vaaddu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -1733,7 +1733,7 @@ vuint32m2_t test_vaaddu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vaaddu.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vaaddu.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vaaddu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -1744,7 +1744,7 @@ vuint32m2_t test_vaaddu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vaaddu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vaaddu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vaaddu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -1756,7 +1756,7 @@ vuint32m4_t test_vaaddu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vaaddu.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vaaddu.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vaaddu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -1767,7 +1767,7 @@ vuint32m4_t test_vaaddu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vaaddu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vaaddu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vaaddu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -1779,7 +1779,7 @@ vuint32m8_t test_vaaddu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vaaddu.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vaaddu.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vaaddu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -1790,7 +1790,7 @@ vuint32m8_t test_vaaddu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vaaddu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vaaddu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vaaddu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -1802,7 +1802,7 @@ vuint64m1_t test_vaaddu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vaaddu.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vaaddu.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vaaddu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -1813,7 +1813,7 @@ vuint64m1_t test_vaaddu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vaaddu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vaaddu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vaaddu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -1825,7 +1825,7 @@ vuint64m2_t test_vaaddu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vaaddu.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vaaddu.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vaaddu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -1836,7 +1836,7 @@ vuint64m2_t test_vaaddu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vaaddu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vaaddu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vaaddu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -1848,7 +1848,7 @@ vuint64m4_t test_vaaddu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vaaddu.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vaaddu.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vaaddu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -1859,7 +1859,7 @@ vuint64m4_t test_vaaddu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vaaddu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vaaddu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vaaddu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
@@ -1871,7 +1871,7 @@ vuint64m8_t test_vaaddu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vaaddu_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vaaddu.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vaaddu.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vaaddu_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vadd.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vadd.c
index bcd1557..ae3a230 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vadd.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vadd.c
@@ -888,7 +888,7 @@ vuint64m8_t test_vadd_vx_u64m8(vuint64m8_t op1, uint64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vadd.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vadd.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vadd_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, vint8mf8_t op2, size_t vl) {
@@ -898,7 +898,7 @@ vint8mf8_t test_vadd_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vadd.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vadd.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vadd_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, int8_t op2, size_t vl) {
@@ -908,7 +908,7 @@ vint8mf8_t test_vadd_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vadd.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vadd.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vadd_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, vint8mf4_t op2, size_t vl) {
@@ -918,7 +918,7 @@ vint8mf4_t test_vadd_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vadd.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vadd.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vadd_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, int8_t op2, size_t vl) {
@@ -928,7 +928,7 @@ vint8mf4_t test_vadd_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vadd.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vadd.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vadd_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, vint8mf2_t op2, size_t vl) {
@@ -938,7 +938,7 @@ vint8mf2_t test_vadd_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vadd.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vadd.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vadd_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, int8_t op2, size_t vl) {
@@ -948,7 +948,7 @@ vint8mf2_t test_vadd_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vadd.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vadd.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vadd_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, vint8m1_t op2, size_t vl) {
@@ -958,7 +958,7 @@ vint8m1_t test_vadd_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vadd.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vadd.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vadd_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, int8_t op2, size_t vl) {
@@ -968,7 +968,7 @@ vint8m1_t test_vadd_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vadd.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vadd.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vadd_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, vint8m2_t op2, size_t vl) {
@@ -978,7 +978,7 @@ vint8m2_t test_vadd_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vadd.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vadd.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vadd_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, int8_t op2, size_t vl) {
@@ -988,7 +988,7 @@ vint8m2_t test_vadd_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vadd.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vadd.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vadd_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, vint8m4_t op2, size_t vl) {
@@ -998,7 +998,7 @@ vint8m4_t test_vadd_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vadd.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vadd.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vadd_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, int8_t op2, size_t vl) {
@@ -1008,7 +1008,7 @@ vint8m4_t test_vadd_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vadd.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vadd.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vadd_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, vint8m8_t op2, size_t vl) {
@@ -1018,7 +1018,7 @@ vint8m8_t test_vadd_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vadd.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vadd.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vadd_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, int8_t op2, size_t vl) {
@@ -1028,7 +1028,7 @@ vint8m8_t test_vadd_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vadd.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vadd.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vadd_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, vint16mf4_t op2, size_t vl) {
@@ -1038,7 +1038,7 @@ vint16mf4_t test_vadd_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vadd.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vadd.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vadd_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, int16_t op2, size_t vl) {
@@ -1048,7 +1048,7 @@ vint16mf4_t test_vadd_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vadd.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vadd.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vadd_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, vint16mf2_t op2, size_t vl) {
@@ -1058,7 +1058,7 @@ vint16mf2_t test_vadd_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vadd.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vadd.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vadd_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, int16_t op2, size_t vl) {
@@ -1068,7 +1068,7 @@ vint16mf2_t test_vadd_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vadd.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vadd.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vadd_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, vint16m1_t op2, size_t vl) {
@@ -1078,7 +1078,7 @@ vint16m1_t test_vadd_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vadd.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vadd.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vadd_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, int16_t op2, size_t vl) {
@@ -1088,7 +1088,7 @@ vint16m1_t test_vadd_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vadd.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vadd.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vadd_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, vint16m2_t op2, size_t vl) {
@@ -1098,7 +1098,7 @@ vint16m2_t test_vadd_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vadd.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vadd.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vadd_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, int16_t op2, size_t vl) {
@@ -1108,7 +1108,7 @@ vint16m2_t test_vadd_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vadd.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vadd.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vadd_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, vint16m4_t op2, size_t vl) {
@@ -1118,7 +1118,7 @@ vint16m4_t test_vadd_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vadd.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vadd.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vadd_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, int16_t op2, size_t vl) {
@@ -1128,7 +1128,7 @@ vint16m4_t test_vadd_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vadd.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vadd.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vadd_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, vint16m8_t op2, size_t vl) {
@@ -1138,7 +1138,7 @@ vint16m8_t test_vadd_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vadd.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vadd.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vadd_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, int16_t op2, size_t vl) {
@@ -1148,7 +1148,7 @@ vint16m8_t test_vadd_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vadd.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vadd.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vadd_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, vint32mf2_t op2, size_t vl) {
@@ -1158,7 +1158,7 @@ vint32mf2_t test_vadd_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vadd.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vadd.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vadd_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, int32_t op2, size_t vl) {
@@ -1168,7 +1168,7 @@ vint32mf2_t test_vadd_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vadd.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vadd.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vadd_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, vint32m1_t op2, size_t vl) {
@@ -1178,7 +1178,7 @@ vint32m1_t test_vadd_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vadd.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vadd.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vadd_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, int32_t op2, size_t vl) {
@@ -1188,7 +1188,7 @@ vint32m1_t test_vadd_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vadd.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vadd.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vadd_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, vint32m2_t op2, size_t vl) {
@@ -1198,7 +1198,7 @@ vint32m2_t test_vadd_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vadd.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vadd.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vadd_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, int32_t op2, size_t vl) {
@@ -1208,7 +1208,7 @@ vint32m2_t test_vadd_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vadd.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vadd.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vadd_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, vint32m4_t op2, size_t vl) {
@@ -1218,7 +1218,7 @@ vint32m4_t test_vadd_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vadd.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vadd.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vadd_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, int32_t op2, size_t vl) {
@@ -1228,7 +1228,7 @@ vint32m4_t test_vadd_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vadd.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vadd.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vadd_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, vint32m8_t op2, size_t vl) {
@@ -1238,7 +1238,7 @@ vint32m8_t test_vadd_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vadd.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vadd.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vadd_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, int32_t op2, size_t vl) {
@@ -1248,7 +1248,7 @@ vint32m8_t test_vadd_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vadd.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vadd.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vadd_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, vint64m1_t op2, size_t vl) {
@@ -1258,7 +1258,7 @@ vint64m1_t test_vadd_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vadd.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vadd.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vadd_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, int64_t op2, size_t vl) {
@@ -1268,7 +1268,7 @@ vint64m1_t test_vadd_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vadd.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vadd.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vadd_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, vint64m2_t op2, size_t vl) {
@@ -1278,7 +1278,7 @@ vint64m2_t test_vadd_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vadd.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vadd.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vadd_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, int64_t op2, size_t vl) {
@@ -1288,7 +1288,7 @@ vint64m2_t test_vadd_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vadd.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vadd.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vadd_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, vint64m4_t op2, size_t vl) {
@@ -1298,7 +1298,7 @@ vint64m4_t test_vadd_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vadd.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vadd.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vadd_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, int64_t op2, size_t vl) {
@@ -1308,7 +1308,7 @@ vint64m4_t test_vadd_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vadd.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vadd.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vadd_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, vint64m8_t op2, size_t vl) {
@@ -1318,7 +1318,7 @@ vint64m8_t test_vadd_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vadd.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vadd.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vadd_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, int64_t op2, size_t vl) {
@@ -1328,7 +1328,7 @@ vint64m8_t test_vadd_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vadd.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vadd.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vadd_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -1338,7 +1338,7 @@ vuint8mf8_t test_vadd_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vadd.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vadd.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vadd_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, uint8_t op2, size_t vl) {
@@ -1348,7 +1348,7 @@ vuint8mf8_t test_vadd_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vadd.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vadd.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vadd_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, vuint8mf4_t op2, size_t vl) {
@@ -1358,7 +1358,7 @@ vuint8mf4_t test_vadd_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vadd.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vadd.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vadd_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, uint8_t op2, size_t vl) {
@@ -1368,7 +1368,7 @@ vuint8mf4_t test_vadd_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vadd.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vadd.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vadd_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, vuint8mf2_t op2, size_t vl) {
@@ -1378,7 +1378,7 @@ vuint8mf2_t test_vadd_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vadd.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vadd.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vadd_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, uint8_t op2, size_t vl) {
@@ -1388,7 +1388,7 @@ vuint8mf2_t test_vadd_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vadd.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vadd.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vadd_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, vuint8m1_t op2, size_t vl) {
@@ -1398,7 +1398,7 @@ vuint8m1_t test_vadd_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t o
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vadd.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vadd.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vadd_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, uint8_t op2, size_t vl) {
@@ -1408,7 +1408,7 @@ vuint8m1_t test_vadd_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t o
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vadd.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vadd.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vadd_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, vuint8m2_t op2, size_t vl) {
@@ -1418,7 +1418,7 @@ vuint8m2_t test_vadd_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t o
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vadd.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vadd.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vadd_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, uint8_t op2, size_t vl) {
@@ -1428,7 +1428,7 @@ vuint8m2_t test_vadd_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t o
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vadd.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vadd.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vadd_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, vuint8m4_t op2, size_t vl) {
@@ -1438,7 +1438,7 @@ vuint8m4_t test_vadd_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t o
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vadd.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vadd.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vadd_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, uint8_t op2, size_t vl) {
@@ -1448,7 +1448,7 @@ vuint8m4_t test_vadd_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t o
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vadd.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vadd.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vadd_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, vuint8m8_t op2, size_t vl) {
@@ -1458,7 +1458,7 @@ vuint8m8_t test_vadd_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t o
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vadd.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vadd.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vadd_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, uint8_t op2, size_t vl) {
@@ -1468,7 +1468,7 @@ vuint8m8_t test_vadd_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t o
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vadd.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vadd.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vadd_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, vuint16mf4_t op2, size_t vl) {
@@ -1478,7 +1478,7 @@ vuint16mf4_t test_vadd_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vadd.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vadd.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vadd_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, uint16_t op2, size_t vl) {
@@ -1488,7 +1488,7 @@ vuint16mf4_t test_vadd_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vadd.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vadd.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vadd_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, vuint16mf2_t op2, size_t vl) {
@@ -1498,7 +1498,7 @@ vuint16mf2_t test_vadd_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vadd.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vadd.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vadd_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, uint16_t op2, size_t vl) {
@@ -1508,7 +1508,7 @@ vuint16mf2_t test_vadd_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vadd.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vadd.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vadd_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, vuint16m1_t op2, size_t vl) {
@@ -1518,7 +1518,7 @@ vuint16m1_t test_vadd_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vadd.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vadd.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vadd_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, uint16_t op2, size_t vl) {
@@ -1528,7 +1528,7 @@ vuint16m1_t test_vadd_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vadd.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vadd.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vadd_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, vuint16m2_t op2, size_t vl) {
@@ -1538,7 +1538,7 @@ vuint16m2_t test_vadd_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vadd.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vadd.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vadd_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, uint16_t op2, size_t vl) {
@@ -1548,7 +1548,7 @@ vuint16m2_t test_vadd_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vadd.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vadd.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vadd_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, vuint16m4_t op2, size_t vl) {
@@ -1558,7 +1558,7 @@ vuint16m4_t test_vadd_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vadd.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vadd.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vadd_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, uint16_t op2, size_t vl) {
@@ -1568,7 +1568,7 @@ vuint16m4_t test_vadd_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vadd.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vadd.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vadd_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, vuint16m8_t op2, size_t vl) {
@@ -1578,7 +1578,7 @@ vuint16m8_t test_vadd_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vadd.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vadd.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vadd_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, uint16_t op2, size_t vl) {
@@ -1588,7 +1588,7 @@ vuint16m8_t test_vadd_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vadd.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vadd.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vadd_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, vuint32mf2_t op2, size_t vl) {
@@ -1598,7 +1598,7 @@ vuint32mf2_t test_vadd_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vadd.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vadd.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vadd_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, uint32_t op2, size_t vl) {
@@ -1608,7 +1608,7 @@ vuint32mf2_t test_vadd_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vadd.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vadd.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vadd_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, vuint32m1_t op2, size_t vl) {
@@ -1618,7 +1618,7 @@ vuint32m1_t test_vadd_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vadd.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vadd.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vadd_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, uint32_t op2, size_t vl) {
@@ -1628,7 +1628,7 @@ vuint32m1_t test_vadd_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vadd.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vadd.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vadd_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, vuint32m2_t op2, size_t vl) {
@@ -1638,7 +1638,7 @@ vuint32m2_t test_vadd_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vadd.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vadd.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vadd_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, uint32_t op2, size_t vl) {
@@ -1648,7 +1648,7 @@ vuint32m2_t test_vadd_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vadd.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vadd.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vadd_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, vuint32m4_t op2, size_t vl) {
@@ -1658,7 +1658,7 @@ vuint32m4_t test_vadd_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vadd.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vadd.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vadd_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, uint32_t op2, size_t vl) {
@@ -1668,7 +1668,7 @@ vuint32m4_t test_vadd_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vadd.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vadd.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vadd_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, vuint32m8_t op2, size_t vl) {
@@ -1678,7 +1678,7 @@ vuint32m8_t test_vadd_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vadd.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vadd.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vadd_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, uint32_t op2, size_t vl) {
@@ -1688,7 +1688,7 @@ vuint32m8_t test_vadd_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vadd.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vadd.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vadd_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, vuint64m1_t op2, size_t vl) {
@@ -1698,7 +1698,7 @@ vuint64m1_t test_vadd_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vadd.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vadd.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vadd_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, uint64_t op2, size_t vl) {
@@ -1708,7 +1708,7 @@ vuint64m1_t test_vadd_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vadd.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vadd.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vadd_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, vuint64m2_t op2, size_t vl) {
@@ -1718,7 +1718,7 @@ vuint64m2_t test_vadd_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vadd.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vadd.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vadd_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, uint64_t op2, size_t vl) {
@@ -1728,7 +1728,7 @@ vuint64m2_t test_vadd_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vadd.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vadd.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vadd_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, vuint64m4_t op2, size_t vl) {
@@ -1738,7 +1738,7 @@ vuint64m4_t test_vadd_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vadd.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vadd.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vadd_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, uint64_t op2, size_t vl) {
@@ -1748,7 +1748,7 @@ vuint64m4_t test_vadd_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vadd_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vadd.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vadd.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vadd_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, vuint64m8_t op2, size_t vl) {
@@ -1758,7 +1758,7 @@ vuint64m8_t test_vadd_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8
 //
 // CHECK-RV64-LABEL: @test_vadd_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vadd.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vadd.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vadd_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, uint64_t op2, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamoadd.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamoadd.c
index 92a9918..49ceb9a 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamoadd.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamoadd.c
@@ -778,7 +778,7 @@ vuint64m8_t test_vamoaddei64_v_u64m8 (uint64_t *base, vuint64m8_t bindex, vuint6
 // CHECK-RV64-LABEL: @test_vamoaddei8_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoadd.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoadd.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamoaddei8_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint8mf8_t bindex, vint32mf2_t value, size_t vl) {
@@ -789,7 +789,7 @@ vint32mf2_t test_vamoaddei8_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint8mf8
 // CHECK-RV64-LABEL: @test_vamoaddei8_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoadd.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoadd.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamoaddei8_v_i32m1_m (vbool32_t mask, int32_t *base, vuint8mf4_t bindex, vint32m1_t value, size_t vl) {
@@ -800,7 +800,7 @@ vint32m1_t test_vamoaddei8_v_i32m1_m (vbool32_t mask, int32_t *base, vuint8mf4_t
 // CHECK-RV64-LABEL: @test_vamoaddei8_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoadd.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoadd.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamoaddei8_v_i32m2_m (vbool16_t mask, int32_t *base, vuint8mf2_t bindex, vint32m2_t value, size_t vl) {
@@ -811,7 +811,7 @@ vint32m2_t test_vamoaddei8_v_i32m2_m (vbool16_t mask, int32_t *base, vuint8mf2_t
 // CHECK-RV64-LABEL: @test_vamoaddei8_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoadd.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoadd.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamoaddei8_v_i32m4_m (vbool8_t mask, int32_t *base, vuint8m1_t bindex, vint32m4_t value, size_t vl) {
@@ -822,7 +822,7 @@ vint32m4_t test_vamoaddei8_v_i32m4_m (vbool8_t mask, int32_t *base, vuint8m1_t b
 // CHECK-RV64-LABEL: @test_vamoaddei8_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoadd.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoadd.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vamoaddei8_v_i32m8_m (vbool4_t mask, int32_t *base, vuint8m2_t bindex, vint32m8_t value, size_t vl) {
@@ -833,7 +833,7 @@ vint32m8_t test_vamoaddei8_v_i32m8_m (vbool4_t mask, int32_t *base, vuint8m2_t b
 // CHECK-RV64-LABEL: @test_vamoaddei16_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoadd.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoadd.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamoaddei16_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint16mf4_t bindex, vint32mf2_t value, size_t vl) {
@@ -844,7 +844,7 @@ vint32mf2_t test_vamoaddei16_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamoaddei16_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoadd.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoadd.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamoaddei16_v_i32m1_m (vbool32_t mask, int32_t *base, vuint16mf2_t bindex, vint32m1_t value, size_t vl) {
@@ -855,7 +855,7 @@ vint32m1_t test_vamoaddei16_v_i32m1_m (vbool32_t mask, int32_t *base, vuint16mf2
 // CHECK-RV64-LABEL: @test_vamoaddei16_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoadd.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoadd.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamoaddei16_v_i32m2_m (vbool16_t mask, int32_t *base, vuint16m1_t bindex, vint32m2_t value, size_t vl) {
@@ -866,7 +866,7 @@ vint32m2_t test_vamoaddei16_v_i32m2_m (vbool16_t mask, int32_t *base, vuint16m1_
 // CHECK-RV64-LABEL: @test_vamoaddei16_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoadd.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoadd.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamoaddei16_v_i32m4_m (vbool8_t mask, int32_t *base, vuint16m2_t bindex, vint32m4_t value, size_t vl) {
@@ -877,7 +877,7 @@ vint32m4_t test_vamoaddei16_v_i32m4_m (vbool8_t mask, int32_t *base, vuint16m2_t
 // CHECK-RV64-LABEL: @test_vamoaddei16_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoadd.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoadd.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vamoaddei16_v_i32m8_m (vbool4_t mask, int32_t *base, vuint16m4_t bindex, vint32m8_t value, size_t vl) {
@@ -888,7 +888,7 @@ vint32m8_t test_vamoaddei16_v_i32m8_m (vbool4_t mask, int32_t *base, vuint16m4_t
 // CHECK-RV64-LABEL: @test_vamoaddei32_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoadd.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoadd.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamoaddei32_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint32mf2_t bindex, vint32mf2_t value, size_t vl) {
@@ -899,7 +899,7 @@ vint32mf2_t test_vamoaddei32_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamoaddei32_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoadd.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoadd.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamoaddei32_v_i32m1_m (vbool32_t mask, int32_t *base, vuint32m1_t bindex, vint32m1_t value, size_t vl) {
@@ -910,7 +910,7 @@ vint32m1_t test_vamoaddei32_v_i32m1_m (vbool32_t mask, int32_t *base, vuint32m1_
 // CHECK-RV64-LABEL: @test_vamoaddei32_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoadd.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoadd.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamoaddei32_v_i32m2_m (vbool16_t mask, int32_t *base, vuint32m2_t bindex, vint32m2_t value, size_t vl) {
@@ -921,7 +921,7 @@ vint32m2_t test_vamoaddei32_v_i32m2_m (vbool16_t mask, int32_t *base, vuint32m2_
 // CHECK-RV64-LABEL: @test_vamoaddei32_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoadd.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoadd.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamoaddei32_v_i32m4_m (vbool8_t mask, int32_t *base, vuint32m4_t bindex, vint32m4_t value, size_t vl) {
@@ -932,7 +932,7 @@ vint32m4_t test_vamoaddei32_v_i32m4_m (vbool8_t mask, int32_t *base, vuint32m4_t
 // CHECK-RV64-LABEL: @test_vamoaddei32_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoadd.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoadd.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vamoaddei32_v_i32m8_m (vbool4_t mask, int32_t *base, vuint32m8_t bindex, vint32m8_t value, size_t vl) {
@@ -943,7 +943,7 @@ vint32m8_t test_vamoaddei32_v_i32m8_m (vbool4_t mask, int32_t *base, vuint32m8_t
 // CHECK-RV64-LABEL: @test_vamoaddei64_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoadd.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoadd.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamoaddei64_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint64m1_t bindex, vint32mf2_t value, size_t vl) {
@@ -954,7 +954,7 @@ vint32mf2_t test_vamoaddei64_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamoaddei64_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoadd.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoadd.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamoaddei64_v_i32m1_m (vbool32_t mask, int32_t *base, vuint64m2_t bindex, vint32m1_t value, size_t vl) {
@@ -965,7 +965,7 @@ vint32m1_t test_vamoaddei64_v_i32m1_m (vbool32_t mask, int32_t *base, vuint64m2_
 // CHECK-RV64-LABEL: @test_vamoaddei64_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoadd.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoadd.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamoaddei64_v_i32m2_m (vbool16_t mask, int32_t *base, vuint64m4_t bindex, vint32m2_t value, size_t vl) {
@@ -976,7 +976,7 @@ vint32m2_t test_vamoaddei64_v_i32m2_m (vbool16_t mask, int32_t *base, vuint64m4_
 // CHECK-RV64-LABEL: @test_vamoaddei64_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoadd.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoadd.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamoaddei64_v_i32m4_m (vbool8_t mask, int32_t *base, vuint64m8_t bindex, vint32m4_t value, size_t vl) {
@@ -987,7 +987,7 @@ vint32m4_t test_vamoaddei64_v_i32m4_m (vbool8_t mask, int32_t *base, vuint64m8_t
 // CHECK-RV64-LABEL: @test_vamoaddei8_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoadd.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoadd.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamoaddei8_v_i64m1_m (vbool64_t mask, int64_t *base, vuint8mf8_t bindex, vint64m1_t value, size_t vl) {
@@ -998,7 +998,7 @@ vint64m1_t test_vamoaddei8_v_i64m1_m (vbool64_t mask, int64_t *base, vuint8mf8_t
 // CHECK-RV64-LABEL: @test_vamoaddei8_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoadd.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoadd.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamoaddei8_v_i64m2_m (vbool32_t mask, int64_t *base, vuint8mf4_t bindex, vint64m2_t value, size_t vl) {
@@ -1009,7 +1009,7 @@ vint64m2_t test_vamoaddei8_v_i64m2_m (vbool32_t mask, int64_t *base, vuint8mf4_t
 // CHECK-RV64-LABEL: @test_vamoaddei8_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoadd.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoadd.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamoaddei8_v_i64m4_m (vbool16_t mask, int64_t *base, vuint8mf2_t bindex, vint64m4_t value, size_t vl) {
@@ -1020,7 +1020,7 @@ vint64m4_t test_vamoaddei8_v_i64m4_m (vbool16_t mask, int64_t *base, vuint8mf2_t
 // CHECK-RV64-LABEL: @test_vamoaddei8_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoadd.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoadd.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamoaddei8_v_i64m8_m (vbool8_t mask, int64_t *base, vuint8m1_t bindex, vint64m8_t value, size_t vl) {
@@ -1031,7 +1031,7 @@ vint64m8_t test_vamoaddei8_v_i64m8_m (vbool8_t mask, int64_t *base, vuint8m1_t b
 // CHECK-RV64-LABEL: @test_vamoaddei16_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoadd.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoadd.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamoaddei16_v_i64m1_m (vbool64_t mask, int64_t *base, vuint16mf4_t bindex, vint64m1_t value, size_t vl) {
@@ -1042,7 +1042,7 @@ vint64m1_t test_vamoaddei16_v_i64m1_m (vbool64_t mask, int64_t *base, vuint16mf4
 // CHECK-RV64-LABEL: @test_vamoaddei16_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoadd.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoadd.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamoaddei16_v_i64m2_m (vbool32_t mask, int64_t *base, vuint16mf2_t bindex, vint64m2_t value, size_t vl) {
@@ -1053,7 +1053,7 @@ vint64m2_t test_vamoaddei16_v_i64m2_m (vbool32_t mask, int64_t *base, vuint16mf2
 // CHECK-RV64-LABEL: @test_vamoaddei16_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoadd.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoadd.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamoaddei16_v_i64m4_m (vbool16_t mask, int64_t *base, vuint16m1_t bindex, vint64m4_t value, size_t vl) {
@@ -1064,7 +1064,7 @@ vint64m4_t test_vamoaddei16_v_i64m4_m (vbool16_t mask, int64_t *base, vuint16m1_
 // CHECK-RV64-LABEL: @test_vamoaddei16_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoadd.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoadd.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamoaddei16_v_i64m8_m (vbool8_t mask, int64_t *base, vuint16m2_t bindex, vint64m8_t value, size_t vl) {
@@ -1075,7 +1075,7 @@ vint64m8_t test_vamoaddei16_v_i64m8_m (vbool8_t mask, int64_t *base, vuint16m2_t
 // CHECK-RV64-LABEL: @test_vamoaddei32_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoadd.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoadd.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamoaddei32_v_i64m1_m (vbool64_t mask, int64_t *base, vuint32mf2_t bindex, vint64m1_t value, size_t vl) {
@@ -1086,7 +1086,7 @@ vint64m1_t test_vamoaddei32_v_i64m1_m (vbool64_t mask, int64_t *base, vuint32mf2
 // CHECK-RV64-LABEL: @test_vamoaddei32_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoadd.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoadd.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamoaddei32_v_i64m2_m (vbool32_t mask, int64_t *base, vuint32m1_t bindex, vint64m2_t value, size_t vl) {
@@ -1097,7 +1097,7 @@ vint64m2_t test_vamoaddei32_v_i64m2_m (vbool32_t mask, int64_t *base, vuint32m1_
 // CHECK-RV64-LABEL: @test_vamoaddei32_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoadd.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoadd.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamoaddei32_v_i64m4_m (vbool16_t mask, int64_t *base, vuint32m2_t bindex, vint64m4_t value, size_t vl) {
@@ -1108,7 +1108,7 @@ vint64m4_t test_vamoaddei32_v_i64m4_m (vbool16_t mask, int64_t *base, vuint32m2_
 // CHECK-RV64-LABEL: @test_vamoaddei32_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoadd.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoadd.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamoaddei32_v_i64m8_m (vbool8_t mask, int64_t *base, vuint32m4_t bindex, vint64m8_t value, size_t vl) {
@@ -1119,7 +1119,7 @@ vint64m8_t test_vamoaddei32_v_i64m8_m (vbool8_t mask, int64_t *base, vuint32m4_t
 // CHECK-RV64-LABEL: @test_vamoaddei64_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoadd.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoadd.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamoaddei64_v_i64m1_m (vbool64_t mask, int64_t *base, vuint64m1_t bindex, vint64m1_t value, size_t vl) {
@@ -1130,7 +1130,7 @@ vint64m1_t test_vamoaddei64_v_i64m1_m (vbool64_t mask, int64_t *base, vuint64m1_
 // CHECK-RV64-LABEL: @test_vamoaddei64_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoadd.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoadd.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamoaddei64_v_i64m2_m (vbool32_t mask, int64_t *base, vuint64m2_t bindex, vint64m2_t value, size_t vl) {
@@ -1141,7 +1141,7 @@ vint64m2_t test_vamoaddei64_v_i64m2_m (vbool32_t mask, int64_t *base, vuint64m2_
 // CHECK-RV64-LABEL: @test_vamoaddei64_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoadd.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoadd.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamoaddei64_v_i64m4_m (vbool16_t mask, int64_t *base, vuint64m4_t bindex, vint64m4_t value, size_t vl) {
@@ -1152,7 +1152,7 @@ vint64m4_t test_vamoaddei64_v_i64m4_m (vbool16_t mask, int64_t *base, vuint64m4_
 // CHECK-RV64-LABEL: @test_vamoaddei64_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoadd.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoadd.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamoaddei64_v_i64m8_m (vbool8_t mask, int64_t *base, vuint64m8_t bindex, vint64m8_t value, size_t vl) {
@@ -1163,7 +1163,7 @@ vint64m8_t test_vamoaddei64_v_i64m8_m (vbool8_t mask, int64_t *base, vuint64m8_t
 // CHECK-RV64-LABEL: @test_vamoaddei8_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoadd.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoadd.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamoaddei8_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint8mf8_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1174,7 +1174,7 @@ vuint32mf2_t test_vamoaddei8_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint8m
 // CHECK-RV64-LABEL: @test_vamoaddei8_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoadd.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoadd.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamoaddei8_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint8mf4_t bindex, vuint32m1_t value, size_t vl) {
@@ -1185,7 +1185,7 @@ vuint32m1_t test_vamoaddei8_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint8mf4
 // CHECK-RV64-LABEL: @test_vamoaddei8_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoadd.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoadd.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamoaddei8_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint8mf2_t bindex, vuint32m2_t value, size_t vl) {
@@ -1196,7 +1196,7 @@ vuint32m2_t test_vamoaddei8_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint8mf2
 // CHECK-RV64-LABEL: @test_vamoaddei8_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoadd.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoadd.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamoaddei8_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint8m1_t bindex, vuint32m4_t value, size_t vl) {
@@ -1207,7 +1207,7 @@ vuint32m4_t test_vamoaddei8_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint8m1_t
 // CHECK-RV64-LABEL: @test_vamoaddei8_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoadd.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoadd.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vamoaddei8_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint8m2_t bindex, vuint32m8_t value, size_t vl) {
@@ -1218,7 +1218,7 @@ vuint32m8_t test_vamoaddei8_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint8m2_t
 // CHECK-RV64-LABEL: @test_vamoaddei16_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoadd.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoadd.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamoaddei16_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint16mf4_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1229,7 +1229,7 @@ vuint32mf2_t test_vamoaddei16_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint1
 // CHECK-RV64-LABEL: @test_vamoaddei16_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoadd.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoadd.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamoaddei16_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint16mf2_t bindex, vuint32m1_t value, size_t vl) {
@@ -1240,7 +1240,7 @@ vuint32m1_t test_vamoaddei16_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamoaddei16_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoadd.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoadd.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamoaddei16_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint16m1_t bindex, vuint32m2_t value, size_t vl) {
@@ -1251,7 +1251,7 @@ vuint32m2_t test_vamoaddei16_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamoaddei16_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoadd.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoadd.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamoaddei16_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint16m2_t bindex, vuint32m4_t value, size_t vl) {
@@ -1262,7 +1262,7 @@ vuint32m4_t test_vamoaddei16_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint16m2
 // CHECK-RV64-LABEL: @test_vamoaddei16_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoadd.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoadd.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vamoaddei16_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint16m4_t bindex, vuint32m8_t value, size_t vl) {
@@ -1273,7 +1273,7 @@ vuint32m8_t test_vamoaddei16_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint16m4
 // CHECK-RV64-LABEL: @test_vamoaddei32_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoadd.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoadd.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamoaddei32_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint32mf2_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1284,7 +1284,7 @@ vuint32mf2_t test_vamoaddei32_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint3
 // CHECK-RV64-LABEL: @test_vamoaddei32_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoadd.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoadd.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamoaddei32_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint32m1_t bindex, vuint32m1_t value, size_t vl) {
@@ -1295,7 +1295,7 @@ vuint32m1_t test_vamoaddei32_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamoaddei32_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoadd.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoadd.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamoaddei32_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint32m2_t bindex, vuint32m2_t value, size_t vl) {
@@ -1306,7 +1306,7 @@ vuint32m2_t test_vamoaddei32_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamoaddei32_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoadd.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoadd.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamoaddei32_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint32m4_t bindex, vuint32m4_t value, size_t vl) {
@@ -1317,7 +1317,7 @@ vuint32m4_t test_vamoaddei32_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint32m4
 // CHECK-RV64-LABEL: @test_vamoaddei32_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoadd.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoadd.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vamoaddei32_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint32m8_t bindex, vuint32m8_t value, size_t vl) {
@@ -1328,7 +1328,7 @@ vuint32m8_t test_vamoaddei32_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint32m8
 // CHECK-RV64-LABEL: @test_vamoaddei64_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoadd.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoadd.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamoaddei64_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint64m1_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1339,7 +1339,7 @@ vuint32mf2_t test_vamoaddei64_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint6
 // CHECK-RV64-LABEL: @test_vamoaddei64_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoadd.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoadd.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamoaddei64_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint64m2_t bindex, vuint32m1_t value, size_t vl) {
@@ -1350,7 +1350,7 @@ vuint32m1_t test_vamoaddei64_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamoaddei64_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoadd.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoadd.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamoaddei64_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint64m4_t bindex, vuint32m2_t value, size_t vl) {
@@ -1361,7 +1361,7 @@ vuint32m2_t test_vamoaddei64_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamoaddei64_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoadd.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoadd.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamoaddei64_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint64m8_t bindex, vuint32m4_t value, size_t vl) {
@@ -1372,7 +1372,7 @@ vuint32m4_t test_vamoaddei64_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint64m8
 // CHECK-RV64-LABEL: @test_vamoaddei8_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoadd.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoadd.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamoaddei8_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint8mf8_t bindex, vuint64m1_t value, size_t vl) {
@@ -1383,7 +1383,7 @@ vuint64m1_t test_vamoaddei8_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint8mf8
 // CHECK-RV64-LABEL: @test_vamoaddei8_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoadd.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoadd.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamoaddei8_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint8mf4_t bindex, vuint64m2_t value, size_t vl) {
@@ -1394,7 +1394,7 @@ vuint64m2_t test_vamoaddei8_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint8mf4
 // CHECK-RV64-LABEL: @test_vamoaddei8_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoadd.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoadd.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamoaddei8_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint8mf2_t bindex, vuint64m4_t value, size_t vl) {
@@ -1405,7 +1405,7 @@ vuint64m4_t test_vamoaddei8_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint8mf2
 // CHECK-RV64-LABEL: @test_vamoaddei8_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoadd.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoadd.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamoaddei8_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint8m1_t bindex, vuint64m8_t value, size_t vl) {
@@ -1416,7 +1416,7 @@ vuint64m8_t test_vamoaddei8_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint8m1_t
 // CHECK-RV64-LABEL: @test_vamoaddei16_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoadd.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoadd.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamoaddei16_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint16mf4_t bindex, vuint64m1_t value, size_t vl) {
@@ -1427,7 +1427,7 @@ vuint64m1_t test_vamoaddei16_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamoaddei16_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoadd.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoadd.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamoaddei16_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint16mf2_t bindex, vuint64m2_t value, size_t vl) {
@@ -1438,7 +1438,7 @@ vuint64m2_t test_vamoaddei16_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamoaddei16_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoadd.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoadd.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamoaddei16_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint16m1_t bindex, vuint64m4_t value, size_t vl) {
@@ -1449,7 +1449,7 @@ vuint64m4_t test_vamoaddei16_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamoaddei16_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoadd.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoadd.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamoaddei16_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint16m2_t bindex, vuint64m8_t value, size_t vl) {
@@ -1460,7 +1460,7 @@ vuint64m8_t test_vamoaddei16_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint16m2
 // CHECK-RV64-LABEL: @test_vamoaddei32_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoadd.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoadd.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamoaddei32_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint32mf2_t bindex, vuint64m1_t value, size_t vl) {
@@ -1471,7 +1471,7 @@ vuint64m1_t test_vamoaddei32_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamoaddei32_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoadd.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoadd.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamoaddei32_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint32m1_t bindex, vuint64m2_t value, size_t vl) {
@@ -1482,7 +1482,7 @@ vuint64m2_t test_vamoaddei32_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamoaddei32_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoadd.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoadd.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamoaddei32_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint32m2_t bindex, vuint64m4_t value, size_t vl) {
@@ -1493,7 +1493,7 @@ vuint64m4_t test_vamoaddei32_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamoaddei32_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoadd.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoadd.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamoaddei32_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint32m4_t bindex, vuint64m8_t value, size_t vl) {
@@ -1504,7 +1504,7 @@ vuint64m8_t test_vamoaddei32_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint32m4
 // CHECK-RV64-LABEL: @test_vamoaddei64_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoadd.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoadd.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamoaddei64_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint64m1_t bindex, vuint64m1_t value, size_t vl) {
@@ -1515,7 +1515,7 @@ vuint64m1_t test_vamoaddei64_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamoaddei64_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoadd.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoadd.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamoaddei64_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint64m2_t bindex, vuint64m2_t value, size_t vl) {
@@ -1526,7 +1526,7 @@ vuint64m2_t test_vamoaddei64_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamoaddei64_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoadd.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoadd.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamoaddei64_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint64m4_t bindex, vuint64m4_t value, size_t vl) {
@@ -1537,7 +1537,7 @@ vuint64m4_t test_vamoaddei64_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamoaddei64_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoadd.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoadd.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamoaddei64_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint64m8_t bindex, vuint64m8_t value, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamoand.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamoand.c
index 0fb76be..0a71bf8 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamoand.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamoand.c
@@ -778,7 +778,7 @@ vuint64m8_t test_vamoandei64_v_u64m8 (uint64_t *base, vuint64m8_t bindex, vuint6
 // CHECK-RV64-LABEL: @test_vamoandei8_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoand.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoand.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamoandei8_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint8mf8_t bindex, vint32mf2_t value, size_t vl) {
@@ -789,7 +789,7 @@ vint32mf2_t test_vamoandei8_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint8mf8
 // CHECK-RV64-LABEL: @test_vamoandei8_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoand.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoand.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamoandei8_v_i32m1_m (vbool32_t mask, int32_t *base, vuint8mf4_t bindex, vint32m1_t value, size_t vl) {
@@ -800,7 +800,7 @@ vint32m1_t test_vamoandei8_v_i32m1_m (vbool32_t mask, int32_t *base, vuint8mf4_t
 // CHECK-RV64-LABEL: @test_vamoandei8_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoand.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoand.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamoandei8_v_i32m2_m (vbool16_t mask, int32_t *base, vuint8mf2_t bindex, vint32m2_t value, size_t vl) {
@@ -811,7 +811,7 @@ vint32m2_t test_vamoandei8_v_i32m2_m (vbool16_t mask, int32_t *base, vuint8mf2_t
 // CHECK-RV64-LABEL: @test_vamoandei8_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoand.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoand.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamoandei8_v_i32m4_m (vbool8_t mask, int32_t *base, vuint8m1_t bindex, vint32m4_t value, size_t vl) {
@@ -822,7 +822,7 @@ vint32m4_t test_vamoandei8_v_i32m4_m (vbool8_t mask, int32_t *base, vuint8m1_t b
 // CHECK-RV64-LABEL: @test_vamoandei8_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoand.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoand.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vamoandei8_v_i32m8_m (vbool4_t mask, int32_t *base, vuint8m2_t bindex, vint32m8_t value, size_t vl) {
@@ -833,7 +833,7 @@ vint32m8_t test_vamoandei8_v_i32m8_m (vbool4_t mask, int32_t *base, vuint8m2_t b
 // CHECK-RV64-LABEL: @test_vamoandei16_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoand.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoand.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamoandei16_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint16mf4_t bindex, vint32mf2_t value, size_t vl) {
@@ -844,7 +844,7 @@ vint32mf2_t test_vamoandei16_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamoandei16_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoand.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoand.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamoandei16_v_i32m1_m (vbool32_t mask, int32_t *base, vuint16mf2_t bindex, vint32m1_t value, size_t vl) {
@@ -855,7 +855,7 @@ vint32m1_t test_vamoandei16_v_i32m1_m (vbool32_t mask, int32_t *base, vuint16mf2
 // CHECK-RV64-LABEL: @test_vamoandei16_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoand.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoand.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamoandei16_v_i32m2_m (vbool16_t mask, int32_t *base, vuint16m1_t bindex, vint32m2_t value, size_t vl) {
@@ -866,7 +866,7 @@ vint32m2_t test_vamoandei16_v_i32m2_m (vbool16_t mask, int32_t *base, vuint16m1_
 // CHECK-RV64-LABEL: @test_vamoandei16_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoand.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoand.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamoandei16_v_i32m4_m (vbool8_t mask, int32_t *base, vuint16m2_t bindex, vint32m4_t value, size_t vl) {
@@ -877,7 +877,7 @@ vint32m4_t test_vamoandei16_v_i32m4_m (vbool8_t mask, int32_t *base, vuint16m2_t
 // CHECK-RV64-LABEL: @test_vamoandei16_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoand.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoand.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vamoandei16_v_i32m8_m (vbool4_t mask, int32_t *base, vuint16m4_t bindex, vint32m8_t value, size_t vl) {
@@ -888,7 +888,7 @@ vint32m8_t test_vamoandei16_v_i32m8_m (vbool4_t mask, int32_t *base, vuint16m4_t
 // CHECK-RV64-LABEL: @test_vamoandei32_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoand.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoand.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamoandei32_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint32mf2_t bindex, vint32mf2_t value, size_t vl) {
@@ -899,7 +899,7 @@ vint32mf2_t test_vamoandei32_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamoandei32_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoand.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoand.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamoandei32_v_i32m1_m (vbool32_t mask, int32_t *base, vuint32m1_t bindex, vint32m1_t value, size_t vl) {
@@ -910,7 +910,7 @@ vint32m1_t test_vamoandei32_v_i32m1_m (vbool32_t mask, int32_t *base, vuint32m1_
 // CHECK-RV64-LABEL: @test_vamoandei32_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoand.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoand.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamoandei32_v_i32m2_m (vbool16_t mask, int32_t *base, vuint32m2_t bindex, vint32m2_t value, size_t vl) {
@@ -921,7 +921,7 @@ vint32m2_t test_vamoandei32_v_i32m2_m (vbool16_t mask, int32_t *base, vuint32m2_
 // CHECK-RV64-LABEL: @test_vamoandei32_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoand.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoand.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamoandei32_v_i32m4_m (vbool8_t mask, int32_t *base, vuint32m4_t bindex, vint32m4_t value, size_t vl) {
@@ -932,7 +932,7 @@ vint32m4_t test_vamoandei32_v_i32m4_m (vbool8_t mask, int32_t *base, vuint32m4_t
 // CHECK-RV64-LABEL: @test_vamoandei32_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoand.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoand.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vamoandei32_v_i32m8_m (vbool4_t mask, int32_t *base, vuint32m8_t bindex, vint32m8_t value, size_t vl) {
@@ -943,7 +943,7 @@ vint32m8_t test_vamoandei32_v_i32m8_m (vbool4_t mask, int32_t *base, vuint32m8_t
 // CHECK-RV64-LABEL: @test_vamoandei64_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoand.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoand.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamoandei64_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint64m1_t bindex, vint32mf2_t value, size_t vl) {
@@ -954,7 +954,7 @@ vint32mf2_t test_vamoandei64_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamoandei64_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoand.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoand.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamoandei64_v_i32m1_m (vbool32_t mask, int32_t *base, vuint64m2_t bindex, vint32m1_t value, size_t vl) {
@@ -965,7 +965,7 @@ vint32m1_t test_vamoandei64_v_i32m1_m (vbool32_t mask, int32_t *base, vuint64m2_
 // CHECK-RV64-LABEL: @test_vamoandei64_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoand.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoand.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamoandei64_v_i32m2_m (vbool16_t mask, int32_t *base, vuint64m4_t bindex, vint32m2_t value, size_t vl) {
@@ -976,7 +976,7 @@ vint32m2_t test_vamoandei64_v_i32m2_m (vbool16_t mask, int32_t *base, vuint64m4_
 // CHECK-RV64-LABEL: @test_vamoandei64_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoand.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoand.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamoandei64_v_i32m4_m (vbool8_t mask, int32_t *base, vuint64m8_t bindex, vint32m4_t value, size_t vl) {
@@ -987,7 +987,7 @@ vint32m4_t test_vamoandei64_v_i32m4_m (vbool8_t mask, int32_t *base, vuint64m8_t
 // CHECK-RV64-LABEL: @test_vamoandei8_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoand.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoand.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamoandei8_v_i64m1_m (vbool64_t mask, int64_t *base, vuint8mf8_t bindex, vint64m1_t value, size_t vl) {
@@ -998,7 +998,7 @@ vint64m1_t test_vamoandei8_v_i64m1_m (vbool64_t mask, int64_t *base, vuint8mf8_t
 // CHECK-RV64-LABEL: @test_vamoandei8_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoand.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoand.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamoandei8_v_i64m2_m (vbool32_t mask, int64_t *base, vuint8mf4_t bindex, vint64m2_t value, size_t vl) {
@@ -1009,7 +1009,7 @@ vint64m2_t test_vamoandei8_v_i64m2_m (vbool32_t mask, int64_t *base, vuint8mf4_t
 // CHECK-RV64-LABEL: @test_vamoandei8_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoand.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoand.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamoandei8_v_i64m4_m (vbool16_t mask, int64_t *base, vuint8mf2_t bindex, vint64m4_t value, size_t vl) {
@@ -1020,7 +1020,7 @@ vint64m4_t test_vamoandei8_v_i64m4_m (vbool16_t mask, int64_t *base, vuint8mf2_t
 // CHECK-RV64-LABEL: @test_vamoandei8_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoand.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoand.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamoandei8_v_i64m8_m (vbool8_t mask, int64_t *base, vuint8m1_t bindex, vint64m8_t value, size_t vl) {
@@ -1031,7 +1031,7 @@ vint64m8_t test_vamoandei8_v_i64m8_m (vbool8_t mask, int64_t *base, vuint8m1_t b
 // CHECK-RV64-LABEL: @test_vamoandei16_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoand.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoand.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamoandei16_v_i64m1_m (vbool64_t mask, int64_t *base, vuint16mf4_t bindex, vint64m1_t value, size_t vl) {
@@ -1042,7 +1042,7 @@ vint64m1_t test_vamoandei16_v_i64m1_m (vbool64_t mask, int64_t *base, vuint16mf4
 // CHECK-RV64-LABEL: @test_vamoandei16_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoand.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoand.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamoandei16_v_i64m2_m (vbool32_t mask, int64_t *base, vuint16mf2_t bindex, vint64m2_t value, size_t vl) {
@@ -1053,7 +1053,7 @@ vint64m2_t test_vamoandei16_v_i64m2_m (vbool32_t mask, int64_t *base, vuint16mf2
 // CHECK-RV64-LABEL: @test_vamoandei16_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoand.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoand.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamoandei16_v_i64m4_m (vbool16_t mask, int64_t *base, vuint16m1_t bindex, vint64m4_t value, size_t vl) {
@@ -1064,7 +1064,7 @@ vint64m4_t test_vamoandei16_v_i64m4_m (vbool16_t mask, int64_t *base, vuint16m1_
 // CHECK-RV64-LABEL: @test_vamoandei16_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoand.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoand.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamoandei16_v_i64m8_m (vbool8_t mask, int64_t *base, vuint16m2_t bindex, vint64m8_t value, size_t vl) {
@@ -1075,7 +1075,7 @@ vint64m8_t test_vamoandei16_v_i64m8_m (vbool8_t mask, int64_t *base, vuint16m2_t
 // CHECK-RV64-LABEL: @test_vamoandei32_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoand.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoand.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamoandei32_v_i64m1_m (vbool64_t mask, int64_t *base, vuint32mf2_t bindex, vint64m1_t value, size_t vl) {
@@ -1086,7 +1086,7 @@ vint64m1_t test_vamoandei32_v_i64m1_m (vbool64_t mask, int64_t *base, vuint32mf2
 // CHECK-RV64-LABEL: @test_vamoandei32_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoand.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoand.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamoandei32_v_i64m2_m (vbool32_t mask, int64_t *base, vuint32m1_t bindex, vint64m2_t value, size_t vl) {
@@ -1097,7 +1097,7 @@ vint64m2_t test_vamoandei32_v_i64m2_m (vbool32_t mask, int64_t *base, vuint32m1_
 // CHECK-RV64-LABEL: @test_vamoandei32_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoand.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoand.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamoandei32_v_i64m4_m (vbool16_t mask, int64_t *base, vuint32m2_t bindex, vint64m4_t value, size_t vl) {
@@ -1108,7 +1108,7 @@ vint64m4_t test_vamoandei32_v_i64m4_m (vbool16_t mask, int64_t *base, vuint32m2_
 // CHECK-RV64-LABEL: @test_vamoandei32_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoand.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoand.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamoandei32_v_i64m8_m (vbool8_t mask, int64_t *base, vuint32m4_t bindex, vint64m8_t value, size_t vl) {
@@ -1119,7 +1119,7 @@ vint64m8_t test_vamoandei32_v_i64m8_m (vbool8_t mask, int64_t *base, vuint32m4_t
 // CHECK-RV64-LABEL: @test_vamoandei64_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoand.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoand.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamoandei64_v_i64m1_m (vbool64_t mask, int64_t *base, vuint64m1_t bindex, vint64m1_t value, size_t vl) {
@@ -1130,7 +1130,7 @@ vint64m1_t test_vamoandei64_v_i64m1_m (vbool64_t mask, int64_t *base, vuint64m1_
 // CHECK-RV64-LABEL: @test_vamoandei64_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoand.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoand.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamoandei64_v_i64m2_m (vbool32_t mask, int64_t *base, vuint64m2_t bindex, vint64m2_t value, size_t vl) {
@@ -1141,7 +1141,7 @@ vint64m2_t test_vamoandei64_v_i64m2_m (vbool32_t mask, int64_t *base, vuint64m2_
 // CHECK-RV64-LABEL: @test_vamoandei64_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoand.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoand.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamoandei64_v_i64m4_m (vbool16_t mask, int64_t *base, vuint64m4_t bindex, vint64m4_t value, size_t vl) {
@@ -1152,7 +1152,7 @@ vint64m4_t test_vamoandei64_v_i64m4_m (vbool16_t mask, int64_t *base, vuint64m4_
 // CHECK-RV64-LABEL: @test_vamoandei64_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoand.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoand.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamoandei64_v_i64m8_m (vbool8_t mask, int64_t *base, vuint64m8_t bindex, vint64m8_t value, size_t vl) {
@@ -1163,7 +1163,7 @@ vint64m8_t test_vamoandei64_v_i64m8_m (vbool8_t mask, int64_t *base, vuint64m8_t
 // CHECK-RV64-LABEL: @test_vamoandei8_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoand.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoand.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamoandei8_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint8mf8_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1174,7 +1174,7 @@ vuint32mf2_t test_vamoandei8_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint8m
 // CHECK-RV64-LABEL: @test_vamoandei8_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoand.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoand.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamoandei8_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint8mf4_t bindex, vuint32m1_t value, size_t vl) {
@@ -1185,7 +1185,7 @@ vuint32m1_t test_vamoandei8_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint8mf4
 // CHECK-RV64-LABEL: @test_vamoandei8_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoand.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoand.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamoandei8_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint8mf2_t bindex, vuint32m2_t value, size_t vl) {
@@ -1196,7 +1196,7 @@ vuint32m2_t test_vamoandei8_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint8mf2
 // CHECK-RV64-LABEL: @test_vamoandei8_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoand.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoand.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamoandei8_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint8m1_t bindex, vuint32m4_t value, size_t vl) {
@@ -1207,7 +1207,7 @@ vuint32m4_t test_vamoandei8_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint8m1_t
 // CHECK-RV64-LABEL: @test_vamoandei8_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoand.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoand.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vamoandei8_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint8m2_t bindex, vuint32m8_t value, size_t vl) {
@@ -1218,7 +1218,7 @@ vuint32m8_t test_vamoandei8_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint8m2_t
 // CHECK-RV64-LABEL: @test_vamoandei16_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoand.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoand.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamoandei16_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint16mf4_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1229,7 +1229,7 @@ vuint32mf2_t test_vamoandei16_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint1
 // CHECK-RV64-LABEL: @test_vamoandei16_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoand.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoand.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamoandei16_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint16mf2_t bindex, vuint32m1_t value, size_t vl) {
@@ -1240,7 +1240,7 @@ vuint32m1_t test_vamoandei16_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamoandei16_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoand.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoand.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamoandei16_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint16m1_t bindex, vuint32m2_t value, size_t vl) {
@@ -1251,7 +1251,7 @@ vuint32m2_t test_vamoandei16_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamoandei16_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoand.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoand.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamoandei16_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint16m2_t bindex, vuint32m4_t value, size_t vl) {
@@ -1262,7 +1262,7 @@ vuint32m4_t test_vamoandei16_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint16m2
 // CHECK-RV64-LABEL: @test_vamoandei16_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoand.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoand.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vamoandei16_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint16m4_t bindex, vuint32m8_t value, size_t vl) {
@@ -1273,7 +1273,7 @@ vuint32m8_t test_vamoandei16_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint16m4
 // CHECK-RV64-LABEL: @test_vamoandei32_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoand.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoand.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamoandei32_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint32mf2_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1284,7 +1284,7 @@ vuint32mf2_t test_vamoandei32_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint3
 // CHECK-RV64-LABEL: @test_vamoandei32_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoand.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoand.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamoandei32_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint32m1_t bindex, vuint32m1_t value, size_t vl) {
@@ -1295,7 +1295,7 @@ vuint32m1_t test_vamoandei32_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamoandei32_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoand.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoand.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamoandei32_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint32m2_t bindex, vuint32m2_t value, size_t vl) {
@@ -1306,7 +1306,7 @@ vuint32m2_t test_vamoandei32_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamoandei32_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoand.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoand.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamoandei32_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint32m4_t bindex, vuint32m4_t value, size_t vl) {
@@ -1317,7 +1317,7 @@ vuint32m4_t test_vamoandei32_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint32m4
 // CHECK-RV64-LABEL: @test_vamoandei32_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoand.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoand.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vamoandei32_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint32m8_t bindex, vuint32m8_t value, size_t vl) {
@@ -1328,7 +1328,7 @@ vuint32m8_t test_vamoandei32_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint32m8
 // CHECK-RV64-LABEL: @test_vamoandei64_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoand.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoand.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamoandei64_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint64m1_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1339,7 +1339,7 @@ vuint32mf2_t test_vamoandei64_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint6
 // CHECK-RV64-LABEL: @test_vamoandei64_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoand.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoand.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamoandei64_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint64m2_t bindex, vuint32m1_t value, size_t vl) {
@@ -1350,7 +1350,7 @@ vuint32m1_t test_vamoandei64_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamoandei64_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoand.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoand.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamoandei64_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint64m4_t bindex, vuint32m2_t value, size_t vl) {
@@ -1361,7 +1361,7 @@ vuint32m2_t test_vamoandei64_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamoandei64_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoand.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoand.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamoandei64_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint64m8_t bindex, vuint32m4_t value, size_t vl) {
@@ -1372,7 +1372,7 @@ vuint32m4_t test_vamoandei64_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint64m8
 // CHECK-RV64-LABEL: @test_vamoandei8_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoand.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoand.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamoandei8_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint8mf8_t bindex, vuint64m1_t value, size_t vl) {
@@ -1383,7 +1383,7 @@ vuint64m1_t test_vamoandei8_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint8mf8
 // CHECK-RV64-LABEL: @test_vamoandei8_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoand.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoand.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamoandei8_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint8mf4_t bindex, vuint64m2_t value, size_t vl) {
@@ -1394,7 +1394,7 @@ vuint64m2_t test_vamoandei8_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint8mf4
 // CHECK-RV64-LABEL: @test_vamoandei8_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoand.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoand.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamoandei8_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint8mf2_t bindex, vuint64m4_t value, size_t vl) {
@@ -1405,7 +1405,7 @@ vuint64m4_t test_vamoandei8_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint8mf2
 // CHECK-RV64-LABEL: @test_vamoandei8_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoand.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoand.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamoandei8_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint8m1_t bindex, vuint64m8_t value, size_t vl) {
@@ -1416,7 +1416,7 @@ vuint64m8_t test_vamoandei8_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint8m1_t
 // CHECK-RV64-LABEL: @test_vamoandei16_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoand.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoand.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamoandei16_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint16mf4_t bindex, vuint64m1_t value, size_t vl) {
@@ -1427,7 +1427,7 @@ vuint64m1_t test_vamoandei16_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamoandei16_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoand.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoand.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamoandei16_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint16mf2_t bindex, vuint64m2_t value, size_t vl) {
@@ -1438,7 +1438,7 @@ vuint64m2_t test_vamoandei16_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamoandei16_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoand.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoand.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamoandei16_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint16m1_t bindex, vuint64m4_t value, size_t vl) {
@@ -1449,7 +1449,7 @@ vuint64m4_t test_vamoandei16_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamoandei16_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoand.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoand.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamoandei16_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint16m2_t bindex, vuint64m8_t value, size_t vl) {
@@ -1460,7 +1460,7 @@ vuint64m8_t test_vamoandei16_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint16m2
 // CHECK-RV64-LABEL: @test_vamoandei32_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoand.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoand.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamoandei32_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint32mf2_t bindex, vuint64m1_t value, size_t vl) {
@@ -1471,7 +1471,7 @@ vuint64m1_t test_vamoandei32_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamoandei32_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoand.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoand.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamoandei32_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint32m1_t bindex, vuint64m2_t value, size_t vl) {
@@ -1482,7 +1482,7 @@ vuint64m2_t test_vamoandei32_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamoandei32_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoand.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoand.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamoandei32_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint32m2_t bindex, vuint64m4_t value, size_t vl) {
@@ -1493,7 +1493,7 @@ vuint64m4_t test_vamoandei32_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamoandei32_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoand.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoand.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamoandei32_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint32m4_t bindex, vuint64m8_t value, size_t vl) {
@@ -1504,7 +1504,7 @@ vuint64m8_t test_vamoandei32_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint32m4
 // CHECK-RV64-LABEL: @test_vamoandei64_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoand.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoand.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamoandei64_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint64m1_t bindex, vuint64m1_t value, size_t vl) {
@@ -1515,7 +1515,7 @@ vuint64m1_t test_vamoandei64_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamoandei64_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoand.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoand.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamoandei64_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint64m2_t bindex, vuint64m2_t value, size_t vl) {
@@ -1526,7 +1526,7 @@ vuint64m2_t test_vamoandei64_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamoandei64_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoand.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoand.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamoandei64_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint64m4_t bindex, vuint64m4_t value, size_t vl) {
@@ -1537,7 +1537,7 @@ vuint64m4_t test_vamoandei64_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamoandei64_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoand.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoand.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamoandei64_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint64m8_t bindex, vuint64m8_t value, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamomax.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamomax.c
index 35f11cc..4863c11b 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamomax.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamomax.c
@@ -778,7 +778,7 @@ vuint64m8_t test_vamomaxuei64_v_u64m8 (uint64_t *base, vuint64m8_t bindex, vuint
 // CHECK-RV64-LABEL: @test_vamomaxei8_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomax.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomax.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamomaxei8_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint8mf8_t bindex, vint32mf2_t value, size_t vl) {
@@ -789,7 +789,7 @@ vint32mf2_t test_vamomaxei8_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint8mf8
 // CHECK-RV64-LABEL: @test_vamomaxei8_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomax.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomax.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamomaxei8_v_i32m1_m (vbool32_t mask, int32_t *base, vuint8mf4_t bindex, vint32m1_t value, size_t vl) {
@@ -800,7 +800,7 @@ vint32m1_t test_vamomaxei8_v_i32m1_m (vbool32_t mask, int32_t *base, vuint8mf4_t
 // CHECK-RV64-LABEL: @test_vamomaxei8_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomax.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomax.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamomaxei8_v_i32m2_m (vbool16_t mask, int32_t *base, vuint8mf2_t bindex, vint32m2_t value, size_t vl) {
@@ -811,7 +811,7 @@ vint32m2_t test_vamomaxei8_v_i32m2_m (vbool16_t mask, int32_t *base, vuint8mf2_t
 // CHECK-RV64-LABEL: @test_vamomaxei8_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomax.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomax.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamomaxei8_v_i32m4_m (vbool8_t mask, int32_t *base, vuint8m1_t bindex, vint32m4_t value, size_t vl) {
@@ -822,7 +822,7 @@ vint32m4_t test_vamomaxei8_v_i32m4_m (vbool8_t mask, int32_t *base, vuint8m1_t b
 // CHECK-RV64-LABEL: @test_vamomaxei8_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamomax.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamomax.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vamomaxei8_v_i32m8_m (vbool4_t mask, int32_t *base, vuint8m2_t bindex, vint32m8_t value, size_t vl) {
@@ -833,7 +833,7 @@ vint32m8_t test_vamomaxei8_v_i32m8_m (vbool4_t mask, int32_t *base, vuint8m2_t b
 // CHECK-RV64-LABEL: @test_vamomaxei16_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomax.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomax.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamomaxei16_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint16mf4_t bindex, vint32mf2_t value, size_t vl) {
@@ -844,7 +844,7 @@ vint32mf2_t test_vamomaxei16_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamomaxei16_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomax.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomax.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamomaxei16_v_i32m1_m (vbool32_t mask, int32_t *base, vuint16mf2_t bindex, vint32m1_t value, size_t vl) {
@@ -855,7 +855,7 @@ vint32m1_t test_vamomaxei16_v_i32m1_m (vbool32_t mask, int32_t *base, vuint16mf2
 // CHECK-RV64-LABEL: @test_vamomaxei16_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomax.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomax.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamomaxei16_v_i32m2_m (vbool16_t mask, int32_t *base, vuint16m1_t bindex, vint32m2_t value, size_t vl) {
@@ -866,7 +866,7 @@ vint32m2_t test_vamomaxei16_v_i32m2_m (vbool16_t mask, int32_t *base, vuint16m1_
 // CHECK-RV64-LABEL: @test_vamomaxei16_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomax.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomax.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamomaxei16_v_i32m4_m (vbool8_t mask, int32_t *base, vuint16m2_t bindex, vint32m4_t value, size_t vl) {
@@ -877,7 +877,7 @@ vint32m4_t test_vamomaxei16_v_i32m4_m (vbool8_t mask, int32_t *base, vuint16m2_t
 // CHECK-RV64-LABEL: @test_vamomaxei16_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamomax.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamomax.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vamomaxei16_v_i32m8_m (vbool4_t mask, int32_t *base, vuint16m4_t bindex, vint32m8_t value, size_t vl) {
@@ -888,7 +888,7 @@ vint32m8_t test_vamomaxei16_v_i32m8_m (vbool4_t mask, int32_t *base, vuint16m4_t
 // CHECK-RV64-LABEL: @test_vamomaxei32_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomax.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomax.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamomaxei32_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint32mf2_t bindex, vint32mf2_t value, size_t vl) {
@@ -899,7 +899,7 @@ vint32mf2_t test_vamomaxei32_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamomaxei32_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomax.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomax.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamomaxei32_v_i32m1_m (vbool32_t mask, int32_t *base, vuint32m1_t bindex, vint32m1_t value, size_t vl) {
@@ -910,7 +910,7 @@ vint32m1_t test_vamomaxei32_v_i32m1_m (vbool32_t mask, int32_t *base, vuint32m1_
 // CHECK-RV64-LABEL: @test_vamomaxei32_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomax.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomax.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamomaxei32_v_i32m2_m (vbool16_t mask, int32_t *base, vuint32m2_t bindex, vint32m2_t value, size_t vl) {
@@ -921,7 +921,7 @@ vint32m2_t test_vamomaxei32_v_i32m2_m (vbool16_t mask, int32_t *base, vuint32m2_
 // CHECK-RV64-LABEL: @test_vamomaxei32_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomax.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomax.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamomaxei32_v_i32m4_m (vbool8_t mask, int32_t *base, vuint32m4_t bindex, vint32m4_t value, size_t vl) {
@@ -932,7 +932,7 @@ vint32m4_t test_vamomaxei32_v_i32m4_m (vbool8_t mask, int32_t *base, vuint32m4_t
 // CHECK-RV64-LABEL: @test_vamomaxei32_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamomax.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamomax.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vamomaxei32_v_i32m8_m (vbool4_t mask, int32_t *base, vuint32m8_t bindex, vint32m8_t value, size_t vl) {
@@ -943,7 +943,7 @@ vint32m8_t test_vamomaxei32_v_i32m8_m (vbool4_t mask, int32_t *base, vuint32m8_t
 // CHECK-RV64-LABEL: @test_vamomaxei64_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomax.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomax.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamomaxei64_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint64m1_t bindex, vint32mf2_t value, size_t vl) {
@@ -954,7 +954,7 @@ vint32mf2_t test_vamomaxei64_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamomaxei64_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomax.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomax.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamomaxei64_v_i32m1_m (vbool32_t mask, int32_t *base, vuint64m2_t bindex, vint32m1_t value, size_t vl) {
@@ -965,7 +965,7 @@ vint32m1_t test_vamomaxei64_v_i32m1_m (vbool32_t mask, int32_t *base, vuint64m2_
 // CHECK-RV64-LABEL: @test_vamomaxei64_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomax.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomax.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamomaxei64_v_i32m2_m (vbool16_t mask, int32_t *base, vuint64m4_t bindex, vint32m2_t value, size_t vl) {
@@ -976,7 +976,7 @@ vint32m2_t test_vamomaxei64_v_i32m2_m (vbool16_t mask, int32_t *base, vuint64m4_
 // CHECK-RV64-LABEL: @test_vamomaxei64_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomax.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomax.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamomaxei64_v_i32m4_m (vbool8_t mask, int32_t *base, vuint64m8_t bindex, vint32m4_t value, size_t vl) {
@@ -987,7 +987,7 @@ vint32m4_t test_vamomaxei64_v_i32m4_m (vbool8_t mask, int32_t *base, vuint64m8_t
 // CHECK-RV64-LABEL: @test_vamomaxei8_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomax.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomax.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamomaxei8_v_i64m1_m (vbool64_t mask, int64_t *base, vuint8mf8_t bindex, vint64m1_t value, size_t vl) {
@@ -998,7 +998,7 @@ vint64m1_t test_vamomaxei8_v_i64m1_m (vbool64_t mask, int64_t *base, vuint8mf8_t
 // CHECK-RV64-LABEL: @test_vamomaxei8_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomax.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomax.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamomaxei8_v_i64m2_m (vbool32_t mask, int64_t *base, vuint8mf4_t bindex, vint64m2_t value, size_t vl) {
@@ -1009,7 +1009,7 @@ vint64m2_t test_vamomaxei8_v_i64m2_m (vbool32_t mask, int64_t *base, vuint8mf4_t
 // CHECK-RV64-LABEL: @test_vamomaxei8_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomax.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomax.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamomaxei8_v_i64m4_m (vbool16_t mask, int64_t *base, vuint8mf2_t bindex, vint64m4_t value, size_t vl) {
@@ -1020,7 +1020,7 @@ vint64m4_t test_vamomaxei8_v_i64m4_m (vbool16_t mask, int64_t *base, vuint8mf2_t
 // CHECK-RV64-LABEL: @test_vamomaxei8_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomax.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomax.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamomaxei8_v_i64m8_m (vbool8_t mask, int64_t *base, vuint8m1_t bindex, vint64m8_t value, size_t vl) {
@@ -1031,7 +1031,7 @@ vint64m8_t test_vamomaxei8_v_i64m8_m (vbool8_t mask, int64_t *base, vuint8m1_t b
 // CHECK-RV64-LABEL: @test_vamomaxei16_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomax.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomax.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamomaxei16_v_i64m1_m (vbool64_t mask, int64_t *base, vuint16mf4_t bindex, vint64m1_t value, size_t vl) {
@@ -1042,7 +1042,7 @@ vint64m1_t test_vamomaxei16_v_i64m1_m (vbool64_t mask, int64_t *base, vuint16mf4
 // CHECK-RV64-LABEL: @test_vamomaxei16_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomax.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomax.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamomaxei16_v_i64m2_m (vbool32_t mask, int64_t *base, vuint16mf2_t bindex, vint64m2_t value, size_t vl) {
@@ -1053,7 +1053,7 @@ vint64m2_t test_vamomaxei16_v_i64m2_m (vbool32_t mask, int64_t *base, vuint16mf2
 // CHECK-RV64-LABEL: @test_vamomaxei16_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomax.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomax.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamomaxei16_v_i64m4_m (vbool16_t mask, int64_t *base, vuint16m1_t bindex, vint64m4_t value, size_t vl) {
@@ -1064,7 +1064,7 @@ vint64m4_t test_vamomaxei16_v_i64m4_m (vbool16_t mask, int64_t *base, vuint16m1_
 // CHECK-RV64-LABEL: @test_vamomaxei16_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomax.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomax.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamomaxei16_v_i64m8_m (vbool8_t mask, int64_t *base, vuint16m2_t bindex, vint64m8_t value, size_t vl) {
@@ -1075,7 +1075,7 @@ vint64m8_t test_vamomaxei16_v_i64m8_m (vbool8_t mask, int64_t *base, vuint16m2_t
 // CHECK-RV64-LABEL: @test_vamomaxei32_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomax.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomax.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamomaxei32_v_i64m1_m (vbool64_t mask, int64_t *base, vuint32mf2_t bindex, vint64m1_t value, size_t vl) {
@@ -1086,7 +1086,7 @@ vint64m1_t test_vamomaxei32_v_i64m1_m (vbool64_t mask, int64_t *base, vuint32mf2
 // CHECK-RV64-LABEL: @test_vamomaxei32_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomax.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomax.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamomaxei32_v_i64m2_m (vbool32_t mask, int64_t *base, vuint32m1_t bindex, vint64m2_t value, size_t vl) {
@@ -1097,7 +1097,7 @@ vint64m2_t test_vamomaxei32_v_i64m2_m (vbool32_t mask, int64_t *base, vuint32m1_
 // CHECK-RV64-LABEL: @test_vamomaxei32_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomax.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomax.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamomaxei32_v_i64m4_m (vbool16_t mask, int64_t *base, vuint32m2_t bindex, vint64m4_t value, size_t vl) {
@@ -1108,7 +1108,7 @@ vint64m4_t test_vamomaxei32_v_i64m4_m (vbool16_t mask, int64_t *base, vuint32m2_
 // CHECK-RV64-LABEL: @test_vamomaxei32_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomax.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomax.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamomaxei32_v_i64m8_m (vbool8_t mask, int64_t *base, vuint32m4_t bindex, vint64m8_t value, size_t vl) {
@@ -1119,7 +1119,7 @@ vint64m8_t test_vamomaxei32_v_i64m8_m (vbool8_t mask, int64_t *base, vuint32m4_t
 // CHECK-RV64-LABEL: @test_vamomaxei64_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomax.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomax.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamomaxei64_v_i64m1_m (vbool64_t mask, int64_t *base, vuint64m1_t bindex, vint64m1_t value, size_t vl) {
@@ -1130,7 +1130,7 @@ vint64m1_t test_vamomaxei64_v_i64m1_m (vbool64_t mask, int64_t *base, vuint64m1_
 // CHECK-RV64-LABEL: @test_vamomaxei64_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomax.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomax.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamomaxei64_v_i64m2_m (vbool32_t mask, int64_t *base, vuint64m2_t bindex, vint64m2_t value, size_t vl) {
@@ -1141,7 +1141,7 @@ vint64m2_t test_vamomaxei64_v_i64m2_m (vbool32_t mask, int64_t *base, vuint64m2_
 // CHECK-RV64-LABEL: @test_vamomaxei64_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomax.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomax.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamomaxei64_v_i64m4_m (vbool16_t mask, int64_t *base, vuint64m4_t bindex, vint64m4_t value, size_t vl) {
@@ -1152,7 +1152,7 @@ vint64m4_t test_vamomaxei64_v_i64m4_m (vbool16_t mask, int64_t *base, vuint64m4_
 // CHECK-RV64-LABEL: @test_vamomaxei64_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomax.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomax.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamomaxei64_v_i64m8_m (vbool8_t mask, int64_t *base, vuint64m8_t bindex, vint64m8_t value, size_t vl) {
@@ -1163,7 +1163,7 @@ vint64m8_t test_vamomaxei64_v_i64m8_m (vbool8_t mask, int64_t *base, vuint64m8_t
 // CHECK-RV64-LABEL: @test_vamomaxuei8_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomaxu.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomaxu.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamomaxuei8_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint8mf8_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1174,7 +1174,7 @@ vuint32mf2_t test_vamomaxuei8_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint8
 // CHECK-RV64-LABEL: @test_vamomaxuei8_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomaxu.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomaxu.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamomaxuei8_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint8mf4_t bindex, vuint32m1_t value, size_t vl) {
@@ -1185,7 +1185,7 @@ vuint32m1_t test_vamomaxuei8_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint8mf
 // CHECK-RV64-LABEL: @test_vamomaxuei8_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomaxu.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomaxu.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamomaxuei8_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint8mf2_t bindex, vuint32m2_t value, size_t vl) {
@@ -1196,7 +1196,7 @@ vuint32m2_t test_vamomaxuei8_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint8mf
 // CHECK-RV64-LABEL: @test_vamomaxuei8_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomaxu.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomaxu.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamomaxuei8_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint8m1_t bindex, vuint32m4_t value, size_t vl) {
@@ -1207,7 +1207,7 @@ vuint32m4_t test_vamomaxuei8_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint8m1_
 // CHECK-RV64-LABEL: @test_vamomaxuei8_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamomaxu.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamomaxu.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vamomaxuei8_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint8m2_t bindex, vuint32m8_t value, size_t vl) {
@@ -1218,7 +1218,7 @@ vuint32m8_t test_vamomaxuei8_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint8m2_
 // CHECK-RV64-LABEL: @test_vamomaxuei16_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomaxu.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomaxu.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamomaxuei16_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint16mf4_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1229,7 +1229,7 @@ vuint32mf2_t test_vamomaxuei16_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint
 // CHECK-RV64-LABEL: @test_vamomaxuei16_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomaxu.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomaxu.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamomaxuei16_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint16mf2_t bindex, vuint32m1_t value, size_t vl) {
@@ -1240,7 +1240,7 @@ vuint32m1_t test_vamomaxuei16_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint16
 // CHECK-RV64-LABEL: @test_vamomaxuei16_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomaxu.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomaxu.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamomaxuei16_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint16m1_t bindex, vuint32m2_t value, size_t vl) {
@@ -1251,7 +1251,7 @@ vuint32m2_t test_vamomaxuei16_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint16
 // CHECK-RV64-LABEL: @test_vamomaxuei16_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomaxu.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomaxu.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamomaxuei16_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint16m2_t bindex, vuint32m4_t value, size_t vl) {
@@ -1262,7 +1262,7 @@ vuint32m4_t test_vamomaxuei16_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamomaxuei16_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamomaxu.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamomaxu.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vamomaxuei16_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint16m4_t bindex, vuint32m8_t value, size_t vl) {
@@ -1273,7 +1273,7 @@ vuint32m8_t test_vamomaxuei16_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamomaxuei32_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomaxu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomaxu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamomaxuei32_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint32mf2_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1284,7 +1284,7 @@ vuint32mf2_t test_vamomaxuei32_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint
 // CHECK-RV64-LABEL: @test_vamomaxuei32_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomaxu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomaxu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamomaxuei32_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint32m1_t bindex, vuint32m1_t value, size_t vl) {
@@ -1295,7 +1295,7 @@ vuint32m1_t test_vamomaxuei32_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint32
 // CHECK-RV64-LABEL: @test_vamomaxuei32_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomaxu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomaxu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamomaxuei32_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint32m2_t bindex, vuint32m2_t value, size_t vl) {
@@ -1306,7 +1306,7 @@ vuint32m2_t test_vamomaxuei32_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint32
 // CHECK-RV64-LABEL: @test_vamomaxuei32_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomaxu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomaxu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamomaxuei32_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint32m4_t bindex, vuint32m4_t value, size_t vl) {
@@ -1317,7 +1317,7 @@ vuint32m4_t test_vamomaxuei32_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamomaxuei32_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamomaxu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamomaxu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vamomaxuei32_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint32m8_t bindex, vuint32m8_t value, size_t vl) {
@@ -1328,7 +1328,7 @@ vuint32m8_t test_vamomaxuei32_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamomaxuei64_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomaxu.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomaxu.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamomaxuei64_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint64m1_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1339,7 +1339,7 @@ vuint32mf2_t test_vamomaxuei64_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint
 // CHECK-RV64-LABEL: @test_vamomaxuei64_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomaxu.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomaxu.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamomaxuei64_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint64m2_t bindex, vuint32m1_t value, size_t vl) {
@@ -1350,7 +1350,7 @@ vuint32m1_t test_vamomaxuei64_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint64
 // CHECK-RV64-LABEL: @test_vamomaxuei64_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomaxu.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomaxu.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamomaxuei64_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint64m4_t bindex, vuint32m2_t value, size_t vl) {
@@ -1361,7 +1361,7 @@ vuint32m2_t test_vamomaxuei64_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint64
 // CHECK-RV64-LABEL: @test_vamomaxuei64_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomaxu.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomaxu.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamomaxuei64_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint64m8_t bindex, vuint32m4_t value, size_t vl) {
@@ -1372,7 +1372,7 @@ vuint32m4_t test_vamomaxuei64_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamomaxuei8_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomaxu.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomaxu.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamomaxuei8_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint8mf8_t bindex, vuint64m1_t value, size_t vl) {
@@ -1383,7 +1383,7 @@ vuint64m1_t test_vamomaxuei8_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint8mf
 // CHECK-RV64-LABEL: @test_vamomaxuei8_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomaxu.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomaxu.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamomaxuei8_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint8mf4_t bindex, vuint64m2_t value, size_t vl) {
@@ -1394,7 +1394,7 @@ vuint64m2_t test_vamomaxuei8_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint8mf
 // CHECK-RV64-LABEL: @test_vamomaxuei8_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomaxu.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomaxu.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamomaxuei8_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint8mf2_t bindex, vuint64m4_t value, size_t vl) {
@@ -1405,7 +1405,7 @@ vuint64m4_t test_vamomaxuei8_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint8mf
 // CHECK-RV64-LABEL: @test_vamomaxuei8_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomaxu.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomaxu.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamomaxuei8_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint8m1_t bindex, vuint64m8_t value, size_t vl) {
@@ -1416,7 +1416,7 @@ vuint64m8_t test_vamomaxuei8_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint8m1_
 // CHECK-RV64-LABEL: @test_vamomaxuei16_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomaxu.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomaxu.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamomaxuei16_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint16mf4_t bindex, vuint64m1_t value, size_t vl) {
@@ -1427,7 +1427,7 @@ vuint64m1_t test_vamomaxuei16_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint16
 // CHECK-RV64-LABEL: @test_vamomaxuei16_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomaxu.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomaxu.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamomaxuei16_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint16mf2_t bindex, vuint64m2_t value, size_t vl) {
@@ -1438,7 +1438,7 @@ vuint64m2_t test_vamomaxuei16_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint16
 // CHECK-RV64-LABEL: @test_vamomaxuei16_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomaxu.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomaxu.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamomaxuei16_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint16m1_t bindex, vuint64m4_t value, size_t vl) {
@@ -1449,7 +1449,7 @@ vuint64m4_t test_vamomaxuei16_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint16
 // CHECK-RV64-LABEL: @test_vamomaxuei16_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomaxu.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomaxu.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamomaxuei16_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint16m2_t bindex, vuint64m8_t value, size_t vl) {
@@ -1460,7 +1460,7 @@ vuint64m8_t test_vamomaxuei16_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamomaxuei32_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomaxu.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomaxu.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamomaxuei32_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint32mf2_t bindex, vuint64m1_t value, size_t vl) {
@@ -1471,7 +1471,7 @@ vuint64m1_t test_vamomaxuei32_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint32
 // CHECK-RV64-LABEL: @test_vamomaxuei32_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomaxu.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomaxu.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamomaxuei32_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint32m1_t bindex, vuint64m2_t value, size_t vl) {
@@ -1482,7 +1482,7 @@ vuint64m2_t test_vamomaxuei32_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint32
 // CHECK-RV64-LABEL: @test_vamomaxuei32_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomaxu.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomaxu.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamomaxuei32_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint32m2_t bindex, vuint64m4_t value, size_t vl) {
@@ -1493,7 +1493,7 @@ vuint64m4_t test_vamomaxuei32_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint32
 // CHECK-RV64-LABEL: @test_vamomaxuei32_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomaxu.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomaxu.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamomaxuei32_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint32m4_t bindex, vuint64m8_t value, size_t vl) {
@@ -1504,7 +1504,7 @@ vuint64m8_t test_vamomaxuei32_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamomaxuei64_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomaxu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomaxu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamomaxuei64_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint64m1_t bindex, vuint64m1_t value, size_t vl) {
@@ -1515,7 +1515,7 @@ vuint64m1_t test_vamomaxuei64_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint64
 // CHECK-RV64-LABEL: @test_vamomaxuei64_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomaxu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomaxu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamomaxuei64_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint64m2_t bindex, vuint64m2_t value, size_t vl) {
@@ -1526,7 +1526,7 @@ vuint64m2_t test_vamomaxuei64_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint64
 // CHECK-RV64-LABEL: @test_vamomaxuei64_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomaxu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomaxu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamomaxuei64_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint64m4_t bindex, vuint64m4_t value, size_t vl) {
@@ -1537,7 +1537,7 @@ vuint64m4_t test_vamomaxuei64_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint64
 // CHECK-RV64-LABEL: @test_vamomaxuei64_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomaxu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomaxu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamomaxuei64_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint64m8_t bindex, vuint64m8_t value, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamomin.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamomin.c
index f875de1..eba57e5 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamomin.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamomin.c
@@ -778,7 +778,7 @@ vuint64m8_t test_vamominuei64_v_u64m8 (uint64_t *base, vuint64m8_t bindex, vuint
 // CHECK-RV64-LABEL: @test_vamominei8_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomin.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomin.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamominei8_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint8mf8_t bindex, vint32mf2_t value, size_t vl) {
@@ -789,7 +789,7 @@ vint32mf2_t test_vamominei8_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint8mf8
 // CHECK-RV64-LABEL: @test_vamominei8_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomin.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomin.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamominei8_v_i32m1_m (vbool32_t mask, int32_t *base, vuint8mf4_t bindex, vint32m1_t value, size_t vl) {
@@ -800,7 +800,7 @@ vint32m1_t test_vamominei8_v_i32m1_m (vbool32_t mask, int32_t *base, vuint8mf4_t
 // CHECK-RV64-LABEL: @test_vamominei8_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomin.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomin.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamominei8_v_i32m2_m (vbool16_t mask, int32_t *base, vuint8mf2_t bindex, vint32m2_t value, size_t vl) {
@@ -811,7 +811,7 @@ vint32m2_t test_vamominei8_v_i32m2_m (vbool16_t mask, int32_t *base, vuint8mf2_t
 // CHECK-RV64-LABEL: @test_vamominei8_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomin.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomin.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamominei8_v_i32m4_m (vbool8_t mask, int32_t *base, vuint8m1_t bindex, vint32m4_t value, size_t vl) {
@@ -822,7 +822,7 @@ vint32m4_t test_vamominei8_v_i32m4_m (vbool8_t mask, int32_t *base, vuint8m1_t b
 // CHECK-RV64-LABEL: @test_vamominei8_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamomin.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamomin.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vamominei8_v_i32m8_m (vbool4_t mask, int32_t *base, vuint8m2_t bindex, vint32m8_t value, size_t vl) {
@@ -833,7 +833,7 @@ vint32m8_t test_vamominei8_v_i32m8_m (vbool4_t mask, int32_t *base, vuint8m2_t b
 // CHECK-RV64-LABEL: @test_vamominei16_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomin.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomin.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamominei16_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint16mf4_t bindex, vint32mf2_t value, size_t vl) {
@@ -844,7 +844,7 @@ vint32mf2_t test_vamominei16_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamominei16_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomin.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomin.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamominei16_v_i32m1_m (vbool32_t mask, int32_t *base, vuint16mf2_t bindex, vint32m1_t value, size_t vl) {
@@ -855,7 +855,7 @@ vint32m1_t test_vamominei16_v_i32m1_m (vbool32_t mask, int32_t *base, vuint16mf2
 // CHECK-RV64-LABEL: @test_vamominei16_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomin.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomin.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamominei16_v_i32m2_m (vbool16_t mask, int32_t *base, vuint16m1_t bindex, vint32m2_t value, size_t vl) {
@@ -866,7 +866,7 @@ vint32m2_t test_vamominei16_v_i32m2_m (vbool16_t mask, int32_t *base, vuint16m1_
 // CHECK-RV64-LABEL: @test_vamominei16_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomin.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomin.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamominei16_v_i32m4_m (vbool8_t mask, int32_t *base, vuint16m2_t bindex, vint32m4_t value, size_t vl) {
@@ -877,7 +877,7 @@ vint32m4_t test_vamominei16_v_i32m4_m (vbool8_t mask, int32_t *base, vuint16m2_t
 // CHECK-RV64-LABEL: @test_vamominei16_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamomin.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamomin.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vamominei16_v_i32m8_m (vbool4_t mask, int32_t *base, vuint16m4_t bindex, vint32m8_t value, size_t vl) {
@@ -888,7 +888,7 @@ vint32m8_t test_vamominei16_v_i32m8_m (vbool4_t mask, int32_t *base, vuint16m4_t
 // CHECK-RV64-LABEL: @test_vamominei32_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomin.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomin.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamominei32_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint32mf2_t bindex, vint32mf2_t value, size_t vl) {
@@ -899,7 +899,7 @@ vint32mf2_t test_vamominei32_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamominei32_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomin.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomin.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamominei32_v_i32m1_m (vbool32_t mask, int32_t *base, vuint32m1_t bindex, vint32m1_t value, size_t vl) {
@@ -910,7 +910,7 @@ vint32m1_t test_vamominei32_v_i32m1_m (vbool32_t mask, int32_t *base, vuint32m1_
 // CHECK-RV64-LABEL: @test_vamominei32_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomin.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomin.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamominei32_v_i32m2_m (vbool16_t mask, int32_t *base, vuint32m2_t bindex, vint32m2_t value, size_t vl) {
@@ -921,7 +921,7 @@ vint32m2_t test_vamominei32_v_i32m2_m (vbool16_t mask, int32_t *base, vuint32m2_
 // CHECK-RV64-LABEL: @test_vamominei32_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomin.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomin.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamominei32_v_i32m4_m (vbool8_t mask, int32_t *base, vuint32m4_t bindex, vint32m4_t value, size_t vl) {
@@ -932,7 +932,7 @@ vint32m4_t test_vamominei32_v_i32m4_m (vbool8_t mask, int32_t *base, vuint32m4_t
 // CHECK-RV64-LABEL: @test_vamominei32_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamomin.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamomin.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vamominei32_v_i32m8_m (vbool4_t mask, int32_t *base, vuint32m8_t bindex, vint32m8_t value, size_t vl) {
@@ -943,7 +943,7 @@ vint32m8_t test_vamominei32_v_i32m8_m (vbool4_t mask, int32_t *base, vuint32m8_t
 // CHECK-RV64-LABEL: @test_vamominei64_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomin.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamomin.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamominei64_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint64m1_t bindex, vint32mf2_t value, size_t vl) {
@@ -954,7 +954,7 @@ vint32mf2_t test_vamominei64_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamominei64_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomin.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamomin.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamominei64_v_i32m1_m (vbool32_t mask, int32_t *base, vuint64m2_t bindex, vint32m1_t value, size_t vl) {
@@ -965,7 +965,7 @@ vint32m1_t test_vamominei64_v_i32m1_m (vbool32_t mask, int32_t *base, vuint64m2_
 // CHECK-RV64-LABEL: @test_vamominei64_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomin.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamomin.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamominei64_v_i32m2_m (vbool16_t mask, int32_t *base, vuint64m4_t bindex, vint32m2_t value, size_t vl) {
@@ -976,7 +976,7 @@ vint32m2_t test_vamominei64_v_i32m2_m (vbool16_t mask, int32_t *base, vuint64m4_
 // CHECK-RV64-LABEL: @test_vamominei64_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomin.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamomin.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamominei64_v_i32m4_m (vbool8_t mask, int32_t *base, vuint64m8_t bindex, vint32m4_t value, size_t vl) {
@@ -987,7 +987,7 @@ vint32m4_t test_vamominei64_v_i32m4_m (vbool8_t mask, int32_t *base, vuint64m8_t
 // CHECK-RV64-LABEL: @test_vamominei8_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomin.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomin.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamominei8_v_i64m1_m (vbool64_t mask, int64_t *base, vuint8mf8_t bindex, vint64m1_t value, size_t vl) {
@@ -998,7 +998,7 @@ vint64m1_t test_vamominei8_v_i64m1_m (vbool64_t mask, int64_t *base, vuint8mf8_t
 // CHECK-RV64-LABEL: @test_vamominei8_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomin.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomin.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamominei8_v_i64m2_m (vbool32_t mask, int64_t *base, vuint8mf4_t bindex, vint64m2_t value, size_t vl) {
@@ -1009,7 +1009,7 @@ vint64m2_t test_vamominei8_v_i64m2_m (vbool32_t mask, int64_t *base, vuint8mf4_t
 // CHECK-RV64-LABEL: @test_vamominei8_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomin.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomin.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamominei8_v_i64m4_m (vbool16_t mask, int64_t *base, vuint8mf2_t bindex, vint64m4_t value, size_t vl) {
@@ -1020,7 +1020,7 @@ vint64m4_t test_vamominei8_v_i64m4_m (vbool16_t mask, int64_t *base, vuint8mf2_t
 // CHECK-RV64-LABEL: @test_vamominei8_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomin.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomin.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamominei8_v_i64m8_m (vbool8_t mask, int64_t *base, vuint8m1_t bindex, vint64m8_t value, size_t vl) {
@@ -1031,7 +1031,7 @@ vint64m8_t test_vamominei8_v_i64m8_m (vbool8_t mask, int64_t *base, vuint8m1_t b
 // CHECK-RV64-LABEL: @test_vamominei16_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomin.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomin.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamominei16_v_i64m1_m (vbool64_t mask, int64_t *base, vuint16mf4_t bindex, vint64m1_t value, size_t vl) {
@@ -1042,7 +1042,7 @@ vint64m1_t test_vamominei16_v_i64m1_m (vbool64_t mask, int64_t *base, vuint16mf4
 // CHECK-RV64-LABEL: @test_vamominei16_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomin.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomin.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamominei16_v_i64m2_m (vbool32_t mask, int64_t *base, vuint16mf2_t bindex, vint64m2_t value, size_t vl) {
@@ -1053,7 +1053,7 @@ vint64m2_t test_vamominei16_v_i64m2_m (vbool32_t mask, int64_t *base, vuint16mf2
 // CHECK-RV64-LABEL: @test_vamominei16_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomin.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomin.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamominei16_v_i64m4_m (vbool16_t mask, int64_t *base, vuint16m1_t bindex, vint64m4_t value, size_t vl) {
@@ -1064,7 +1064,7 @@ vint64m4_t test_vamominei16_v_i64m4_m (vbool16_t mask, int64_t *base, vuint16m1_
 // CHECK-RV64-LABEL: @test_vamominei16_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomin.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomin.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamominei16_v_i64m8_m (vbool8_t mask, int64_t *base, vuint16m2_t bindex, vint64m8_t value, size_t vl) {
@@ -1075,7 +1075,7 @@ vint64m8_t test_vamominei16_v_i64m8_m (vbool8_t mask, int64_t *base, vuint16m2_t
 // CHECK-RV64-LABEL: @test_vamominei32_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomin.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomin.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamominei32_v_i64m1_m (vbool64_t mask, int64_t *base, vuint32mf2_t bindex, vint64m1_t value, size_t vl) {
@@ -1086,7 +1086,7 @@ vint64m1_t test_vamominei32_v_i64m1_m (vbool64_t mask, int64_t *base, vuint32mf2
 // CHECK-RV64-LABEL: @test_vamominei32_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomin.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomin.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamominei32_v_i64m2_m (vbool32_t mask, int64_t *base, vuint32m1_t bindex, vint64m2_t value, size_t vl) {
@@ -1097,7 +1097,7 @@ vint64m2_t test_vamominei32_v_i64m2_m (vbool32_t mask, int64_t *base, vuint32m1_
 // CHECK-RV64-LABEL: @test_vamominei32_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomin.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomin.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamominei32_v_i64m4_m (vbool16_t mask, int64_t *base, vuint32m2_t bindex, vint64m4_t value, size_t vl) {
@@ -1108,7 +1108,7 @@ vint64m4_t test_vamominei32_v_i64m4_m (vbool16_t mask, int64_t *base, vuint32m2_
 // CHECK-RV64-LABEL: @test_vamominei32_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomin.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomin.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamominei32_v_i64m8_m (vbool8_t mask, int64_t *base, vuint32m4_t bindex, vint64m8_t value, size_t vl) {
@@ -1119,7 +1119,7 @@ vint64m8_t test_vamominei32_v_i64m8_m (vbool8_t mask, int64_t *base, vuint32m4_t
 // CHECK-RV64-LABEL: @test_vamominei64_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomin.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamomin.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamominei64_v_i64m1_m (vbool64_t mask, int64_t *base, vuint64m1_t bindex, vint64m1_t value, size_t vl) {
@@ -1130,7 +1130,7 @@ vint64m1_t test_vamominei64_v_i64m1_m (vbool64_t mask, int64_t *base, vuint64m1_
 // CHECK-RV64-LABEL: @test_vamominei64_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomin.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamomin.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamominei64_v_i64m2_m (vbool32_t mask, int64_t *base, vuint64m2_t bindex, vint64m2_t value, size_t vl) {
@@ -1141,7 +1141,7 @@ vint64m2_t test_vamominei64_v_i64m2_m (vbool32_t mask, int64_t *base, vuint64m2_
 // CHECK-RV64-LABEL: @test_vamominei64_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomin.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamomin.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamominei64_v_i64m4_m (vbool16_t mask, int64_t *base, vuint64m4_t bindex, vint64m4_t value, size_t vl) {
@@ -1152,7 +1152,7 @@ vint64m4_t test_vamominei64_v_i64m4_m (vbool16_t mask, int64_t *base, vuint64m4_
 // CHECK-RV64-LABEL: @test_vamominei64_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomin.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamomin.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamominei64_v_i64m8_m (vbool8_t mask, int64_t *base, vuint64m8_t bindex, vint64m8_t value, size_t vl) {
@@ -1163,7 +1163,7 @@ vint64m8_t test_vamominei64_v_i64m8_m (vbool8_t mask, int64_t *base, vuint64m8_t
 // CHECK-RV64-LABEL: @test_vamominuei8_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamominu.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamominu.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamominuei8_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint8mf8_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1174,7 +1174,7 @@ vuint32mf2_t test_vamominuei8_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint8
 // CHECK-RV64-LABEL: @test_vamominuei8_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamominu.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamominu.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamominuei8_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint8mf4_t bindex, vuint32m1_t value, size_t vl) {
@@ -1185,7 +1185,7 @@ vuint32m1_t test_vamominuei8_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint8mf
 // CHECK-RV64-LABEL: @test_vamominuei8_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamominu.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamominu.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamominuei8_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint8mf2_t bindex, vuint32m2_t value, size_t vl) {
@@ -1196,7 +1196,7 @@ vuint32m2_t test_vamominuei8_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint8mf
 // CHECK-RV64-LABEL: @test_vamominuei8_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamominu.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamominu.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamominuei8_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint8m1_t bindex, vuint32m4_t value, size_t vl) {
@@ -1207,7 +1207,7 @@ vuint32m4_t test_vamominuei8_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint8m1_
 // CHECK-RV64-LABEL: @test_vamominuei8_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamominu.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamominu.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vamominuei8_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint8m2_t bindex, vuint32m8_t value, size_t vl) {
@@ -1218,7 +1218,7 @@ vuint32m8_t test_vamominuei8_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint8m2_
 // CHECK-RV64-LABEL: @test_vamominuei16_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamominu.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamominu.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamominuei16_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint16mf4_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1229,7 +1229,7 @@ vuint32mf2_t test_vamominuei16_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint
 // CHECK-RV64-LABEL: @test_vamominuei16_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamominu.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamominu.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamominuei16_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint16mf2_t bindex, vuint32m1_t value, size_t vl) {
@@ -1240,7 +1240,7 @@ vuint32m1_t test_vamominuei16_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint16
 // CHECK-RV64-LABEL: @test_vamominuei16_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamominu.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamominu.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamominuei16_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint16m1_t bindex, vuint32m2_t value, size_t vl) {
@@ -1251,7 +1251,7 @@ vuint32m2_t test_vamominuei16_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint16
 // CHECK-RV64-LABEL: @test_vamominuei16_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamominu.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamominu.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamominuei16_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint16m2_t bindex, vuint32m4_t value, size_t vl) {
@@ -1262,7 +1262,7 @@ vuint32m4_t test_vamominuei16_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamominuei16_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamominu.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamominu.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vamominuei16_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint16m4_t bindex, vuint32m8_t value, size_t vl) {
@@ -1273,7 +1273,7 @@ vuint32m8_t test_vamominuei16_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamominuei32_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamominu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamominu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamominuei32_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint32mf2_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1284,7 +1284,7 @@ vuint32mf2_t test_vamominuei32_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint
 // CHECK-RV64-LABEL: @test_vamominuei32_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamominu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamominu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamominuei32_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint32m1_t bindex, vuint32m1_t value, size_t vl) {
@@ -1295,7 +1295,7 @@ vuint32m1_t test_vamominuei32_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint32
 // CHECK-RV64-LABEL: @test_vamominuei32_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamominu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamominu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamominuei32_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint32m2_t bindex, vuint32m2_t value, size_t vl) {
@@ -1306,7 +1306,7 @@ vuint32m2_t test_vamominuei32_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint32
 // CHECK-RV64-LABEL: @test_vamominuei32_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamominu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamominu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamominuei32_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint32m4_t bindex, vuint32m4_t value, size_t vl) {
@@ -1317,7 +1317,7 @@ vuint32m4_t test_vamominuei32_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamominuei32_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamominu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamominu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vamominuei32_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint32m8_t bindex, vuint32m8_t value, size_t vl) {
@@ -1328,7 +1328,7 @@ vuint32m8_t test_vamominuei32_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamominuei64_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamominu.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamominu.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamominuei64_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint64m1_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1339,7 +1339,7 @@ vuint32mf2_t test_vamominuei64_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint
 // CHECK-RV64-LABEL: @test_vamominuei64_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamominu.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamominu.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamominuei64_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint64m2_t bindex, vuint32m1_t value, size_t vl) {
@@ -1350,7 +1350,7 @@ vuint32m1_t test_vamominuei64_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint64
 // CHECK-RV64-LABEL: @test_vamominuei64_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamominu.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamominu.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamominuei64_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint64m4_t bindex, vuint32m2_t value, size_t vl) {
@@ -1361,7 +1361,7 @@ vuint32m2_t test_vamominuei64_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint64
 // CHECK-RV64-LABEL: @test_vamominuei64_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamominu.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamominu.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamominuei64_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint64m8_t bindex, vuint32m4_t value, size_t vl) {
@@ -1372,7 +1372,7 @@ vuint32m4_t test_vamominuei64_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamominuei8_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamominu.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamominu.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamominuei8_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint8mf8_t bindex, vuint64m1_t value, size_t vl) {
@@ -1383,7 +1383,7 @@ vuint64m1_t test_vamominuei8_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint8mf
 // CHECK-RV64-LABEL: @test_vamominuei8_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamominu.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamominu.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamominuei8_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint8mf4_t bindex, vuint64m2_t value, size_t vl) {
@@ -1394,7 +1394,7 @@ vuint64m2_t test_vamominuei8_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint8mf
 // CHECK-RV64-LABEL: @test_vamominuei8_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamominu.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamominu.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamominuei8_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint8mf2_t bindex, vuint64m4_t value, size_t vl) {
@@ -1405,7 +1405,7 @@ vuint64m4_t test_vamominuei8_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint8mf
 // CHECK-RV64-LABEL: @test_vamominuei8_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamominu.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamominu.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamominuei8_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint8m1_t bindex, vuint64m8_t value, size_t vl) {
@@ -1416,7 +1416,7 @@ vuint64m8_t test_vamominuei8_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint8m1_
 // CHECK-RV64-LABEL: @test_vamominuei16_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamominu.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamominu.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamominuei16_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint16mf4_t bindex, vuint64m1_t value, size_t vl) {
@@ -1427,7 +1427,7 @@ vuint64m1_t test_vamominuei16_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint16
 // CHECK-RV64-LABEL: @test_vamominuei16_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamominu.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamominu.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamominuei16_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint16mf2_t bindex, vuint64m2_t value, size_t vl) {
@@ -1438,7 +1438,7 @@ vuint64m2_t test_vamominuei16_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint16
 // CHECK-RV64-LABEL: @test_vamominuei16_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamominu.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamominu.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamominuei16_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint16m1_t bindex, vuint64m4_t value, size_t vl) {
@@ -1449,7 +1449,7 @@ vuint64m4_t test_vamominuei16_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint16
 // CHECK-RV64-LABEL: @test_vamominuei16_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamominu.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamominu.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamominuei16_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint16m2_t bindex, vuint64m8_t value, size_t vl) {
@@ -1460,7 +1460,7 @@ vuint64m8_t test_vamominuei16_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamominuei32_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamominu.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamominu.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamominuei32_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint32mf2_t bindex, vuint64m1_t value, size_t vl) {
@@ -1471,7 +1471,7 @@ vuint64m1_t test_vamominuei32_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint32
 // CHECK-RV64-LABEL: @test_vamominuei32_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamominu.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamominu.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamominuei32_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint32m1_t bindex, vuint64m2_t value, size_t vl) {
@@ -1482,7 +1482,7 @@ vuint64m2_t test_vamominuei32_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint32
 // CHECK-RV64-LABEL: @test_vamominuei32_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamominu.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamominu.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamominuei32_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint32m2_t bindex, vuint64m4_t value, size_t vl) {
@@ -1493,7 +1493,7 @@ vuint64m4_t test_vamominuei32_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint32
 // CHECK-RV64-LABEL: @test_vamominuei32_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamominu.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamominu.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamominuei32_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint32m4_t bindex, vuint64m8_t value, size_t vl) {
@@ -1504,7 +1504,7 @@ vuint64m8_t test_vamominuei32_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamominuei64_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamominu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamominu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamominuei64_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint64m1_t bindex, vuint64m1_t value, size_t vl) {
@@ -1515,7 +1515,7 @@ vuint64m1_t test_vamominuei64_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint64
 // CHECK-RV64-LABEL: @test_vamominuei64_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamominu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamominu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamominuei64_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint64m2_t bindex, vuint64m2_t value, size_t vl) {
@@ -1526,7 +1526,7 @@ vuint64m2_t test_vamominuei64_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint64
 // CHECK-RV64-LABEL: @test_vamominuei64_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamominu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamominu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamominuei64_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint64m4_t bindex, vuint64m4_t value, size_t vl) {
@@ -1537,7 +1537,7 @@ vuint64m4_t test_vamominuei64_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint64
 // CHECK-RV64-LABEL: @test_vamominuei64_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamominu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamominu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamominuei64_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint64m8_t bindex, vuint64m8_t value, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamoor.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamoor.c
index 379f154..56cc72e0 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamoor.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamoor.c
@@ -778,7 +778,7 @@ vuint64m8_t test_vamoorei64_v_u64m8 (uint64_t *base, vuint64m8_t bindex, vuint64
 // CHECK-RV64-LABEL: @test_vamoorei8_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoor.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoor.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamoorei8_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint8mf8_t bindex, vint32mf2_t value, size_t vl) {
@@ -789,7 +789,7 @@ vint32mf2_t test_vamoorei8_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint8mf8_
 // CHECK-RV64-LABEL: @test_vamoorei8_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoor.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoor.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamoorei8_v_i32m1_m (vbool32_t mask, int32_t *base, vuint8mf4_t bindex, vint32m1_t value, size_t vl) {
@@ -800,7 +800,7 @@ vint32m1_t test_vamoorei8_v_i32m1_m (vbool32_t mask, int32_t *base, vuint8mf4_t
 // CHECK-RV64-LABEL: @test_vamoorei8_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoor.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoor.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamoorei8_v_i32m2_m (vbool16_t mask, int32_t *base, vuint8mf2_t bindex, vint32m2_t value, size_t vl) {
@@ -811,7 +811,7 @@ vint32m2_t test_vamoorei8_v_i32m2_m (vbool16_t mask, int32_t *base, vuint8mf2_t
 // CHECK-RV64-LABEL: @test_vamoorei8_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoor.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoor.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamoorei8_v_i32m4_m (vbool8_t mask, int32_t *base, vuint8m1_t bindex, vint32m4_t value, size_t vl) {
@@ -822,7 +822,7 @@ vint32m4_t test_vamoorei8_v_i32m4_m (vbool8_t mask, int32_t *base, vuint8m1_t bi
 // CHECK-RV64-LABEL: @test_vamoorei8_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoor.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoor.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vamoorei8_v_i32m8_m (vbool4_t mask, int32_t *base, vuint8m2_t bindex, vint32m8_t value, size_t vl) {
@@ -833,7 +833,7 @@ vint32m8_t test_vamoorei8_v_i32m8_m (vbool4_t mask, int32_t *base, vuint8m2_t bi
 // CHECK-RV64-LABEL: @test_vamoorei16_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoor.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoor.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamoorei16_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint16mf4_t bindex, vint32mf2_t value, size_t vl) {
@@ -844,7 +844,7 @@ vint32mf2_t test_vamoorei16_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint16mf
 // CHECK-RV64-LABEL: @test_vamoorei16_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoor.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoor.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamoorei16_v_i32m1_m (vbool32_t mask, int32_t *base, vuint16mf2_t bindex, vint32m1_t value, size_t vl) {
@@ -855,7 +855,7 @@ vint32m1_t test_vamoorei16_v_i32m1_m (vbool32_t mask, int32_t *base, vuint16mf2_
 // CHECK-RV64-LABEL: @test_vamoorei16_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoor.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoor.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamoorei16_v_i32m2_m (vbool16_t mask, int32_t *base, vuint16m1_t bindex, vint32m2_t value, size_t vl) {
@@ -866,7 +866,7 @@ vint32m2_t test_vamoorei16_v_i32m2_m (vbool16_t mask, int32_t *base, vuint16m1_t
 // CHECK-RV64-LABEL: @test_vamoorei16_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoor.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoor.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamoorei16_v_i32m4_m (vbool8_t mask, int32_t *base, vuint16m2_t bindex, vint32m4_t value, size_t vl) {
@@ -877,7 +877,7 @@ vint32m4_t test_vamoorei16_v_i32m4_m (vbool8_t mask, int32_t *base, vuint16m2_t
 // CHECK-RV64-LABEL: @test_vamoorei16_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoor.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoor.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vamoorei16_v_i32m8_m (vbool4_t mask, int32_t *base, vuint16m4_t bindex, vint32m8_t value, size_t vl) {
@@ -888,7 +888,7 @@ vint32m8_t test_vamoorei16_v_i32m8_m (vbool4_t mask, int32_t *base, vuint16m4_t
 // CHECK-RV64-LABEL: @test_vamoorei32_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoor.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoor.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamoorei32_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint32mf2_t bindex, vint32mf2_t value, size_t vl) {
@@ -899,7 +899,7 @@ vint32mf2_t test_vamoorei32_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint32mf
 // CHECK-RV64-LABEL: @test_vamoorei32_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamoorei32_v_i32m1_m (vbool32_t mask, int32_t *base, vuint32m1_t bindex, vint32m1_t value, size_t vl) {
@@ -910,7 +910,7 @@ vint32m1_t test_vamoorei32_v_i32m1_m (vbool32_t mask, int32_t *base, vuint32m1_t
 // CHECK-RV64-LABEL: @test_vamoorei32_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoor.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoor.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamoorei32_v_i32m2_m (vbool16_t mask, int32_t *base, vuint32m2_t bindex, vint32m2_t value, size_t vl) {
@@ -921,7 +921,7 @@ vint32m2_t test_vamoorei32_v_i32m2_m (vbool16_t mask, int32_t *base, vuint32m2_t
 // CHECK-RV64-LABEL: @test_vamoorei32_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoor.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoor.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamoorei32_v_i32m4_m (vbool8_t mask, int32_t *base, vuint32m4_t bindex, vint32m4_t value, size_t vl) {
@@ -932,7 +932,7 @@ vint32m4_t test_vamoorei32_v_i32m4_m (vbool8_t mask, int32_t *base, vuint32m4_t
 // CHECK-RV64-LABEL: @test_vamoorei32_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoor.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoor.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vamoorei32_v_i32m8_m (vbool4_t mask, int32_t *base, vuint32m8_t bindex, vint32m8_t value, size_t vl) {
@@ -943,7 +943,7 @@ vint32m8_t test_vamoorei32_v_i32m8_m (vbool4_t mask, int32_t *base, vuint32m8_t
 // CHECK-RV64-LABEL: @test_vamoorei64_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoor.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoor.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamoorei64_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint64m1_t bindex, vint32mf2_t value, size_t vl) {
@@ -954,7 +954,7 @@ vint32mf2_t test_vamoorei64_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint64m1
 // CHECK-RV64-LABEL: @test_vamoorei64_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoor.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoor.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamoorei64_v_i32m1_m (vbool32_t mask, int32_t *base, vuint64m2_t bindex, vint32m1_t value, size_t vl) {
@@ -965,7 +965,7 @@ vint32m1_t test_vamoorei64_v_i32m1_m (vbool32_t mask, int32_t *base, vuint64m2_t
 // CHECK-RV64-LABEL: @test_vamoorei64_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoor.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoor.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamoorei64_v_i32m2_m (vbool16_t mask, int32_t *base, vuint64m4_t bindex, vint32m2_t value, size_t vl) {
@@ -976,7 +976,7 @@ vint32m2_t test_vamoorei64_v_i32m2_m (vbool16_t mask, int32_t *base, vuint64m4_t
 // CHECK-RV64-LABEL: @test_vamoorei64_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoor.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoor.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamoorei64_v_i32m4_m (vbool8_t mask, int32_t *base, vuint64m8_t bindex, vint32m4_t value, size_t vl) {
@@ -987,7 +987,7 @@ vint32m4_t test_vamoorei64_v_i32m4_m (vbool8_t mask, int32_t *base, vuint64m8_t
 // CHECK-RV64-LABEL: @test_vamoorei8_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoor.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoor.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamoorei8_v_i64m1_m (vbool64_t mask, int64_t *base, vuint8mf8_t bindex, vint64m1_t value, size_t vl) {
@@ -998,7 +998,7 @@ vint64m1_t test_vamoorei8_v_i64m1_m (vbool64_t mask, int64_t *base, vuint8mf8_t
 // CHECK-RV64-LABEL: @test_vamoorei8_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoor.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoor.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamoorei8_v_i64m2_m (vbool32_t mask, int64_t *base, vuint8mf4_t bindex, vint64m2_t value, size_t vl) {
@@ -1009,7 +1009,7 @@ vint64m2_t test_vamoorei8_v_i64m2_m (vbool32_t mask, int64_t *base, vuint8mf4_t
 // CHECK-RV64-LABEL: @test_vamoorei8_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoor.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoor.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamoorei8_v_i64m4_m (vbool16_t mask, int64_t *base, vuint8mf2_t bindex, vint64m4_t value, size_t vl) {
@@ -1020,7 +1020,7 @@ vint64m4_t test_vamoorei8_v_i64m4_m (vbool16_t mask, int64_t *base, vuint8mf2_t
 // CHECK-RV64-LABEL: @test_vamoorei8_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoor.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoor.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamoorei8_v_i64m8_m (vbool8_t mask, int64_t *base, vuint8m1_t bindex, vint64m8_t value, size_t vl) {
@@ -1031,7 +1031,7 @@ vint64m8_t test_vamoorei8_v_i64m8_m (vbool8_t mask, int64_t *base, vuint8m1_t bi
 // CHECK-RV64-LABEL: @test_vamoorei16_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoor.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoor.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamoorei16_v_i64m1_m (vbool64_t mask, int64_t *base, vuint16mf4_t bindex, vint64m1_t value, size_t vl) {
@@ -1042,7 +1042,7 @@ vint64m1_t test_vamoorei16_v_i64m1_m (vbool64_t mask, int64_t *base, vuint16mf4_
 // CHECK-RV64-LABEL: @test_vamoorei16_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoor.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoor.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamoorei16_v_i64m2_m (vbool32_t mask, int64_t *base, vuint16mf2_t bindex, vint64m2_t value, size_t vl) {
@@ -1053,7 +1053,7 @@ vint64m2_t test_vamoorei16_v_i64m2_m (vbool32_t mask, int64_t *base, vuint16mf2_
 // CHECK-RV64-LABEL: @test_vamoorei16_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoor.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoor.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamoorei16_v_i64m4_m (vbool16_t mask, int64_t *base, vuint16m1_t bindex, vint64m4_t value, size_t vl) {
@@ -1064,7 +1064,7 @@ vint64m4_t test_vamoorei16_v_i64m4_m (vbool16_t mask, int64_t *base, vuint16m1_t
 // CHECK-RV64-LABEL: @test_vamoorei16_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoor.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoor.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamoorei16_v_i64m8_m (vbool8_t mask, int64_t *base, vuint16m2_t bindex, vint64m8_t value, size_t vl) {
@@ -1075,7 +1075,7 @@ vint64m8_t test_vamoorei16_v_i64m8_m (vbool8_t mask, int64_t *base, vuint16m2_t
 // CHECK-RV64-LABEL: @test_vamoorei32_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoor.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoor.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamoorei32_v_i64m1_m (vbool64_t mask, int64_t *base, vuint32mf2_t bindex, vint64m1_t value, size_t vl) {
@@ -1086,7 +1086,7 @@ vint64m1_t test_vamoorei32_v_i64m1_m (vbool64_t mask, int64_t *base, vuint32mf2_
 // CHECK-RV64-LABEL: @test_vamoorei32_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoor.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoor.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamoorei32_v_i64m2_m (vbool32_t mask, int64_t *base, vuint32m1_t bindex, vint64m2_t value, size_t vl) {
@@ -1097,7 +1097,7 @@ vint64m2_t test_vamoorei32_v_i64m2_m (vbool32_t mask, int64_t *base, vuint32m1_t
 // CHECK-RV64-LABEL: @test_vamoorei32_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoor.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoor.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamoorei32_v_i64m4_m (vbool16_t mask, int64_t *base, vuint32m2_t bindex, vint64m4_t value, size_t vl) {
@@ -1108,7 +1108,7 @@ vint64m4_t test_vamoorei32_v_i64m4_m (vbool16_t mask, int64_t *base, vuint32m2_t
 // CHECK-RV64-LABEL: @test_vamoorei32_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoor.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoor.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamoorei32_v_i64m8_m (vbool8_t mask, int64_t *base, vuint32m4_t bindex, vint64m8_t value, size_t vl) {
@@ -1119,7 +1119,7 @@ vint64m8_t test_vamoorei32_v_i64m8_m (vbool8_t mask, int64_t *base, vuint32m4_t
 // CHECK-RV64-LABEL: @test_vamoorei64_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamoorei64_v_i64m1_m (vbool64_t mask, int64_t *base, vuint64m1_t bindex, vint64m1_t value, size_t vl) {
@@ -1130,7 +1130,7 @@ vint64m1_t test_vamoorei64_v_i64m1_m (vbool64_t mask, int64_t *base, vuint64m1_t
 // CHECK-RV64-LABEL: @test_vamoorei64_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoor.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoor.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamoorei64_v_i64m2_m (vbool32_t mask, int64_t *base, vuint64m2_t bindex, vint64m2_t value, size_t vl) {
@@ -1141,7 +1141,7 @@ vint64m2_t test_vamoorei64_v_i64m2_m (vbool32_t mask, int64_t *base, vuint64m2_t
 // CHECK-RV64-LABEL: @test_vamoorei64_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoor.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoor.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamoorei64_v_i64m4_m (vbool16_t mask, int64_t *base, vuint64m4_t bindex, vint64m4_t value, size_t vl) {
@@ -1152,7 +1152,7 @@ vint64m4_t test_vamoorei64_v_i64m4_m (vbool16_t mask, int64_t *base, vuint64m4_t
 // CHECK-RV64-LABEL: @test_vamoorei64_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoor.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoor.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamoorei64_v_i64m8_m (vbool8_t mask, int64_t *base, vuint64m8_t bindex, vint64m8_t value, size_t vl) {
@@ -1163,7 +1163,7 @@ vint64m8_t test_vamoorei64_v_i64m8_m (vbool8_t mask, int64_t *base, vuint64m8_t
 // CHECK-RV64-LABEL: @test_vamoorei8_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoor.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoor.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamoorei8_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint8mf8_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1174,7 +1174,7 @@ vuint32mf2_t test_vamoorei8_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint8mf
 // CHECK-RV64-LABEL: @test_vamoorei8_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoor.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoor.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamoorei8_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint8mf4_t bindex, vuint32m1_t value, size_t vl) {
@@ -1185,7 +1185,7 @@ vuint32m1_t test_vamoorei8_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint8mf4_
 // CHECK-RV64-LABEL: @test_vamoorei8_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoor.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoor.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamoorei8_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint8mf2_t bindex, vuint32m2_t value, size_t vl) {
@@ -1196,7 +1196,7 @@ vuint32m2_t test_vamoorei8_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint8mf2_
 // CHECK-RV64-LABEL: @test_vamoorei8_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoor.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoor.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamoorei8_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint8m1_t bindex, vuint32m4_t value, size_t vl) {
@@ -1207,7 +1207,7 @@ vuint32m4_t test_vamoorei8_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint8m1_t
 // CHECK-RV64-LABEL: @test_vamoorei8_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoor.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoor.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vamoorei8_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint8m2_t bindex, vuint32m8_t value, size_t vl) {
@@ -1218,7 +1218,7 @@ vuint32m8_t test_vamoorei8_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint8m2_t
 // CHECK-RV64-LABEL: @test_vamoorei16_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoor.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoor.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamoorei16_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint16mf4_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1229,7 +1229,7 @@ vuint32mf2_t test_vamoorei16_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint16
 // CHECK-RV64-LABEL: @test_vamoorei16_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoor.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoor.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamoorei16_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint16mf2_t bindex, vuint32m1_t value, size_t vl) {
@@ -1240,7 +1240,7 @@ vuint32m1_t test_vamoorei16_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint16mf
 // CHECK-RV64-LABEL: @test_vamoorei16_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoor.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoor.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamoorei16_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint16m1_t bindex, vuint32m2_t value, size_t vl) {
@@ -1251,7 +1251,7 @@ vuint32m2_t test_vamoorei16_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint16m1
 // CHECK-RV64-LABEL: @test_vamoorei16_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoor.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoor.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamoorei16_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint16m2_t bindex, vuint32m4_t value, size_t vl) {
@@ -1262,7 +1262,7 @@ vuint32m4_t test_vamoorei16_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint16m2_
 // CHECK-RV64-LABEL: @test_vamoorei16_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoor.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoor.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vamoorei16_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint16m4_t bindex, vuint32m8_t value, size_t vl) {
@@ -1273,7 +1273,7 @@ vuint32m8_t test_vamoorei16_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint16m4_
 // CHECK-RV64-LABEL: @test_vamoorei32_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoor.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoor.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamoorei32_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint32mf2_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1284,7 +1284,7 @@ vuint32mf2_t test_vamoorei32_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint32
 // CHECK-RV64-LABEL: @test_vamoorei32_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamoorei32_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint32m1_t bindex, vuint32m1_t value, size_t vl) {
@@ -1295,7 +1295,7 @@ vuint32m1_t test_vamoorei32_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint32m1
 // CHECK-RV64-LABEL: @test_vamoorei32_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoor.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoor.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamoorei32_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint32m2_t bindex, vuint32m2_t value, size_t vl) {
@@ -1306,7 +1306,7 @@ vuint32m2_t test_vamoorei32_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint32m2
 // CHECK-RV64-LABEL: @test_vamoorei32_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoor.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoor.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamoorei32_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint32m4_t bindex, vuint32m4_t value, size_t vl) {
@@ -1317,7 +1317,7 @@ vuint32m4_t test_vamoorei32_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint32m4_
 // CHECK-RV64-LABEL: @test_vamoorei32_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoor.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoor.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vamoorei32_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint32m8_t bindex, vuint32m8_t value, size_t vl) {
@@ -1328,7 +1328,7 @@ vuint32m8_t test_vamoorei32_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint32m8_
 // CHECK-RV64-LABEL: @test_vamoorei64_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoor.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoor.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamoorei64_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint64m1_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1339,7 +1339,7 @@ vuint32mf2_t test_vamoorei64_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint64
 // CHECK-RV64-LABEL: @test_vamoorei64_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoor.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoor.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamoorei64_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint64m2_t bindex, vuint32m1_t value, size_t vl) {
@@ -1350,7 +1350,7 @@ vuint32m1_t test_vamoorei64_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint64m2
 // CHECK-RV64-LABEL: @test_vamoorei64_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoor.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoor.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamoorei64_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint64m4_t bindex, vuint32m2_t value, size_t vl) {
@@ -1361,7 +1361,7 @@ vuint32m2_t test_vamoorei64_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint64m4
 // CHECK-RV64-LABEL: @test_vamoorei64_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoor.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoor.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamoorei64_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint64m8_t bindex, vuint32m4_t value, size_t vl) {
@@ -1372,7 +1372,7 @@ vuint32m4_t test_vamoorei64_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint64m8_
 // CHECK-RV64-LABEL: @test_vamoorei8_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoor.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoor.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamoorei8_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint8mf8_t bindex, vuint64m1_t value, size_t vl) {
@@ -1383,7 +1383,7 @@ vuint64m1_t test_vamoorei8_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint8mf8_
 // CHECK-RV64-LABEL: @test_vamoorei8_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoor.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoor.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamoorei8_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint8mf4_t bindex, vuint64m2_t value, size_t vl) {
@@ -1394,7 +1394,7 @@ vuint64m2_t test_vamoorei8_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint8mf4_
 // CHECK-RV64-LABEL: @test_vamoorei8_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoor.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoor.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamoorei8_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint8mf2_t bindex, vuint64m4_t value, size_t vl) {
@@ -1405,7 +1405,7 @@ vuint64m4_t test_vamoorei8_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint8mf2_
 // CHECK-RV64-LABEL: @test_vamoorei8_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoor.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoor.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamoorei8_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint8m1_t bindex, vuint64m8_t value, size_t vl) {
@@ -1416,7 +1416,7 @@ vuint64m8_t test_vamoorei8_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint8m1_t
 // CHECK-RV64-LABEL: @test_vamoorei16_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoor.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoor.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamoorei16_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint16mf4_t bindex, vuint64m1_t value, size_t vl) {
@@ -1427,7 +1427,7 @@ vuint64m1_t test_vamoorei16_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint16mf
 // CHECK-RV64-LABEL: @test_vamoorei16_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoor.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoor.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamoorei16_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint16mf2_t bindex, vuint64m2_t value, size_t vl) {
@@ -1438,7 +1438,7 @@ vuint64m2_t test_vamoorei16_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint16mf
 // CHECK-RV64-LABEL: @test_vamoorei16_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoor.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoor.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamoorei16_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint16m1_t bindex, vuint64m4_t value, size_t vl) {
@@ -1449,7 +1449,7 @@ vuint64m4_t test_vamoorei16_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint16m1
 // CHECK-RV64-LABEL: @test_vamoorei16_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoor.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoor.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamoorei16_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint16m2_t bindex, vuint64m8_t value, size_t vl) {
@@ -1460,7 +1460,7 @@ vuint64m8_t test_vamoorei16_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint16m2_
 // CHECK-RV64-LABEL: @test_vamoorei32_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoor.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoor.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamoorei32_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint32mf2_t bindex, vuint64m1_t value, size_t vl) {
@@ -1471,7 +1471,7 @@ vuint64m1_t test_vamoorei32_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint32mf
 // CHECK-RV64-LABEL: @test_vamoorei32_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoor.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoor.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamoorei32_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint32m1_t bindex, vuint64m2_t value, size_t vl) {
@@ -1482,7 +1482,7 @@ vuint64m2_t test_vamoorei32_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint32m1
 // CHECK-RV64-LABEL: @test_vamoorei32_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoor.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoor.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamoorei32_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint32m2_t bindex, vuint64m4_t value, size_t vl) {
@@ -1493,7 +1493,7 @@ vuint64m4_t test_vamoorei32_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint32m2
 // CHECK-RV64-LABEL: @test_vamoorei32_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoor.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoor.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamoorei32_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint32m4_t bindex, vuint64m8_t value, size_t vl) {
@@ -1504,7 +1504,7 @@ vuint64m8_t test_vamoorei32_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint32m4_
 // CHECK-RV64-LABEL: @test_vamoorei64_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamoorei64_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint64m1_t bindex, vuint64m1_t value, size_t vl) {
@@ -1515,7 +1515,7 @@ vuint64m1_t test_vamoorei64_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint64m1
 // CHECK-RV64-LABEL: @test_vamoorei64_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoor.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoor.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamoorei64_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint64m2_t bindex, vuint64m2_t value, size_t vl) {
@@ -1526,7 +1526,7 @@ vuint64m2_t test_vamoorei64_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint64m2
 // CHECK-RV64-LABEL: @test_vamoorei64_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoor.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoor.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamoorei64_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint64m4_t bindex, vuint64m4_t value, size_t vl) {
@@ -1537,7 +1537,7 @@ vuint64m4_t test_vamoorei64_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint64m4
 // CHECK-RV64-LABEL: @test_vamoorei64_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoor.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoor.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamoorei64_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint64m8_t bindex, vuint64m8_t value, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamoswap.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamoswap.c
index fe8957e..9a7d81d 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamoswap.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamoswap.c
@@ -1269,7 +1269,7 @@ vfloat64m8_t test_vamoswapei64_v_f64m8(double *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoswap.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoswap.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamoswapei8_v_i32mf2_m(vbool64_t mask, int32_t *base,
@@ -1282,7 +1282,7 @@ vint32mf2_t test_vamoswapei8_v_i32mf2_m(vbool64_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoswap.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoswap.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamoswapei8_v_i32m1_m(vbool32_t mask, int32_t *base,
@@ -1295,7 +1295,7 @@ vint32m1_t test_vamoswapei8_v_i32m1_m(vbool32_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoswap.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoswap.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamoswapei8_v_i32m2_m(vbool16_t mask, int32_t *base,
@@ -1308,7 +1308,7 @@ vint32m2_t test_vamoswapei8_v_i32m2_m(vbool16_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoswap.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoswap.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamoswapei8_v_i32m4_m(vbool8_t mask, int32_t *base,
@@ -1321,7 +1321,7 @@ vint32m4_t test_vamoswapei8_v_i32m4_m(vbool8_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoswap.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoswap.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vamoswapei8_v_i32m8_m(vbool4_t mask, int32_t *base,
@@ -1334,7 +1334,7 @@ vint32m8_t test_vamoswapei8_v_i32m8_m(vbool4_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoswap.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoswap.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamoswapei16_v_i32mf2_m(vbool64_t mask, int32_t *base,
@@ -1347,7 +1347,7 @@ vint32mf2_t test_vamoswapei16_v_i32mf2_m(vbool64_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoswap.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoswap.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamoswapei16_v_i32m1_m(vbool32_t mask, int32_t *base,
@@ -1360,7 +1360,7 @@ vint32m1_t test_vamoswapei16_v_i32m1_m(vbool32_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoswap.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoswap.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamoswapei16_v_i32m2_m(vbool16_t mask, int32_t *base,
@@ -1373,7 +1373,7 @@ vint32m2_t test_vamoswapei16_v_i32m2_m(vbool16_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoswap.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoswap.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamoswapei16_v_i32m4_m(vbool8_t mask, int32_t *base,
@@ -1386,7 +1386,7 @@ vint32m4_t test_vamoswapei16_v_i32m4_m(vbool8_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoswap.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoswap.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vamoswapei16_v_i32m8_m(vbool4_t mask, int32_t *base,
@@ -1399,7 +1399,7 @@ vint32m8_t test_vamoswapei16_v_i32m8_m(vbool4_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoswap.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoswap.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamoswapei32_v_i32mf2_m(vbool64_t mask, int32_t *base,
@@ -1412,7 +1412,7 @@ vint32mf2_t test_vamoswapei32_v_i32mf2_m(vbool64_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoswap.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoswap.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamoswapei32_v_i32m1_m(vbool32_t mask, int32_t *base,
@@ -1425,7 +1425,7 @@ vint32m1_t test_vamoswapei32_v_i32m1_m(vbool32_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoswap.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoswap.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamoswapei32_v_i32m2_m(vbool16_t mask, int32_t *base,
@@ -1438,7 +1438,7 @@ vint32m2_t test_vamoswapei32_v_i32m2_m(vbool16_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoswap.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoswap.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamoswapei32_v_i32m4_m(vbool8_t mask, int32_t *base,
@@ -1451,7 +1451,7 @@ vint32m4_t test_vamoswapei32_v_i32m4_m(vbool8_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoswap.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoswap.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vamoswapei32_v_i32m8_m(vbool4_t mask, int32_t *base,
@@ -1464,7 +1464,7 @@ vint32m8_t test_vamoswapei32_v_i32m8_m(vbool4_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoswap.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoswap.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamoswapei64_v_i32mf2_m(vbool64_t mask, int32_t *base,
@@ -1477,7 +1477,7 @@ vint32mf2_t test_vamoswapei64_v_i32mf2_m(vbool64_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoswap.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoswap.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamoswapei64_v_i32m1_m(vbool32_t mask, int32_t *base,
@@ -1490,7 +1490,7 @@ vint32m1_t test_vamoswapei64_v_i32m1_m(vbool32_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoswap.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoswap.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamoswapei64_v_i32m2_m(vbool16_t mask, int32_t *base,
@@ -1503,7 +1503,7 @@ vint32m2_t test_vamoswapei64_v_i32m2_m(vbool16_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoswap.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoswap.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamoswapei64_v_i32m4_m(vbool8_t mask, int32_t *base,
@@ -1516,7 +1516,7 @@ vint32m4_t test_vamoswapei64_v_i32m4_m(vbool8_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoswap.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoswap.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamoswapei8_v_i64m1_m(vbool64_t mask, int64_t *base,
@@ -1529,7 +1529,7 @@ vint64m1_t test_vamoswapei8_v_i64m1_m(vbool64_t mask, int64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoswap.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoswap.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamoswapei8_v_i64m2_m(vbool32_t mask, int64_t *base,
@@ -1542,7 +1542,7 @@ vint64m2_t test_vamoswapei8_v_i64m2_m(vbool32_t mask, int64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoswap.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoswap.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamoswapei8_v_i64m4_m(vbool16_t mask, int64_t *base,
@@ -1555,7 +1555,7 @@ vint64m4_t test_vamoswapei8_v_i64m4_m(vbool16_t mask, int64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoswap.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoswap.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamoswapei8_v_i64m8_m(vbool8_t mask, int64_t *base,
@@ -1568,7 +1568,7 @@ vint64m8_t test_vamoswapei8_v_i64m8_m(vbool8_t mask, int64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoswap.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoswap.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamoswapei16_v_i64m1_m(vbool64_t mask, int64_t *base,
@@ -1581,7 +1581,7 @@ vint64m1_t test_vamoswapei16_v_i64m1_m(vbool64_t mask, int64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoswap.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoswap.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamoswapei16_v_i64m2_m(vbool32_t mask, int64_t *base,
@@ -1594,7 +1594,7 @@ vint64m2_t test_vamoswapei16_v_i64m2_m(vbool32_t mask, int64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoswap.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoswap.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamoswapei16_v_i64m4_m(vbool16_t mask, int64_t *base,
@@ -1607,7 +1607,7 @@ vint64m4_t test_vamoswapei16_v_i64m4_m(vbool16_t mask, int64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoswap.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoswap.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamoswapei16_v_i64m8_m(vbool8_t mask, int64_t *base,
@@ -1620,7 +1620,7 @@ vint64m8_t test_vamoswapei16_v_i64m8_m(vbool8_t mask, int64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoswap.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoswap.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamoswapei32_v_i64m1_m(vbool64_t mask, int64_t *base,
@@ -1633,7 +1633,7 @@ vint64m1_t test_vamoswapei32_v_i64m1_m(vbool64_t mask, int64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoswap.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoswap.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamoswapei32_v_i64m2_m(vbool32_t mask, int64_t *base,
@@ -1646,7 +1646,7 @@ vint64m2_t test_vamoswapei32_v_i64m2_m(vbool32_t mask, int64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoswap.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoswap.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamoswapei32_v_i64m4_m(vbool16_t mask, int64_t *base,
@@ -1659,7 +1659,7 @@ vint64m4_t test_vamoswapei32_v_i64m4_m(vbool16_t mask, int64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoswap.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoswap.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamoswapei32_v_i64m8_m(vbool8_t mask, int64_t *base,
@@ -1672,7 +1672,7 @@ vint64m8_t test_vamoswapei32_v_i64m8_m(vbool8_t mask, int64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoswap.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoswap.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamoswapei64_v_i64m1_m(vbool64_t mask, int64_t *base,
@@ -1685,7 +1685,7 @@ vint64m1_t test_vamoswapei64_v_i64m1_m(vbool64_t mask, int64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoswap.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoswap.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamoswapei64_v_i64m2_m(vbool32_t mask, int64_t *base,
@@ -1698,7 +1698,7 @@ vint64m2_t test_vamoswapei64_v_i64m2_m(vbool32_t mask, int64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoswap.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoswap.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamoswapei64_v_i64m4_m(vbool16_t mask, int64_t *base,
@@ -1711,7 +1711,7 @@ vint64m4_t test_vamoswapei64_v_i64m4_m(vbool16_t mask, int64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoswap.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoswap.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamoswapei64_v_i64m8_m(vbool8_t mask, int64_t *base,
@@ -1724,7 +1724,7 @@ vint64m8_t test_vamoswapei64_v_i64m8_m(vbool8_t mask, int64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoswap.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoswap.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamoswapei8_v_u32mf2_m(vbool64_t mask, uint32_t *base,
@@ -1737,7 +1737,7 @@ vuint32mf2_t test_vamoswapei8_v_u32mf2_m(vbool64_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoswap.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoswap.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamoswapei8_v_u32m1_m(vbool32_t mask, uint32_t *base,
@@ -1750,7 +1750,7 @@ vuint32m1_t test_vamoswapei8_v_u32m1_m(vbool32_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoswap.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoswap.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamoswapei8_v_u32m2_m(vbool16_t mask, uint32_t *base,
@@ -1763,7 +1763,7 @@ vuint32m2_t test_vamoswapei8_v_u32m2_m(vbool16_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoswap.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoswap.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamoswapei8_v_u32m4_m(vbool8_t mask, uint32_t *base,
@@ -1776,7 +1776,7 @@ vuint32m4_t test_vamoswapei8_v_u32m4_m(vbool8_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoswap.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoswap.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vamoswapei8_v_u32m8_m(vbool4_t mask, uint32_t *base,
@@ -1789,7 +1789,7 @@ vuint32m8_t test_vamoswapei8_v_u32m8_m(vbool4_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoswap.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoswap.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamoswapei16_v_u32mf2_m(vbool64_t mask, uint32_t *base,
@@ -1802,7 +1802,7 @@ vuint32mf2_t test_vamoswapei16_v_u32mf2_m(vbool64_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoswap.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoswap.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamoswapei16_v_u32m1_m(vbool32_t mask, uint32_t *base,
@@ -1815,7 +1815,7 @@ vuint32m1_t test_vamoswapei16_v_u32m1_m(vbool32_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoswap.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoswap.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamoswapei16_v_u32m2_m(vbool16_t mask, uint32_t *base,
@@ -1828,7 +1828,7 @@ vuint32m2_t test_vamoswapei16_v_u32m2_m(vbool16_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoswap.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoswap.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamoswapei16_v_u32m4_m(vbool8_t mask, uint32_t *base,
@@ -1841,7 +1841,7 @@ vuint32m4_t test_vamoswapei16_v_u32m4_m(vbool8_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoswap.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoswap.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vamoswapei16_v_u32m8_m(vbool4_t mask, uint32_t *base,
@@ -1854,7 +1854,7 @@ vuint32m8_t test_vamoswapei16_v_u32m8_m(vbool4_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoswap.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoswap.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamoswapei32_v_u32mf2_m(vbool64_t mask, uint32_t *base,
@@ -1867,7 +1867,7 @@ vuint32mf2_t test_vamoswapei32_v_u32mf2_m(vbool64_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoswap.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoswap.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamoswapei32_v_u32m1_m(vbool32_t mask, uint32_t *base,
@@ -1880,7 +1880,7 @@ vuint32m1_t test_vamoswapei32_v_u32m1_m(vbool32_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoswap.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoswap.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamoswapei32_v_u32m2_m(vbool16_t mask, uint32_t *base,
@@ -1893,7 +1893,7 @@ vuint32m2_t test_vamoswapei32_v_u32m2_m(vbool16_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoswap.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoswap.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamoswapei32_v_u32m4_m(vbool8_t mask, uint32_t *base,
@@ -1906,7 +1906,7 @@ vuint32m4_t test_vamoswapei32_v_u32m4_m(vbool8_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoswap.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoswap.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vamoswapei32_v_u32m8_m(vbool4_t mask, uint32_t *base,
@@ -1919,7 +1919,7 @@ vuint32m8_t test_vamoswapei32_v_u32m8_m(vbool4_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoswap.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoswap.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamoswapei64_v_u32mf2_m(vbool64_t mask, uint32_t *base,
@@ -1932,7 +1932,7 @@ vuint32mf2_t test_vamoswapei64_v_u32mf2_m(vbool64_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoswap.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoswap.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamoswapei64_v_u32m1_m(vbool32_t mask, uint32_t *base,
@@ -1945,7 +1945,7 @@ vuint32m1_t test_vamoswapei64_v_u32m1_m(vbool32_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoswap.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoswap.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamoswapei64_v_u32m2_m(vbool16_t mask, uint32_t *base,
@@ -1958,7 +1958,7 @@ vuint32m2_t test_vamoswapei64_v_u32m2_m(vbool16_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoswap.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoswap.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamoswapei64_v_u32m4_m(vbool8_t mask, uint32_t *base,
@@ -1971,7 +1971,7 @@ vuint32m4_t test_vamoswapei64_v_u32m4_m(vbool8_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoswap.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoswap.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamoswapei8_v_u64m1_m(vbool64_t mask, uint64_t *base,
@@ -1984,7 +1984,7 @@ vuint64m1_t test_vamoswapei8_v_u64m1_m(vbool64_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoswap.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoswap.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamoswapei8_v_u64m2_m(vbool32_t mask, uint64_t *base,
@@ -1997,7 +1997,7 @@ vuint64m2_t test_vamoswapei8_v_u64m2_m(vbool32_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoswap.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoswap.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamoswapei8_v_u64m4_m(vbool16_t mask, uint64_t *base,
@@ -2010,7 +2010,7 @@ vuint64m4_t test_vamoswapei8_v_u64m4_m(vbool16_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoswap.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoswap.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamoswapei8_v_u64m8_m(vbool8_t mask, uint64_t *base,
@@ -2023,7 +2023,7 @@ vuint64m8_t test_vamoswapei8_v_u64m8_m(vbool8_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoswap.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoswap.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamoswapei16_v_u64m1_m(vbool64_t mask, uint64_t *base,
@@ -2036,7 +2036,7 @@ vuint64m1_t test_vamoswapei16_v_u64m1_m(vbool64_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoswap.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoswap.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamoswapei16_v_u64m2_m(vbool32_t mask, uint64_t *base,
@@ -2049,7 +2049,7 @@ vuint64m2_t test_vamoswapei16_v_u64m2_m(vbool32_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoswap.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoswap.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamoswapei16_v_u64m4_m(vbool16_t mask, uint64_t *base,
@@ -2062,7 +2062,7 @@ vuint64m4_t test_vamoswapei16_v_u64m4_m(vbool16_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoswap.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoswap.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamoswapei16_v_u64m8_m(vbool8_t mask, uint64_t *base,
@@ -2075,7 +2075,7 @@ vuint64m8_t test_vamoswapei16_v_u64m8_m(vbool8_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoswap.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoswap.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamoswapei32_v_u64m1_m(vbool64_t mask, uint64_t *base,
@@ -2088,7 +2088,7 @@ vuint64m1_t test_vamoswapei32_v_u64m1_m(vbool64_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoswap.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoswap.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamoswapei32_v_u64m2_m(vbool32_t mask, uint64_t *base,
@@ -2101,7 +2101,7 @@ vuint64m2_t test_vamoswapei32_v_u64m2_m(vbool32_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoswap.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoswap.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamoswapei32_v_u64m4_m(vbool16_t mask, uint64_t *base,
@@ -2114,7 +2114,7 @@ vuint64m4_t test_vamoswapei32_v_u64m4_m(vbool16_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoswap.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoswap.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamoswapei32_v_u64m8_m(vbool8_t mask, uint64_t *base,
@@ -2127,7 +2127,7 @@ vuint64m8_t test_vamoswapei32_v_u64m8_m(vbool8_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoswap.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoswap.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamoswapei64_v_u64m1_m(vbool64_t mask, uint64_t *base,
@@ -2140,7 +2140,7 @@ vuint64m1_t test_vamoswapei64_v_u64m1_m(vbool64_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoswap.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoswap.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamoswapei64_v_u64m2_m(vbool32_t mask, uint64_t *base,
@@ -2153,7 +2153,7 @@ vuint64m2_t test_vamoswapei64_v_u64m2_m(vbool32_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoswap.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoswap.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamoswapei64_v_u64m4_m(vbool16_t mask, uint64_t *base,
@@ -2166,7 +2166,7 @@ vuint64m4_t test_vamoswapei64_v_u64m4_m(vbool16_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoswap.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoswap.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamoswapei64_v_u64m8_m(vbool8_t mask, uint64_t *base,
@@ -2179,7 +2179,7 @@ vuint64m8_t test_vamoswapei64_v_u64m8_m(vbool8_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vamoswap.mask.nxv1f32.nxv1i8.i64(<vscale x 1 x float>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vamoswap.mask.nxv1f32.nxv1i8.i64(<vscale x 1 x float>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP1]]
 //
 vfloat32mf2_t test_vamoswapei8_v_f32mf2_m(vbool64_t mask, float *base,
@@ -2192,7 +2192,7 @@ vfloat32mf2_t test_vamoswapei8_v_f32mf2_m(vbool64_t mask, float *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vamoswap.mask.nxv2f32.nxv2i8.i64(<vscale x 2 x float>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vamoswap.mask.nxv2f32.nxv2i8.i64(<vscale x 2 x float>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP1]]
 //
 vfloat32m1_t test_vamoswapei8_v_f32m1_m(vbool32_t mask, float *base,
@@ -2205,7 +2205,7 @@ vfloat32m1_t test_vamoswapei8_v_f32m1_m(vbool32_t mask, float *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vamoswap.mask.nxv4f32.nxv4i8.i64(<vscale x 4 x float>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vamoswap.mask.nxv4f32.nxv4i8.i64(<vscale x 4 x float>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 vfloat32m2_t test_vamoswapei8_v_f32m2_m(vbool16_t mask, float *base,
@@ -2218,7 +2218,7 @@ vfloat32m2_t test_vamoswapei8_v_f32m2_m(vbool16_t mask, float *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vamoswap.mask.nxv8f32.nxv8i8.i64(<vscale x 8 x float>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vamoswap.mask.nxv8f32.nxv8i8.i64(<vscale x 8 x float>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP1]]
 //
 vfloat32m4_t test_vamoswapei8_v_f32m4_m(vbool8_t mask, float *base,
@@ -2231,7 +2231,7 @@ vfloat32m4_t test_vamoswapei8_v_f32m4_m(vbool8_t mask, float *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vamoswap.mask.nxv16f32.nxv16i8.i64(<vscale x 16 x float>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vamoswap.mask.nxv16f32.nxv16i8.i64(<vscale x 16 x float>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP1]]
 //
 vfloat32m8_t test_vamoswapei8_v_f32m8_m(vbool4_t mask, float *base,
@@ -2244,7 +2244,7 @@ vfloat32m8_t test_vamoswapei8_v_f32m8_m(vbool4_t mask, float *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vamoswap.mask.nxv1f32.nxv1i16.i64(<vscale x 1 x float>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vamoswap.mask.nxv1f32.nxv1i16.i64(<vscale x 1 x float>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP1]]
 //
 vfloat32mf2_t test_vamoswapei16_v_f32mf2_m(vbool64_t mask, float *base,
@@ -2257,7 +2257,7 @@ vfloat32mf2_t test_vamoswapei16_v_f32mf2_m(vbool64_t mask, float *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vamoswap.mask.nxv2f32.nxv2i16.i64(<vscale x 2 x float>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vamoswap.mask.nxv2f32.nxv2i16.i64(<vscale x 2 x float>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP1]]
 //
 vfloat32m1_t test_vamoswapei16_v_f32m1_m(vbool32_t mask, float *base,
@@ -2270,7 +2270,7 @@ vfloat32m1_t test_vamoswapei16_v_f32m1_m(vbool32_t mask, float *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vamoswap.mask.nxv4f32.nxv4i16.i64(<vscale x 4 x float>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vamoswap.mask.nxv4f32.nxv4i16.i64(<vscale x 4 x float>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 vfloat32m2_t test_vamoswapei16_v_f32m2_m(vbool16_t mask, float *base,
@@ -2283,7 +2283,7 @@ vfloat32m2_t test_vamoswapei16_v_f32m2_m(vbool16_t mask, float *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vamoswap.mask.nxv8f32.nxv8i16.i64(<vscale x 8 x float>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vamoswap.mask.nxv8f32.nxv8i16.i64(<vscale x 8 x float>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP1]]
 //
 vfloat32m4_t test_vamoswapei16_v_f32m4_m(vbool8_t mask, float *base,
@@ -2296,7 +2296,7 @@ vfloat32m4_t test_vamoswapei16_v_f32m4_m(vbool8_t mask, float *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vamoswap.mask.nxv16f32.nxv16i16.i64(<vscale x 16 x float>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vamoswap.mask.nxv16f32.nxv16i16.i64(<vscale x 16 x float>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP1]]
 //
 vfloat32m8_t test_vamoswapei16_v_f32m8_m(vbool4_t mask, float *base,
@@ -2309,7 +2309,7 @@ vfloat32m8_t test_vamoswapei16_v_f32m8_m(vbool4_t mask, float *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vamoswap.mask.nxv1f32.nxv1i32.i64(<vscale x 1 x float>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vamoswap.mask.nxv1f32.nxv1i32.i64(<vscale x 1 x float>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP1]]
 //
 vfloat32mf2_t test_vamoswapei32_v_f32mf2_m(vbool64_t mask, float *base,
@@ -2322,7 +2322,7 @@ vfloat32mf2_t test_vamoswapei32_v_f32mf2_m(vbool64_t mask, float *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vamoswap.mask.nxv2f32.nxv2i32.i64(<vscale x 2 x float>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vamoswap.mask.nxv2f32.nxv2i32.i64(<vscale x 2 x float>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP1]]
 //
 vfloat32m1_t test_vamoswapei32_v_f32m1_m(vbool32_t mask, float *base,
@@ -2335,7 +2335,7 @@ vfloat32m1_t test_vamoswapei32_v_f32m1_m(vbool32_t mask, float *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vamoswap.mask.nxv4f32.nxv4i32.i64(<vscale x 4 x float>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vamoswap.mask.nxv4f32.nxv4i32.i64(<vscale x 4 x float>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 vfloat32m2_t test_vamoswapei32_v_f32m2_m(vbool16_t mask, float *base,
@@ -2348,7 +2348,7 @@ vfloat32m2_t test_vamoswapei32_v_f32m2_m(vbool16_t mask, float *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vamoswap.mask.nxv8f32.nxv8i32.i64(<vscale x 8 x float>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vamoswap.mask.nxv8f32.nxv8i32.i64(<vscale x 8 x float>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP1]]
 //
 vfloat32m4_t test_vamoswapei32_v_f32m4_m(vbool8_t mask, float *base,
@@ -2361,7 +2361,7 @@ vfloat32m4_t test_vamoswapei32_v_f32m4_m(vbool8_t mask, float *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vamoswap.mask.nxv16f32.nxv16i32.i64(<vscale x 16 x float>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vamoswap.mask.nxv16f32.nxv16i32.i64(<vscale x 16 x float>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP1]]
 //
 vfloat32m8_t test_vamoswapei32_v_f32m8_m(vbool4_t mask, float *base,
@@ -2374,7 +2374,7 @@ vfloat32m8_t test_vamoswapei32_v_f32m8_m(vbool4_t mask, float *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vamoswap.mask.nxv1f32.nxv1i64.i64(<vscale x 1 x float>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vamoswap.mask.nxv1f32.nxv1i64.i64(<vscale x 1 x float>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP1]]
 //
 vfloat32mf2_t test_vamoswapei64_v_f32mf2_m(vbool64_t mask, float *base,
@@ -2387,7 +2387,7 @@ vfloat32mf2_t test_vamoswapei64_v_f32mf2_m(vbool64_t mask, float *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vamoswap.mask.nxv2f32.nxv2i64.i64(<vscale x 2 x float>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vamoswap.mask.nxv2f32.nxv2i64.i64(<vscale x 2 x float>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP1]]
 //
 vfloat32m1_t test_vamoswapei64_v_f32m1_m(vbool32_t mask, float *base,
@@ -2400,7 +2400,7 @@ vfloat32m1_t test_vamoswapei64_v_f32m1_m(vbool32_t mask, float *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vamoswap.mask.nxv4f32.nxv4i64.i64(<vscale x 4 x float>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vamoswap.mask.nxv4f32.nxv4i64.i64(<vscale x 4 x float>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 vfloat32m2_t test_vamoswapei64_v_f32m2_m(vbool16_t mask, float *base,
@@ -2413,7 +2413,7 @@ vfloat32m2_t test_vamoswapei64_v_f32m2_m(vbool16_t mask, float *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vamoswap.mask.nxv8f32.nxv8i64.i64(<vscale x 8 x float>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vamoswap.mask.nxv8f32.nxv8i64.i64(<vscale x 8 x float>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP1]]
 //
 vfloat32m4_t test_vamoswapei64_v_f32m4_m(vbool8_t mask, float *base,
@@ -2426,7 +2426,7 @@ vfloat32m4_t test_vamoswapei64_v_f32m4_m(vbool8_t mask, float *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vamoswap.mask.nxv1f64.nxv1i8.i64(<vscale x 1 x double>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vamoswap.mask.nxv1f64.nxv1i8.i64(<vscale x 1 x double>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP1]]
 //
 vfloat64m1_t test_vamoswapei8_v_f64m1_m(vbool64_t mask, double *base,
@@ -2439,7 +2439,7 @@ vfloat64m1_t test_vamoswapei8_v_f64m1_m(vbool64_t mask, double *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vamoswap.mask.nxv2f64.nxv2i8.i64(<vscale x 2 x double>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vamoswap.mask.nxv2f64.nxv2i8.i64(<vscale x 2 x double>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 vfloat64m2_t test_vamoswapei8_v_f64m2_m(vbool32_t mask, double *base,
@@ -2452,7 +2452,7 @@ vfloat64m2_t test_vamoswapei8_v_f64m2_m(vbool32_t mask, double *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vamoswap.mask.nxv4f64.nxv4i8.i64(<vscale x 4 x double>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vamoswap.mask.nxv4f64.nxv4i8.i64(<vscale x 4 x double>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP1]]
 //
 vfloat64m4_t test_vamoswapei8_v_f64m4_m(vbool16_t mask, double *base,
@@ -2465,7 +2465,7 @@ vfloat64m4_t test_vamoswapei8_v_f64m4_m(vbool16_t mask, double *base,
 // CHECK-RV64-LABEL: @test_vamoswapei8_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vamoswap.mask.nxv8f64.nxv8i8.i64(<vscale x 8 x double>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vamoswap.mask.nxv8f64.nxv8i8.i64(<vscale x 8 x double>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP1]]
 //
 vfloat64m8_t test_vamoswapei8_v_f64m8_m(vbool8_t mask, double *base,
@@ -2478,7 +2478,7 @@ vfloat64m8_t test_vamoswapei8_v_f64m8_m(vbool8_t mask, double *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vamoswap.mask.nxv1f64.nxv1i16.i64(<vscale x 1 x double>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vamoswap.mask.nxv1f64.nxv1i16.i64(<vscale x 1 x double>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP1]]
 //
 vfloat64m1_t test_vamoswapei16_v_f64m1_m(vbool64_t mask, double *base,
@@ -2491,7 +2491,7 @@ vfloat64m1_t test_vamoswapei16_v_f64m1_m(vbool64_t mask, double *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vamoswap.mask.nxv2f64.nxv2i16.i64(<vscale x 2 x double>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vamoswap.mask.nxv2f64.nxv2i16.i64(<vscale x 2 x double>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 vfloat64m2_t test_vamoswapei16_v_f64m2_m(vbool32_t mask, double *base,
@@ -2504,7 +2504,7 @@ vfloat64m2_t test_vamoswapei16_v_f64m2_m(vbool32_t mask, double *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vamoswap.mask.nxv4f64.nxv4i16.i64(<vscale x 4 x double>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vamoswap.mask.nxv4f64.nxv4i16.i64(<vscale x 4 x double>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP1]]
 //
 vfloat64m4_t test_vamoswapei16_v_f64m4_m(vbool16_t mask, double *base,
@@ -2517,7 +2517,7 @@ vfloat64m4_t test_vamoswapei16_v_f64m4_m(vbool16_t mask, double *base,
 // CHECK-RV64-LABEL: @test_vamoswapei16_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vamoswap.mask.nxv8f64.nxv8i16.i64(<vscale x 8 x double>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vamoswap.mask.nxv8f64.nxv8i16.i64(<vscale x 8 x double>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP1]]
 //
 vfloat64m8_t test_vamoswapei16_v_f64m8_m(vbool8_t mask, double *base,
@@ -2530,7 +2530,7 @@ vfloat64m8_t test_vamoswapei16_v_f64m8_m(vbool8_t mask, double *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vamoswap.mask.nxv1f64.nxv1i32.i64(<vscale x 1 x double>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vamoswap.mask.nxv1f64.nxv1i32.i64(<vscale x 1 x double>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP1]]
 //
 vfloat64m1_t test_vamoswapei32_v_f64m1_m(vbool64_t mask, double *base,
@@ -2543,7 +2543,7 @@ vfloat64m1_t test_vamoswapei32_v_f64m1_m(vbool64_t mask, double *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vamoswap.mask.nxv2f64.nxv2i32.i64(<vscale x 2 x double>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vamoswap.mask.nxv2f64.nxv2i32.i64(<vscale x 2 x double>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 vfloat64m2_t test_vamoswapei32_v_f64m2_m(vbool32_t mask, double *base,
@@ -2556,7 +2556,7 @@ vfloat64m2_t test_vamoswapei32_v_f64m2_m(vbool32_t mask, double *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vamoswap.mask.nxv4f64.nxv4i32.i64(<vscale x 4 x double>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vamoswap.mask.nxv4f64.nxv4i32.i64(<vscale x 4 x double>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP1]]
 //
 vfloat64m4_t test_vamoswapei32_v_f64m4_m(vbool16_t mask, double *base,
@@ -2569,7 +2569,7 @@ vfloat64m4_t test_vamoswapei32_v_f64m4_m(vbool16_t mask, double *base,
 // CHECK-RV64-LABEL: @test_vamoswapei32_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vamoswap.mask.nxv8f64.nxv8i32.i64(<vscale x 8 x double>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vamoswap.mask.nxv8f64.nxv8i32.i64(<vscale x 8 x double>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP1]]
 //
 vfloat64m8_t test_vamoswapei32_v_f64m8_m(vbool8_t mask, double *base,
@@ -2582,7 +2582,7 @@ vfloat64m8_t test_vamoswapei32_v_f64m8_m(vbool8_t mask, double *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vamoswap.mask.nxv1f64.nxv1i64.i64(<vscale x 1 x double>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vamoswap.mask.nxv1f64.nxv1i64.i64(<vscale x 1 x double>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP1]]
 //
 vfloat64m1_t test_vamoswapei64_v_f64m1_m(vbool64_t mask, double *base,
@@ -2595,7 +2595,7 @@ vfloat64m1_t test_vamoswapei64_v_f64m1_m(vbool64_t mask, double *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vamoswap.mask.nxv2f64.nxv2i64.i64(<vscale x 2 x double>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vamoswap.mask.nxv2f64.nxv2i64.i64(<vscale x 2 x double>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 vfloat64m2_t test_vamoswapei64_v_f64m2_m(vbool32_t mask, double *base,
@@ -2608,7 +2608,7 @@ vfloat64m2_t test_vamoswapei64_v_f64m2_m(vbool32_t mask, double *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vamoswap.mask.nxv4f64.nxv4i64.i64(<vscale x 4 x double>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vamoswap.mask.nxv4f64.nxv4i64.i64(<vscale x 4 x double>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP1]]
 //
 vfloat64m4_t test_vamoswapei64_v_f64m4_m(vbool16_t mask, double *base,
@@ -2621,7 +2621,7 @@ vfloat64m4_t test_vamoswapei64_v_f64m4_m(vbool16_t mask, double *base,
 // CHECK-RV64-LABEL: @test_vamoswapei64_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vamoswap.mask.nxv8f64.nxv8i64.i64(<vscale x 8 x double>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vamoswap.mask.nxv8f64.nxv8i64.i64(<vscale x 8 x double>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP1]]
 //
 vfloat64m8_t test_vamoswapei64_v_f64m8_m(vbool8_t mask, double *base,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamoxor.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamoxor.c
index b510bf5..26e405e 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamoxor.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vamoxor.c
@@ -778,7 +778,7 @@ vuint64m8_t test_vamoxorei64_v_u64m8 (uint64_t *base, vuint64m8_t bindex, vuint6
 // CHECK-RV64-LABEL: @test_vamoxorei8_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoxor.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoxor.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamoxorei8_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint8mf8_t bindex, vint32mf2_t value, size_t vl) {
@@ -789,7 +789,7 @@ vint32mf2_t test_vamoxorei8_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint8mf8
 // CHECK-RV64-LABEL: @test_vamoxorei8_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoxor.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoxor.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamoxorei8_v_i32m1_m (vbool32_t mask, int32_t *base, vuint8mf4_t bindex, vint32m1_t value, size_t vl) {
@@ -800,7 +800,7 @@ vint32m1_t test_vamoxorei8_v_i32m1_m (vbool32_t mask, int32_t *base, vuint8mf4_t
 // CHECK-RV64-LABEL: @test_vamoxorei8_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoxor.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoxor.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamoxorei8_v_i32m2_m (vbool16_t mask, int32_t *base, vuint8mf2_t bindex, vint32m2_t value, size_t vl) {
@@ -811,7 +811,7 @@ vint32m2_t test_vamoxorei8_v_i32m2_m (vbool16_t mask, int32_t *base, vuint8mf2_t
 // CHECK-RV64-LABEL: @test_vamoxorei8_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoxor.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoxor.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamoxorei8_v_i32m4_m (vbool8_t mask, int32_t *base, vuint8m1_t bindex, vint32m4_t value, size_t vl) {
@@ -822,7 +822,7 @@ vint32m4_t test_vamoxorei8_v_i32m4_m (vbool8_t mask, int32_t *base, vuint8m1_t b
 // CHECK-RV64-LABEL: @test_vamoxorei8_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoxor.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoxor.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vamoxorei8_v_i32m8_m (vbool4_t mask, int32_t *base, vuint8m2_t bindex, vint32m8_t value, size_t vl) {
@@ -833,7 +833,7 @@ vint32m8_t test_vamoxorei8_v_i32m8_m (vbool4_t mask, int32_t *base, vuint8m2_t b
 // CHECK-RV64-LABEL: @test_vamoxorei16_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoxor.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoxor.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamoxorei16_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint16mf4_t bindex, vint32mf2_t value, size_t vl) {
@@ -844,7 +844,7 @@ vint32mf2_t test_vamoxorei16_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamoxorei16_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoxor.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoxor.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamoxorei16_v_i32m1_m (vbool32_t mask, int32_t *base, vuint16mf2_t bindex, vint32m1_t value, size_t vl) {
@@ -855,7 +855,7 @@ vint32m1_t test_vamoxorei16_v_i32m1_m (vbool32_t mask, int32_t *base, vuint16mf2
 // CHECK-RV64-LABEL: @test_vamoxorei16_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoxor.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoxor.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamoxorei16_v_i32m2_m (vbool16_t mask, int32_t *base, vuint16m1_t bindex, vint32m2_t value, size_t vl) {
@@ -866,7 +866,7 @@ vint32m2_t test_vamoxorei16_v_i32m2_m (vbool16_t mask, int32_t *base, vuint16m1_
 // CHECK-RV64-LABEL: @test_vamoxorei16_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoxor.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoxor.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamoxorei16_v_i32m4_m (vbool8_t mask, int32_t *base, vuint16m2_t bindex, vint32m4_t value, size_t vl) {
@@ -877,7 +877,7 @@ vint32m4_t test_vamoxorei16_v_i32m4_m (vbool8_t mask, int32_t *base, vuint16m2_t
 // CHECK-RV64-LABEL: @test_vamoxorei16_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoxor.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoxor.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vamoxorei16_v_i32m8_m (vbool4_t mask, int32_t *base, vuint16m4_t bindex, vint32m8_t value, size_t vl) {
@@ -888,7 +888,7 @@ vint32m8_t test_vamoxorei16_v_i32m8_m (vbool4_t mask, int32_t *base, vuint16m4_t
 // CHECK-RV64-LABEL: @test_vamoxorei32_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoxor.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoxor.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamoxorei32_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint32mf2_t bindex, vint32mf2_t value, size_t vl) {
@@ -899,7 +899,7 @@ vint32mf2_t test_vamoxorei32_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamoxorei32_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoxor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoxor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamoxorei32_v_i32m1_m (vbool32_t mask, int32_t *base, vuint32m1_t bindex, vint32m1_t value, size_t vl) {
@@ -910,7 +910,7 @@ vint32m1_t test_vamoxorei32_v_i32m1_m (vbool32_t mask, int32_t *base, vuint32m1_
 // CHECK-RV64-LABEL: @test_vamoxorei32_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoxor.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoxor.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamoxorei32_v_i32m2_m (vbool16_t mask, int32_t *base, vuint32m2_t bindex, vint32m2_t value, size_t vl) {
@@ -921,7 +921,7 @@ vint32m2_t test_vamoxorei32_v_i32m2_m (vbool16_t mask, int32_t *base, vuint32m2_
 // CHECK-RV64-LABEL: @test_vamoxorei32_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoxor.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoxor.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamoxorei32_v_i32m4_m (vbool8_t mask, int32_t *base, vuint32m4_t bindex, vint32m4_t value, size_t vl) {
@@ -932,7 +932,7 @@ vint32m4_t test_vamoxorei32_v_i32m4_m (vbool8_t mask, int32_t *base, vuint32m4_t
 // CHECK-RV64-LABEL: @test_vamoxorei32_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoxor.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoxor.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vamoxorei32_v_i32m8_m (vbool4_t mask, int32_t *base, vuint32m8_t bindex, vint32m8_t value, size_t vl) {
@@ -943,7 +943,7 @@ vint32m8_t test_vamoxorei32_v_i32m8_m (vbool4_t mask, int32_t *base, vuint32m8_t
 // CHECK-RV64-LABEL: @test_vamoxorei64_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoxor.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoxor.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vamoxorei64_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint64m1_t bindex, vint32mf2_t value, size_t vl) {
@@ -954,7 +954,7 @@ vint32mf2_t test_vamoxorei64_v_i32mf2_m (vbool64_t mask, int32_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamoxorei64_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoxor.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoxor.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vamoxorei64_v_i32m1_m (vbool32_t mask, int32_t *base, vuint64m2_t bindex, vint32m1_t value, size_t vl) {
@@ -965,7 +965,7 @@ vint32m1_t test_vamoxorei64_v_i32m1_m (vbool32_t mask, int32_t *base, vuint64m2_
 // CHECK-RV64-LABEL: @test_vamoxorei64_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoxor.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoxor.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vamoxorei64_v_i32m2_m (vbool16_t mask, int32_t *base, vuint64m4_t bindex, vint32m2_t value, size_t vl) {
@@ -976,7 +976,7 @@ vint32m2_t test_vamoxorei64_v_i32m2_m (vbool16_t mask, int32_t *base, vuint64m4_
 // CHECK-RV64-LABEL: @test_vamoxorei64_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoxor.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoxor.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vamoxorei64_v_i32m4_m (vbool8_t mask, int32_t *base, vuint64m8_t bindex, vint32m4_t value, size_t vl) {
@@ -987,7 +987,7 @@ vint32m4_t test_vamoxorei64_v_i32m4_m (vbool8_t mask, int32_t *base, vuint64m8_t
 // CHECK-RV64-LABEL: @test_vamoxorei8_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoxor.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoxor.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamoxorei8_v_i64m1_m (vbool64_t mask, int64_t *base, vuint8mf8_t bindex, vint64m1_t value, size_t vl) {
@@ -998,7 +998,7 @@ vint64m1_t test_vamoxorei8_v_i64m1_m (vbool64_t mask, int64_t *base, vuint8mf8_t
 // CHECK-RV64-LABEL: @test_vamoxorei8_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoxor.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoxor.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamoxorei8_v_i64m2_m (vbool32_t mask, int64_t *base, vuint8mf4_t bindex, vint64m2_t value, size_t vl) {
@@ -1009,7 +1009,7 @@ vint64m2_t test_vamoxorei8_v_i64m2_m (vbool32_t mask, int64_t *base, vuint8mf4_t
 // CHECK-RV64-LABEL: @test_vamoxorei8_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoxor.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoxor.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamoxorei8_v_i64m4_m (vbool16_t mask, int64_t *base, vuint8mf2_t bindex, vint64m4_t value, size_t vl) {
@@ -1020,7 +1020,7 @@ vint64m4_t test_vamoxorei8_v_i64m4_m (vbool16_t mask, int64_t *base, vuint8mf2_t
 // CHECK-RV64-LABEL: @test_vamoxorei8_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoxor.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoxor.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamoxorei8_v_i64m8_m (vbool8_t mask, int64_t *base, vuint8m1_t bindex, vint64m8_t value, size_t vl) {
@@ -1031,7 +1031,7 @@ vint64m8_t test_vamoxorei8_v_i64m8_m (vbool8_t mask, int64_t *base, vuint8m1_t b
 // CHECK-RV64-LABEL: @test_vamoxorei16_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoxor.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoxor.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamoxorei16_v_i64m1_m (vbool64_t mask, int64_t *base, vuint16mf4_t bindex, vint64m1_t value, size_t vl) {
@@ -1042,7 +1042,7 @@ vint64m1_t test_vamoxorei16_v_i64m1_m (vbool64_t mask, int64_t *base, vuint16mf4
 // CHECK-RV64-LABEL: @test_vamoxorei16_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoxor.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoxor.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamoxorei16_v_i64m2_m (vbool32_t mask, int64_t *base, vuint16mf2_t bindex, vint64m2_t value, size_t vl) {
@@ -1053,7 +1053,7 @@ vint64m2_t test_vamoxorei16_v_i64m2_m (vbool32_t mask, int64_t *base, vuint16mf2
 // CHECK-RV64-LABEL: @test_vamoxorei16_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoxor.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoxor.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamoxorei16_v_i64m4_m (vbool16_t mask, int64_t *base, vuint16m1_t bindex, vint64m4_t value, size_t vl) {
@@ -1064,7 +1064,7 @@ vint64m4_t test_vamoxorei16_v_i64m4_m (vbool16_t mask, int64_t *base, vuint16m1_
 // CHECK-RV64-LABEL: @test_vamoxorei16_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoxor.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoxor.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamoxorei16_v_i64m8_m (vbool8_t mask, int64_t *base, vuint16m2_t bindex, vint64m8_t value, size_t vl) {
@@ -1075,7 +1075,7 @@ vint64m8_t test_vamoxorei16_v_i64m8_m (vbool8_t mask, int64_t *base, vuint16m2_t
 // CHECK-RV64-LABEL: @test_vamoxorei32_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoxor.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoxor.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamoxorei32_v_i64m1_m (vbool64_t mask, int64_t *base, vuint32mf2_t bindex, vint64m1_t value, size_t vl) {
@@ -1086,7 +1086,7 @@ vint64m1_t test_vamoxorei32_v_i64m1_m (vbool64_t mask, int64_t *base, vuint32mf2
 // CHECK-RV64-LABEL: @test_vamoxorei32_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoxor.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoxor.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamoxorei32_v_i64m2_m (vbool32_t mask, int64_t *base, vuint32m1_t bindex, vint64m2_t value, size_t vl) {
@@ -1097,7 +1097,7 @@ vint64m2_t test_vamoxorei32_v_i64m2_m (vbool32_t mask, int64_t *base, vuint32m1_
 // CHECK-RV64-LABEL: @test_vamoxorei32_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoxor.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoxor.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamoxorei32_v_i64m4_m (vbool16_t mask, int64_t *base, vuint32m2_t bindex, vint64m4_t value, size_t vl) {
@@ -1108,7 +1108,7 @@ vint64m4_t test_vamoxorei32_v_i64m4_m (vbool16_t mask, int64_t *base, vuint32m2_
 // CHECK-RV64-LABEL: @test_vamoxorei32_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoxor.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoxor.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamoxorei32_v_i64m8_m (vbool8_t mask, int64_t *base, vuint32m4_t bindex, vint64m8_t value, size_t vl) {
@@ -1119,7 +1119,7 @@ vint64m8_t test_vamoxorei32_v_i64m8_m (vbool8_t mask, int64_t *base, vuint32m4_t
 // CHECK-RV64-LABEL: @test_vamoxorei64_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoxor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoxor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vamoxorei64_v_i64m1_m (vbool64_t mask, int64_t *base, vuint64m1_t bindex, vint64m1_t value, size_t vl) {
@@ -1130,7 +1130,7 @@ vint64m1_t test_vamoxorei64_v_i64m1_m (vbool64_t mask, int64_t *base, vuint64m1_
 // CHECK-RV64-LABEL: @test_vamoxorei64_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoxor.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoxor.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vamoxorei64_v_i64m2_m (vbool32_t mask, int64_t *base, vuint64m2_t bindex, vint64m2_t value, size_t vl) {
@@ -1141,7 +1141,7 @@ vint64m2_t test_vamoxorei64_v_i64m2_m (vbool32_t mask, int64_t *base, vuint64m2_
 // CHECK-RV64-LABEL: @test_vamoxorei64_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoxor.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoxor.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vamoxorei64_v_i64m4_m (vbool16_t mask, int64_t *base, vuint64m4_t bindex, vint64m4_t value, size_t vl) {
@@ -1152,7 +1152,7 @@ vint64m4_t test_vamoxorei64_v_i64m4_m (vbool16_t mask, int64_t *base, vuint64m4_
 // CHECK-RV64-LABEL: @test_vamoxorei64_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoxor.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoxor.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vamoxorei64_v_i64m8_m (vbool8_t mask, int64_t *base, vuint64m8_t bindex, vint64m8_t value, size_t vl) {
@@ -1163,7 +1163,7 @@ vint64m8_t test_vamoxorei64_v_i64m8_m (vbool8_t mask, int64_t *base, vuint64m8_t
 // CHECK-RV64-LABEL: @test_vamoxorei8_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoxor.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoxor.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamoxorei8_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint8mf8_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1174,7 +1174,7 @@ vuint32mf2_t test_vamoxorei8_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint8m
 // CHECK-RV64-LABEL: @test_vamoxorei8_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoxor.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoxor.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamoxorei8_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint8mf4_t bindex, vuint32m1_t value, size_t vl) {
@@ -1185,7 +1185,7 @@ vuint32m1_t test_vamoxorei8_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint8mf4
 // CHECK-RV64-LABEL: @test_vamoxorei8_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoxor.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoxor.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamoxorei8_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint8mf2_t bindex, vuint32m2_t value, size_t vl) {
@@ -1196,7 +1196,7 @@ vuint32m2_t test_vamoxorei8_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint8mf2
 // CHECK-RV64-LABEL: @test_vamoxorei8_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoxor.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoxor.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamoxorei8_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint8m1_t bindex, vuint32m4_t value, size_t vl) {
@@ -1207,7 +1207,7 @@ vuint32m4_t test_vamoxorei8_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint8m1_t
 // CHECK-RV64-LABEL: @test_vamoxorei8_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoxor.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoxor.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vamoxorei8_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint8m2_t bindex, vuint32m8_t value, size_t vl) {
@@ -1218,7 +1218,7 @@ vuint32m8_t test_vamoxorei8_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint8m2_t
 // CHECK-RV64-LABEL: @test_vamoxorei16_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoxor.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoxor.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamoxorei16_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint16mf4_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1229,7 +1229,7 @@ vuint32mf2_t test_vamoxorei16_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint1
 // CHECK-RV64-LABEL: @test_vamoxorei16_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoxor.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoxor.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamoxorei16_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint16mf2_t bindex, vuint32m1_t value, size_t vl) {
@@ -1240,7 +1240,7 @@ vuint32m1_t test_vamoxorei16_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamoxorei16_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoxor.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoxor.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamoxorei16_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint16m1_t bindex, vuint32m2_t value, size_t vl) {
@@ -1251,7 +1251,7 @@ vuint32m2_t test_vamoxorei16_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamoxorei16_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoxor.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoxor.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamoxorei16_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint16m2_t bindex, vuint32m4_t value, size_t vl) {
@@ -1262,7 +1262,7 @@ vuint32m4_t test_vamoxorei16_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint16m2
 // CHECK-RV64-LABEL: @test_vamoxorei16_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoxor.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoxor.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vamoxorei16_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint16m4_t bindex, vuint32m8_t value, size_t vl) {
@@ -1273,7 +1273,7 @@ vuint32m8_t test_vamoxorei16_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint16m4
 // CHECK-RV64-LABEL: @test_vamoxorei32_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoxor.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoxor.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamoxorei32_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint32mf2_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1284,7 +1284,7 @@ vuint32mf2_t test_vamoxorei32_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint3
 // CHECK-RV64-LABEL: @test_vamoxorei32_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoxor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoxor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamoxorei32_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint32m1_t bindex, vuint32m1_t value, size_t vl) {
@@ -1295,7 +1295,7 @@ vuint32m1_t test_vamoxorei32_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamoxorei32_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoxor.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoxor.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamoxorei32_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint32m2_t bindex, vuint32m2_t value, size_t vl) {
@@ -1306,7 +1306,7 @@ vuint32m2_t test_vamoxorei32_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamoxorei32_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoxor.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoxor.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamoxorei32_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint32m4_t bindex, vuint32m4_t value, size_t vl) {
@@ -1317,7 +1317,7 @@ vuint32m4_t test_vamoxorei32_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint32m4
 // CHECK-RV64-LABEL: @test_vamoxorei32_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoxor.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vamoxor.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vamoxorei32_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint32m8_t bindex, vuint32m8_t value, size_t vl) {
@@ -1328,7 +1328,7 @@ vuint32m8_t test_vamoxorei32_v_u32m8_m (vbool4_t mask, uint32_t *base, vuint32m8
 // CHECK-RV64-LABEL: @test_vamoxorei64_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoxor.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vamoxor.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vamoxorei64_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint64m1_t bindex, vuint32mf2_t value, size_t vl) {
@@ -1339,7 +1339,7 @@ vuint32mf2_t test_vamoxorei64_v_u32mf2_m (vbool64_t mask, uint32_t *base, vuint6
 // CHECK-RV64-LABEL: @test_vamoxorei64_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoxor.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vamoxor.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vamoxorei64_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint64m2_t bindex, vuint32m1_t value, size_t vl) {
@@ -1350,7 +1350,7 @@ vuint32m1_t test_vamoxorei64_v_u32m1_m (vbool32_t mask, uint32_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamoxorei64_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoxor.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vamoxor.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vamoxorei64_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint64m4_t bindex, vuint32m2_t value, size_t vl) {
@@ -1361,7 +1361,7 @@ vuint32m2_t test_vamoxorei64_v_u32m2_m (vbool16_t mask, uint32_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamoxorei64_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoxor.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vamoxor.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vamoxorei64_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint64m8_t bindex, vuint32m4_t value, size_t vl) {
@@ -1372,7 +1372,7 @@ vuint32m4_t test_vamoxorei64_v_u32m4_m (vbool8_t mask, uint32_t *base, vuint64m8
 // CHECK-RV64-LABEL: @test_vamoxorei8_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoxor.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoxor.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamoxorei8_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint8mf8_t bindex, vuint64m1_t value, size_t vl) {
@@ -1383,7 +1383,7 @@ vuint64m1_t test_vamoxorei8_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint8mf8
 // CHECK-RV64-LABEL: @test_vamoxorei8_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoxor.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoxor.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamoxorei8_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint8mf4_t bindex, vuint64m2_t value, size_t vl) {
@@ -1394,7 +1394,7 @@ vuint64m2_t test_vamoxorei8_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint8mf4
 // CHECK-RV64-LABEL: @test_vamoxorei8_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoxor.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoxor.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamoxorei8_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint8mf2_t bindex, vuint64m4_t value, size_t vl) {
@@ -1405,7 +1405,7 @@ vuint64m4_t test_vamoxorei8_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint8mf2
 // CHECK-RV64-LABEL: @test_vamoxorei8_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoxor.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoxor.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamoxorei8_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint8m1_t bindex, vuint64m8_t value, size_t vl) {
@@ -1416,7 +1416,7 @@ vuint64m8_t test_vamoxorei8_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint8m1_t
 // CHECK-RV64-LABEL: @test_vamoxorei16_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoxor.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoxor.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamoxorei16_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint16mf4_t bindex, vuint64m1_t value, size_t vl) {
@@ -1427,7 +1427,7 @@ vuint64m1_t test_vamoxorei16_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamoxorei16_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoxor.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoxor.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamoxorei16_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint16mf2_t bindex, vuint64m2_t value, size_t vl) {
@@ -1438,7 +1438,7 @@ vuint64m2_t test_vamoxorei16_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamoxorei16_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoxor.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoxor.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamoxorei16_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint16m1_t bindex, vuint64m4_t value, size_t vl) {
@@ -1449,7 +1449,7 @@ vuint64m4_t test_vamoxorei16_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint16m
 // CHECK-RV64-LABEL: @test_vamoxorei16_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoxor.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoxor.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamoxorei16_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint16m2_t bindex, vuint64m8_t value, size_t vl) {
@@ -1460,7 +1460,7 @@ vuint64m8_t test_vamoxorei16_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint16m2
 // CHECK-RV64-LABEL: @test_vamoxorei32_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoxor.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoxor.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamoxorei32_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint32mf2_t bindex, vuint64m1_t value, size_t vl) {
@@ -1471,7 +1471,7 @@ vuint64m1_t test_vamoxorei32_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamoxorei32_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoxor.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoxor.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamoxorei32_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint32m1_t bindex, vuint64m2_t value, size_t vl) {
@@ -1482,7 +1482,7 @@ vuint64m2_t test_vamoxorei32_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamoxorei32_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoxor.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoxor.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamoxorei32_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint32m2_t bindex, vuint64m4_t value, size_t vl) {
@@ -1493,7 +1493,7 @@ vuint64m4_t test_vamoxorei32_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint32m
 // CHECK-RV64-LABEL: @test_vamoxorei32_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoxor.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoxor.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamoxorei32_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint32m4_t bindex, vuint64m8_t value, size_t vl) {
@@ -1504,7 +1504,7 @@ vuint64m8_t test_vamoxorei32_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint32m4
 // CHECK-RV64-LABEL: @test_vamoxorei64_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoxor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vamoxor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vamoxorei64_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint64m1_t bindex, vuint64m1_t value, size_t vl) {
@@ -1515,7 +1515,7 @@ vuint64m1_t test_vamoxorei64_v_u64m1_m (vbool64_t mask, uint64_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamoxorei64_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoxor.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vamoxor.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vamoxorei64_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint64m2_t bindex, vuint64m2_t value, size_t vl) {
@@ -1526,7 +1526,7 @@ vuint64m2_t test_vamoxorei64_v_u64m2_m (vbool32_t mask, uint64_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamoxorei64_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoxor.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vamoxor.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vamoxorei64_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint64m4_t bindex, vuint64m4_t value, size_t vl) {
@@ -1537,7 +1537,7 @@ vuint64m4_t test_vamoxorei64_v_u64m4_m (vbool16_t mask, uint64_t *base, vuint64m
 // CHECK-RV64-LABEL: @test_vamoxorei64_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoxor.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vamoxor.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vamoxorei64_v_u64m8_m (vbool8_t mask, uint64_t *base, vuint64m8_t bindex, vuint64m8_t value, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vand.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vand.c
index c943e90..c67e269 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vand.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vand.c
@@ -887,7 +887,7 @@ vuint64m8_t test_vand_vx_u64m8(vuint64m8_t op1, uint64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vand.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vand.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vand_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, vint8mf8_t op2, size_t vl) {
@@ -897,7 +897,7 @@ vint8mf8_t test_vand_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vand.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vand.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vand_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, int8_t op2, size_t vl) {
@@ -907,7 +907,7 @@ vint8mf8_t test_vand_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vand.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vand.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vand_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, vint8mf4_t op2, size_t vl) {
@@ -917,7 +917,7 @@ vint8mf4_t test_vand_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vand.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vand.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vand_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, int8_t op2, size_t vl) {
@@ -927,7 +927,7 @@ vint8mf4_t test_vand_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vand.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vand.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vand_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, vint8mf2_t op2, size_t vl) {
@@ -937,7 +937,7 @@ vint8mf2_t test_vand_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vand.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vand.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vand_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, int8_t op2, size_t vl) {
@@ -947,7 +947,7 @@ vint8mf2_t test_vand_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vand.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vand.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vand_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, vint8m1_t op2, size_t vl) {
@@ -957,7 +957,7 @@ vint8m1_t test_vand_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vand.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vand.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vand_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, int8_t op2, size_t vl) {
@@ -967,7 +967,7 @@ vint8m1_t test_vand_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vand.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vand.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vand_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, vint8m2_t op2, size_t vl) {
@@ -977,7 +977,7 @@ vint8m2_t test_vand_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vand.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vand.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vand_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, int8_t op2, size_t vl) {
@@ -987,7 +987,7 @@ vint8m2_t test_vand_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vand.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vand.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vand_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, vint8m4_t op2, size_t vl) {
@@ -997,7 +997,7 @@ vint8m4_t test_vand_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vand.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vand.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vand_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, int8_t op2, size_t vl) {
@@ -1007,7 +1007,7 @@ vint8m4_t test_vand_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vand.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vand.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vand_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, vint8m8_t op2, size_t vl) {
@@ -1017,7 +1017,7 @@ vint8m8_t test_vand_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vand.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vand.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vand_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, int8_t op2, size_t vl) {
@@ -1027,7 +1027,7 @@ vint8m8_t test_vand_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vand.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vand.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vand_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, vint16mf4_t op2, size_t vl) {
@@ -1037,7 +1037,7 @@ vint16mf4_t test_vand_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vand.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vand.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vand_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, int16_t op2, size_t vl) {
@@ -1047,7 +1047,7 @@ vint16mf4_t test_vand_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vand.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vand.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vand_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, vint16mf2_t op2, size_t vl) {
@@ -1057,7 +1057,7 @@ vint16mf2_t test_vand_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vand.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vand.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vand_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, int16_t op2, size_t vl) {
@@ -1067,7 +1067,7 @@ vint16mf2_t test_vand_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vand.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vand.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vand_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, vint16m1_t op2, size_t vl) {
@@ -1077,7 +1077,7 @@ vint16m1_t test_vand_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vand.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vand.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vand_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, int16_t op2, size_t vl) {
@@ -1087,7 +1087,7 @@ vint16m1_t test_vand_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vand.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vand.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vand_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, vint16m2_t op2, size_t vl) {
@@ -1097,7 +1097,7 @@ vint16m2_t test_vand_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vand.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vand.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vand_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, int16_t op2, size_t vl) {
@@ -1107,7 +1107,7 @@ vint16m2_t test_vand_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vand.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vand.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vand_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, vint16m4_t op2, size_t vl) {
@@ -1117,7 +1117,7 @@ vint16m4_t test_vand_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vand.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vand.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vand_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, int16_t op2, size_t vl) {
@@ -1127,7 +1127,7 @@ vint16m4_t test_vand_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vand.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vand.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vand_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, vint16m8_t op2, size_t vl) {
@@ -1137,7 +1137,7 @@ vint16m8_t test_vand_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vand.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vand.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vand_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, int16_t op2, size_t vl) {
@@ -1147,7 +1147,7 @@ vint16m8_t test_vand_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vand.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vand.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vand_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, vint32mf2_t op2, size_t vl) {
@@ -1157,7 +1157,7 @@ vint32mf2_t test_vand_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vand.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vand.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vand_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, int32_t op2, size_t vl) {
@@ -1167,7 +1167,7 @@ vint32mf2_t test_vand_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vand.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vand.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vand_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, vint32m1_t op2, size_t vl) {
@@ -1177,7 +1177,7 @@ vint32m1_t test_vand_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vand.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vand.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vand_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, int32_t op2, size_t vl) {
@@ -1187,7 +1187,7 @@ vint32m1_t test_vand_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vand.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vand.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vand_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, vint32m2_t op2, size_t vl) {
@@ -1197,7 +1197,7 @@ vint32m2_t test_vand_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vand.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vand.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vand_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, int32_t op2, size_t vl) {
@@ -1207,7 +1207,7 @@ vint32m2_t test_vand_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vand.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vand.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vand_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, vint32m4_t op2, size_t vl) {
@@ -1217,7 +1217,7 @@ vint32m4_t test_vand_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vand.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vand.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vand_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, int32_t op2, size_t vl) {
@@ -1227,7 +1227,7 @@ vint32m4_t test_vand_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vand.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vand.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vand_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, vint32m8_t op2, size_t vl) {
@@ -1237,7 +1237,7 @@ vint32m8_t test_vand_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vand.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vand.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vand_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, int32_t op2, size_t vl) {
@@ -1247,7 +1247,7 @@ vint32m8_t test_vand_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vand.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vand.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vand_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, vint64m1_t op2, size_t vl) {
@@ -1257,7 +1257,7 @@ vint64m1_t test_vand_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vand.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vand.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vand_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, int64_t op2, size_t vl) {
@@ -1267,7 +1267,7 @@ vint64m1_t test_vand_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vand.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vand.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vand_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, vint64m2_t op2, size_t vl) {
@@ -1277,7 +1277,7 @@ vint64m2_t test_vand_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vand.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vand.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vand_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, int64_t op2, size_t vl) {
@@ -1287,7 +1287,7 @@ vint64m2_t test_vand_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vand.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vand.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vand_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, vint64m4_t op2, size_t vl) {
@@ -1297,7 +1297,7 @@ vint64m4_t test_vand_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vand.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vand.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vand_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, int64_t op2, size_t vl) {
@@ -1307,7 +1307,7 @@ vint64m4_t test_vand_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vand_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vand.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vand.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vand_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, vint64m8_t op2, size_t vl) {
@@ -1317,7 +1317,7 @@ vint64m8_t test_vand_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vand_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vand.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vand.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vand_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, int64_t op2, size_t vl) {
@@ -1327,7 +1327,7 @@ vint64m8_t test_vand_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vand.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vand.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vand_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -1337,7 +1337,7 @@ vuint8mf8_t test_vand_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vand.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vand.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vand_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, uint8_t op2, size_t vl) {
@@ -1347,7 +1347,7 @@ vuint8mf8_t test_vand_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vand.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vand.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vand_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, vuint8mf4_t op2, size_t vl) {
@@ -1357,7 +1357,7 @@ vuint8mf4_t test_vand_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vand.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vand.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vand_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, uint8_t op2, size_t vl) {
@@ -1367,7 +1367,7 @@ vuint8mf4_t test_vand_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vand.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vand.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vand_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, vuint8mf2_t op2, size_t vl) {
@@ -1377,7 +1377,7 @@ vuint8mf2_t test_vand_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vand.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vand.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vand_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, uint8_t op2, size_t vl) {
@@ -1387,7 +1387,7 @@ vuint8mf2_t test_vand_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vand.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vand.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vand_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, vuint8m1_t op2, size_t vl) {
@@ -1397,7 +1397,7 @@ vuint8m1_t test_vand_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t o
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vand.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vand.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vand_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, uint8_t op2, size_t vl) {
@@ -1407,7 +1407,7 @@ vuint8m1_t test_vand_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t o
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vand.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vand.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vand_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, vuint8m2_t op2, size_t vl) {
@@ -1417,7 +1417,7 @@ vuint8m2_t test_vand_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t o
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vand.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vand.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vand_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, uint8_t op2, size_t vl) {
@@ -1427,7 +1427,7 @@ vuint8m2_t test_vand_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t o
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vand.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vand.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vand_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, vuint8m4_t op2, size_t vl) {
@@ -1437,7 +1437,7 @@ vuint8m4_t test_vand_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t o
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vand.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vand.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vand_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, uint8_t op2, size_t vl) {
@@ -1447,7 +1447,7 @@ vuint8m4_t test_vand_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t o
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vand.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vand.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vand_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, vuint8m8_t op2, size_t vl) {
@@ -1457,7 +1457,7 @@ vuint8m8_t test_vand_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t o
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vand.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vand.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vand_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, uint8_t op2, size_t vl) {
@@ -1467,7 +1467,7 @@ vuint8m8_t test_vand_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t o
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vand.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vand.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vand_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, vuint16mf4_t op2, size_t vl) {
@@ -1477,7 +1477,7 @@ vuint16mf4_t test_vand_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vand.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vand.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vand_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, uint16_t op2, size_t vl) {
@@ -1487,7 +1487,7 @@ vuint16mf4_t test_vand_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vand.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vand.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vand_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, vuint16mf2_t op2, size_t vl) {
@@ -1497,7 +1497,7 @@ vuint16mf2_t test_vand_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vand.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vand.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vand_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, uint16_t op2, size_t vl) {
@@ -1507,7 +1507,7 @@ vuint16mf2_t test_vand_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vand.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vand.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vand_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, vuint16m1_t op2, size_t vl) {
@@ -1517,7 +1517,7 @@ vuint16m1_t test_vand_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vand.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vand.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vand_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, uint16_t op2, size_t vl) {
@@ -1527,7 +1527,7 @@ vuint16m1_t test_vand_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vand.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vand.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vand_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, vuint16m2_t op2, size_t vl) {
@@ -1537,7 +1537,7 @@ vuint16m2_t test_vand_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vand.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vand.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vand_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, uint16_t op2, size_t vl) {
@@ -1547,7 +1547,7 @@ vuint16m2_t test_vand_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vand.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vand.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vand_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, vuint16m4_t op2, size_t vl) {
@@ -1557,7 +1557,7 @@ vuint16m4_t test_vand_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vand.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vand.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vand_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, uint16_t op2, size_t vl) {
@@ -1567,7 +1567,7 @@ vuint16m4_t test_vand_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vand.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vand.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vand_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, vuint16m8_t op2, size_t vl) {
@@ -1577,7 +1577,7 @@ vuint16m8_t test_vand_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vand.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vand.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vand_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, uint16_t op2, size_t vl) {
@@ -1587,7 +1587,7 @@ vuint16m8_t test_vand_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vand.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vand.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vand_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, vuint32mf2_t op2, size_t vl) {
@@ -1597,7 +1597,7 @@ vuint32mf2_t test_vand_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vand.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vand.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vand_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, uint32_t op2, size_t vl) {
@@ -1607,7 +1607,7 @@ vuint32mf2_t test_vand_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vand.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vand.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vand_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, vuint32m1_t op2, size_t vl) {
@@ -1617,7 +1617,7 @@ vuint32m1_t test_vand_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vand.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vand.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vand_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, uint32_t op2, size_t vl) {
@@ -1627,7 +1627,7 @@ vuint32m1_t test_vand_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vand.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vand.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vand_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, vuint32m2_t op2, size_t vl) {
@@ -1637,7 +1637,7 @@ vuint32m2_t test_vand_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vand.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vand.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vand_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, uint32_t op2, size_t vl) {
@@ -1647,7 +1647,7 @@ vuint32m2_t test_vand_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vand.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vand.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vand_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, vuint32m4_t op2, size_t vl) {
@@ -1657,7 +1657,7 @@ vuint32m4_t test_vand_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vand.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vand.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vand_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, uint32_t op2, size_t vl) {
@@ -1667,7 +1667,7 @@ vuint32m4_t test_vand_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vand.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vand.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vand_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, vuint32m8_t op2, size_t vl) {
@@ -1677,7 +1677,7 @@ vuint32m8_t test_vand_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vand.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vand.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vand_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, uint32_t op2, size_t vl) {
@@ -1687,7 +1687,7 @@ vuint32m8_t test_vand_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vand.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vand.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vand_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, vuint64m1_t op2, size_t vl) {
@@ -1697,7 +1697,7 @@ vuint64m1_t test_vand_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vand.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vand.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vand_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, uint64_t op2, size_t vl) {
@@ -1707,7 +1707,7 @@ vuint64m1_t test_vand_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vand.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vand.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vand_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, vuint64m2_t op2, size_t vl) {
@@ -1717,7 +1717,7 @@ vuint64m2_t test_vand_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vand.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vand.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vand_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, uint64_t op2, size_t vl) {
@@ -1727,7 +1727,7 @@ vuint64m2_t test_vand_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vand.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vand.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vand_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, vuint64m4_t op2, size_t vl) {
@@ -1737,7 +1737,7 @@ vuint64m4_t test_vand_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vand.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vand.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vand_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, uint64_t op2, size_t vl) {
@@ -1747,7 +1747,7 @@ vuint64m4_t test_vand_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vand_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vand.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vand.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vand_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, vuint64m8_t op2, size_t vl) {
@@ -1757,7 +1757,7 @@ vuint64m8_t test_vand_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8
 //
 // CHECK-RV64-LABEL: @test_vand_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vand.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vand.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vand_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, uint64_t op2, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vasub.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vasub.c
index 4eb6dc1..1695e2b 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vasub.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vasub.c
@@ -890,7 +890,7 @@ vuint64m8_t test_vasubu_vx_u64m8(vuint64m8_t op1, uint64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vasub.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vasub.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vasub_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
@@ -901,7 +901,7 @@ vint8mf8_t test_vasub_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vasub.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vasub.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vasub_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
@@ -912,7 +912,7 @@ vint8mf8_t test_vasub_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vasub.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vasub.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vasub_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
@@ -923,7 +923,7 @@ vint8mf4_t test_vasub_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vasub.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vasub.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vasub_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
@@ -934,7 +934,7 @@ vint8mf4_t test_vasub_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vasub.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vasub.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vasub_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
@@ -945,7 +945,7 @@ vint8mf2_t test_vasub_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vasub.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vasub.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vasub_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
@@ -956,7 +956,7 @@ vint8mf2_t test_vasub_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vasub.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vasub.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vasub_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
@@ -967,7 +967,7 @@ vint8m1_t test_vasub_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vasub.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vasub.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vasub_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
@@ -978,7 +978,7 @@ vint8m1_t test_vasub_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vasub.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vasub.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vasub_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
@@ -989,7 +989,7 @@ vint8m2_t test_vasub_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vasub.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vasub.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vasub_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
@@ -1000,7 +1000,7 @@ vint8m2_t test_vasub_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vasub.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vasub.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vasub_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
@@ -1011,7 +1011,7 @@ vint8m4_t test_vasub_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vasub.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vasub.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vasub_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
@@ -1022,7 +1022,7 @@ vint8m4_t test_vasub_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vasub.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vasub.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vasub_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
@@ -1033,7 +1033,7 @@ vint8m8_t test_vasub_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vasub.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vasub.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vasub_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
@@ -1044,7 +1044,7 @@ vint8m8_t test_vasub_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vasub.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vasub.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vasub_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -1056,7 +1056,7 @@ vint16mf4_t test_vasub_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vasub.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vasub.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vasub_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -1067,7 +1067,7 @@ vint16mf4_t test_vasub_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vasub.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vasub.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vasub_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -1079,7 +1079,7 @@ vint16mf2_t test_vasub_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vasub.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vasub.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vasub_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -1090,7 +1090,7 @@ vint16mf2_t test_vasub_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vasub.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vasub.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vasub_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -1101,7 +1101,7 @@ vint16m1_t test_vasub_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vasub.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vasub.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vasub_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -1112,7 +1112,7 @@ vint16m1_t test_vasub_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vasub.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vasub.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vasub_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -1123,7 +1123,7 @@ vint16m2_t test_vasub_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vasub.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vasub.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vasub_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -1134,7 +1134,7 @@ vint16m2_t test_vasub_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vasub.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vasub.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vasub_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -1145,7 +1145,7 @@ vint16m4_t test_vasub_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vasub.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vasub.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vasub_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -1156,7 +1156,7 @@ vint16m4_t test_vasub_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vasub.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vasub.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vasub_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -1167,7 +1167,7 @@ vint16m8_t test_vasub_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vasub.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vasub.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vasub_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -1178,7 +1178,7 @@ vint16m8_t test_vasub_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vasub.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vasub.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vasub_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -1190,7 +1190,7 @@ vint32mf2_t test_vasub_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vasub.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vasub.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vasub_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -1201,7 +1201,7 @@ vint32mf2_t test_vasub_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vasub.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vasub.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vasub_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -1212,7 +1212,7 @@ vint32m1_t test_vasub_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vasub.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vasub.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vasub_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -1223,7 +1223,7 @@ vint32m1_t test_vasub_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vasub.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vasub.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vasub_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -1234,7 +1234,7 @@ vint32m2_t test_vasub_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vasub.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vasub.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vasub_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -1245,7 +1245,7 @@ vint32m2_t test_vasub_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vasub.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vasub.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vasub_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -1256,7 +1256,7 @@ vint32m4_t test_vasub_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vasub.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vasub.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vasub_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -1267,7 +1267,7 @@ vint32m4_t test_vasub_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vasub.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vasub.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vasub_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -1278,7 +1278,7 @@ vint32m8_t test_vasub_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vasub.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vasub.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vasub_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -1289,7 +1289,7 @@ vint32m8_t test_vasub_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vasub.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vasub.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vasub_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -1300,7 +1300,7 @@ vint64m1_t test_vasub_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vasub.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vasub.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vasub_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -1311,7 +1311,7 @@ vint64m1_t test_vasub_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vasub.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vasub.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vasub_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -1322,7 +1322,7 @@ vint64m2_t test_vasub_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vasub.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vasub.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vasub_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -1333,7 +1333,7 @@ vint64m2_t test_vasub_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vasub.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vasub.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vasub_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -1344,7 +1344,7 @@ vint64m4_t test_vasub_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vasub.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vasub.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vasub_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -1355,7 +1355,7 @@ vint64m4_t test_vasub_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vasub.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vasub.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vasub_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -1366,7 +1366,7 @@ vint64m8_t test_vasub_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasub_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vasub.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vasub.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vasub_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -1377,7 +1377,7 @@ vint64m8_t test_vasub_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vasubu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vasubu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vasubu_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
@@ -1389,7 +1389,7 @@ vuint8mf8_t test_vasubu_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vasubu.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vasubu.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vasubu_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
@@ -1400,7 +1400,7 @@ vuint8mf8_t test_vasubu_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vasubu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vasubu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vasubu_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
@@ -1412,7 +1412,7 @@ vuint8mf4_t test_vasubu_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vasubu.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vasubu.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vasubu_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
@@ -1423,7 +1423,7 @@ vuint8mf4_t test_vasubu_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vasubu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vasubu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vasubu_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
@@ -1435,7 +1435,7 @@ vuint8mf2_t test_vasubu_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vasubu.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vasubu.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vasubu_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
@@ -1446,7 +1446,7 @@ vuint8mf2_t test_vasubu_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vasubu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vasubu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vasubu_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
@@ -1457,7 +1457,7 @@ vuint8m1_t test_vasubu_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vasubu.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vasubu.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vasubu_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
@@ -1468,7 +1468,7 @@ vuint8m1_t test_vasubu_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vasubu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vasubu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vasubu_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
@@ -1479,7 +1479,7 @@ vuint8m2_t test_vasubu_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vasubu.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vasubu.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vasubu_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
@@ -1490,7 +1490,7 @@ vuint8m2_t test_vasubu_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vasubu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vasubu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vasubu_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
@@ -1501,7 +1501,7 @@ vuint8m4_t test_vasubu_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vasubu.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vasubu.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vasubu_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
@@ -1512,7 +1512,7 @@ vuint8m4_t test_vasubu_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vasubu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vasubu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vasubu_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
@@ -1523,7 +1523,7 @@ vuint8m8_t test_vasubu_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vasubu.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vasubu.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vasubu_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
@@ -1534,7 +1534,7 @@ vuint8m8_t test_vasubu_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vasubu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vasubu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vasubu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -1546,7 +1546,7 @@ vuint16mf4_t test_vasubu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vasubu.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vasubu.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vasubu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -1558,7 +1558,7 @@ vuint16mf4_t test_vasubu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vasubu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vasubu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vasubu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -1570,7 +1570,7 @@ vuint16mf2_t test_vasubu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vasubu.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vasubu.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vasubu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -1582,7 +1582,7 @@ vuint16mf2_t test_vasubu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vasubu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vasubu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vasubu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -1594,7 +1594,7 @@ vuint16m1_t test_vasubu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vasubu.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vasubu.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vasubu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -1605,7 +1605,7 @@ vuint16m1_t test_vasubu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vasubu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vasubu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vasubu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -1617,7 +1617,7 @@ vuint16m2_t test_vasubu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vasubu.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vasubu.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vasubu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -1628,7 +1628,7 @@ vuint16m2_t test_vasubu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vasubu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vasubu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vasubu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -1640,7 +1640,7 @@ vuint16m4_t test_vasubu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vasubu.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vasubu.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vasubu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -1651,7 +1651,7 @@ vuint16m4_t test_vasubu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vasubu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vasubu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vasubu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -1663,7 +1663,7 @@ vuint16m8_t test_vasubu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vasubu.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vasubu.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vasubu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -1674,7 +1674,7 @@ vuint16m8_t test_vasubu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vasubu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vasubu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vasubu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -1686,7 +1686,7 @@ vuint32mf2_t test_vasubu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vasubu.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vasubu.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vasubu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -1698,7 +1698,7 @@ vuint32mf2_t test_vasubu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vasubu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vasubu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vasubu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -1710,7 +1710,7 @@ vuint32m1_t test_vasubu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vasubu.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vasubu.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vasubu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -1721,7 +1721,7 @@ vuint32m1_t test_vasubu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vasubu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vasubu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vasubu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -1733,7 +1733,7 @@ vuint32m2_t test_vasubu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vasubu.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vasubu.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vasubu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -1744,7 +1744,7 @@ vuint32m2_t test_vasubu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vasubu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vasubu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vasubu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -1756,7 +1756,7 @@ vuint32m4_t test_vasubu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vasubu.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vasubu.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vasubu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -1767,7 +1767,7 @@ vuint32m4_t test_vasubu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vasubu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vasubu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vasubu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -1779,7 +1779,7 @@ vuint32m8_t test_vasubu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vasubu.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vasubu.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vasubu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -1790,7 +1790,7 @@ vuint32m8_t test_vasubu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vasubu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vasubu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vasubu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -1802,7 +1802,7 @@ vuint64m1_t test_vasubu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vasubu.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vasubu.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vasubu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -1813,7 +1813,7 @@ vuint64m1_t test_vasubu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vasubu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vasubu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vasubu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -1825,7 +1825,7 @@ vuint64m2_t test_vasubu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vasubu.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vasubu.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vasubu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -1836,7 +1836,7 @@ vuint64m2_t test_vasubu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vasubu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vasubu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vasubu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -1848,7 +1848,7 @@ vuint64m4_t test_vasubu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vasubu.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vasubu.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vasubu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -1859,7 +1859,7 @@ vuint64m4_t test_vasubu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vasubu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vasubu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vasubu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
@@ -1871,7 +1871,7 @@ vuint64m8_t test_vasubu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vasubu_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vasubu.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vasubu.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vasubu_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vcompress.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vcompress.c
index 049e69f..cd2c0b0 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vcompress.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vcompress.c
@@ -7,7 +7,7 @@
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i8mf8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vcompress.nxv1i8.i64(<vscale x 1 x i8> [[DEST:%.*]], <vscale x 1 x i8> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vcompress.nxv1i8.i64(<vscale x 1 x i8> [[DEST:%.*]], <vscale x 1 x i8> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vcompress_vm_i8mf8 (vbool64_t mask, vint8mf8_t dest, vint8mf8_t src, size_t vl) {
@@ -16,7 +16,7 @@ vint8mf8_t test_vcompress_vm_i8mf8 (vbool64_t mask, vint8mf8_t dest, vint8mf8_t
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i8mf4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vcompress.nxv2i8.i64(<vscale x 2 x i8> [[DEST:%.*]], <vscale x 2 x i8> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vcompress.nxv2i8.i64(<vscale x 2 x i8> [[DEST:%.*]], <vscale x 2 x i8> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vcompress_vm_i8mf4 (vbool32_t mask, vint8mf4_t dest, vint8mf4_t src, size_t vl) {
@@ -25,7 +25,7 @@ vint8mf4_t test_vcompress_vm_i8mf4 (vbool32_t mask, vint8mf4_t dest, vint8mf4_t
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i8mf2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vcompress.nxv4i8.i64(<vscale x 4 x i8> [[DEST:%.*]], <vscale x 4 x i8> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vcompress.nxv4i8.i64(<vscale x 4 x i8> [[DEST:%.*]], <vscale x 4 x i8> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vcompress_vm_i8mf2 (vbool16_t mask, vint8mf2_t dest, vint8mf2_t src, size_t vl) {
@@ -34,7 +34,7 @@ vint8mf2_t test_vcompress_vm_i8mf2 (vbool16_t mask, vint8mf2_t dest, vint8mf2_t
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i8m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vcompress.nxv8i8.i64(<vscale x 8 x i8> [[DEST:%.*]], <vscale x 8 x i8> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vcompress.nxv8i8.i64(<vscale x 8 x i8> [[DEST:%.*]], <vscale x 8 x i8> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vcompress_vm_i8m1 (vbool8_t mask, vint8m1_t dest, vint8m1_t src, size_t vl) {
@@ -43,7 +43,7 @@ vint8m1_t test_vcompress_vm_i8m1 (vbool8_t mask, vint8m1_t dest, vint8m1_t src,
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i8m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vcompress.nxv16i8.i64(<vscale x 16 x i8> [[DEST:%.*]], <vscale x 16 x i8> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vcompress.nxv16i8.i64(<vscale x 16 x i8> [[DEST:%.*]], <vscale x 16 x i8> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vcompress_vm_i8m2 (vbool4_t mask, vint8m2_t dest, vint8m2_t src, size_t vl) {
@@ -52,7 +52,7 @@ vint8m2_t test_vcompress_vm_i8m2 (vbool4_t mask, vint8m2_t dest, vint8m2_t src,
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i8m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vcompress.nxv32i8.i64(<vscale x 32 x i8> [[DEST:%.*]], <vscale x 32 x i8> [[SRC:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vcompress.nxv32i8.i64(<vscale x 32 x i8> [[DEST:%.*]], <vscale x 32 x i8> [[SRC:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vcompress_vm_i8m4 (vbool2_t mask, vint8m4_t dest, vint8m4_t src, size_t vl) {
@@ -61,7 +61,7 @@ vint8m4_t test_vcompress_vm_i8m4 (vbool2_t mask, vint8m4_t dest, vint8m4_t src,
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i8m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vcompress.nxv64i8.i64(<vscale x 64 x i8> [[DEST:%.*]], <vscale x 64 x i8> [[SRC:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vcompress.nxv64i8.i64(<vscale x 64 x i8> [[DEST:%.*]], <vscale x 64 x i8> [[SRC:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vcompress_vm_i8m8 (vbool1_t mask, vint8m8_t dest, vint8m8_t src, size_t vl) {
@@ -70,7 +70,7 @@ vint8m8_t test_vcompress_vm_i8m8 (vbool1_t mask, vint8m8_t dest, vint8m8_t src,
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i16mf4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vcompress.nxv1i16.i64(<vscale x 1 x i16> [[DEST:%.*]], <vscale x 1 x i16> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vcompress.nxv1i16.i64(<vscale x 1 x i16> [[DEST:%.*]], <vscale x 1 x i16> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vcompress_vm_i16mf4 (vbool64_t mask, vint16mf4_t dest, vint16mf4_t src, size_t vl) {
@@ -79,7 +79,7 @@ vint16mf4_t test_vcompress_vm_i16mf4 (vbool64_t mask, vint16mf4_t dest, vint16mf
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i16mf2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vcompress.nxv2i16.i64(<vscale x 2 x i16> [[DEST:%.*]], <vscale x 2 x i16> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vcompress.nxv2i16.i64(<vscale x 2 x i16> [[DEST:%.*]], <vscale x 2 x i16> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vcompress_vm_i16mf2 (vbool32_t mask, vint16mf2_t dest, vint16mf2_t src, size_t vl) {
@@ -88,7 +88,7 @@ vint16mf2_t test_vcompress_vm_i16mf2 (vbool32_t mask, vint16mf2_t dest, vint16mf
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i16m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vcompress.nxv4i16.i64(<vscale x 4 x i16> [[DEST:%.*]], <vscale x 4 x i16> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vcompress.nxv4i16.i64(<vscale x 4 x i16> [[DEST:%.*]], <vscale x 4 x i16> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vcompress_vm_i16m1 (vbool16_t mask, vint16m1_t dest, vint16m1_t src, size_t vl) {
@@ -97,7 +97,7 @@ vint16m1_t test_vcompress_vm_i16m1 (vbool16_t mask, vint16m1_t dest, vint16m1_t
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i16m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vcompress.nxv8i16.i64(<vscale x 8 x i16> [[DEST:%.*]], <vscale x 8 x i16> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vcompress.nxv8i16.i64(<vscale x 8 x i16> [[DEST:%.*]], <vscale x 8 x i16> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vcompress_vm_i16m2 (vbool8_t mask, vint16m2_t dest, vint16m2_t src, size_t vl) {
@@ -106,7 +106,7 @@ vint16m2_t test_vcompress_vm_i16m2 (vbool8_t mask, vint16m2_t dest, vint16m2_t s
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i16m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vcompress.nxv16i16.i64(<vscale x 16 x i16> [[DEST:%.*]], <vscale x 16 x i16> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vcompress.nxv16i16.i64(<vscale x 16 x i16> [[DEST:%.*]], <vscale x 16 x i16> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vcompress_vm_i16m4 (vbool4_t mask, vint16m4_t dest, vint16m4_t src, size_t vl) {
@@ -115,7 +115,7 @@ vint16m4_t test_vcompress_vm_i16m4 (vbool4_t mask, vint16m4_t dest, vint16m4_t s
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i16m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vcompress.nxv32i16.i64(<vscale x 32 x i16> [[DEST:%.*]], <vscale x 32 x i16> [[SRC:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vcompress.nxv32i16.i64(<vscale x 32 x i16> [[DEST:%.*]], <vscale x 32 x i16> [[SRC:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vcompress_vm_i16m8 (vbool2_t mask, vint16m8_t dest, vint16m8_t src, size_t vl) {
@@ -124,7 +124,7 @@ vint16m8_t test_vcompress_vm_i16m8 (vbool2_t mask, vint16m8_t dest, vint16m8_t s
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i32mf2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vcompress.nxv1i32.i64(<vscale x 1 x i32> [[DEST:%.*]], <vscale x 1 x i32> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vcompress.nxv1i32.i64(<vscale x 1 x i32> [[DEST:%.*]], <vscale x 1 x i32> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vcompress_vm_i32mf2 (vbool64_t mask, vint32mf2_t dest, vint32mf2_t src, size_t vl) {
@@ -133,7 +133,7 @@ vint32mf2_t test_vcompress_vm_i32mf2 (vbool64_t mask, vint32mf2_t dest, vint32mf
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i32m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vcompress.nxv2i32.i64(<vscale x 2 x i32> [[DEST:%.*]], <vscale x 2 x i32> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vcompress.nxv2i32.i64(<vscale x 2 x i32> [[DEST:%.*]], <vscale x 2 x i32> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vcompress_vm_i32m1 (vbool32_t mask, vint32m1_t dest, vint32m1_t src, size_t vl) {
@@ -142,7 +142,7 @@ vint32m1_t test_vcompress_vm_i32m1 (vbool32_t mask, vint32m1_t dest, vint32m1_t
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i32m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vcompress.nxv4i32.i64(<vscale x 4 x i32> [[DEST:%.*]], <vscale x 4 x i32> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vcompress.nxv4i32.i64(<vscale x 4 x i32> [[DEST:%.*]], <vscale x 4 x i32> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vcompress_vm_i32m2 (vbool16_t mask, vint32m2_t dest, vint32m2_t src, size_t vl) {
@@ -151,7 +151,7 @@ vint32m2_t test_vcompress_vm_i32m2 (vbool16_t mask, vint32m2_t dest, vint32m2_t
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i32m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vcompress.nxv8i32.i64(<vscale x 8 x i32> [[DEST:%.*]], <vscale x 8 x i32> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vcompress.nxv8i32.i64(<vscale x 8 x i32> [[DEST:%.*]], <vscale x 8 x i32> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vcompress_vm_i32m4 (vbool8_t mask, vint32m4_t dest, vint32m4_t src, size_t vl) {
@@ -160,7 +160,7 @@ vint32m4_t test_vcompress_vm_i32m4 (vbool8_t mask, vint32m4_t dest, vint32m4_t s
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i32m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vcompress.nxv16i32.i64(<vscale x 16 x i32> [[DEST:%.*]], <vscale x 16 x i32> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vcompress.nxv16i32.i64(<vscale x 16 x i32> [[DEST:%.*]], <vscale x 16 x i32> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vcompress_vm_i32m8 (vbool4_t mask, vint32m8_t dest, vint32m8_t src, size_t vl) {
@@ -169,7 +169,7 @@ vint32m8_t test_vcompress_vm_i32m8 (vbool4_t mask, vint32m8_t dest, vint32m8_t s
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i64m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vcompress.nxv1i64.i64(<vscale x 1 x i64> [[DEST:%.*]], <vscale x 1 x i64> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vcompress.nxv1i64.i64(<vscale x 1 x i64> [[DEST:%.*]], <vscale x 1 x i64> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vcompress_vm_i64m1 (vbool64_t mask, vint64m1_t dest, vint64m1_t src, size_t vl) {
@@ -178,7 +178,7 @@ vint64m1_t test_vcompress_vm_i64m1 (vbool64_t mask, vint64m1_t dest, vint64m1_t
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i64m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vcompress.nxv2i64.i64(<vscale x 2 x i64> [[DEST:%.*]], <vscale x 2 x i64> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vcompress.nxv2i64.i64(<vscale x 2 x i64> [[DEST:%.*]], <vscale x 2 x i64> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vcompress_vm_i64m2 (vbool32_t mask, vint64m2_t dest, vint64m2_t src, size_t vl) {
@@ -187,7 +187,7 @@ vint64m2_t test_vcompress_vm_i64m2 (vbool32_t mask, vint64m2_t dest, vint64m2_t
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i64m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vcompress.nxv4i64.i64(<vscale x 4 x i64> [[DEST:%.*]], <vscale x 4 x i64> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vcompress.nxv4i64.i64(<vscale x 4 x i64> [[DEST:%.*]], <vscale x 4 x i64> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vcompress_vm_i64m4 (vbool16_t mask, vint64m4_t dest, vint64m4_t src, size_t vl) {
@@ -196,7 +196,7 @@ vint64m4_t test_vcompress_vm_i64m4 (vbool16_t mask, vint64m4_t dest, vint64m4_t
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_i64m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vcompress.nxv8i64.i64(<vscale x 8 x i64> [[DEST:%.*]], <vscale x 8 x i64> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vcompress.nxv8i64.i64(<vscale x 8 x i64> [[DEST:%.*]], <vscale x 8 x i64> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vcompress_vm_i64m8 (vbool8_t mask, vint64m8_t dest, vint64m8_t src, size_t vl) {
@@ -205,7 +205,7 @@ vint64m8_t test_vcompress_vm_i64m8 (vbool8_t mask, vint64m8_t dest, vint64m8_t s
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u8mf8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vcompress.nxv1i8.i64(<vscale x 1 x i8> [[DEST:%.*]], <vscale x 1 x i8> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vcompress.nxv1i8.i64(<vscale x 1 x i8> [[DEST:%.*]], <vscale x 1 x i8> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vcompress_vm_u8mf8 (vbool64_t mask, vuint8mf8_t dest, vuint8mf8_t src, size_t vl) {
@@ -214,7 +214,7 @@ vuint8mf8_t test_vcompress_vm_u8mf8 (vbool64_t mask, vuint8mf8_t dest, vuint8mf8
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u8mf4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vcompress.nxv2i8.i64(<vscale x 2 x i8> [[DEST:%.*]], <vscale x 2 x i8> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vcompress.nxv2i8.i64(<vscale x 2 x i8> [[DEST:%.*]], <vscale x 2 x i8> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vcompress_vm_u8mf4 (vbool32_t mask, vuint8mf4_t dest, vuint8mf4_t src, size_t vl) {
@@ -223,7 +223,7 @@ vuint8mf4_t test_vcompress_vm_u8mf4 (vbool32_t mask, vuint8mf4_t dest, vuint8mf4
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u8mf2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vcompress.nxv4i8.i64(<vscale x 4 x i8> [[DEST:%.*]], <vscale x 4 x i8> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vcompress.nxv4i8.i64(<vscale x 4 x i8> [[DEST:%.*]], <vscale x 4 x i8> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vcompress_vm_u8mf2 (vbool16_t mask, vuint8mf2_t dest, vuint8mf2_t src, size_t vl) {
@@ -232,7 +232,7 @@ vuint8mf2_t test_vcompress_vm_u8mf2 (vbool16_t mask, vuint8mf2_t dest, vuint8mf2
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u8m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vcompress.nxv8i8.i64(<vscale x 8 x i8> [[DEST:%.*]], <vscale x 8 x i8> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vcompress.nxv8i8.i64(<vscale x 8 x i8> [[DEST:%.*]], <vscale x 8 x i8> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vcompress_vm_u8m1 (vbool8_t mask, vuint8m1_t dest, vuint8m1_t src, size_t vl) {
@@ -241,7 +241,7 @@ vuint8m1_t test_vcompress_vm_u8m1 (vbool8_t mask, vuint8m1_t dest, vuint8m1_t sr
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u8m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vcompress.nxv16i8.i64(<vscale x 16 x i8> [[DEST:%.*]], <vscale x 16 x i8> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vcompress.nxv16i8.i64(<vscale x 16 x i8> [[DEST:%.*]], <vscale x 16 x i8> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vcompress_vm_u8m2 (vbool4_t mask, vuint8m2_t dest, vuint8m2_t src, size_t vl) {
@@ -250,7 +250,7 @@ vuint8m2_t test_vcompress_vm_u8m2 (vbool4_t mask, vuint8m2_t dest, vuint8m2_t sr
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u8m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vcompress.nxv32i8.i64(<vscale x 32 x i8> [[DEST:%.*]], <vscale x 32 x i8> [[SRC:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vcompress.nxv32i8.i64(<vscale x 32 x i8> [[DEST:%.*]], <vscale x 32 x i8> [[SRC:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vcompress_vm_u8m4 (vbool2_t mask, vuint8m4_t dest, vuint8m4_t src, size_t vl) {
@@ -259,7 +259,7 @@ vuint8m4_t test_vcompress_vm_u8m4 (vbool2_t mask, vuint8m4_t dest, vuint8m4_t sr
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u8m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vcompress.nxv64i8.i64(<vscale x 64 x i8> [[DEST:%.*]], <vscale x 64 x i8> [[SRC:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vcompress.nxv64i8.i64(<vscale x 64 x i8> [[DEST:%.*]], <vscale x 64 x i8> [[SRC:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vcompress_vm_u8m8 (vbool1_t mask, vuint8m8_t dest, vuint8m8_t src, size_t vl) {
@@ -268,7 +268,7 @@ vuint8m8_t test_vcompress_vm_u8m8 (vbool1_t mask, vuint8m8_t dest, vuint8m8_t sr
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u16mf4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vcompress.nxv1i16.i64(<vscale x 1 x i16> [[DEST:%.*]], <vscale x 1 x i16> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vcompress.nxv1i16.i64(<vscale x 1 x i16> [[DEST:%.*]], <vscale x 1 x i16> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vcompress_vm_u16mf4 (vbool64_t mask, vuint16mf4_t dest, vuint16mf4_t src, size_t vl) {
@@ -277,7 +277,7 @@ vuint16mf4_t test_vcompress_vm_u16mf4 (vbool64_t mask, vuint16mf4_t dest, vuint1
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u16mf2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vcompress.nxv2i16.i64(<vscale x 2 x i16> [[DEST:%.*]], <vscale x 2 x i16> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vcompress.nxv2i16.i64(<vscale x 2 x i16> [[DEST:%.*]], <vscale x 2 x i16> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vcompress_vm_u16mf2 (vbool32_t mask, vuint16mf2_t dest, vuint16mf2_t src, size_t vl) {
@@ -286,7 +286,7 @@ vuint16mf2_t test_vcompress_vm_u16mf2 (vbool32_t mask, vuint16mf2_t dest, vuint1
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u16m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vcompress.nxv4i16.i64(<vscale x 4 x i16> [[DEST:%.*]], <vscale x 4 x i16> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vcompress.nxv4i16.i64(<vscale x 4 x i16> [[DEST:%.*]], <vscale x 4 x i16> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vcompress_vm_u16m1 (vbool16_t mask, vuint16m1_t dest, vuint16m1_t src, size_t vl) {
@@ -295,7 +295,7 @@ vuint16m1_t test_vcompress_vm_u16m1 (vbool16_t mask, vuint16m1_t dest, vuint16m1
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u16m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vcompress.nxv8i16.i64(<vscale x 8 x i16> [[DEST:%.*]], <vscale x 8 x i16> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vcompress.nxv8i16.i64(<vscale x 8 x i16> [[DEST:%.*]], <vscale x 8 x i16> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vcompress_vm_u16m2 (vbool8_t mask, vuint16m2_t dest, vuint16m2_t src, size_t vl) {
@@ -304,7 +304,7 @@ vuint16m2_t test_vcompress_vm_u16m2 (vbool8_t mask, vuint16m2_t dest, vuint16m2_
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u16m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vcompress.nxv16i16.i64(<vscale x 16 x i16> [[DEST:%.*]], <vscale x 16 x i16> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vcompress.nxv16i16.i64(<vscale x 16 x i16> [[DEST:%.*]], <vscale x 16 x i16> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vcompress_vm_u16m4 (vbool4_t mask, vuint16m4_t dest, vuint16m4_t src, size_t vl) {
@@ -313,7 +313,7 @@ vuint16m4_t test_vcompress_vm_u16m4 (vbool4_t mask, vuint16m4_t dest, vuint16m4_
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u16m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vcompress.nxv32i16.i64(<vscale x 32 x i16> [[DEST:%.*]], <vscale x 32 x i16> [[SRC:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vcompress.nxv32i16.i64(<vscale x 32 x i16> [[DEST:%.*]], <vscale x 32 x i16> [[SRC:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vcompress_vm_u16m8 (vbool2_t mask, vuint16m8_t dest, vuint16m8_t src, size_t vl) {
@@ -322,7 +322,7 @@ vuint16m8_t test_vcompress_vm_u16m8 (vbool2_t mask, vuint16m8_t dest, vuint16m8_
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u32mf2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vcompress.nxv1i32.i64(<vscale x 1 x i32> [[DEST:%.*]], <vscale x 1 x i32> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vcompress.nxv1i32.i64(<vscale x 1 x i32> [[DEST:%.*]], <vscale x 1 x i32> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vcompress_vm_u32mf2 (vbool64_t mask, vuint32mf2_t dest, vuint32mf2_t src, size_t vl) {
@@ -331,7 +331,7 @@ vuint32mf2_t test_vcompress_vm_u32mf2 (vbool64_t mask, vuint32mf2_t dest, vuint3
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u32m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vcompress.nxv2i32.i64(<vscale x 2 x i32> [[DEST:%.*]], <vscale x 2 x i32> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vcompress.nxv2i32.i64(<vscale x 2 x i32> [[DEST:%.*]], <vscale x 2 x i32> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vcompress_vm_u32m1 (vbool32_t mask, vuint32m1_t dest, vuint32m1_t src, size_t vl) {
@@ -340,7 +340,7 @@ vuint32m1_t test_vcompress_vm_u32m1 (vbool32_t mask, vuint32m1_t dest, vuint32m1
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u32m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vcompress.nxv4i32.i64(<vscale x 4 x i32> [[DEST:%.*]], <vscale x 4 x i32> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vcompress.nxv4i32.i64(<vscale x 4 x i32> [[DEST:%.*]], <vscale x 4 x i32> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vcompress_vm_u32m2 (vbool16_t mask, vuint32m2_t dest, vuint32m2_t src, size_t vl) {
@@ -349,7 +349,7 @@ vuint32m2_t test_vcompress_vm_u32m2 (vbool16_t mask, vuint32m2_t dest, vuint32m2
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u32m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vcompress.nxv8i32.i64(<vscale x 8 x i32> [[DEST:%.*]], <vscale x 8 x i32> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vcompress.nxv8i32.i64(<vscale x 8 x i32> [[DEST:%.*]], <vscale x 8 x i32> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vcompress_vm_u32m4 (vbool8_t mask, vuint32m4_t dest, vuint32m4_t src, size_t vl) {
@@ -358,7 +358,7 @@ vuint32m4_t test_vcompress_vm_u32m4 (vbool8_t mask, vuint32m4_t dest, vuint32m4_
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u32m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vcompress.nxv16i32.i64(<vscale x 16 x i32> [[DEST:%.*]], <vscale x 16 x i32> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vcompress.nxv16i32.i64(<vscale x 16 x i32> [[DEST:%.*]], <vscale x 16 x i32> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vcompress_vm_u32m8 (vbool4_t mask, vuint32m8_t dest, vuint32m8_t src, size_t vl) {
@@ -367,7 +367,7 @@ vuint32m8_t test_vcompress_vm_u32m8 (vbool4_t mask, vuint32m8_t dest, vuint32m8_
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u64m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vcompress.nxv1i64.i64(<vscale x 1 x i64> [[DEST:%.*]], <vscale x 1 x i64> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vcompress.nxv1i64.i64(<vscale x 1 x i64> [[DEST:%.*]], <vscale x 1 x i64> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vcompress_vm_u64m1 (vbool64_t mask, vuint64m1_t dest, vuint64m1_t src, size_t vl) {
@@ -376,7 +376,7 @@ vuint64m1_t test_vcompress_vm_u64m1 (vbool64_t mask, vuint64m1_t dest, vuint64m1
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u64m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vcompress.nxv2i64.i64(<vscale x 2 x i64> [[DEST:%.*]], <vscale x 2 x i64> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vcompress.nxv2i64.i64(<vscale x 2 x i64> [[DEST:%.*]], <vscale x 2 x i64> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vcompress_vm_u64m2 (vbool32_t mask, vuint64m2_t dest, vuint64m2_t src, size_t vl) {
@@ -385,7 +385,7 @@ vuint64m2_t test_vcompress_vm_u64m2 (vbool32_t mask, vuint64m2_t dest, vuint64m2
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u64m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vcompress.nxv4i64.i64(<vscale x 4 x i64> [[DEST:%.*]], <vscale x 4 x i64> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vcompress.nxv4i64.i64(<vscale x 4 x i64> [[DEST:%.*]], <vscale x 4 x i64> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vcompress_vm_u64m4 (vbool16_t mask, vuint64m4_t dest, vuint64m4_t src, size_t vl) {
@@ -394,7 +394,7 @@ vuint64m4_t test_vcompress_vm_u64m4 (vbool16_t mask, vuint64m4_t dest, vuint64m4
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_u64m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vcompress.nxv8i64.i64(<vscale x 8 x i64> [[DEST:%.*]], <vscale x 8 x i64> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vcompress.nxv8i64.i64(<vscale x 8 x i64> [[DEST:%.*]], <vscale x 8 x i64> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vcompress_vm_u64m8 (vbool8_t mask, vuint64m8_t dest, vuint64m8_t src, size_t vl) {
@@ -403,7 +403,7 @@ vuint64m8_t test_vcompress_vm_u64m8 (vbool8_t mask, vuint64m8_t dest, vuint64m8_
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_f32mf2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vcompress.nxv1f32.i64(<vscale x 1 x float> [[DEST:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vcompress.nxv1f32.i64(<vscale x 1 x float> [[DEST:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vcompress_vm_f32mf2 (vbool64_t mask, vfloat32mf2_t dest, vfloat32mf2_t src, size_t vl) {
@@ -412,7 +412,7 @@ vfloat32mf2_t test_vcompress_vm_f32mf2 (vbool64_t mask, vfloat32mf2_t dest, vflo
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_f32m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vcompress.nxv2f32.i64(<vscale x 2 x float> [[DEST:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vcompress.nxv2f32.i64(<vscale x 2 x float> [[DEST:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vcompress_vm_f32m1 (vbool32_t mask, vfloat32m1_t dest, vfloat32m1_t src, size_t vl) {
@@ -421,7 +421,7 @@ vfloat32m1_t test_vcompress_vm_f32m1 (vbool32_t mask, vfloat32m1_t dest, vfloat3
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_f32m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vcompress.nxv4f32.i64(<vscale x 4 x float> [[DEST:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vcompress.nxv4f32.i64(<vscale x 4 x float> [[DEST:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vcompress_vm_f32m2 (vbool16_t mask, vfloat32m2_t dest, vfloat32m2_t src, size_t vl) {
@@ -430,7 +430,7 @@ vfloat32m2_t test_vcompress_vm_f32m2 (vbool16_t mask, vfloat32m2_t dest, vfloat3
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_f32m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vcompress.nxv8f32.i64(<vscale x 8 x float> [[DEST:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vcompress.nxv8f32.i64(<vscale x 8 x float> [[DEST:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vcompress_vm_f32m4 (vbool8_t mask, vfloat32m4_t dest, vfloat32m4_t src, size_t vl) {
@@ -439,7 +439,7 @@ vfloat32m4_t test_vcompress_vm_f32m4 (vbool8_t mask, vfloat32m4_t dest, vfloat32
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_f32m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vcompress.nxv16f32.i64(<vscale x 16 x float> [[DEST:%.*]], <vscale x 16 x float> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vcompress.nxv16f32.i64(<vscale x 16 x float> [[DEST:%.*]], <vscale x 16 x float> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vcompress_vm_f32m8 (vbool4_t mask, vfloat32m8_t dest, vfloat32m8_t src, size_t vl) {
@@ -448,7 +448,7 @@ vfloat32m8_t test_vcompress_vm_f32m8 (vbool4_t mask, vfloat32m8_t dest, vfloat32
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_f64m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vcompress.nxv1f64.i64(<vscale x 1 x double> [[DEST:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vcompress.nxv1f64.i64(<vscale x 1 x double> [[DEST:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vcompress_vm_f64m1 (vbool64_t mask, vfloat64m1_t dest, vfloat64m1_t src, size_t vl) {
@@ -457,7 +457,7 @@ vfloat64m1_t test_vcompress_vm_f64m1 (vbool64_t mask, vfloat64m1_t dest, vfloat6
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_f64m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vcompress.nxv2f64.i64(<vscale x 2 x double> [[DEST:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vcompress.nxv2f64.i64(<vscale x 2 x double> [[DEST:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vcompress_vm_f64m2 (vbool32_t mask, vfloat64m2_t dest, vfloat64m2_t src, size_t vl) {
@@ -466,7 +466,7 @@ vfloat64m2_t test_vcompress_vm_f64m2 (vbool32_t mask, vfloat64m2_t dest, vfloat6
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_f64m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vcompress.nxv4f64.i64(<vscale x 4 x double> [[DEST:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vcompress.nxv4f64.i64(<vscale x 4 x double> [[DEST:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vcompress_vm_f64m4 (vbool16_t mask, vfloat64m4_t dest, vfloat64m4_t src, size_t vl) {
@@ -475,7 +475,7 @@ vfloat64m4_t test_vcompress_vm_f64m4 (vbool16_t mask, vfloat64m4_t dest, vfloat6
 
 // CHECK-RV64-LABEL: @test_vcompress_vm_f64m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vcompress.nxv8f64.i64(<vscale x 8 x double> [[DEST:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vcompress.nxv8f64.i64(<vscale x 8 x double> [[DEST:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vcompress_vm_f64m8 (vbool8_t mask, vfloat64m8_t dest, vfloat64m8_t src, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vdiv.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vdiv.c
index edd61e2..8788f07 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vdiv.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vdiv.c
@@ -887,7 +887,7 @@ vuint64m8_t test_vdivu_vx_u64m8(vuint64m8_t op1, uint64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vdiv.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vdiv.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vdiv_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, vint8mf8_t op2, size_t vl) {
@@ -897,7 +897,7 @@ vint8mf8_t test_vdiv_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vdiv.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vdiv.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vdiv_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, int8_t op2, size_t vl) {
@@ -907,7 +907,7 @@ vint8mf8_t test_vdiv_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vdiv.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vdiv.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vdiv_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, vint8mf4_t op2, size_t vl) {
@@ -917,7 +917,7 @@ vint8mf4_t test_vdiv_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vdiv.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vdiv.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vdiv_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, int8_t op2, size_t vl) {
@@ -927,7 +927,7 @@ vint8mf4_t test_vdiv_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vdiv.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vdiv.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vdiv_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, vint8mf2_t op2, size_t vl) {
@@ -937,7 +937,7 @@ vint8mf2_t test_vdiv_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vdiv.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vdiv.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vdiv_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, int8_t op2, size_t vl) {
@@ -947,7 +947,7 @@ vint8mf2_t test_vdiv_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vdiv.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vdiv.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vdiv_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, vint8m1_t op2, size_t vl) {
@@ -957,7 +957,7 @@ vint8m1_t test_vdiv_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vdiv.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vdiv.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vdiv_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, int8_t op2, size_t vl) {
@@ -967,7 +967,7 @@ vint8m1_t test_vdiv_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vdiv.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vdiv.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vdiv_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, vint8m2_t op2, size_t vl) {
@@ -977,7 +977,7 @@ vint8m2_t test_vdiv_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vdiv.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vdiv.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vdiv_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, int8_t op2, size_t vl) {
@@ -987,7 +987,7 @@ vint8m2_t test_vdiv_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vdiv.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vdiv.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vdiv_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, vint8m4_t op2, size_t vl) {
@@ -997,7 +997,7 @@ vint8m4_t test_vdiv_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vdiv.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vdiv.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vdiv_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, int8_t op2, size_t vl) {
@@ -1007,7 +1007,7 @@ vint8m4_t test_vdiv_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vdiv.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vdiv.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vdiv_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, vint8m8_t op2, size_t vl) {
@@ -1017,7 +1017,7 @@ vint8m8_t test_vdiv_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vdiv.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vdiv.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vdiv_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, int8_t op2, size_t vl) {
@@ -1027,7 +1027,7 @@ vint8m8_t test_vdiv_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vdiv.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vdiv.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vdiv_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, vint16mf4_t op2, size_t vl) {
@@ -1037,7 +1037,7 @@ vint16mf4_t test_vdiv_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vdiv.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vdiv.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vdiv_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, int16_t op2, size_t vl) {
@@ -1047,7 +1047,7 @@ vint16mf4_t test_vdiv_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vdiv.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vdiv.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vdiv_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, vint16mf2_t op2, size_t vl) {
@@ -1057,7 +1057,7 @@ vint16mf2_t test_vdiv_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vdiv.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vdiv.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vdiv_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, int16_t op2, size_t vl) {
@@ -1067,7 +1067,7 @@ vint16mf2_t test_vdiv_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vdiv.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vdiv.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vdiv_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, vint16m1_t op2, size_t vl) {
@@ -1077,7 +1077,7 @@ vint16m1_t test_vdiv_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vdiv.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vdiv.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vdiv_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, int16_t op2, size_t vl) {
@@ -1087,7 +1087,7 @@ vint16m1_t test_vdiv_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vdiv.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vdiv.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vdiv_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, vint16m2_t op2, size_t vl) {
@@ -1097,7 +1097,7 @@ vint16m2_t test_vdiv_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vdiv.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vdiv.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vdiv_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, int16_t op2, size_t vl) {
@@ -1107,7 +1107,7 @@ vint16m2_t test_vdiv_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vdiv.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vdiv.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vdiv_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, vint16m4_t op2, size_t vl) {
@@ -1117,7 +1117,7 @@ vint16m4_t test_vdiv_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vdiv.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vdiv.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vdiv_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, int16_t op2, size_t vl) {
@@ -1127,7 +1127,7 @@ vint16m4_t test_vdiv_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vdiv.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vdiv.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vdiv_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, vint16m8_t op2, size_t vl) {
@@ -1137,7 +1137,7 @@ vint16m8_t test_vdiv_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vdiv.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vdiv.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vdiv_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, int16_t op2, size_t vl) {
@@ -1147,7 +1147,7 @@ vint16m8_t test_vdiv_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vdiv.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vdiv.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vdiv_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, vint32mf2_t op2, size_t vl) {
@@ -1157,7 +1157,7 @@ vint32mf2_t test_vdiv_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vdiv.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vdiv.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vdiv_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, int32_t op2, size_t vl) {
@@ -1167,7 +1167,7 @@ vint32mf2_t test_vdiv_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vdiv.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vdiv.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vdiv_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, vint32m1_t op2, size_t vl) {
@@ -1177,7 +1177,7 @@ vint32m1_t test_vdiv_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vdiv.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vdiv.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vdiv_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, int32_t op2, size_t vl) {
@@ -1187,7 +1187,7 @@ vint32m1_t test_vdiv_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vdiv.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vdiv.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vdiv_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, vint32m2_t op2, size_t vl) {
@@ -1197,7 +1197,7 @@ vint32m2_t test_vdiv_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vdiv.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vdiv.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vdiv_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, int32_t op2, size_t vl) {
@@ -1207,7 +1207,7 @@ vint32m2_t test_vdiv_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vdiv.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vdiv.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vdiv_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, vint32m4_t op2, size_t vl) {
@@ -1217,7 +1217,7 @@ vint32m4_t test_vdiv_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vdiv.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vdiv.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vdiv_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, int32_t op2, size_t vl) {
@@ -1227,7 +1227,7 @@ vint32m4_t test_vdiv_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vdiv.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vdiv.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vdiv_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, vint32m8_t op2, size_t vl) {
@@ -1237,7 +1237,7 @@ vint32m8_t test_vdiv_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vdiv.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vdiv.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vdiv_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, int32_t op2, size_t vl) {
@@ -1247,7 +1247,7 @@ vint32m8_t test_vdiv_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vdiv.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vdiv.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vdiv_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, vint64m1_t op2, size_t vl) {
@@ -1257,7 +1257,7 @@ vint64m1_t test_vdiv_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vdiv.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vdiv.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vdiv_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, int64_t op2, size_t vl) {
@@ -1267,7 +1267,7 @@ vint64m1_t test_vdiv_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vdiv.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vdiv.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vdiv_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, vint64m2_t op2, size_t vl) {
@@ -1277,7 +1277,7 @@ vint64m2_t test_vdiv_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vdiv.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vdiv.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vdiv_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, int64_t op2, size_t vl) {
@@ -1287,7 +1287,7 @@ vint64m2_t test_vdiv_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vdiv.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vdiv.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vdiv_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, vint64m4_t op2, size_t vl) {
@@ -1297,7 +1297,7 @@ vint64m4_t test_vdiv_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vdiv.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vdiv.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vdiv_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, int64_t op2, size_t vl) {
@@ -1307,7 +1307,7 @@ vint64m4_t test_vdiv_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vdiv.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vdiv.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vdiv_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, vint64m8_t op2, size_t vl) {
@@ -1317,7 +1317,7 @@ vint64m8_t test_vdiv_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vdiv_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vdiv.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vdiv.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vdiv_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, int64_t op2, size_t vl) {
@@ -1327,7 +1327,7 @@ vint64m8_t test_vdiv_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vdivu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vdivu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vdivu_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -1337,7 +1337,7 @@ vuint8mf8_t test_vdivu_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vdivu.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vdivu.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vdivu_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, uint8_t op2, size_t vl) {
@@ -1347,7 +1347,7 @@ vuint8mf8_t test_vdivu_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vdivu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vdivu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vdivu_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, vuint8mf4_t op2, size_t vl) {
@@ -1357,7 +1357,7 @@ vuint8mf4_t test_vdivu_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vdivu.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vdivu.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vdivu_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, uint8_t op2, size_t vl) {
@@ -1367,7 +1367,7 @@ vuint8mf4_t test_vdivu_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vdivu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vdivu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vdivu_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, vuint8mf2_t op2, size_t vl) {
@@ -1377,7 +1377,7 @@ vuint8mf2_t test_vdivu_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vdivu.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vdivu.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vdivu_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, uint8_t op2, size_t vl) {
@@ -1387,7 +1387,7 @@ vuint8mf2_t test_vdivu_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vdivu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vdivu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vdivu_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, vuint8m1_t op2, size_t vl) {
@@ -1397,7 +1397,7 @@ vuint8m1_t test_vdivu_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vdivu.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vdivu.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vdivu_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, uint8_t op2, size_t vl) {
@@ -1407,7 +1407,7 @@ vuint8m1_t test_vdivu_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vdivu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vdivu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vdivu_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, vuint8m2_t op2, size_t vl) {
@@ -1417,7 +1417,7 @@ vuint8m2_t test_vdivu_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vdivu.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vdivu.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vdivu_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, uint8_t op2, size_t vl) {
@@ -1427,7 +1427,7 @@ vuint8m2_t test_vdivu_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vdivu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vdivu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vdivu_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, vuint8m4_t op2, size_t vl) {
@@ -1437,7 +1437,7 @@ vuint8m4_t test_vdivu_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vdivu.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vdivu.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vdivu_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, uint8_t op2, size_t vl) {
@@ -1447,7 +1447,7 @@ vuint8m4_t test_vdivu_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vdivu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vdivu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vdivu_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, vuint8m8_t op2, size_t vl) {
@@ -1457,7 +1457,7 @@ vuint8m8_t test_vdivu_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vdivu.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vdivu.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vdivu_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, uint8_t op2, size_t vl) {
@@ -1467,7 +1467,7 @@ vuint8m8_t test_vdivu_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vdivu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vdivu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vdivu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, vuint16mf4_t op2, size_t vl) {
@@ -1477,7 +1477,7 @@ vuint16mf4_t test_vdivu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vdivu.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vdivu.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vdivu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, uint16_t op2, size_t vl) {
@@ -1487,7 +1487,7 @@ vuint16mf4_t test_vdivu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vdivu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vdivu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vdivu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, vuint16mf2_t op2, size_t vl) {
@@ -1497,7 +1497,7 @@ vuint16mf2_t test_vdivu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vdivu.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vdivu.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vdivu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, uint16_t op2, size_t vl) {
@@ -1507,7 +1507,7 @@ vuint16mf2_t test_vdivu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vdivu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vdivu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vdivu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, vuint16m1_t op2, size_t vl) {
@@ -1517,7 +1517,7 @@ vuint16m1_t test_vdivu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vdivu.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vdivu.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vdivu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, uint16_t op2, size_t vl) {
@@ -1527,7 +1527,7 @@ vuint16m1_t test_vdivu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vdivu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vdivu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vdivu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, vuint16m2_t op2, size_t vl) {
@@ -1537,7 +1537,7 @@ vuint16m2_t test_vdivu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vdivu.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vdivu.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vdivu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, uint16_t op2, size_t vl) {
@@ -1547,7 +1547,7 @@ vuint16m2_t test_vdivu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vdivu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vdivu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vdivu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, vuint16m4_t op2, size_t vl) {
@@ -1557,7 +1557,7 @@ vuint16m4_t test_vdivu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vdivu.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vdivu.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vdivu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, uint16_t op2, size_t vl) {
@@ -1567,7 +1567,7 @@ vuint16m4_t test_vdivu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vdivu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vdivu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vdivu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, vuint16m8_t op2, size_t vl) {
@@ -1577,7 +1577,7 @@ vuint16m8_t test_vdivu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vdivu.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vdivu.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vdivu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, uint16_t op2, size_t vl) {
@@ -1587,7 +1587,7 @@ vuint16m8_t test_vdivu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vdivu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vdivu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vdivu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, vuint32mf2_t op2, size_t vl) {
@@ -1597,7 +1597,7 @@ vuint32mf2_t test_vdivu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vdivu.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vdivu.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vdivu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, uint32_t op2, size_t vl) {
@@ -1607,7 +1607,7 @@ vuint32mf2_t test_vdivu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vdivu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vdivu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vdivu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, vuint32m1_t op2, size_t vl) {
@@ -1617,7 +1617,7 @@ vuint32m1_t test_vdivu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vdivu.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vdivu.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vdivu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, uint32_t op2, size_t vl) {
@@ -1627,7 +1627,7 @@ vuint32m1_t test_vdivu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vdivu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vdivu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vdivu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, vuint32m2_t op2, size_t vl) {
@@ -1637,7 +1637,7 @@ vuint32m2_t test_vdivu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vdivu.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vdivu.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vdivu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, uint32_t op2, size_t vl) {
@@ -1647,7 +1647,7 @@ vuint32m2_t test_vdivu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vdivu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vdivu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vdivu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, vuint32m4_t op2, size_t vl) {
@@ -1657,7 +1657,7 @@ vuint32m4_t test_vdivu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vdivu.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vdivu.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vdivu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, uint32_t op2, size_t vl) {
@@ -1667,7 +1667,7 @@ vuint32m4_t test_vdivu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vdivu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vdivu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vdivu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, vuint32m8_t op2, size_t vl) {
@@ -1677,7 +1677,7 @@ vuint32m8_t test_vdivu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vdivu.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vdivu.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vdivu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, uint32_t op2, size_t vl) {
@@ -1687,7 +1687,7 @@ vuint32m8_t test_vdivu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vdivu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vdivu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vdivu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, vuint64m1_t op2, size_t vl) {
@@ -1697,7 +1697,7 @@ vuint64m1_t test_vdivu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vdivu.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vdivu.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vdivu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, uint64_t op2, size_t vl) {
@@ -1707,7 +1707,7 @@ vuint64m1_t test_vdivu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vdivu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vdivu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vdivu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, vuint64m2_t op2, size_t vl) {
@@ -1717,7 +1717,7 @@ vuint64m2_t test_vdivu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vdivu.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vdivu.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vdivu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, uint64_t op2, size_t vl) {
@@ -1727,7 +1727,7 @@ vuint64m2_t test_vdivu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vdivu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vdivu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vdivu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, vuint64m4_t op2, size_t vl) {
@@ -1737,7 +1737,7 @@ vuint64m4_t test_vdivu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vdivu.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vdivu.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vdivu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, uint64_t op2, size_t vl) {
@@ -1747,7 +1747,7 @@ vuint64m4_t test_vdivu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vdivu_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vdivu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vdivu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vdivu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, vuint64m8_t op2, size_t vl) {
@@ -1757,7 +1757,7 @@ vuint64m8_t test_vdivu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vdivu_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vdivu.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vdivu.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vdivu_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, uint64_t op2, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfabs.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfabs.c
index 431c1fa..86a7e7049 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfabs.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfabs.c
@@ -98,7 +98,7 @@ vfloat64m8_t test_vfabs_v_f64m8 (vfloat64m8_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfabs_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsgnjx.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP1]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsgnjx.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP1]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfabs_v_f32mf2_m (vbool64_t mask, vfloat32mf2_t maskedoff, vfloat32mf2_t op1, size_t vl) {
@@ -108,7 +108,7 @@ vfloat32mf2_t test_vfabs_v_f32mf2_m (vbool64_t mask, vfloat32mf2_t maskedoff, vf
 //
 // CHECK-RV64-LABEL: @test_vfabs_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsgnjx.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP1]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsgnjx.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP1]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfabs_v_f32m1_m (vbool32_t mask, vfloat32m1_t maskedoff, vfloat32m1_t op1, size_t vl) {
@@ -118,7 +118,7 @@ vfloat32m1_t test_vfabs_v_f32m1_m (vbool32_t mask, vfloat32m1_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vfabs_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsgnjx.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP1]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsgnjx.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP1]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfabs_v_f32m2_m (vbool16_t mask, vfloat32m2_t maskedoff, vfloat32m2_t op1, size_t vl) {
@@ -128,7 +128,7 @@ vfloat32m2_t test_vfabs_v_f32m2_m (vbool16_t mask, vfloat32m2_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vfabs_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsgnjx.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP1]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsgnjx.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP1]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfabs_v_f32m4_m (vbool8_t mask, vfloat32m4_t maskedoff, vfloat32m4_t op1, size_t vl) {
@@ -138,7 +138,7 @@ vfloat32m4_t test_vfabs_v_f32m4_m (vbool8_t mask, vfloat32m4_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vfabs_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsgnjx.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP1]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsgnjx.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP1]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfabs_v_f32m8_m (vbool4_t mask, vfloat32m8_t maskedoff, vfloat32m8_t op1, size_t vl) {
@@ -148,7 +148,7 @@ vfloat32m8_t test_vfabs_v_f32m8_m (vbool4_t mask, vfloat32m8_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vfabs_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsgnjx.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP1]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsgnjx.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP1]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfabs_v_f64m1_m (vbool64_t mask, vfloat64m1_t maskedoff, vfloat64m1_t op1, size_t vl) {
@@ -158,7 +158,7 @@ vfloat64m1_t test_vfabs_v_f64m1_m (vbool64_t mask, vfloat64m1_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vfabs_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsgnjx.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP1]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsgnjx.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP1]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfabs_v_f64m2_m (vbool32_t mask, vfloat64m2_t maskedoff, vfloat64m2_t op1, size_t vl) {
@@ -168,7 +168,7 @@ vfloat64m2_t test_vfabs_v_f64m2_m (vbool32_t mask, vfloat64m2_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vfabs_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsgnjx.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP1]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsgnjx.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP1]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfabs_v_f64m4_m (vbool16_t mask, vfloat64m4_t maskedoff, vfloat64m4_t op1, size_t vl) {
@@ -178,7 +178,7 @@ vfloat64m4_t test_vfabs_v_f64m4_m (vbool16_t mask, vfloat64m4_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vfabs_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsgnjx.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP1]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsgnjx.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP1]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfabs_v_f64m8_m (vbool8_t mask, vfloat64m8_t maskedoff, vfloat64m8_t op1, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfadd.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfadd.c
index b9efbd7..40129bb 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfadd.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfadd.c
@@ -188,7 +188,7 @@ vfloat64m8_t test_vfadd_vf_f64m8(vfloat64m8_t op1, double op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfadd_vv_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfadd.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfadd.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfadd_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff, vfloat32mf2_t op1, vfloat32mf2_t op2, size_t vl) {
@@ -198,7 +198,7 @@ vfloat32mf2_t test_vfadd_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff, vf
 //
 // CHECK-RV64-LABEL: @test_vfadd_vf_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfadd.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfadd.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfadd_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff, vfloat32mf2_t op1, float op2, size_t vl) {
@@ -208,7 +208,7 @@ vfloat32mf2_t test_vfadd_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff, vf
 //
 // CHECK-RV64-LABEL: @test_vfadd_vv_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfadd.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfadd.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfadd_vv_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, vfloat32m1_t op1, vfloat32m1_t op2, size_t vl) {
@@ -218,7 +218,7 @@ vfloat32m1_t test_vfadd_vv_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vfadd_vf_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfadd.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfadd.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfadd_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, vfloat32m1_t op1, float op2, size_t vl) {
@@ -228,7 +228,7 @@ vfloat32m1_t test_vfadd_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vfadd_vv_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfadd.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfadd.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfadd_vv_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, vfloat32m2_t op1, vfloat32m2_t op2, size_t vl) {
@@ -238,7 +238,7 @@ vfloat32m2_t test_vfadd_vv_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vfadd_vf_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfadd.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfadd.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfadd_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, vfloat32m2_t op1, float op2, size_t vl) {
@@ -248,7 +248,7 @@ vfloat32m2_t test_vfadd_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vfadd_vv_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfadd.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfadd.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfadd_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, vfloat32m4_t op1, vfloat32m4_t op2, size_t vl) {
@@ -258,7 +258,7 @@ vfloat32m4_t test_vfadd_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vfadd_vf_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfadd.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfadd.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfadd_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, vfloat32m4_t op1, float op2, size_t vl) {
@@ -268,7 +268,7 @@ vfloat32m4_t test_vfadd_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vfadd_vv_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfadd.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfadd.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfadd_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff, vfloat32m8_t op1, vfloat32m8_t op2, size_t vl) {
@@ -278,7 +278,7 @@ vfloat32m8_t test_vfadd_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vfadd_vf_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfadd.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfadd.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfadd_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff, vfloat32m8_t op1, float op2, size_t vl) {
@@ -288,7 +288,7 @@ vfloat32m8_t test_vfadd_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vfadd_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfadd.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfadd.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfadd_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, vfloat64m1_t op1, vfloat64m1_t op2, size_t vl) {
@@ -298,7 +298,7 @@ vfloat64m1_t test_vfadd_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vfadd_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfadd.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfadd.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfadd_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, vfloat64m1_t op1, double op2, size_t vl) {
@@ -308,7 +308,7 @@ vfloat64m1_t test_vfadd_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vfadd_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfadd.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfadd.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfadd_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, vfloat64m2_t op1, vfloat64m2_t op2, size_t vl) {
@@ -318,7 +318,7 @@ vfloat64m2_t test_vfadd_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vfadd_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfadd.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfadd.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfadd_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, vfloat64m2_t op1, double op2, size_t vl) {
@@ -328,7 +328,7 @@ vfloat64m2_t test_vfadd_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vfadd_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfadd.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfadd.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfadd_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, vfloat64m4_t op1, vfloat64m4_t op2, size_t vl) {
@@ -338,7 +338,7 @@ vfloat64m4_t test_vfadd_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vfadd_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfadd.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfadd.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfadd_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, vfloat64m4_t op1, double op2, size_t vl) {
@@ -348,7 +348,7 @@ vfloat64m4_t test_vfadd_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vfadd_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfadd.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfadd.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfadd_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff, vfloat64m8_t op1, vfloat64m8_t op2, size_t vl) {
@@ -358,7 +358,7 @@ vfloat64m8_t test_vfadd_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vfadd_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfadd.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfadd.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfadd_vf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff, vfloat64m8_t op1, double op2, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfclass.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfclass.c
index 927a459..131182f 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfclass.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfclass.c
@@ -98,7 +98,7 @@ vuint64m8_t test_vfclass_v_u64m8(vfloat64m8_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfclass_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vfclass.mask.nxv1f32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vfclass.mask.nxv1f32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vfclass_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -109,7 +109,7 @@ vuint32mf2_t test_vfclass_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfclass_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vfclass.mask.nxv2f32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vfclass.mask.nxv2f32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vfclass_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -120,7 +120,7 @@ vuint32m1_t test_vfclass_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfclass_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vfclass.mask.nxv4f32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vfclass.mask.nxv4f32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vfclass_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -131,7 +131,7 @@ vuint32m2_t test_vfclass_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfclass_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vfclass.mask.nxv8f32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vfclass.mask.nxv8f32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vfclass_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -142,7 +142,7 @@ vuint32m4_t test_vfclass_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfclass_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vfclass.mask.nxv16f32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vfclass.mask.nxv16f32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vfclass_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -153,7 +153,7 @@ vuint32m8_t test_vfclass_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfclass_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vfclass.mask.nxv1f64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vfclass.mask.nxv1f64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vfclass_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -164,7 +164,7 @@ vuint64m1_t test_vfclass_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfclass_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vfclass.mask.nxv2f64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vfclass.mask.nxv2f64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vfclass_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -175,7 +175,7 @@ vuint64m2_t test_vfclass_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfclass_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vfclass.mask.nxv4f64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vfclass.mask.nxv4f64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vfclass_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -186,7 +186,7 @@ vuint64m4_t test_vfclass_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfclass_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vfclass.mask.nxv8f64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vfclass.mask.nxv8f64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vfclass_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfcvt.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfcvt.c
index 14f9dee..bd50cad 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfcvt.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfcvt.c
@@ -548,7 +548,7 @@ vfloat64m8_t test_vfcvt_f_xu_v_f64m8(vuint64m8_t src, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfcvt_x_f_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vfcvt.x.f.v.mask.nxv1i32.nxv1f32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vfcvt.x.f.v.mask.nxv1i32.nxv1f32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vfcvt_x_f_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -559,7 +559,7 @@ vint32mf2_t test_vfcvt_x_f_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_rtz_x_f_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vfcvt.rtz.x.f.v.mask.nxv1i32.nxv1f32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vfcvt.rtz.x.f.v.mask.nxv1i32.nxv1f32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vfcvt_rtz_x_f_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -570,7 +570,7 @@ vint32mf2_t test_vfcvt_rtz_x_f_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_x_f_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vfcvt.x.f.v.mask.nxv2i32.nxv2f32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vfcvt.x.f.v.mask.nxv2i32.nxv2f32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vfcvt_x_f_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -581,7 +581,7 @@ vint32m1_t test_vfcvt_x_f_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_rtz_x_f_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vfcvt.rtz.x.f.v.mask.nxv2i32.nxv2f32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vfcvt.rtz.x.f.v.mask.nxv2i32.nxv2f32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vfcvt_rtz_x_f_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -592,7 +592,7 @@ vint32m1_t test_vfcvt_rtz_x_f_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_x_f_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vfcvt.x.f.v.mask.nxv4i32.nxv4f32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vfcvt.x.f.v.mask.nxv4i32.nxv4f32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vfcvt_x_f_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -603,7 +603,7 @@ vint32m2_t test_vfcvt_x_f_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_rtz_x_f_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vfcvt.rtz.x.f.v.mask.nxv4i32.nxv4f32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vfcvt.rtz.x.f.v.mask.nxv4i32.nxv4f32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vfcvt_rtz_x_f_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -614,7 +614,7 @@ vint32m2_t test_vfcvt_rtz_x_f_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_x_f_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vfcvt.x.f.v.mask.nxv8i32.nxv8f32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vfcvt.x.f.v.mask.nxv8i32.nxv8f32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vfcvt_x_f_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -625,7 +625,7 @@ vint32m4_t test_vfcvt_x_f_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_rtz_x_f_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vfcvt.rtz.x.f.v.mask.nxv8i32.nxv8f32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vfcvt.rtz.x.f.v.mask.nxv8i32.nxv8f32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vfcvt_rtz_x_f_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -636,7 +636,7 @@ vint32m4_t test_vfcvt_rtz_x_f_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_x_f_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vfcvt.x.f.v.mask.nxv16i32.nxv16f32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vfcvt.x.f.v.mask.nxv16i32.nxv16f32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vfcvt_x_f_v_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -647,7 +647,7 @@ vint32m8_t test_vfcvt_x_f_v_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_rtz_x_f_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vfcvt.rtz.x.f.v.mask.nxv16i32.nxv16f32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vfcvt.rtz.x.f.v.mask.nxv16i32.nxv16f32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vfcvt_rtz_x_f_v_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -658,7 +658,7 @@ vint32m8_t test_vfcvt_rtz_x_f_v_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_xu_f_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vfcvt.xu.f.v.mask.nxv1i32.nxv1f32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vfcvt.xu.f.v.mask.nxv1i32.nxv1f32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vfcvt_xu_f_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -669,7 +669,7 @@ vuint32mf2_t test_vfcvt_xu_f_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_rtz_xu_f_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vfcvt.rtz.xu.f.v.mask.nxv1i32.nxv1f32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vfcvt.rtz.xu.f.v.mask.nxv1i32.nxv1f32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vfcvt_rtz_xu_f_v_u32mf2_m(vbool64_t mask,
@@ -681,7 +681,7 @@ vuint32mf2_t test_vfcvt_rtz_xu_f_v_u32mf2_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_xu_f_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vfcvt.xu.f.v.mask.nxv2i32.nxv2f32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vfcvt.xu.f.v.mask.nxv2i32.nxv2f32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vfcvt_xu_f_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -692,7 +692,7 @@ vuint32m1_t test_vfcvt_xu_f_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_rtz_xu_f_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vfcvt.rtz.xu.f.v.mask.nxv2i32.nxv2f32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vfcvt.rtz.xu.f.v.mask.nxv2i32.nxv2f32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vfcvt_rtz_xu_f_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -703,7 +703,7 @@ vuint32m1_t test_vfcvt_rtz_xu_f_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_xu_f_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vfcvt.xu.f.v.mask.nxv4i32.nxv4f32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vfcvt.xu.f.v.mask.nxv4i32.nxv4f32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vfcvt_xu_f_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -714,7 +714,7 @@ vuint32m2_t test_vfcvt_xu_f_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_rtz_xu_f_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vfcvt.rtz.xu.f.v.mask.nxv4i32.nxv4f32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vfcvt.rtz.xu.f.v.mask.nxv4i32.nxv4f32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vfcvt_rtz_xu_f_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -725,7 +725,7 @@ vuint32m2_t test_vfcvt_rtz_xu_f_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_xu_f_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vfcvt.xu.f.v.mask.nxv8i32.nxv8f32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vfcvt.xu.f.v.mask.nxv8i32.nxv8f32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vfcvt_xu_f_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -736,7 +736,7 @@ vuint32m4_t test_vfcvt_xu_f_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_rtz_xu_f_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vfcvt.rtz.xu.f.v.mask.nxv8i32.nxv8f32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vfcvt.rtz.xu.f.v.mask.nxv8i32.nxv8f32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vfcvt_rtz_xu_f_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -747,7 +747,7 @@ vuint32m4_t test_vfcvt_rtz_xu_f_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_xu_f_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vfcvt.xu.f.v.mask.nxv16i32.nxv16f32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vfcvt.xu.f.v.mask.nxv16i32.nxv16f32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vfcvt_xu_f_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -758,7 +758,7 @@ vuint32m8_t test_vfcvt_xu_f_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_rtz_xu_f_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vfcvt.rtz.xu.f.v.mask.nxv16i32.nxv16f32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vfcvt.rtz.xu.f.v.mask.nxv16i32.nxv16f32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vfcvt_rtz_xu_f_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -769,7 +769,7 @@ vuint32m8_t test_vfcvt_rtz_xu_f_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_f_x_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfcvt.f.x.v.mask.nxv1f32.nxv1i32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfcvt.f.x.v.mask.nxv1f32.nxv1i32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfcvt_f_x_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -780,7 +780,7 @@ vfloat32mf2_t test_vfcvt_f_x_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_f_x_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfcvt.f.x.v.mask.nxv2f32.nxv2i32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfcvt.f.x.v.mask.nxv2f32.nxv2i32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfcvt_f_x_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -791,7 +791,7 @@ vfloat32m1_t test_vfcvt_f_x_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_f_x_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfcvt.f.x.v.mask.nxv4f32.nxv4i32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfcvt.f.x.v.mask.nxv4f32.nxv4i32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfcvt_f_x_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -802,7 +802,7 @@ vfloat32m2_t test_vfcvt_f_x_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_f_x_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfcvt.f.x.v.mask.nxv8f32.nxv8i32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfcvt.f.x.v.mask.nxv8f32.nxv8i32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfcvt_f_x_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -813,7 +813,7 @@ vfloat32m4_t test_vfcvt_f_x_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_f_x_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfcvt.f.x.v.mask.nxv16f32.nxv16i32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfcvt.f.x.v.mask.nxv16f32.nxv16i32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfcvt_f_x_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -824,7 +824,7 @@ vfloat32m8_t test_vfcvt_f_x_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_f_xu_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfcvt.f.xu.v.mask.nxv1f32.nxv1i32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfcvt.f.xu.v.mask.nxv1f32.nxv1i32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfcvt_f_xu_v_f32mf2_m(vbool64_t mask,
@@ -836,7 +836,7 @@ vfloat32mf2_t test_vfcvt_f_xu_v_f32mf2_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_f_xu_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfcvt.f.xu.v.mask.nxv2f32.nxv2i32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfcvt.f.xu.v.mask.nxv2f32.nxv2i32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfcvt_f_xu_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -847,7 +847,7 @@ vfloat32m1_t test_vfcvt_f_xu_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_f_xu_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfcvt.f.xu.v.mask.nxv4f32.nxv4i32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfcvt.f.xu.v.mask.nxv4f32.nxv4i32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfcvt_f_xu_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -858,7 +858,7 @@ vfloat32m2_t test_vfcvt_f_xu_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_f_xu_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfcvt.f.xu.v.mask.nxv8f32.nxv8i32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfcvt.f.xu.v.mask.nxv8f32.nxv8i32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfcvt_f_xu_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -869,7 +869,7 @@ vfloat32m4_t test_vfcvt_f_xu_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_f_xu_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfcvt.f.xu.v.mask.nxv16f32.nxv16i32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfcvt.f.xu.v.mask.nxv16f32.nxv16i32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfcvt_f_xu_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -880,7 +880,7 @@ vfloat32m8_t test_vfcvt_f_xu_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_x_f_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vfcvt.x.f.v.mask.nxv1i64.nxv1f64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vfcvt.x.f.v.mask.nxv1i64.nxv1f64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vfcvt_x_f_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -891,7 +891,7 @@ vint64m1_t test_vfcvt_x_f_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_rtz_x_f_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vfcvt.rtz.x.f.v.mask.nxv1i64.nxv1f64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vfcvt.rtz.x.f.v.mask.nxv1i64.nxv1f64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vfcvt_rtz_x_f_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -902,7 +902,7 @@ vint64m1_t test_vfcvt_rtz_x_f_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_x_f_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vfcvt.x.f.v.mask.nxv2i64.nxv2f64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vfcvt.x.f.v.mask.nxv2i64.nxv2f64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vfcvt_x_f_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -913,7 +913,7 @@ vint64m2_t test_vfcvt_x_f_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_rtz_x_f_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vfcvt.rtz.x.f.v.mask.nxv2i64.nxv2f64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vfcvt.rtz.x.f.v.mask.nxv2i64.nxv2f64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vfcvt_rtz_x_f_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -924,7 +924,7 @@ vint64m2_t test_vfcvt_rtz_x_f_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_x_f_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vfcvt.x.f.v.mask.nxv4i64.nxv4f64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vfcvt.x.f.v.mask.nxv4i64.nxv4f64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vfcvt_x_f_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -935,7 +935,7 @@ vint64m4_t test_vfcvt_x_f_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_rtz_x_f_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vfcvt.rtz.x.f.v.mask.nxv4i64.nxv4f64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vfcvt.rtz.x.f.v.mask.nxv4i64.nxv4f64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vfcvt_rtz_x_f_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -946,7 +946,7 @@ vint64m4_t test_vfcvt_rtz_x_f_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_x_f_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vfcvt.x.f.v.mask.nxv8i64.nxv8f64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vfcvt.x.f.v.mask.nxv8i64.nxv8f64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vfcvt_x_f_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -957,7 +957,7 @@ vint64m8_t test_vfcvt_x_f_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_rtz_x_f_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vfcvt.rtz.x.f.v.mask.nxv8i64.nxv8f64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vfcvt.rtz.x.f.v.mask.nxv8i64.nxv8f64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vfcvt_rtz_x_f_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -968,7 +968,7 @@ vint64m8_t test_vfcvt_rtz_x_f_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_xu_f_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vfcvt.xu.f.v.mask.nxv1i64.nxv1f64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vfcvt.xu.f.v.mask.nxv1i64.nxv1f64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vfcvt_xu_f_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -979,7 +979,7 @@ vuint64m1_t test_vfcvt_xu_f_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_rtz_xu_f_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vfcvt.rtz.xu.f.v.mask.nxv1i64.nxv1f64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vfcvt.rtz.xu.f.v.mask.nxv1i64.nxv1f64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vfcvt_rtz_xu_f_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -990,7 +990,7 @@ vuint64m1_t test_vfcvt_rtz_xu_f_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_xu_f_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vfcvt.xu.f.v.mask.nxv2i64.nxv2f64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vfcvt.xu.f.v.mask.nxv2i64.nxv2f64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vfcvt_xu_f_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -1001,7 +1001,7 @@ vuint64m2_t test_vfcvt_xu_f_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_rtz_xu_f_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vfcvt.rtz.xu.f.v.mask.nxv2i64.nxv2f64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vfcvt.rtz.xu.f.v.mask.nxv2i64.nxv2f64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vfcvt_rtz_xu_f_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -1012,7 +1012,7 @@ vuint64m2_t test_vfcvt_rtz_xu_f_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_xu_f_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vfcvt.xu.f.v.mask.nxv4i64.nxv4f64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vfcvt.xu.f.v.mask.nxv4i64.nxv4f64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vfcvt_xu_f_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -1023,7 +1023,7 @@ vuint64m4_t test_vfcvt_xu_f_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_rtz_xu_f_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vfcvt.rtz.xu.f.v.mask.nxv4i64.nxv4f64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vfcvt.rtz.xu.f.v.mask.nxv4i64.nxv4f64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vfcvt_rtz_xu_f_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -1034,7 +1034,7 @@ vuint64m4_t test_vfcvt_rtz_xu_f_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_xu_f_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vfcvt.xu.f.v.mask.nxv8i64.nxv8f64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vfcvt.xu.f.v.mask.nxv8i64.nxv8f64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vfcvt_xu_f_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
@@ -1045,7 +1045,7 @@ vuint64m8_t test_vfcvt_xu_f_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_rtz_xu_f_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vfcvt.rtz.xu.f.v.mask.nxv8i64.nxv8f64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vfcvt.rtz.xu.f.v.mask.nxv8i64.nxv8f64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vfcvt_rtz_xu_f_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
@@ -1056,7 +1056,7 @@ vuint64m8_t test_vfcvt_rtz_xu_f_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_f_x_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfcvt.f.x.v.mask.nxv1f64.nxv1i64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfcvt.f.x.v.mask.nxv1f64.nxv1i64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfcvt_f_x_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -1067,7 +1067,7 @@ vfloat64m1_t test_vfcvt_f_x_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_f_x_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfcvt.f.x.v.mask.nxv2f64.nxv2i64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfcvt.f.x.v.mask.nxv2f64.nxv2i64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfcvt_f_x_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -1078,7 +1078,7 @@ vfloat64m2_t test_vfcvt_f_x_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_f_x_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfcvt.f.x.v.mask.nxv4f64.nxv4i64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfcvt.f.x.v.mask.nxv4f64.nxv4i64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfcvt_f_x_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -1089,7 +1089,7 @@ vfloat64m4_t test_vfcvt_f_x_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_f_x_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfcvt.f.x.v.mask.nxv8f64.nxv8i64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfcvt.f.x.v.mask.nxv8f64.nxv8i64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfcvt_f_x_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -1100,7 +1100,7 @@ vfloat64m8_t test_vfcvt_f_x_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_f_xu_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfcvt.f.xu.v.mask.nxv1f64.nxv1i64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfcvt.f.xu.v.mask.nxv1f64.nxv1i64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfcvt_f_xu_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -1111,7 +1111,7 @@ vfloat64m1_t test_vfcvt_f_xu_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_f_xu_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfcvt.f.xu.v.mask.nxv2f64.nxv2i64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfcvt.f.xu.v.mask.nxv2f64.nxv2i64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfcvt_f_xu_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -1122,7 +1122,7 @@ vfloat64m2_t test_vfcvt_f_xu_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_f_xu_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfcvt.f.xu.v.mask.nxv4f64.nxv4i64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfcvt.f.xu.v.mask.nxv4f64.nxv4i64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfcvt_f_xu_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -1133,7 +1133,7 @@ vfloat64m4_t test_vfcvt_f_xu_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfcvt_f_xu_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfcvt.f.xu.v.mask.nxv8f64.nxv8i64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfcvt.f.xu.v.mask.nxv8f64.nxv8i64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfcvt_f_xu_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfdiv.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfdiv.c
index 4a2b144..5fdf00e 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfdiv.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfdiv.c
@@ -197,7 +197,7 @@ vfloat64m8_t test_vfdiv_vf_f64m8(vfloat64m8_t op1, double op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfdiv_vv_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfdiv.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfdiv.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfdiv_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -209,7 +209,7 @@ vfloat32mf2_t test_vfdiv_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfdiv_vf_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfdiv.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfdiv.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfdiv_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -220,7 +220,7 @@ vfloat32mf2_t test_vfdiv_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfdiv_vv_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfdiv.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfdiv.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfdiv_vv_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -232,7 +232,7 @@ vfloat32m1_t test_vfdiv_vv_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfdiv_vf_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfdiv.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfdiv.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfdiv_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -243,7 +243,7 @@ vfloat32m1_t test_vfdiv_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfdiv_vv_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfdiv.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfdiv.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfdiv_vv_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -255,7 +255,7 @@ vfloat32m2_t test_vfdiv_vv_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfdiv_vf_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfdiv.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfdiv.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfdiv_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -266,7 +266,7 @@ vfloat32m2_t test_vfdiv_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfdiv_vv_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfdiv.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfdiv.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfdiv_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -278,7 +278,7 @@ vfloat32m4_t test_vfdiv_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfdiv_vf_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfdiv.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfdiv.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfdiv_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -289,7 +289,7 @@ vfloat32m4_t test_vfdiv_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfdiv_vv_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfdiv.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfdiv.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfdiv_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -301,7 +301,7 @@ vfloat32m8_t test_vfdiv_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfdiv_vf_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfdiv.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfdiv.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfdiv_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -312,7 +312,7 @@ vfloat32m8_t test_vfdiv_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfdiv_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfdiv.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfdiv.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfdiv_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -324,7 +324,7 @@ vfloat64m1_t test_vfdiv_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfdiv_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfdiv.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfdiv.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfdiv_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -335,7 +335,7 @@ vfloat64m1_t test_vfdiv_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfdiv_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfdiv.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfdiv.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfdiv_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -347,7 +347,7 @@ vfloat64m2_t test_vfdiv_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfdiv_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfdiv.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfdiv.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfdiv_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -358,7 +358,7 @@ vfloat64m2_t test_vfdiv_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfdiv_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfdiv.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfdiv.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfdiv_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -370,7 +370,7 @@ vfloat64m4_t test_vfdiv_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfdiv_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfdiv.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfdiv.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfdiv_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -381,7 +381,7 @@ vfloat64m4_t test_vfdiv_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfdiv_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfdiv.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfdiv.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfdiv_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -393,7 +393,7 @@ vfloat64m8_t test_vfdiv_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfdiv_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfdiv.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfdiv.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfdiv_vf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfirst.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfirst.c
index 5005d4b..70d2264 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfirst.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfirst.c
@@ -63,7 +63,7 @@ long test_vfirst_m_b64(vbool64_t op1, size_t vl) { return vfirst(op1, vl); }
 //
 // CHECK-RV64-LABEL: @test_vfirst_m_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vfirst.mask.nxv64i1.i64(<vscale x 64 x i1> [[OP1:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vfirst.mask.nxv64i1.i64(<vscale x 64 x i1> [[OP1:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret i64 [[TMP0]]
 //
 long test_vfirst_m_b1_m(vbool1_t mask, vbool1_t op1, size_t vl) {
@@ -73,7 +73,7 @@ long test_vfirst_m_b1_m(vbool1_t mask, vbool1_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfirst_m_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vfirst.mask.nxv32i1.i64(<vscale x 32 x i1> [[OP1:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vfirst.mask.nxv32i1.i64(<vscale x 32 x i1> [[OP1:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret i64 [[TMP0]]
 //
 long test_vfirst_m_b2_m(vbool2_t mask, vbool2_t op1, size_t vl) {
@@ -83,7 +83,7 @@ long test_vfirst_m_b2_m(vbool2_t mask, vbool2_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfirst_m_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vfirst.mask.nxv16i1.i64(<vscale x 16 x i1> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vfirst.mask.nxv16i1.i64(<vscale x 16 x i1> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret i64 [[TMP0]]
 //
 long test_vfirst_m_b4_m(vbool4_t mask, vbool4_t op1, size_t vl) {
@@ -93,7 +93,7 @@ long test_vfirst_m_b4_m(vbool4_t mask, vbool4_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfirst_m_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vfirst.mask.nxv8i1.i64(<vscale x 8 x i1> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vfirst.mask.nxv8i1.i64(<vscale x 8 x i1> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret i64 [[TMP0]]
 //
 long test_vfirst_m_b8_m(vbool8_t mask, vbool8_t op1, size_t vl) {
@@ -103,7 +103,7 @@ long test_vfirst_m_b8_m(vbool8_t mask, vbool8_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfirst_m_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vfirst.mask.nxv4i1.i64(<vscale x 4 x i1> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vfirst.mask.nxv4i1.i64(<vscale x 4 x i1> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret i64 [[TMP0]]
 //
 long test_vfirst_m_b16_m(vbool16_t mask, vbool16_t op1, size_t vl) {
@@ -113,7 +113,7 @@ long test_vfirst_m_b16_m(vbool16_t mask, vbool16_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfirst_m_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vfirst.mask.nxv2i1.i64(<vscale x 2 x i1> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vfirst.mask.nxv2i1.i64(<vscale x 2 x i1> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret i64 [[TMP0]]
 //
 long test_vfirst_m_b32_m(vbool32_t mask, vbool32_t op1, size_t vl) {
@@ -123,7 +123,7 @@ long test_vfirst_m_b32_m(vbool32_t mask, vbool32_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfirst_m_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vfirst.mask.nxv1i1.i64(<vscale x 1 x i1> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vfirst.mask.nxv1i1.i64(<vscale x 1 x i1> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret i64 [[TMP0]]
 //
 long test_vfirst_m_b64_m(vbool64_t mask, vbool64_t op1, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmacc.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmacc.c
index de64c4c..54174fa 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmacc.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmacc.c
@@ -206,7 +206,7 @@ vfloat64m8_t test_vfmacc_vf_f64m8(vfloat64m8_t acc, double op1,
 //
 // CHECK-RV64-LABEL: @test_vfmacc_vv_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmacc.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmacc.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfmacc_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
@@ -218,7 +218,7 @@ vfloat32mf2_t test_vfmacc_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmacc_vf_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmacc.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmacc.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfmacc_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
@@ -229,7 +229,7 @@ vfloat32mf2_t test_vfmacc_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmacc_vv_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmacc.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmacc.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfmacc_vv_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
@@ -241,7 +241,7 @@ vfloat32m1_t test_vfmacc_vv_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmacc_vf_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmacc.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmacc.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfmacc_vf_f32m1_m(vbool32_t mask, vfloat32m1_t acc, float op1,
@@ -252,7 +252,7 @@ vfloat32m1_t test_vfmacc_vf_f32m1_m(vbool32_t mask, vfloat32m1_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfmacc_vv_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmacc.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmacc.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfmacc_vv_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
@@ -264,7 +264,7 @@ vfloat32m2_t test_vfmacc_vv_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmacc_vf_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmacc.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmacc.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfmacc_vf_f32m2_m(vbool16_t mask, vfloat32m2_t acc, float op1,
@@ -275,7 +275,7 @@ vfloat32m2_t test_vfmacc_vf_f32m2_m(vbool16_t mask, vfloat32m2_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfmacc_vv_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmacc.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmacc.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfmacc_vv_f32m4_m(vbool8_t mask, vfloat32m4_t acc,
@@ -287,7 +287,7 @@ vfloat32m4_t test_vfmacc_vv_f32m4_m(vbool8_t mask, vfloat32m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmacc_vf_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmacc.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmacc.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfmacc_vf_f32m4_m(vbool8_t mask, vfloat32m4_t acc, float op1,
@@ -298,7 +298,7 @@ vfloat32m4_t test_vfmacc_vf_f32m4_m(vbool8_t mask, vfloat32m4_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfmacc_vv_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmacc.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[ACC:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmacc.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[ACC:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfmacc_vv_f32m8_m(vbool4_t mask, vfloat32m8_t acc,
@@ -310,7 +310,7 @@ vfloat32m8_t test_vfmacc_vv_f32m8_m(vbool4_t mask, vfloat32m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmacc_vf_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmacc.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmacc.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfmacc_vf_f32m8_m(vbool4_t mask, vfloat32m8_t acc, float op1,
@@ -321,7 +321,7 @@ vfloat32m8_t test_vfmacc_vf_f32m8_m(vbool4_t mask, vfloat32m8_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfmacc_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmacc.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmacc.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfmacc_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -333,7 +333,7 @@ vfloat64m1_t test_vfmacc_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmacc_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmacc.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmacc.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfmacc_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -344,7 +344,7 @@ vfloat64m1_t test_vfmacc_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmacc_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmacc.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmacc.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfmacc_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -356,7 +356,7 @@ vfloat64m2_t test_vfmacc_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmacc_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmacc.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmacc.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfmacc_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -367,7 +367,7 @@ vfloat64m2_t test_vfmacc_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmacc_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmacc.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmacc.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfmacc_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -379,7 +379,7 @@ vfloat64m4_t test_vfmacc_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmacc_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmacc.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmacc.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfmacc_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -390,7 +390,7 @@ vfloat64m4_t test_vfmacc_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmacc_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmacc.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmacc.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfmacc_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
@@ -402,7 +402,7 @@ vfloat64m8_t test_vfmacc_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmacc_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmacc.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmacc.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfmacc_vf_f64m8_m(vbool8_t mask, vfloat64m8_t acc, double op1,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmadd.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmadd.c
index 43f636d..fda3ee2 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmadd.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmadd.c
@@ -206,7 +206,7 @@ vfloat64m8_t test_vfmadd_vf_f64m8(vfloat64m8_t acc, double op1,
 //
 // CHECK-RV64-LABEL: @test_vfmadd_vv_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmadd.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmadd.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfmadd_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
@@ -218,7 +218,7 @@ vfloat32mf2_t test_vfmadd_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmadd_vf_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmadd.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmadd.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfmadd_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
@@ -229,7 +229,7 @@ vfloat32mf2_t test_vfmadd_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmadd_vv_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmadd.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmadd.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfmadd_vv_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
@@ -241,7 +241,7 @@ vfloat32m1_t test_vfmadd_vv_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmadd_vf_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmadd.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmadd.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfmadd_vf_f32m1_m(vbool32_t mask, vfloat32m1_t acc, float op1,
@@ -252,7 +252,7 @@ vfloat32m1_t test_vfmadd_vf_f32m1_m(vbool32_t mask, vfloat32m1_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfmadd_vv_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmadd.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmadd.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfmadd_vv_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
@@ -264,7 +264,7 @@ vfloat32m2_t test_vfmadd_vv_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmadd_vf_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmadd.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmadd.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfmadd_vf_f32m2_m(vbool16_t mask, vfloat32m2_t acc, float op1,
@@ -275,7 +275,7 @@ vfloat32m2_t test_vfmadd_vf_f32m2_m(vbool16_t mask, vfloat32m2_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfmadd_vv_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmadd.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmadd.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfmadd_vv_f32m4_m(vbool8_t mask, vfloat32m4_t acc,
@@ -287,7 +287,7 @@ vfloat32m4_t test_vfmadd_vv_f32m4_m(vbool8_t mask, vfloat32m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmadd_vf_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmadd.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmadd.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfmadd_vf_f32m4_m(vbool8_t mask, vfloat32m4_t acc, float op1,
@@ -298,7 +298,7 @@ vfloat32m4_t test_vfmadd_vf_f32m4_m(vbool8_t mask, vfloat32m4_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfmadd_vv_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmadd.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[ACC:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmadd.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[ACC:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfmadd_vv_f32m8_m(vbool4_t mask, vfloat32m8_t acc,
@@ -310,7 +310,7 @@ vfloat32m8_t test_vfmadd_vv_f32m8_m(vbool4_t mask, vfloat32m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmadd_vf_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmadd.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmadd.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfmadd_vf_f32m8_m(vbool4_t mask, vfloat32m8_t acc, float op1,
@@ -321,7 +321,7 @@ vfloat32m8_t test_vfmadd_vf_f32m8_m(vbool4_t mask, vfloat32m8_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfmadd_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmadd.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmadd.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfmadd_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -333,7 +333,7 @@ vfloat64m1_t test_vfmadd_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmadd_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmadd.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmadd.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfmadd_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -344,7 +344,7 @@ vfloat64m1_t test_vfmadd_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmadd_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmadd.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmadd.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfmadd_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -356,7 +356,7 @@ vfloat64m2_t test_vfmadd_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmadd_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmadd.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmadd.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfmadd_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -367,7 +367,7 @@ vfloat64m2_t test_vfmadd_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmadd_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmadd.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmadd.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfmadd_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -379,7 +379,7 @@ vfloat64m4_t test_vfmadd_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmadd_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmadd.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmadd.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfmadd_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -390,7 +390,7 @@ vfloat64m4_t test_vfmadd_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmadd_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmadd.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmadd.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfmadd_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
@@ -402,7 +402,7 @@ vfloat64m8_t test_vfmadd_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmadd_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmadd.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmadd.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfmadd_vf_f64m8_m(vbool8_t mask, vfloat64m8_t acc, double op1,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmax.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmax.c
index 50ea952..c5c516a 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmax.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmax.c
@@ -197,7 +197,7 @@ vfloat64m8_t test_vfmax_vf_f64m8(vfloat64m8_t op1, double op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfmax_vv_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmax.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmax.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfmax_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -209,7 +209,7 @@ vfloat32mf2_t test_vfmax_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmax_vf_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmax.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmax.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfmax_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -220,7 +220,7 @@ vfloat32mf2_t test_vfmax_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmax_vv_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmax.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmax.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfmax_vv_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -232,7 +232,7 @@ vfloat32m1_t test_vfmax_vv_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmax_vf_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmax.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmax.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfmax_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -243,7 +243,7 @@ vfloat32m1_t test_vfmax_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmax_vv_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmax.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmax.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfmax_vv_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -255,7 +255,7 @@ vfloat32m2_t test_vfmax_vv_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmax_vf_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmax.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmax.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfmax_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -266,7 +266,7 @@ vfloat32m2_t test_vfmax_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmax_vv_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmax.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmax.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfmax_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -278,7 +278,7 @@ vfloat32m4_t test_vfmax_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmax_vf_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmax.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmax.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfmax_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -289,7 +289,7 @@ vfloat32m4_t test_vfmax_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmax_vv_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmax.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmax.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfmax_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -301,7 +301,7 @@ vfloat32m8_t test_vfmax_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmax_vf_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmax.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmax.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfmax_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -312,7 +312,7 @@ vfloat32m8_t test_vfmax_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmax_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmax.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmax.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfmax_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -324,7 +324,7 @@ vfloat64m1_t test_vfmax_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmax_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmax.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmax.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfmax_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -335,7 +335,7 @@ vfloat64m1_t test_vfmax_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmax_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmax.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmax.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfmax_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -347,7 +347,7 @@ vfloat64m2_t test_vfmax_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmax_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmax.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmax.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfmax_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -358,7 +358,7 @@ vfloat64m2_t test_vfmax_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmax_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmax.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmax.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfmax_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -370,7 +370,7 @@ vfloat64m4_t test_vfmax_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmax_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmax.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmax.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfmax_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -381,7 +381,7 @@ vfloat64m4_t test_vfmax_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmax_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmax.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmax.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfmax_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -393,7 +393,7 @@ vfloat64m8_t test_vfmax_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmax_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmax.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmax.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfmax_vf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmerge.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmerge.c
index cfeb02c..09bce7c 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmerge.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmerge.c
@@ -8,7 +8,7 @@
 //
 // CHECK-RV64-LABEL: @test_vfmerge_vfm_f32mf2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmerge.nxv1f32.f32.i64(<vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmerge.nxv1f32.f32.i64(<vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfmerge_vfm_f32mf2(vbool64_t mask, vfloat32mf2_t op1,
@@ -19,7 +19,7 @@ vfloat32mf2_t test_vfmerge_vfm_f32mf2(vbool64_t mask, vfloat32mf2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vfmerge_vfm_f32m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmerge.nxv2f32.f32.i64(<vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmerge.nxv2f32.f32.i64(<vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfmerge_vfm_f32m1(vbool32_t mask, vfloat32m1_t op1, float op2,
@@ -30,7 +30,7 @@ vfloat32m1_t test_vfmerge_vfm_f32m1(vbool32_t mask, vfloat32m1_t op1, float op2,
 //
 // CHECK-RV64-LABEL: @test_vfmerge_vfm_f32m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmerge.nxv4f32.f32.i64(<vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmerge.nxv4f32.f32.i64(<vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfmerge_vfm_f32m2(vbool16_t mask, vfloat32m2_t op1, float op2,
@@ -41,7 +41,7 @@ vfloat32m2_t test_vfmerge_vfm_f32m2(vbool16_t mask, vfloat32m2_t op1, float op2,
 //
 // CHECK-RV64-LABEL: @test_vfmerge_vfm_f32m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmerge.nxv8f32.f32.i64(<vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmerge.nxv8f32.f32.i64(<vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfmerge_vfm_f32m4(vbool8_t mask, vfloat32m4_t op1, float op2,
@@ -52,7 +52,7 @@ vfloat32m4_t test_vfmerge_vfm_f32m4(vbool8_t mask, vfloat32m4_t op1, float op2,
 //
 // CHECK-RV64-LABEL: @test_vfmerge_vfm_f32m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmerge.nxv16f32.f32.i64(<vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmerge.nxv16f32.f32.i64(<vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfmerge_vfm_f32m8(vbool4_t mask, vfloat32m8_t op1, float op2,
@@ -63,7 +63,7 @@ vfloat32m8_t test_vfmerge_vfm_f32m8(vbool4_t mask, vfloat32m8_t op1, float op2,
 //
 // CHECK-RV64-LABEL: @test_vfmerge_vfm_f64m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmerge.nxv1f64.f64.i64(<vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmerge.nxv1f64.f64.i64(<vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfmerge_vfm_f64m1(vbool64_t mask, vfloat64m1_t op1,
@@ -74,7 +74,7 @@ vfloat64m1_t test_vfmerge_vfm_f64m1(vbool64_t mask, vfloat64m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vfmerge_vfm_f64m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmerge.nxv2f64.f64.i64(<vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmerge.nxv2f64.f64.i64(<vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfmerge_vfm_f64m2(vbool32_t mask, vfloat64m2_t op1,
@@ -85,7 +85,7 @@ vfloat64m2_t test_vfmerge_vfm_f64m2(vbool32_t mask, vfloat64m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vfmerge_vfm_f64m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmerge.nxv4f64.f64.i64(<vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmerge.nxv4f64.f64.i64(<vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfmerge_vfm_f64m4(vbool16_t mask, vfloat64m4_t op1,
@@ -96,7 +96,7 @@ vfloat64m4_t test_vfmerge_vfm_f64m4(vbool16_t mask, vfloat64m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vfmerge_vfm_f64m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmerge.nxv8f64.f64.i64(<vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmerge.nxv8f64.f64.i64(<vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfmerge_vfm_f64m8(vbool8_t mask, vfloat64m8_t op1, double op2,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmin.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmin.c
index ce48003..46ceaca 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmin.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmin.c
@@ -197,7 +197,7 @@ vfloat64m8_t test_vfmin_vf_f64m8(vfloat64m8_t op1, double op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfmin_vv_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmin.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmin.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfmin_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -209,7 +209,7 @@ vfloat32mf2_t test_vfmin_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmin_vf_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmin.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmin.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfmin_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -220,7 +220,7 @@ vfloat32mf2_t test_vfmin_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmin_vv_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmin.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmin.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfmin_vv_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -232,7 +232,7 @@ vfloat32m1_t test_vfmin_vv_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmin_vf_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmin.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmin.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfmin_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -243,7 +243,7 @@ vfloat32m1_t test_vfmin_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmin_vv_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmin.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmin.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfmin_vv_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -255,7 +255,7 @@ vfloat32m2_t test_vfmin_vv_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmin_vf_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmin.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmin.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfmin_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -266,7 +266,7 @@ vfloat32m2_t test_vfmin_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmin_vv_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmin.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmin.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfmin_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -278,7 +278,7 @@ vfloat32m4_t test_vfmin_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmin_vf_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmin.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmin.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfmin_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -289,7 +289,7 @@ vfloat32m4_t test_vfmin_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmin_vv_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmin.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmin.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfmin_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -301,7 +301,7 @@ vfloat32m8_t test_vfmin_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmin_vf_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmin.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmin.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfmin_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -312,7 +312,7 @@ vfloat32m8_t test_vfmin_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmin_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmin.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmin.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfmin_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -324,7 +324,7 @@ vfloat64m1_t test_vfmin_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmin_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmin.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmin.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfmin_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -335,7 +335,7 @@ vfloat64m1_t test_vfmin_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmin_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmin.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmin.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfmin_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -347,7 +347,7 @@ vfloat64m2_t test_vfmin_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmin_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmin.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmin.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfmin_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -358,7 +358,7 @@ vfloat64m2_t test_vfmin_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmin_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmin.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmin.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfmin_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -370,7 +370,7 @@ vfloat64m4_t test_vfmin_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmin_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmin.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmin.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfmin_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -381,7 +381,7 @@ vfloat64m4_t test_vfmin_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmin_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmin.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmin.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfmin_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -393,7 +393,7 @@ vfloat64m8_t test_vfmin_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmin_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmin.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmin.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfmin_vf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmsac.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmsac.c
index 6af54aa..604da9b 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmsac.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmsac.c
@@ -206,7 +206,7 @@ vfloat64m8_t test_vfmsac_vf_f64m8(vfloat64m8_t acc, double op1,
 //
 // CHECK-RV64-LABEL: @test_vfmsac_vv_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmsac.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmsac.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfmsac_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
@@ -218,7 +218,7 @@ vfloat32mf2_t test_vfmsac_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsac_vf_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmsac.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmsac.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfmsac_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
@@ -229,7 +229,7 @@ vfloat32mf2_t test_vfmsac_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsac_vv_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmsac.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmsac.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfmsac_vv_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
@@ -241,7 +241,7 @@ vfloat32m1_t test_vfmsac_vv_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsac_vf_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmsac.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmsac.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfmsac_vf_f32m1_m(vbool32_t mask, vfloat32m1_t acc, float op1,
@@ -252,7 +252,7 @@ vfloat32m1_t test_vfmsac_vf_f32m1_m(vbool32_t mask, vfloat32m1_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfmsac_vv_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmsac.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmsac.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfmsac_vv_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
@@ -264,7 +264,7 @@ vfloat32m2_t test_vfmsac_vv_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsac_vf_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmsac.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmsac.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfmsac_vf_f32m2_m(vbool16_t mask, vfloat32m2_t acc, float op1,
@@ -275,7 +275,7 @@ vfloat32m2_t test_vfmsac_vf_f32m2_m(vbool16_t mask, vfloat32m2_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfmsac_vv_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmsac.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmsac.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfmsac_vv_f32m4_m(vbool8_t mask, vfloat32m4_t acc,
@@ -287,7 +287,7 @@ vfloat32m4_t test_vfmsac_vv_f32m4_m(vbool8_t mask, vfloat32m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsac_vf_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmsac.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmsac.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfmsac_vf_f32m4_m(vbool8_t mask, vfloat32m4_t acc, float op1,
@@ -298,7 +298,7 @@ vfloat32m4_t test_vfmsac_vf_f32m4_m(vbool8_t mask, vfloat32m4_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfmsac_vv_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmsac.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[ACC:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmsac.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[ACC:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfmsac_vv_f32m8_m(vbool4_t mask, vfloat32m8_t acc,
@@ -310,7 +310,7 @@ vfloat32m8_t test_vfmsac_vv_f32m8_m(vbool4_t mask, vfloat32m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsac_vf_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmsac.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmsac.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfmsac_vf_f32m8_m(vbool4_t mask, vfloat32m8_t acc, float op1,
@@ -321,7 +321,7 @@ vfloat32m8_t test_vfmsac_vf_f32m8_m(vbool4_t mask, vfloat32m8_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfmsac_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmsac.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmsac.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfmsac_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -333,7 +333,7 @@ vfloat64m1_t test_vfmsac_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsac_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmsac.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmsac.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfmsac_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -344,7 +344,7 @@ vfloat64m1_t test_vfmsac_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsac_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmsac.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmsac.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfmsac_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -356,7 +356,7 @@ vfloat64m2_t test_vfmsac_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsac_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmsac.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmsac.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfmsac_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -367,7 +367,7 @@ vfloat64m2_t test_vfmsac_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsac_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmsac.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmsac.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfmsac_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -379,7 +379,7 @@ vfloat64m4_t test_vfmsac_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsac_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmsac.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmsac.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfmsac_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -390,7 +390,7 @@ vfloat64m4_t test_vfmsac_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsac_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmsac.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmsac.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfmsac_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
@@ -402,7 +402,7 @@ vfloat64m8_t test_vfmsac_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsac_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmsac.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmsac.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfmsac_vf_f64m8_m(vbool8_t mask, vfloat64m8_t acc, double op1,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmsub.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmsub.c
index 6e01064..913fe65 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmsub.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmsub.c
@@ -206,7 +206,7 @@ vfloat64m8_t test_vfmsub_vf_f64m8(vfloat64m8_t acc, double op1,
 //
 // CHECK-RV64-LABEL: @test_vfmsub_vv_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmsub.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmsub.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfmsub_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
@@ -218,7 +218,7 @@ vfloat32mf2_t test_vfmsub_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsub_vf_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmsub.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmsub.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfmsub_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
@@ -229,7 +229,7 @@ vfloat32mf2_t test_vfmsub_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsub_vv_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmsub.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmsub.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfmsub_vv_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
@@ -241,7 +241,7 @@ vfloat32m1_t test_vfmsub_vv_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsub_vf_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmsub.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmsub.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfmsub_vf_f32m1_m(vbool32_t mask, vfloat32m1_t acc, float op1,
@@ -252,7 +252,7 @@ vfloat32m1_t test_vfmsub_vf_f32m1_m(vbool32_t mask, vfloat32m1_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfmsub_vv_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmsub.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmsub.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfmsub_vv_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
@@ -264,7 +264,7 @@ vfloat32m2_t test_vfmsub_vv_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsub_vf_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmsub.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmsub.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfmsub_vf_f32m2_m(vbool16_t mask, vfloat32m2_t acc, float op1,
@@ -275,7 +275,7 @@ vfloat32m2_t test_vfmsub_vf_f32m2_m(vbool16_t mask, vfloat32m2_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfmsub_vv_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmsub.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmsub.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfmsub_vv_f32m4_m(vbool8_t mask, vfloat32m4_t acc,
@@ -287,7 +287,7 @@ vfloat32m4_t test_vfmsub_vv_f32m4_m(vbool8_t mask, vfloat32m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsub_vf_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmsub.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmsub.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfmsub_vf_f32m4_m(vbool8_t mask, vfloat32m4_t acc, float op1,
@@ -298,7 +298,7 @@ vfloat32m4_t test_vfmsub_vf_f32m4_m(vbool8_t mask, vfloat32m4_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfmsub_vv_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmsub.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[ACC:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmsub.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[ACC:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfmsub_vv_f32m8_m(vbool4_t mask, vfloat32m8_t acc,
@@ -310,7 +310,7 @@ vfloat32m8_t test_vfmsub_vv_f32m8_m(vbool4_t mask, vfloat32m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsub_vf_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmsub.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmsub.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfmsub_vf_f32m8_m(vbool4_t mask, vfloat32m8_t acc, float op1,
@@ -321,7 +321,7 @@ vfloat32m8_t test_vfmsub_vf_f32m8_m(vbool4_t mask, vfloat32m8_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfmsub_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmsub.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmsub.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfmsub_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -333,7 +333,7 @@ vfloat64m1_t test_vfmsub_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsub_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmsub.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmsub.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfmsub_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -344,7 +344,7 @@ vfloat64m1_t test_vfmsub_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsub_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmsub.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmsub.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfmsub_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -356,7 +356,7 @@ vfloat64m2_t test_vfmsub_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsub_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmsub.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmsub.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfmsub_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -367,7 +367,7 @@ vfloat64m2_t test_vfmsub_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsub_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmsub.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmsub.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfmsub_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -379,7 +379,7 @@ vfloat64m4_t test_vfmsub_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsub_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmsub.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmsub.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfmsub_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -390,7 +390,7 @@ vfloat64m4_t test_vfmsub_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsub_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmsub.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmsub.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfmsub_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
@@ -402,7 +402,7 @@ vfloat64m8_t test_vfmsub_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfmsub_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmsub.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmsub.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfmsub_vf_f64m8_m(vbool8_t mask, vfloat64m8_t acc, double op1,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmul.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmul.c
index 98c1130..4b1d3a8 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmul.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfmul.c
@@ -197,7 +197,7 @@ vfloat64m8_t test_vfmul_vf_f64m8(vfloat64m8_t op1, double op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfmul_vv_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmul.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmul.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfmul_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -209,7 +209,7 @@ vfloat32mf2_t test_vfmul_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmul_vf_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmul.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmul.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfmul_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -220,7 +220,7 @@ vfloat32mf2_t test_vfmul_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmul_vv_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmul.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmul.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfmul_vv_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -232,7 +232,7 @@ vfloat32m1_t test_vfmul_vv_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmul_vf_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmul.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmul.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfmul_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -243,7 +243,7 @@ vfloat32m1_t test_vfmul_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmul_vv_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmul.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmul.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfmul_vv_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -255,7 +255,7 @@ vfloat32m2_t test_vfmul_vv_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmul_vf_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmul.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmul.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfmul_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -266,7 +266,7 @@ vfloat32m2_t test_vfmul_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmul_vv_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmul.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmul.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfmul_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -278,7 +278,7 @@ vfloat32m4_t test_vfmul_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmul_vf_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmul.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmul.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfmul_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -289,7 +289,7 @@ vfloat32m4_t test_vfmul_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmul_vv_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmul.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmul.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfmul_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -301,7 +301,7 @@ vfloat32m8_t test_vfmul_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmul_vf_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmul.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmul.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfmul_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -312,7 +312,7 @@ vfloat32m8_t test_vfmul_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmul_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmul.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmul.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfmul_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -324,7 +324,7 @@ vfloat64m1_t test_vfmul_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmul_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmul.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmul.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfmul_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -335,7 +335,7 @@ vfloat64m1_t test_vfmul_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmul_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmul.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmul.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfmul_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -347,7 +347,7 @@ vfloat64m2_t test_vfmul_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmul_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmul.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmul.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfmul_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -358,7 +358,7 @@ vfloat64m2_t test_vfmul_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmul_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmul.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmul.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfmul_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -370,7 +370,7 @@ vfloat64m4_t test_vfmul_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmul_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmul.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmul.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfmul_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -381,7 +381,7 @@ vfloat64m4_t test_vfmul_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmul_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmul.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmul.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfmul_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -393,7 +393,7 @@ vfloat64m8_t test_vfmul_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfmul_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmul.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmul.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfmul_vf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfncvt.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfncvt.c
index 0223b34..ef9f3fe 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfncvt.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfncvt.c
@@ -528,7 +528,7 @@ vfloat32m4_t test_vfncvt_rod_f_f_w_f32m4(vfloat64m8_t src, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfncvt_x_f_w_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vfncvt.x.f.w.mask.nxv1i16.nxv1f32.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vfncvt.x.f.w.mask.nxv1i16.nxv1f32.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vfncvt_x_f_w_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -539,7 +539,7 @@ vint16mf4_t test_vfncvt_x_f_w_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rtz_x_f_w_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vfncvt.rtz.x.f.w.mask.nxv1i16.nxv1f32.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vfncvt.rtz.x.f.w.mask.nxv1i16.nxv1f32.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vfncvt_rtz_x_f_w_i16mf4_m(vbool64_t mask,
@@ -551,7 +551,7 @@ vint16mf4_t test_vfncvt_rtz_x_f_w_i16mf4_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_x_f_w_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vfncvt.x.f.w.mask.nxv2i16.nxv2f32.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vfncvt.x.f.w.mask.nxv2i16.nxv2f32.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vfncvt_x_f_w_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -562,7 +562,7 @@ vint16mf2_t test_vfncvt_x_f_w_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rtz_x_f_w_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vfncvt.rtz.x.f.w.mask.nxv2i16.nxv2f32.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vfncvt.rtz.x.f.w.mask.nxv2i16.nxv2f32.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vfncvt_rtz_x_f_w_i16mf2_m(vbool32_t mask,
@@ -574,7 +574,7 @@ vint16mf2_t test_vfncvt_rtz_x_f_w_i16mf2_m(vbool32_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_x_f_w_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vfncvt.x.f.w.mask.nxv4i16.nxv4f32.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vfncvt.x.f.w.mask.nxv4i16.nxv4f32.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vfncvt_x_f_w_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -585,7 +585,7 @@ vint16m1_t test_vfncvt_x_f_w_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rtz_x_f_w_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vfncvt.rtz.x.f.w.mask.nxv4i16.nxv4f32.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vfncvt.rtz.x.f.w.mask.nxv4i16.nxv4f32.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vfncvt_rtz_x_f_w_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -596,7 +596,7 @@ vint16m1_t test_vfncvt_rtz_x_f_w_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_x_f_w_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vfncvt.x.f.w.mask.nxv8i16.nxv8f32.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vfncvt.x.f.w.mask.nxv8i16.nxv8f32.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vfncvt_x_f_w_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -607,7 +607,7 @@ vint16m2_t test_vfncvt_x_f_w_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rtz_x_f_w_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vfncvt.rtz.x.f.w.mask.nxv8i16.nxv8f32.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vfncvt.rtz.x.f.w.mask.nxv8i16.nxv8f32.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vfncvt_rtz_x_f_w_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -618,7 +618,7 @@ vint16m2_t test_vfncvt_rtz_x_f_w_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_x_f_w_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vfncvt.x.f.w.mask.nxv16i16.nxv16f32.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vfncvt.x.f.w.mask.nxv16i16.nxv16f32.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vfncvt_x_f_w_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -629,7 +629,7 @@ vint16m4_t test_vfncvt_x_f_w_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rtz_x_f_w_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vfncvt.rtz.x.f.w.mask.nxv16i16.nxv16f32.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vfncvt.rtz.x.f.w.mask.nxv16i16.nxv16f32.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vfncvt_rtz_x_f_w_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -640,7 +640,7 @@ vint16m4_t test_vfncvt_rtz_x_f_w_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_xu_f_w_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vfncvt.xu.f.w.mask.nxv1i16.nxv1f32.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vfncvt.xu.f.w.mask.nxv1i16.nxv1f32.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vfncvt_xu_f_w_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -651,7 +651,7 @@ vuint16mf4_t test_vfncvt_xu_f_w_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rtz_xu_f_w_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vfncvt.rtz.xu.f.w.mask.nxv1i16.nxv1f32.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vfncvt.rtz.xu.f.w.mask.nxv1i16.nxv1f32.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vfncvt_rtz_xu_f_w_u16mf4_m(vbool64_t mask,
@@ -663,7 +663,7 @@ vuint16mf4_t test_vfncvt_rtz_xu_f_w_u16mf4_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_xu_f_w_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vfncvt.xu.f.w.mask.nxv2i16.nxv2f32.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vfncvt.xu.f.w.mask.nxv2i16.nxv2f32.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vfncvt_xu_f_w_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -674,7 +674,7 @@ vuint16mf2_t test_vfncvt_xu_f_w_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rtz_xu_f_w_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vfncvt.rtz.xu.f.w.mask.nxv2i16.nxv2f32.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vfncvt.rtz.xu.f.w.mask.nxv2i16.nxv2f32.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vfncvt_rtz_xu_f_w_u16mf2_m(vbool32_t mask,
@@ -686,7 +686,7 @@ vuint16mf2_t test_vfncvt_rtz_xu_f_w_u16mf2_m(vbool32_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_xu_f_w_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vfncvt.xu.f.w.mask.nxv4i16.nxv4f32.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vfncvt.xu.f.w.mask.nxv4i16.nxv4f32.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vfncvt_xu_f_w_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -697,7 +697,7 @@ vuint16m1_t test_vfncvt_xu_f_w_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rtz_xu_f_w_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vfncvt.rtz.xu.f.w.mask.nxv4i16.nxv4f32.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vfncvt.rtz.xu.f.w.mask.nxv4i16.nxv4f32.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vfncvt_rtz_xu_f_w_u16m1_m(vbool16_t mask,
@@ -709,7 +709,7 @@ vuint16m1_t test_vfncvt_rtz_xu_f_w_u16m1_m(vbool16_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_xu_f_w_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vfncvt.xu.f.w.mask.nxv8i16.nxv8f32.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vfncvt.xu.f.w.mask.nxv8i16.nxv8f32.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vfncvt_xu_f_w_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -720,7 +720,7 @@ vuint16m2_t test_vfncvt_xu_f_w_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rtz_xu_f_w_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vfncvt.rtz.xu.f.w.mask.nxv8i16.nxv8f32.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vfncvt.rtz.xu.f.w.mask.nxv8i16.nxv8f32.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vfncvt_rtz_xu_f_w_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -731,7 +731,7 @@ vuint16m2_t test_vfncvt_rtz_xu_f_w_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_xu_f_w_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vfncvt.xu.f.w.mask.nxv16i16.nxv16f32.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vfncvt.xu.f.w.mask.nxv16i16.nxv16f32.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vfncvt_xu_f_w_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -742,7 +742,7 @@ vuint16m4_t test_vfncvt_xu_f_w_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rtz_xu_f_w_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vfncvt.rtz.xu.f.w.mask.nxv16i16.nxv16f32.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vfncvt.rtz.xu.f.w.mask.nxv16i16.nxv16f32.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vfncvt_rtz_xu_f_w_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -753,7 +753,7 @@ vuint16m4_t test_vfncvt_rtz_xu_f_w_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_x_f_w_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vfncvt.x.f.w.mask.nxv1i32.nxv1f64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vfncvt.x.f.w.mask.nxv1i32.nxv1f64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vfncvt_x_f_w_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -764,7 +764,7 @@ vint32mf2_t test_vfncvt_x_f_w_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rtz_x_f_w_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vfncvt.rtz.x.f.w.mask.nxv1i32.nxv1f64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vfncvt.rtz.x.f.w.mask.nxv1i32.nxv1f64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vfncvt_rtz_x_f_w_i32mf2_m(vbool64_t mask,
@@ -776,7 +776,7 @@ vint32mf2_t test_vfncvt_rtz_x_f_w_i32mf2_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_x_f_w_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vfncvt.x.f.w.mask.nxv2i32.nxv2f64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vfncvt.x.f.w.mask.nxv2i32.nxv2f64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vfncvt_x_f_w_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -787,7 +787,7 @@ vint32m1_t test_vfncvt_x_f_w_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rtz_x_f_w_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vfncvt.rtz.x.f.w.mask.nxv2i32.nxv2f64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vfncvt.rtz.x.f.w.mask.nxv2i32.nxv2f64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vfncvt_rtz_x_f_w_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -798,7 +798,7 @@ vint32m1_t test_vfncvt_rtz_x_f_w_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_x_f_w_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vfncvt.x.f.w.mask.nxv4i32.nxv4f64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vfncvt.x.f.w.mask.nxv4i32.nxv4f64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vfncvt_x_f_w_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -809,7 +809,7 @@ vint32m2_t test_vfncvt_x_f_w_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rtz_x_f_w_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vfncvt.rtz.x.f.w.mask.nxv4i32.nxv4f64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vfncvt.rtz.x.f.w.mask.nxv4i32.nxv4f64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vfncvt_rtz_x_f_w_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -820,7 +820,7 @@ vint32m2_t test_vfncvt_rtz_x_f_w_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_x_f_w_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vfncvt.x.f.w.mask.nxv8i32.nxv8f64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vfncvt.x.f.w.mask.nxv8i32.nxv8f64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vfncvt_x_f_w_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -831,7 +831,7 @@ vint32m4_t test_vfncvt_x_f_w_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rtz_x_f_w_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vfncvt.rtz.x.f.w.mask.nxv8i32.nxv8f64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vfncvt.rtz.x.f.w.mask.nxv8i32.nxv8f64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vfncvt_rtz_x_f_w_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -842,7 +842,7 @@ vint32m4_t test_vfncvt_rtz_x_f_w_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_xu_f_w_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vfncvt.xu.f.w.mask.nxv1i32.nxv1f64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vfncvt.xu.f.w.mask.nxv1i32.nxv1f64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vfncvt_xu_f_w_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -853,7 +853,7 @@ vuint32mf2_t test_vfncvt_xu_f_w_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rtz_xu_f_w_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vfncvt.rtz.xu.f.w.mask.nxv1i32.nxv1f64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vfncvt.rtz.xu.f.w.mask.nxv1i32.nxv1f64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vfncvt_rtz_xu_f_w_u32mf2_m(vbool64_t mask,
@@ -865,7 +865,7 @@ vuint32mf2_t test_vfncvt_rtz_xu_f_w_u32mf2_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_xu_f_w_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vfncvt.xu.f.w.mask.nxv2i32.nxv2f64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vfncvt.xu.f.w.mask.nxv2i32.nxv2f64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vfncvt_xu_f_w_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -876,7 +876,7 @@ vuint32m1_t test_vfncvt_xu_f_w_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rtz_xu_f_w_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vfncvt.rtz.xu.f.w.mask.nxv2i32.nxv2f64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vfncvt.rtz.xu.f.w.mask.nxv2i32.nxv2f64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vfncvt_rtz_xu_f_w_u32m1_m(vbool32_t mask,
@@ -888,7 +888,7 @@ vuint32m1_t test_vfncvt_rtz_xu_f_w_u32m1_m(vbool32_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_xu_f_w_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vfncvt.xu.f.w.mask.nxv4i32.nxv4f64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vfncvt.xu.f.w.mask.nxv4i32.nxv4f64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vfncvt_xu_f_w_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -899,7 +899,7 @@ vuint32m2_t test_vfncvt_xu_f_w_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rtz_xu_f_w_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vfncvt.rtz.xu.f.w.mask.nxv4i32.nxv4f64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vfncvt.rtz.xu.f.w.mask.nxv4i32.nxv4f64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vfncvt_rtz_xu_f_w_u32m2_m(vbool16_t mask,
@@ -911,7 +911,7 @@ vuint32m2_t test_vfncvt_rtz_xu_f_w_u32m2_m(vbool16_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_xu_f_w_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vfncvt.xu.f.w.mask.nxv8i32.nxv8f64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vfncvt.xu.f.w.mask.nxv8i32.nxv8f64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vfncvt_xu_f_w_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -922,7 +922,7 @@ vuint32m4_t test_vfncvt_xu_f_w_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rtz_xu_f_w_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vfncvt.rtz.xu.f.w.mask.nxv8i32.nxv8f64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vfncvt.rtz.xu.f.w.mask.nxv8i32.nxv8f64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vfncvt_rtz_xu_f_w_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -933,7 +933,7 @@ vuint32m4_t test_vfncvt_rtz_xu_f_w_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_f_x_w_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfncvt.f.x.w.mask.nxv1f32.nxv1i64.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfncvt.f.x.w.mask.nxv1f32.nxv1i64.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfncvt_f_x_w_f32mf2_m(vbool64_t mask,
@@ -945,7 +945,7 @@ vfloat32mf2_t test_vfncvt_f_x_w_f32mf2_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_f_x_w_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfncvt.f.x.w.mask.nxv2f32.nxv2i64.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfncvt.f.x.w.mask.nxv2f32.nxv2i64.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfncvt_f_x_w_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -956,7 +956,7 @@ vfloat32m1_t test_vfncvt_f_x_w_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_f_x_w_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfncvt.f.x.w.mask.nxv4f32.nxv4i64.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfncvt.f.x.w.mask.nxv4f32.nxv4i64.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfncvt_f_x_w_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -967,7 +967,7 @@ vfloat32m2_t test_vfncvt_f_x_w_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_f_x_w_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfncvt.f.x.w.mask.nxv8f32.nxv8i64.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfncvt.f.x.w.mask.nxv8f32.nxv8i64.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfncvt_f_x_w_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -978,7 +978,7 @@ vfloat32m4_t test_vfncvt_f_x_w_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_f_xu_w_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfncvt.f.xu.w.mask.nxv1f32.nxv1i64.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfncvt.f.xu.w.mask.nxv1f32.nxv1i64.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfncvt_f_xu_w_f32mf2_m(vbool64_t mask,
@@ -990,7 +990,7 @@ vfloat32mf2_t test_vfncvt_f_xu_w_f32mf2_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_f_xu_w_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfncvt.f.xu.w.mask.nxv2f32.nxv2i64.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfncvt.f.xu.w.mask.nxv2f32.nxv2i64.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfncvt_f_xu_w_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -1001,7 +1001,7 @@ vfloat32m1_t test_vfncvt_f_xu_w_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_f_xu_w_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfncvt.f.xu.w.mask.nxv4f32.nxv4i64.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfncvt.f.xu.w.mask.nxv4f32.nxv4i64.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfncvt_f_xu_w_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -1012,7 +1012,7 @@ vfloat32m2_t test_vfncvt_f_xu_w_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_f_xu_w_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfncvt.f.xu.w.mask.nxv8f32.nxv8i64.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfncvt.f.xu.w.mask.nxv8f32.nxv8i64.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfncvt_f_xu_w_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -1023,7 +1023,7 @@ vfloat32m4_t test_vfncvt_f_xu_w_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_f_f_w_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfncvt.f.f.w.mask.nxv1f32.nxv1f64.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfncvt.f.f.w.mask.nxv1f32.nxv1f64.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfncvt_f_f_w_f32mf2_m(vbool64_t mask,
@@ -1035,7 +1035,7 @@ vfloat32mf2_t test_vfncvt_f_f_w_f32mf2_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rod_f_f_w_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfncvt.rod.f.f.w.mask.nxv1f32.nxv1f64.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfncvt.rod.f.f.w.mask.nxv1f32.nxv1f64.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfncvt_rod_f_f_w_f32mf2_m(vbool64_t mask,
@@ -1047,7 +1047,7 @@ vfloat32mf2_t test_vfncvt_rod_f_f_w_f32mf2_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_f_f_w_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfncvt.f.f.w.mask.nxv2f32.nxv2f64.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfncvt.f.f.w.mask.nxv2f32.nxv2f64.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfncvt_f_f_w_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -1058,7 +1058,7 @@ vfloat32m1_t test_vfncvt_f_f_w_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rod_f_f_w_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfncvt.rod.f.f.w.mask.nxv2f32.nxv2f64.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfncvt.rod.f.f.w.mask.nxv2f32.nxv2f64.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfncvt_rod_f_f_w_f32m1_m(vbool32_t mask,
@@ -1070,7 +1070,7 @@ vfloat32m1_t test_vfncvt_rod_f_f_w_f32m1_m(vbool32_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_f_f_w_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfncvt.f.f.w.mask.nxv4f32.nxv4f64.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfncvt.f.f.w.mask.nxv4f32.nxv4f64.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfncvt_f_f_w_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -1081,7 +1081,7 @@ vfloat32m2_t test_vfncvt_f_f_w_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rod_f_f_w_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfncvt.rod.f.f.w.mask.nxv4f32.nxv4f64.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfncvt.rod.f.f.w.mask.nxv4f32.nxv4f64.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfncvt_rod_f_f_w_f32m2_m(vbool16_t mask,
@@ -1093,7 +1093,7 @@ vfloat32m2_t test_vfncvt_rod_f_f_w_f32m2_m(vbool16_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_f_f_w_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfncvt.f.f.w.mask.nxv8f32.nxv8f64.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfncvt.f.f.w.mask.nxv8f32.nxv8f64.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfncvt_f_f_w_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -1104,7 +1104,7 @@ vfloat32m4_t test_vfncvt_f_f_w_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfncvt_rod_f_f_w_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfncvt.rod.f.f.w.mask.nxv8f32.nxv8f64.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfncvt.rod.f.f.w.mask.nxv8f32.nxv8f64.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfncvt_rod_f_f_w_f32m4_m(vbool8_t mask,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfneg.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfneg.c
index 8ac2d95..cbab1f6 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfneg.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfneg.c
@@ -98,7 +98,7 @@ vfloat64m8_t test_vfneg_v_f64m8 (vfloat64m8_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfneg_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsgnjn.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP1]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsgnjn.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP1]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfneg_v_f32mf2_m (vbool64_t mask, vfloat32mf2_t maskedoff, vfloat32mf2_t op1, size_t vl) {
@@ -108,7 +108,7 @@ vfloat32mf2_t test_vfneg_v_f32mf2_m (vbool64_t mask, vfloat32mf2_t maskedoff, vf
 //
 // CHECK-RV64-LABEL: @test_vfneg_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsgnjn.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP1]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsgnjn.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP1]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfneg_v_f32m1_m (vbool32_t mask, vfloat32m1_t maskedoff, vfloat32m1_t op1, size_t vl) {
@@ -118,7 +118,7 @@ vfloat32m1_t test_vfneg_v_f32m1_m (vbool32_t mask, vfloat32m1_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vfneg_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsgnjn.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP1]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsgnjn.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP1]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfneg_v_f32m2_m (vbool16_t mask, vfloat32m2_t maskedoff, vfloat32m2_t op1, size_t vl) {
@@ -128,7 +128,7 @@ vfloat32m2_t test_vfneg_v_f32m2_m (vbool16_t mask, vfloat32m2_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vfneg_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsgnjn.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP1]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsgnjn.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP1]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfneg_v_f32m4_m (vbool8_t mask, vfloat32m4_t maskedoff, vfloat32m4_t op1, size_t vl) {
@@ -138,7 +138,7 @@ vfloat32m4_t test_vfneg_v_f32m4_m (vbool8_t mask, vfloat32m4_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vfneg_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsgnjn.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP1]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsgnjn.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP1]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfneg_v_f32m8_m (vbool4_t mask, vfloat32m8_t maskedoff, vfloat32m8_t op1, size_t vl) {
@@ -148,7 +148,7 @@ vfloat32m8_t test_vfneg_v_f32m8_m (vbool4_t mask, vfloat32m8_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vfneg_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsgnjn.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP1]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsgnjn.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP1]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfneg_v_f64m1_m (vbool64_t mask, vfloat64m1_t maskedoff, vfloat64m1_t op1, size_t vl) {
@@ -158,7 +158,7 @@ vfloat64m1_t test_vfneg_v_f64m1_m (vbool64_t mask, vfloat64m1_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vfneg_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsgnjn.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP1]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsgnjn.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP1]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfneg_v_f64m2_m (vbool32_t mask, vfloat64m2_t maskedoff, vfloat64m2_t op1, size_t vl) {
@@ -168,7 +168,7 @@ vfloat64m2_t test_vfneg_v_f64m2_m (vbool32_t mask, vfloat64m2_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vfneg_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsgnjn.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP1]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsgnjn.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP1]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfneg_v_f64m4_m (vbool16_t mask, vfloat64m4_t maskedoff, vfloat64m4_t op1, size_t vl) {
@@ -178,7 +178,7 @@ vfloat64m4_t test_vfneg_v_f64m4_m (vbool16_t mask, vfloat64m4_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vfneg_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsgnjn.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP1]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsgnjn.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP1]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfneg_v_f64m8_m (vbool8_t mask, vfloat64m8_t maskedoff, vfloat64m8_t op1, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfnmacc.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfnmacc.c
index 6c6aed7..cebd237 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfnmacc.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfnmacc.c
@@ -206,7 +206,7 @@ vfloat64m8_t test_vfnmacc_vf_f64m8(vfloat64m8_t acc, double op1,
 //
 // CHECK-RV64-LABEL: @test_vfnmacc_vv_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfnmacc.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfnmacc.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfnmacc_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
@@ -218,7 +218,7 @@ vfloat32mf2_t test_vfnmacc_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmacc_vf_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfnmacc.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfnmacc.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfnmacc_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
@@ -230,7 +230,7 @@ vfloat32mf2_t test_vfnmacc_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmacc_vv_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfnmacc.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfnmacc.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfnmacc_vv_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
@@ -242,7 +242,7 @@ vfloat32m1_t test_vfnmacc_vv_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmacc_vf_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfnmacc.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfnmacc.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfnmacc_vf_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
@@ -253,7 +253,7 @@ vfloat32m1_t test_vfnmacc_vf_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmacc_vv_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfnmacc.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfnmacc.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfnmacc_vv_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
@@ -265,7 +265,7 @@ vfloat32m2_t test_vfnmacc_vv_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmacc_vf_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfnmacc.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfnmacc.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfnmacc_vf_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
@@ -276,7 +276,7 @@ vfloat32m2_t test_vfnmacc_vf_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmacc_vv_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfnmacc.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfnmacc.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfnmacc_vv_f32m4_m(vbool8_t mask, vfloat32m4_t acc,
@@ -288,7 +288,7 @@ vfloat32m4_t test_vfnmacc_vv_f32m4_m(vbool8_t mask, vfloat32m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmacc_vf_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfnmacc.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfnmacc.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfnmacc_vf_f32m4_m(vbool8_t mask, vfloat32m4_t acc, float op1,
@@ -299,7 +299,7 @@ vfloat32m4_t test_vfnmacc_vf_f32m4_m(vbool8_t mask, vfloat32m4_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfnmacc_vv_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfnmacc.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[ACC:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfnmacc.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[ACC:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfnmacc_vv_f32m8_m(vbool4_t mask, vfloat32m8_t acc,
@@ -311,7 +311,7 @@ vfloat32m8_t test_vfnmacc_vv_f32m8_m(vbool4_t mask, vfloat32m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmacc_vf_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfnmacc.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfnmacc.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfnmacc_vf_f32m8_m(vbool4_t mask, vfloat32m8_t acc, float op1,
@@ -322,7 +322,7 @@ vfloat32m8_t test_vfnmacc_vf_f32m8_m(vbool4_t mask, vfloat32m8_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfnmacc_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfnmacc.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfnmacc.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfnmacc_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -334,7 +334,7 @@ vfloat64m1_t test_vfnmacc_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmacc_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfnmacc.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfnmacc.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfnmacc_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -345,7 +345,7 @@ vfloat64m1_t test_vfnmacc_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmacc_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfnmacc.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfnmacc.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfnmacc_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -357,7 +357,7 @@ vfloat64m2_t test_vfnmacc_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmacc_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfnmacc.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfnmacc.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfnmacc_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -368,7 +368,7 @@ vfloat64m2_t test_vfnmacc_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmacc_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfnmacc.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfnmacc.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfnmacc_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -380,7 +380,7 @@ vfloat64m4_t test_vfnmacc_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmacc_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfnmacc.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfnmacc.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfnmacc_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -391,7 +391,7 @@ vfloat64m4_t test_vfnmacc_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmacc_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfnmacc.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfnmacc.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfnmacc_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
@@ -403,7 +403,7 @@ vfloat64m8_t test_vfnmacc_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmacc_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfnmacc.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfnmacc.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfnmacc_vf_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfnmadd.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfnmadd.c
index 1ea9a02..56feb08 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfnmadd.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfnmadd.c
@@ -206,7 +206,7 @@ vfloat64m8_t test_vfnmadd_vf_f64m8(vfloat64m8_t acc, double op1,
 //
 // CHECK-RV64-LABEL: @test_vfnmadd_vv_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfnmadd.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfnmadd.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfnmadd_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
@@ -218,7 +218,7 @@ vfloat32mf2_t test_vfnmadd_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmadd_vf_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfnmadd.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfnmadd.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfnmadd_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
@@ -230,7 +230,7 @@ vfloat32mf2_t test_vfnmadd_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmadd_vv_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfnmadd.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfnmadd.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfnmadd_vv_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
@@ -242,7 +242,7 @@ vfloat32m1_t test_vfnmadd_vv_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmadd_vf_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfnmadd.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfnmadd.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfnmadd_vf_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
@@ -253,7 +253,7 @@ vfloat32m1_t test_vfnmadd_vf_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmadd_vv_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfnmadd.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfnmadd.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfnmadd_vv_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
@@ -265,7 +265,7 @@ vfloat32m2_t test_vfnmadd_vv_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmadd_vf_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfnmadd.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfnmadd.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfnmadd_vf_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
@@ -276,7 +276,7 @@ vfloat32m2_t test_vfnmadd_vf_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmadd_vv_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfnmadd.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfnmadd.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfnmadd_vv_f32m4_m(vbool8_t mask, vfloat32m4_t acc,
@@ -288,7 +288,7 @@ vfloat32m4_t test_vfnmadd_vv_f32m4_m(vbool8_t mask, vfloat32m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmadd_vf_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfnmadd.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfnmadd.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfnmadd_vf_f32m4_m(vbool8_t mask, vfloat32m4_t acc, float op1,
@@ -299,7 +299,7 @@ vfloat32m4_t test_vfnmadd_vf_f32m4_m(vbool8_t mask, vfloat32m4_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfnmadd_vv_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfnmadd.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[ACC:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfnmadd.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[ACC:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfnmadd_vv_f32m8_m(vbool4_t mask, vfloat32m8_t acc,
@@ -311,7 +311,7 @@ vfloat32m8_t test_vfnmadd_vv_f32m8_m(vbool4_t mask, vfloat32m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmadd_vf_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfnmadd.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfnmadd.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfnmadd_vf_f32m8_m(vbool4_t mask, vfloat32m8_t acc, float op1,
@@ -322,7 +322,7 @@ vfloat32m8_t test_vfnmadd_vf_f32m8_m(vbool4_t mask, vfloat32m8_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfnmadd_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfnmadd.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfnmadd.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfnmadd_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -334,7 +334,7 @@ vfloat64m1_t test_vfnmadd_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmadd_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfnmadd.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfnmadd.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfnmadd_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -345,7 +345,7 @@ vfloat64m1_t test_vfnmadd_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmadd_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfnmadd.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfnmadd.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfnmadd_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -357,7 +357,7 @@ vfloat64m2_t test_vfnmadd_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmadd_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfnmadd.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfnmadd.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfnmadd_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -368,7 +368,7 @@ vfloat64m2_t test_vfnmadd_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmadd_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfnmadd.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfnmadd.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfnmadd_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -380,7 +380,7 @@ vfloat64m4_t test_vfnmadd_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmadd_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfnmadd.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfnmadd.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfnmadd_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -391,7 +391,7 @@ vfloat64m4_t test_vfnmadd_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmadd_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfnmadd.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfnmadd.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfnmadd_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
@@ -403,7 +403,7 @@ vfloat64m8_t test_vfnmadd_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmadd_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfnmadd.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfnmadd.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfnmadd_vf_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfnmsac.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfnmsac.c
index c930ac2..3bcdfda 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfnmsac.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfnmsac.c
@@ -206,7 +206,7 @@ vfloat64m8_t test_vfnmsac_vf_f64m8(vfloat64m8_t acc, double op1,
 //
 // CHECK-RV64-LABEL: @test_vfnmsac_vv_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfnmsac.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfnmsac.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfnmsac_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
@@ -218,7 +218,7 @@ vfloat32mf2_t test_vfnmsac_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsac_vf_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfnmsac.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfnmsac.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfnmsac_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
@@ -230,7 +230,7 @@ vfloat32mf2_t test_vfnmsac_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsac_vv_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfnmsac.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfnmsac.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfnmsac_vv_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
@@ -242,7 +242,7 @@ vfloat32m1_t test_vfnmsac_vv_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsac_vf_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfnmsac.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfnmsac.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfnmsac_vf_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
@@ -253,7 +253,7 @@ vfloat32m1_t test_vfnmsac_vf_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsac_vv_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfnmsac.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfnmsac.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfnmsac_vv_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
@@ -265,7 +265,7 @@ vfloat32m2_t test_vfnmsac_vv_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsac_vf_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfnmsac.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfnmsac.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfnmsac_vf_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
@@ -276,7 +276,7 @@ vfloat32m2_t test_vfnmsac_vf_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsac_vv_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfnmsac.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfnmsac.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfnmsac_vv_f32m4_m(vbool8_t mask, vfloat32m4_t acc,
@@ -288,7 +288,7 @@ vfloat32m4_t test_vfnmsac_vv_f32m4_m(vbool8_t mask, vfloat32m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsac_vf_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfnmsac.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfnmsac.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfnmsac_vf_f32m4_m(vbool8_t mask, vfloat32m4_t acc, float op1,
@@ -299,7 +299,7 @@ vfloat32m4_t test_vfnmsac_vf_f32m4_m(vbool8_t mask, vfloat32m4_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfnmsac_vv_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfnmsac.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[ACC:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfnmsac.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[ACC:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfnmsac_vv_f32m8_m(vbool4_t mask, vfloat32m8_t acc,
@@ -311,7 +311,7 @@ vfloat32m8_t test_vfnmsac_vv_f32m8_m(vbool4_t mask, vfloat32m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsac_vf_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfnmsac.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfnmsac.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfnmsac_vf_f32m8_m(vbool4_t mask, vfloat32m8_t acc, float op1,
@@ -322,7 +322,7 @@ vfloat32m8_t test_vfnmsac_vf_f32m8_m(vbool4_t mask, vfloat32m8_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfnmsac_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfnmsac.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfnmsac.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfnmsac_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -334,7 +334,7 @@ vfloat64m1_t test_vfnmsac_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsac_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfnmsac.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfnmsac.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfnmsac_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -345,7 +345,7 @@ vfloat64m1_t test_vfnmsac_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsac_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfnmsac.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfnmsac.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfnmsac_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -357,7 +357,7 @@ vfloat64m2_t test_vfnmsac_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsac_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfnmsac.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfnmsac.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfnmsac_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -368,7 +368,7 @@ vfloat64m2_t test_vfnmsac_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsac_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfnmsac.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfnmsac.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfnmsac_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -380,7 +380,7 @@ vfloat64m4_t test_vfnmsac_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsac_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfnmsac.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfnmsac.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfnmsac_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -391,7 +391,7 @@ vfloat64m4_t test_vfnmsac_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsac_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfnmsac.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfnmsac.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfnmsac_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
@@ -403,7 +403,7 @@ vfloat64m8_t test_vfnmsac_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsac_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfnmsac.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfnmsac.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfnmsac_vf_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfnmsub.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfnmsub.c
index 7a43eddc..ba316ca0 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfnmsub.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfnmsub.c
@@ -206,7 +206,7 @@ vfloat64m8_t test_vfnmsub_vf_f64m8(vfloat64m8_t acc, double op1,
 //
 // CHECK-RV64-LABEL: @test_vfnmsub_vv_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfnmsub.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfnmsub.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfnmsub_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
@@ -218,7 +218,7 @@ vfloat32mf2_t test_vfnmsub_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsub_vf_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfnmsub.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfnmsub.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfnmsub_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
@@ -230,7 +230,7 @@ vfloat32mf2_t test_vfnmsub_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsub_vv_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfnmsub.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfnmsub.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfnmsub_vv_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
@@ -242,7 +242,7 @@ vfloat32m1_t test_vfnmsub_vv_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsub_vf_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfnmsub.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfnmsub.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfnmsub_vf_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
@@ -253,7 +253,7 @@ vfloat32m1_t test_vfnmsub_vf_f32m1_m(vbool32_t mask, vfloat32m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsub_vv_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfnmsub.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfnmsub.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfnmsub_vv_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
@@ -265,7 +265,7 @@ vfloat32m2_t test_vfnmsub_vv_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsub_vf_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfnmsub.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfnmsub.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfnmsub_vf_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
@@ -276,7 +276,7 @@ vfloat32m2_t test_vfnmsub_vf_f32m2_m(vbool16_t mask, vfloat32m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsub_vv_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfnmsub.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfnmsub.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfnmsub_vv_f32m4_m(vbool8_t mask, vfloat32m4_t acc,
@@ -288,7 +288,7 @@ vfloat32m4_t test_vfnmsub_vv_f32m4_m(vbool8_t mask, vfloat32m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsub_vf_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfnmsub.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfnmsub.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfnmsub_vf_f32m4_m(vbool8_t mask, vfloat32m4_t acc, float op1,
@@ -299,7 +299,7 @@ vfloat32m4_t test_vfnmsub_vf_f32m4_m(vbool8_t mask, vfloat32m4_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfnmsub_vv_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfnmsub.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[ACC:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfnmsub.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[ACC:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfnmsub_vv_f32m8_m(vbool4_t mask, vfloat32m8_t acc,
@@ -311,7 +311,7 @@ vfloat32m8_t test_vfnmsub_vv_f32m8_m(vbool4_t mask, vfloat32m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsub_vf_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfnmsub.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfnmsub.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfnmsub_vf_f32m8_m(vbool4_t mask, vfloat32m8_t acc, float op1,
@@ -322,7 +322,7 @@ vfloat32m8_t test_vfnmsub_vf_f32m8_m(vbool4_t mask, vfloat32m8_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfnmsub_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfnmsub.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfnmsub.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfnmsub_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -334,7 +334,7 @@ vfloat64m1_t test_vfnmsub_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsub_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfnmsub.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfnmsub.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfnmsub_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -345,7 +345,7 @@ vfloat64m1_t test_vfnmsub_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsub_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfnmsub.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfnmsub.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfnmsub_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -357,7 +357,7 @@ vfloat64m2_t test_vfnmsub_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsub_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfnmsub.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfnmsub.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfnmsub_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -368,7 +368,7 @@ vfloat64m2_t test_vfnmsub_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsub_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfnmsub.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfnmsub.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfnmsub_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -380,7 +380,7 @@ vfloat64m4_t test_vfnmsub_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsub_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfnmsub.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfnmsub.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfnmsub_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -391,7 +391,7 @@ vfloat64m4_t test_vfnmsub_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsub_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfnmsub.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfnmsub.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfnmsub_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
@@ -403,7 +403,7 @@ vfloat64m8_t test_vfnmsub_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfnmsub_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfnmsub.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfnmsub.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[ACC:%.*]], double [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfnmsub_vf_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfrdiv.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfrdiv.c
index 59849b5..db1e672 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfrdiv.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfrdiv.c
@@ -98,7 +98,7 @@ vfloat64m8_t test_vfrdiv_vf_f64m8(vfloat64m8_t op1, double op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfrdiv_vf_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfrdiv.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfrdiv.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfrdiv_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -109,7 +109,7 @@ vfloat32mf2_t test_vfrdiv_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrdiv_vf_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfrdiv.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfrdiv.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfrdiv_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -120,7 +120,7 @@ vfloat32m1_t test_vfrdiv_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrdiv_vf_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfrdiv.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfrdiv.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfrdiv_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -131,7 +131,7 @@ vfloat32m2_t test_vfrdiv_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrdiv_vf_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfrdiv.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfrdiv.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfrdiv_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -142,7 +142,7 @@ vfloat32m4_t test_vfrdiv_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrdiv_vf_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfrdiv.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfrdiv.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfrdiv_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -153,7 +153,7 @@ vfloat32m8_t test_vfrdiv_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrdiv_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfrdiv.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfrdiv.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfrdiv_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -164,7 +164,7 @@ vfloat64m1_t test_vfrdiv_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrdiv_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfrdiv.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfrdiv.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfrdiv_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -175,7 +175,7 @@ vfloat64m2_t test_vfrdiv_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrdiv_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfrdiv.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfrdiv.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfrdiv_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -186,7 +186,7 @@ vfloat64m4_t test_vfrdiv_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrdiv_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfrdiv.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfrdiv.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfrdiv_vf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfrec7.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfrec7.c
index d6dcd73..f044b2e 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfrec7.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfrec7.c
@@ -98,7 +98,7 @@ vfloat64m8_t test_vfrec7_v_f64m8(vfloat64m8_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfrec7_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfrec7.mask.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfrec7.mask.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfrec7_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -109,7 +109,7 @@ vfloat32mf2_t test_vfrec7_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrec7_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfrec7.mask.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfrec7.mask.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfrec7_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -120,7 +120,7 @@ vfloat32m1_t test_vfrec7_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrec7_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfrec7.mask.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfrec7.mask.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfrec7_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -131,7 +131,7 @@ vfloat32m2_t test_vfrec7_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrec7_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfrec7.mask.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfrec7.mask.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfrec7_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -142,7 +142,7 @@ vfloat32m4_t test_vfrec7_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrec7_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfrec7.mask.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfrec7.mask.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfrec7_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -153,7 +153,7 @@ vfloat32m8_t test_vfrec7_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrec7_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfrec7.mask.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfrec7.mask.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfrec7_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -164,7 +164,7 @@ vfloat64m1_t test_vfrec7_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrec7_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfrec7.mask.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfrec7.mask.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfrec7_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -175,7 +175,7 @@ vfloat64m2_t test_vfrec7_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrec7_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfrec7.mask.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfrec7.mask.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfrec7_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -186,7 +186,7 @@ vfloat64m4_t test_vfrec7_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrec7_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfrec7.mask.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfrec7.mask.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfrec7_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfredmax.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfredmax.c
index a96977c..0d74456 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfredmax.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfredmax.c
@@ -108,7 +108,7 @@ vfloat64m1_t test_vfredmax_vs_f64m8_f64m1(vfloat64m1_t dst, vfloat64m8_t vector,
 //
 // CHECK-RV64-LABEL: @test_vfredmax_vs_f32mf2_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredmax.mask.nxv2f32.nxv1f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 1 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredmax.mask.nxv2f32.nxv1f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 1 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfredmax_vs_f32mf2_f32m1_m(vbool64_t mask, vfloat32m1_t dst,
@@ -120,7 +120,7 @@ vfloat32m1_t test_vfredmax_vs_f32mf2_f32m1_m(vbool64_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredmax_vs_f32m1_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredmax.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 2 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredmax.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 2 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfredmax_vs_f32m1_f32m1_m(vbool32_t mask, vfloat32m1_t dst,
@@ -132,7 +132,7 @@ vfloat32m1_t test_vfredmax_vs_f32m1_f32m1_m(vbool32_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredmax_vs_f32m2_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredmax.mask.nxv2f32.nxv4f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 4 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredmax.mask.nxv2f32.nxv4f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 4 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfredmax_vs_f32m2_f32m1_m(vbool16_t mask, vfloat32m1_t dst,
@@ -144,7 +144,7 @@ vfloat32m1_t test_vfredmax_vs_f32m2_f32m1_m(vbool16_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredmax_vs_f32m4_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredmax.mask.nxv2f32.nxv8f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 8 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredmax.mask.nxv2f32.nxv8f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 8 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfredmax_vs_f32m4_f32m1_m(vbool8_t mask, vfloat32m1_t dst,
@@ -156,7 +156,7 @@ vfloat32m1_t test_vfredmax_vs_f32m4_f32m1_m(vbool8_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredmax_vs_f32m8_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredmax.mask.nxv2f32.nxv16f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 16 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredmax.mask.nxv2f32.nxv16f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 16 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfredmax_vs_f32m8_f32m1_m(vbool4_t mask, vfloat32m1_t dst,
@@ -168,7 +168,7 @@ vfloat32m1_t test_vfredmax_vs_f32m8_f32m1_m(vbool4_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredmax_vs_f64m1_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredmax.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 1 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredmax.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 1 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfredmax_vs_f64m1_f64m1_m(vbool64_t mask, vfloat64m1_t dst,
@@ -180,7 +180,7 @@ vfloat64m1_t test_vfredmax_vs_f64m1_f64m1_m(vbool64_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredmax_vs_f64m2_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredmax.mask.nxv1f64.nxv2f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 2 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredmax.mask.nxv1f64.nxv2f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 2 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfredmax_vs_f64m2_f64m1_m(vbool32_t mask, vfloat64m1_t dst,
@@ -192,7 +192,7 @@ vfloat64m1_t test_vfredmax_vs_f64m2_f64m1_m(vbool32_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredmax_vs_f64m4_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredmax.mask.nxv1f64.nxv4f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 4 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredmax.mask.nxv1f64.nxv4f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 4 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfredmax_vs_f64m4_f64m1_m(vbool16_t mask, vfloat64m1_t dst,
@@ -204,7 +204,7 @@ vfloat64m1_t test_vfredmax_vs_f64m4_f64m1_m(vbool16_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredmax_vs_f64m8_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredmax.mask.nxv1f64.nxv8f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 8 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredmax.mask.nxv1f64.nxv8f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 8 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfredmax_vs_f64m8_f64m1_m(vbool8_t mask, vfloat64m1_t dst,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfredmin.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfredmin.c
index 0947610..b3467db 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfredmin.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfredmin.c
@@ -108,7 +108,7 @@ vfloat64m1_t test_vfredmin_vs_f64m8_f64m1(vfloat64m1_t dst, vfloat64m8_t vector,
 //
 // CHECK-RV64-LABEL: @test_vfredmin_vs_f32mf2_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredmin.mask.nxv2f32.nxv1f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 1 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredmin.mask.nxv2f32.nxv1f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 1 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfredmin_vs_f32mf2_f32m1_m(vbool64_t mask, vfloat32m1_t dst,
@@ -120,7 +120,7 @@ vfloat32m1_t test_vfredmin_vs_f32mf2_f32m1_m(vbool64_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredmin_vs_f32m1_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredmin.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 2 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredmin.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 2 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfredmin_vs_f32m1_f32m1_m(vbool32_t mask, vfloat32m1_t dst,
@@ -132,7 +132,7 @@ vfloat32m1_t test_vfredmin_vs_f32m1_f32m1_m(vbool32_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredmin_vs_f32m2_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredmin.mask.nxv2f32.nxv4f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 4 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredmin.mask.nxv2f32.nxv4f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 4 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfredmin_vs_f32m2_f32m1_m(vbool16_t mask, vfloat32m1_t dst,
@@ -144,7 +144,7 @@ vfloat32m1_t test_vfredmin_vs_f32m2_f32m1_m(vbool16_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredmin_vs_f32m4_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredmin.mask.nxv2f32.nxv8f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 8 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredmin.mask.nxv2f32.nxv8f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 8 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfredmin_vs_f32m4_f32m1_m(vbool8_t mask, vfloat32m1_t dst,
@@ -156,7 +156,7 @@ vfloat32m1_t test_vfredmin_vs_f32m4_f32m1_m(vbool8_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredmin_vs_f32m8_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredmin.mask.nxv2f32.nxv16f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 16 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredmin.mask.nxv2f32.nxv16f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 16 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfredmin_vs_f32m8_f32m1_m(vbool4_t mask, vfloat32m1_t dst,
@@ -168,7 +168,7 @@ vfloat32m1_t test_vfredmin_vs_f32m8_f32m1_m(vbool4_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredmin_vs_f64m1_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredmin.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 1 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredmin.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 1 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfredmin_vs_f64m1_f64m1_m(vbool64_t mask, vfloat64m1_t dst,
@@ -180,7 +180,7 @@ vfloat64m1_t test_vfredmin_vs_f64m1_f64m1_m(vbool64_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredmin_vs_f64m2_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredmin.mask.nxv1f64.nxv2f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 2 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredmin.mask.nxv1f64.nxv2f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 2 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfredmin_vs_f64m2_f64m1_m(vbool32_t mask, vfloat64m1_t dst,
@@ -192,7 +192,7 @@ vfloat64m1_t test_vfredmin_vs_f64m2_f64m1_m(vbool32_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredmin_vs_f64m4_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredmin.mask.nxv1f64.nxv4f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 4 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredmin.mask.nxv1f64.nxv4f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 4 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfredmin_vs_f64m4_f64m1_m(vbool16_t mask, vfloat64m1_t dst,
@@ -204,7 +204,7 @@ vfloat64m1_t test_vfredmin_vs_f64m4_f64m1_m(vbool16_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredmin_vs_f64m8_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredmin.mask.nxv1f64.nxv8f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 8 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredmin.mask.nxv1f64.nxv8f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 8 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfredmin_vs_f64m8_f64m1_m(vbool8_t mask, vfloat64m1_t dst,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfredsum.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfredsum.c
index d280047..8c82c78 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfredsum.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfredsum.c
@@ -108,7 +108,7 @@ vfloat64m1_t test_vfredsum_vs_f64m8_f64m1(vfloat64m1_t dst, vfloat64m8_t vector,
 //
 // CHECK-RV64-LABEL: @test_vfredsum_vs_f32mf2_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredsum.mask.nxv2f32.nxv1f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 1 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredsum.mask.nxv2f32.nxv1f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 1 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfredsum_vs_f32mf2_f32m1_m(vbool64_t mask, vfloat32m1_t dst,
@@ -120,7 +120,7 @@ vfloat32m1_t test_vfredsum_vs_f32mf2_f32m1_m(vbool64_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredsum_vs_f32m1_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredsum.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 2 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredsum.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 2 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfredsum_vs_f32m1_f32m1_m(vbool32_t mask, vfloat32m1_t dst,
@@ -132,7 +132,7 @@ vfloat32m1_t test_vfredsum_vs_f32m1_f32m1_m(vbool32_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredsum_vs_f32m2_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredsum.mask.nxv2f32.nxv4f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 4 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredsum.mask.nxv2f32.nxv4f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 4 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfredsum_vs_f32m2_f32m1_m(vbool16_t mask, vfloat32m1_t dst,
@@ -144,7 +144,7 @@ vfloat32m1_t test_vfredsum_vs_f32m2_f32m1_m(vbool16_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredsum_vs_f32m4_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredsum.mask.nxv2f32.nxv8f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 8 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredsum.mask.nxv2f32.nxv8f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 8 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfredsum_vs_f32m4_f32m1_m(vbool8_t mask, vfloat32m1_t dst,
@@ -156,7 +156,7 @@ vfloat32m1_t test_vfredsum_vs_f32m4_f32m1_m(vbool8_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredsum_vs_f32m8_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredsum.mask.nxv2f32.nxv16f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 16 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredsum.mask.nxv2f32.nxv16f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 16 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfredsum_vs_f32m8_f32m1_m(vbool4_t mask, vfloat32m1_t dst,
@@ -168,7 +168,7 @@ vfloat32m1_t test_vfredsum_vs_f32m8_f32m1_m(vbool4_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredsum_vs_f64m1_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredsum.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 1 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredsum.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 1 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfredsum_vs_f64m1_f64m1_m(vbool64_t mask, vfloat64m1_t dst,
@@ -180,7 +180,7 @@ vfloat64m1_t test_vfredsum_vs_f64m1_f64m1_m(vbool64_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredsum_vs_f64m2_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredsum.mask.nxv1f64.nxv2f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 2 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredsum.mask.nxv1f64.nxv2f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 2 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfredsum_vs_f64m2_f64m1_m(vbool32_t mask, vfloat64m1_t dst,
@@ -192,7 +192,7 @@ vfloat64m1_t test_vfredsum_vs_f64m2_f64m1_m(vbool32_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredsum_vs_f64m4_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredsum.mask.nxv1f64.nxv4f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 4 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredsum.mask.nxv1f64.nxv4f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 4 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfredsum_vs_f64m4_f64m1_m(vbool16_t mask, vfloat64m1_t dst,
@@ -204,7 +204,7 @@ vfloat64m1_t test_vfredsum_vs_f64m4_f64m1_m(vbool16_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredsum_vs_f64m8_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredsum.mask.nxv1f64.nxv8f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 8 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredsum.mask.nxv1f64.nxv8f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 8 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfredsum_vs_f64m8_f64m1_m(vbool8_t mask, vfloat64m1_t dst,
@@ -324,7 +324,7 @@ vfloat64m1_t test_vfredosum_vs_f64m8_f64m1(vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredosum_vs_f32mf2_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredosum.mask.nxv2f32.nxv1f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 1 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredosum.mask.nxv2f32.nxv1f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 1 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfredosum_vs_f32mf2_f32m1_m(vbool64_t mask, vfloat32m1_t dst,
@@ -336,7 +336,7 @@ vfloat32m1_t test_vfredosum_vs_f32mf2_f32m1_m(vbool64_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredosum_vs_f32m1_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredosum.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 2 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredosum.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 2 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfredosum_vs_f32m1_f32m1_m(vbool32_t mask, vfloat32m1_t dst,
@@ -348,7 +348,7 @@ vfloat32m1_t test_vfredosum_vs_f32m1_f32m1_m(vbool32_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredosum_vs_f32m2_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredosum.mask.nxv2f32.nxv4f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 4 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredosum.mask.nxv2f32.nxv4f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 4 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfredosum_vs_f32m2_f32m1_m(vbool16_t mask, vfloat32m1_t dst,
@@ -360,7 +360,7 @@ vfloat32m1_t test_vfredosum_vs_f32m2_f32m1_m(vbool16_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredosum_vs_f32m4_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredosum.mask.nxv2f32.nxv8f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 8 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredosum.mask.nxv2f32.nxv8f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 8 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfredosum_vs_f32m4_f32m1_m(vbool8_t mask, vfloat32m1_t dst,
@@ -372,7 +372,7 @@ vfloat32m1_t test_vfredosum_vs_f32m4_f32m1_m(vbool8_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredosum_vs_f32m8_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredosum.mask.nxv2f32.nxv16f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 16 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfredosum.mask.nxv2f32.nxv16f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 16 x float> [[VECTOR:%.*]], <vscale x 2 x float> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfredosum_vs_f32m8_f32m1_m(vbool4_t mask, vfloat32m1_t dst,
@@ -384,7 +384,7 @@ vfloat32m1_t test_vfredosum_vs_f32m8_f32m1_m(vbool4_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredosum_vs_f64m1_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredosum.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 1 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredosum.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 1 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfredosum_vs_f64m1_f64m1_m(vbool64_t mask, vfloat64m1_t dst,
@@ -396,7 +396,7 @@ vfloat64m1_t test_vfredosum_vs_f64m1_f64m1_m(vbool64_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredosum_vs_f64m2_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredosum.mask.nxv1f64.nxv2f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 2 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredosum.mask.nxv1f64.nxv2f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 2 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfredosum_vs_f64m2_f64m1_m(vbool32_t mask, vfloat64m1_t dst,
@@ -408,7 +408,7 @@ vfloat64m1_t test_vfredosum_vs_f64m2_f64m1_m(vbool32_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredosum_vs_f64m4_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredosum.mask.nxv1f64.nxv4f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 4 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredosum.mask.nxv1f64.nxv4f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 4 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfredosum_vs_f64m4_f64m1_m(vbool16_t mask, vfloat64m1_t dst,
@@ -420,7 +420,7 @@ vfloat64m1_t test_vfredosum_vs_f64m4_f64m1_m(vbool16_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfredosum_vs_f64m8_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredosum.mask.nxv1f64.nxv8f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 8 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfredosum.mask.nxv1f64.nxv8f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 8 x double> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfredosum_vs_f64m8_f64m1_m(vbool8_t mask, vfloat64m1_t dst,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfrsqrt7.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfrsqrt7.c
index c712f40..b20cd8c 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfrsqrt7.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfrsqrt7.c
@@ -98,7 +98,7 @@ vfloat64m8_t test_vfrsqrt7_v_f64m8(vfloat64m8_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfrsqrt7_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfrsqrt7.mask.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfrsqrt7.mask.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfrsqrt7_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -109,7 +109,7 @@ vfloat32mf2_t test_vfrsqrt7_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrsqrt7_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfrsqrt7.mask.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfrsqrt7.mask.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfrsqrt7_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -120,7 +120,7 @@ vfloat32m1_t test_vfrsqrt7_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrsqrt7_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfrsqrt7.mask.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfrsqrt7.mask.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfrsqrt7_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -131,7 +131,7 @@ vfloat32m2_t test_vfrsqrt7_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrsqrt7_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfrsqrt7.mask.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfrsqrt7.mask.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfrsqrt7_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -142,7 +142,7 @@ vfloat32m4_t test_vfrsqrt7_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrsqrt7_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfrsqrt7.mask.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfrsqrt7.mask.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfrsqrt7_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -153,7 +153,7 @@ vfloat32m8_t test_vfrsqrt7_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrsqrt7_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfrsqrt7.mask.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfrsqrt7.mask.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfrsqrt7_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -164,7 +164,7 @@ vfloat64m1_t test_vfrsqrt7_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrsqrt7_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfrsqrt7.mask.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfrsqrt7.mask.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfrsqrt7_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -175,7 +175,7 @@ vfloat64m2_t test_vfrsqrt7_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrsqrt7_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfrsqrt7.mask.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfrsqrt7.mask.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfrsqrt7_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -186,7 +186,7 @@ vfloat64m4_t test_vfrsqrt7_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrsqrt7_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfrsqrt7.mask.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfrsqrt7.mask.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfrsqrt7_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfrsub.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfrsub.c
index 95a64ef..883ed13 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfrsub.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfrsub.c
@@ -98,7 +98,7 @@ vfloat64m8_t test_vfrsub_vf_f64m8(vfloat64m8_t op1, double op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfrsub_vf_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfrsub.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfrsub.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfrsub_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -109,7 +109,7 @@ vfloat32mf2_t test_vfrsub_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrsub_vf_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfrsub.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfrsub.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfrsub_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -120,7 +120,7 @@ vfloat32m1_t test_vfrsub_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrsub_vf_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfrsub.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfrsub.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfrsub_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -131,7 +131,7 @@ vfloat32m2_t test_vfrsub_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrsub_vf_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfrsub.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfrsub.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfrsub_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -142,7 +142,7 @@ vfloat32m4_t test_vfrsub_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrsub_vf_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfrsub.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfrsub.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfrsub_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -153,7 +153,7 @@ vfloat32m8_t test_vfrsub_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrsub_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfrsub.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfrsub.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfrsub_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -164,7 +164,7 @@ vfloat64m1_t test_vfrsub_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrsub_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfrsub.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfrsub.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfrsub_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -175,7 +175,7 @@ vfloat64m2_t test_vfrsub_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrsub_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfrsub.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfrsub.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfrsub_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -186,7 +186,7 @@ vfloat64m4_t test_vfrsub_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfrsub_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfrsub.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfrsub.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfrsub_vf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfsgnj.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfsgnj.c
index 279940d..cb8bf8a 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfsgnj.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfsgnj.c
@@ -575,7 +575,7 @@ vfloat64m8_t test_vfsgnjx_vf_f64m8(vfloat64m8_t op1, double op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfsgnj_vv_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsgnj.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsgnj.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfsgnj_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -587,7 +587,7 @@ vfloat32mf2_t test_vfsgnj_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnj_vf_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsgnj.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsgnj.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfsgnj_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -598,7 +598,7 @@ vfloat32mf2_t test_vfsgnj_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnj_vv_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsgnj.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsgnj.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfsgnj_vv_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -610,7 +610,7 @@ vfloat32m1_t test_vfsgnj_vv_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnj_vf_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsgnj.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsgnj.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfsgnj_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -621,7 +621,7 @@ vfloat32m1_t test_vfsgnj_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnj_vv_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsgnj.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsgnj.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfsgnj_vv_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -633,7 +633,7 @@ vfloat32m2_t test_vfsgnj_vv_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnj_vf_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsgnj.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsgnj.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfsgnj_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -644,7 +644,7 @@ vfloat32m2_t test_vfsgnj_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnj_vv_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsgnj.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsgnj.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfsgnj_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -656,7 +656,7 @@ vfloat32m4_t test_vfsgnj_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnj_vf_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsgnj.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsgnj.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfsgnj_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -667,7 +667,7 @@ vfloat32m4_t test_vfsgnj_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnj_vv_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsgnj.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsgnj.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfsgnj_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -679,7 +679,7 @@ vfloat32m8_t test_vfsgnj_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnj_vf_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsgnj.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsgnj.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfsgnj_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -690,7 +690,7 @@ vfloat32m8_t test_vfsgnj_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnj_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsgnj.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsgnj.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfsgnj_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -702,7 +702,7 @@ vfloat64m1_t test_vfsgnj_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnj_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsgnj.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsgnj.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfsgnj_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -713,7 +713,7 @@ vfloat64m1_t test_vfsgnj_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnj_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsgnj.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsgnj.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfsgnj_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -725,7 +725,7 @@ vfloat64m2_t test_vfsgnj_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnj_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsgnj.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsgnj.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfsgnj_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -736,7 +736,7 @@ vfloat64m2_t test_vfsgnj_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnj_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsgnj.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsgnj.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfsgnj_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -748,7 +748,7 @@ vfloat64m4_t test_vfsgnj_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnj_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsgnj.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsgnj.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfsgnj_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -759,7 +759,7 @@ vfloat64m4_t test_vfsgnj_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnj_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsgnj.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsgnj.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfsgnj_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -771,7 +771,7 @@ vfloat64m8_t test_vfsgnj_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnj_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsgnj.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsgnj.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfsgnj_vf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -782,7 +782,7 @@ vfloat64m8_t test_vfsgnj_vf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjn_vv_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsgnjn.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsgnjn.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfsgnjn_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -794,7 +794,7 @@ vfloat32mf2_t test_vfsgnjn_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjn_vf_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsgnjn.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsgnjn.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfsgnjn_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -806,7 +806,7 @@ vfloat32mf2_t test_vfsgnjn_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjn_vv_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsgnjn.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsgnjn.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfsgnjn_vv_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -818,7 +818,7 @@ vfloat32m1_t test_vfsgnjn_vv_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjn_vf_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsgnjn.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsgnjn.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfsgnjn_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -829,7 +829,7 @@ vfloat32m1_t test_vfsgnjn_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjn_vv_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsgnjn.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsgnjn.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfsgnjn_vv_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -841,7 +841,7 @@ vfloat32m2_t test_vfsgnjn_vv_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjn_vf_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsgnjn.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsgnjn.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfsgnjn_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -852,7 +852,7 @@ vfloat32m2_t test_vfsgnjn_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjn_vv_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsgnjn.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsgnjn.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfsgnjn_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -864,7 +864,7 @@ vfloat32m4_t test_vfsgnjn_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjn_vf_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsgnjn.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsgnjn.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfsgnjn_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -875,7 +875,7 @@ vfloat32m4_t test_vfsgnjn_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjn_vv_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsgnjn.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsgnjn.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfsgnjn_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -887,7 +887,7 @@ vfloat32m8_t test_vfsgnjn_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjn_vf_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsgnjn.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsgnjn.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfsgnjn_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -898,7 +898,7 @@ vfloat32m8_t test_vfsgnjn_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjn_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsgnjn.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsgnjn.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfsgnjn_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -910,7 +910,7 @@ vfloat64m1_t test_vfsgnjn_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjn_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsgnjn.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsgnjn.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfsgnjn_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -921,7 +921,7 @@ vfloat64m1_t test_vfsgnjn_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjn_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsgnjn.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsgnjn.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfsgnjn_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -933,7 +933,7 @@ vfloat64m2_t test_vfsgnjn_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjn_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsgnjn.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsgnjn.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfsgnjn_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -944,7 +944,7 @@ vfloat64m2_t test_vfsgnjn_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjn_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsgnjn.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsgnjn.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfsgnjn_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -956,7 +956,7 @@ vfloat64m4_t test_vfsgnjn_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjn_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsgnjn.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsgnjn.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfsgnjn_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -967,7 +967,7 @@ vfloat64m4_t test_vfsgnjn_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjn_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsgnjn.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsgnjn.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfsgnjn_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -979,7 +979,7 @@ vfloat64m8_t test_vfsgnjn_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjn_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsgnjn.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsgnjn.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfsgnjn_vf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -990,7 +990,7 @@ vfloat64m8_t test_vfsgnjn_vf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjx_vv_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsgnjx.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsgnjx.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfsgnjx_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -1002,7 +1002,7 @@ vfloat32mf2_t test_vfsgnjx_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjx_vf_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsgnjx.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsgnjx.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfsgnjx_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -1014,7 +1014,7 @@ vfloat32mf2_t test_vfsgnjx_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjx_vv_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsgnjx.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsgnjx.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfsgnjx_vv_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -1026,7 +1026,7 @@ vfloat32m1_t test_vfsgnjx_vv_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjx_vf_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsgnjx.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsgnjx.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfsgnjx_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -1037,7 +1037,7 @@ vfloat32m1_t test_vfsgnjx_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjx_vv_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsgnjx.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsgnjx.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfsgnjx_vv_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -1049,7 +1049,7 @@ vfloat32m2_t test_vfsgnjx_vv_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjx_vf_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsgnjx.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsgnjx.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfsgnjx_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -1060,7 +1060,7 @@ vfloat32m2_t test_vfsgnjx_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjx_vv_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsgnjx.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsgnjx.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfsgnjx_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -1072,7 +1072,7 @@ vfloat32m4_t test_vfsgnjx_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjx_vf_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsgnjx.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsgnjx.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfsgnjx_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -1083,7 +1083,7 @@ vfloat32m4_t test_vfsgnjx_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjx_vv_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsgnjx.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsgnjx.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfsgnjx_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -1095,7 +1095,7 @@ vfloat32m8_t test_vfsgnjx_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjx_vf_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsgnjx.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsgnjx.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfsgnjx_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -1106,7 +1106,7 @@ vfloat32m8_t test_vfsgnjx_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjx_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsgnjx.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsgnjx.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfsgnjx_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -1118,7 +1118,7 @@ vfloat64m1_t test_vfsgnjx_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjx_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsgnjx.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsgnjx.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfsgnjx_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -1129,7 +1129,7 @@ vfloat64m1_t test_vfsgnjx_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjx_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsgnjx.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsgnjx.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfsgnjx_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -1141,7 +1141,7 @@ vfloat64m2_t test_vfsgnjx_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjx_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsgnjx.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsgnjx.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfsgnjx_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -1152,7 +1152,7 @@ vfloat64m2_t test_vfsgnjx_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjx_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsgnjx.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsgnjx.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfsgnjx_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -1164,7 +1164,7 @@ vfloat64m4_t test_vfsgnjx_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjx_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsgnjx.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsgnjx.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfsgnjx_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -1175,7 +1175,7 @@ vfloat64m4_t test_vfsgnjx_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjx_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsgnjx.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsgnjx.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfsgnjx_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -1187,7 +1187,7 @@ vfloat64m8_t test_vfsgnjx_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsgnjx_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsgnjx.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsgnjx.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfsgnjx_vf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfslide1down.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfslide1down.c
index e25169c..d14f8b7 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfslide1down.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfslide1down.c
@@ -107,7 +107,7 @@ vfloat64m8_t test_vfslide1down_vf_f64m8(vfloat64m8_t src, double value,
 //
 // CHECK-RV64-LABEL: @test_vfslide1down_vf_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfslide1down.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], float [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfslide1down.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], float [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfslide1down_vf_f32mf2_m(vbool64_t mask,
@@ -120,7 +120,7 @@ vfloat32mf2_t test_vfslide1down_vf_f32mf2_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfslide1down_vf_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfslide1down.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], float [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfslide1down.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], float [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfslide1down_vf_f32m1_m(vbool32_t mask,
@@ -133,7 +133,7 @@ vfloat32m1_t test_vfslide1down_vf_f32m1_m(vbool32_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfslide1down_vf_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfslide1down.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], float [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfslide1down.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], float [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfslide1down_vf_f32m2_m(vbool16_t mask,
@@ -146,7 +146,7 @@ vfloat32m2_t test_vfslide1down_vf_f32m2_m(vbool16_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfslide1down_vf_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfslide1down.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], float [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfslide1down.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], float [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfslide1down_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -158,7 +158,7 @@ vfloat32m4_t test_vfslide1down_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfslide1down_vf_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfslide1down.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[SRC:%.*]], float [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfslide1down.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[SRC:%.*]], float [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfslide1down_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -170,7 +170,7 @@ vfloat32m8_t test_vfslide1down_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfslide1down_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfslide1down.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], double [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfslide1down.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], double [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfslide1down_vf_f64m1_m(vbool64_t mask,
@@ -183,7 +183,7 @@ vfloat64m1_t test_vfslide1down_vf_f64m1_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfslide1down_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfslide1down.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], double [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfslide1down.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], double [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfslide1down_vf_f64m2_m(vbool32_t mask,
@@ -196,7 +196,7 @@ vfloat64m2_t test_vfslide1down_vf_f64m2_m(vbool32_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfslide1down_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfslide1down.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], double [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfslide1down.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], double [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfslide1down_vf_f64m4_m(vbool16_t mask,
@@ -209,7 +209,7 @@ vfloat64m4_t test_vfslide1down_vf_f64m4_m(vbool16_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfslide1down_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfslide1down.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], double [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfslide1down.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], double [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfslide1down_vf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfslide1up.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfslide1up.c
index 1768558..ff45b8a 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfslide1up.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfslide1up.c
@@ -107,7 +107,7 @@ vfloat64m8_t test_vfslide1up_vf_f64m8(vfloat64m8_t src, double value,
 //
 // CHECK-RV64-LABEL: @test_vfslide1up_vf_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfslide1up.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], float [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfslide1up.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], float [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfslide1up_vf_f32mf2_m(vbool64_t mask,
@@ -120,7 +120,7 @@ vfloat32mf2_t test_vfslide1up_vf_f32mf2_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfslide1up_vf_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfslide1up.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], float [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfslide1up.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], float [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfslide1up_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -132,7 +132,7 @@ vfloat32m1_t test_vfslide1up_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfslide1up_vf_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfslide1up.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], float [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfslide1up.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], float [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfslide1up_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -144,7 +144,7 @@ vfloat32m2_t test_vfslide1up_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfslide1up_vf_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfslide1up.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], float [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfslide1up.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], float [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfslide1up_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -156,7 +156,7 @@ vfloat32m4_t test_vfslide1up_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfslide1up_vf_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfslide1up.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[SRC:%.*]], float [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfslide1up.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[SRC:%.*]], float [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfslide1up_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -168,7 +168,7 @@ vfloat32m8_t test_vfslide1up_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfslide1up_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfslide1up.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], double [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfslide1up.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[SRC:%.*]], double [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfslide1up_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -180,7 +180,7 @@ vfloat64m1_t test_vfslide1up_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfslide1up_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfslide1up.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], double [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfslide1up.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[SRC:%.*]], double [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfslide1up_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -192,7 +192,7 @@ vfloat64m2_t test_vfslide1up_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfslide1up_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfslide1up.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], double [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfslide1up.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[SRC:%.*]], double [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfslide1up_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -204,7 +204,7 @@ vfloat64m4_t test_vfslide1up_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfslide1up_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfslide1up.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], double [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfslide1up.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[SRC:%.*]], double [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfslide1up_vf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfsqrt.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfsqrt.c
index 1d54c45..e7c49f3 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfsqrt.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfsqrt.c
@@ -98,7 +98,7 @@ vfloat64m8_t test_vfsqrt_v_f64m8(vfloat64m8_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfsqrt_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsqrt.mask.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsqrt.mask.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfsqrt_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -109,7 +109,7 @@ vfloat32mf2_t test_vfsqrt_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsqrt_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsqrt.mask.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsqrt.mask.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfsqrt_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -120,7 +120,7 @@ vfloat32m1_t test_vfsqrt_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsqrt_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsqrt.mask.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsqrt.mask.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfsqrt_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -131,7 +131,7 @@ vfloat32m2_t test_vfsqrt_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsqrt_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsqrt.mask.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsqrt.mask.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfsqrt_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -142,7 +142,7 @@ vfloat32m4_t test_vfsqrt_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsqrt_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsqrt.mask.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsqrt.mask.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfsqrt_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -153,7 +153,7 @@ vfloat32m8_t test_vfsqrt_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsqrt_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsqrt.mask.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsqrt.mask.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfsqrt_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -164,7 +164,7 @@ vfloat64m1_t test_vfsqrt_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsqrt_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsqrt.mask.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsqrt.mask.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfsqrt_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -175,7 +175,7 @@ vfloat64m2_t test_vfsqrt_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsqrt_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsqrt.mask.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsqrt.mask.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfsqrt_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -186,7 +186,7 @@ vfloat64m4_t test_vfsqrt_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsqrt_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsqrt.mask.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsqrt.mask.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfsqrt_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfsub.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfsub.c
index 9088855..43a2489 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfsub.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfsub.c
@@ -197,7 +197,7 @@ vfloat64m8_t test_vfsub_vf_f64m8(vfloat64m8_t op1, double op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfsub_vv_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsub.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsub.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfsub_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -209,7 +209,7 @@ vfloat32mf2_t test_vfsub_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsub_vf_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsub.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfsub.mask.nxv1f32.f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfsub_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -220,7 +220,7 @@ vfloat32mf2_t test_vfsub_vf_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsub_vv_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsub.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsub.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfsub_vv_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -232,7 +232,7 @@ vfloat32m1_t test_vfsub_vv_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsub_vf_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsub.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfsub.mask.nxv2f32.f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfsub_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -243,7 +243,7 @@ vfloat32m1_t test_vfsub_vf_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsub_vv_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsub.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsub.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfsub_vv_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -255,7 +255,7 @@ vfloat32m2_t test_vfsub_vv_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsub_vf_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsub.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfsub.mask.nxv4f32.f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfsub_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -266,7 +266,7 @@ vfloat32m2_t test_vfsub_vf_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsub_vv_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsub.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsub.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfsub_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -278,7 +278,7 @@ vfloat32m4_t test_vfsub_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsub_vf_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsub.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfsub.mask.nxv8f32.f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfsub_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -289,7 +289,7 @@ vfloat32m4_t test_vfsub_vf_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsub_vv_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsub.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsub.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfsub_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -301,7 +301,7 @@ vfloat32m8_t test_vfsub_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsub_vf_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsub.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfsub.mask.nxv16f32.f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfsub_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -312,7 +312,7 @@ vfloat32m8_t test_vfsub_vf_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsub_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsub.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsub.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfsub_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -324,7 +324,7 @@ vfloat64m1_t test_vfsub_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsub_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsub.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfsub.mask.nxv1f64.f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfsub_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -335,7 +335,7 @@ vfloat64m1_t test_vfsub_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsub_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsub.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsub.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfsub_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -347,7 +347,7 @@ vfloat64m2_t test_vfsub_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsub_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsub.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfsub.mask.nxv2f64.f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfsub_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -358,7 +358,7 @@ vfloat64m2_t test_vfsub_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsub_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsub.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsub.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfsub_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -370,7 +370,7 @@ vfloat64m4_t test_vfsub_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsub_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsub.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfsub.mask.nxv4f64.f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfsub_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -381,7 +381,7 @@ vfloat64m4_t test_vfsub_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsub_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsub.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsub.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfsub_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -393,7 +393,7 @@ vfloat64m8_t test_vfsub_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfsub_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsub.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfsub.mask.nxv8f64.f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfsub_vf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwadd.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwadd.c
index 9dd2f87..26f6b81 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwadd.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwadd.c
@@ -176,7 +176,7 @@ vfloat64m8_t test_vfwadd_wf_f64m8(vfloat64m8_t op1, float op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfwadd_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwadd.mask.nxv1f64.nxv1f32.nxv1f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwadd.mask.nxv1f64.nxv1f32.nxv1f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwadd_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -188,7 +188,7 @@ vfloat64m1_t test_vfwadd_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwadd_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwadd.mask.nxv1f64.nxv1f32.f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwadd.mask.nxv1f64.nxv1f32.f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwadd_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -199,7 +199,7 @@ vfloat64m1_t test_vfwadd_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwadd_wv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwadd.w.mask.nxv1f64.nxv1f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwadd.w.mask.nxv1f64.nxv1f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwadd_wv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -211,7 +211,7 @@ vfloat64m1_t test_vfwadd_wv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwadd_wf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwadd.w.mask.nxv1f64.f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwadd.w.mask.nxv1f64.f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwadd_wf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -222,7 +222,7 @@ vfloat64m1_t test_vfwadd_wf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwadd_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwadd.mask.nxv2f64.nxv2f32.nxv2f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwadd.mask.nxv2f64.nxv2f32.nxv2f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfwadd_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -234,7 +234,7 @@ vfloat64m2_t test_vfwadd_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwadd_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwadd.mask.nxv2f64.nxv2f32.f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwadd.mask.nxv2f64.nxv2f32.f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfwadd_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -245,7 +245,7 @@ vfloat64m2_t test_vfwadd_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwadd_wv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwadd.w.mask.nxv2f64.nxv2f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwadd.w.mask.nxv2f64.nxv2f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfwadd_wv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -257,7 +257,7 @@ vfloat64m2_t test_vfwadd_wv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwadd_wf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwadd.w.mask.nxv2f64.f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwadd.w.mask.nxv2f64.f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfwadd_wf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -268,7 +268,7 @@ vfloat64m2_t test_vfwadd_wf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwadd_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwadd.mask.nxv4f64.nxv4f32.nxv4f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwadd.mask.nxv4f64.nxv4f32.nxv4f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfwadd_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -280,7 +280,7 @@ vfloat64m4_t test_vfwadd_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwadd_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwadd.mask.nxv4f64.nxv4f32.f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwadd.mask.nxv4f64.nxv4f32.f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfwadd_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -291,7 +291,7 @@ vfloat64m4_t test_vfwadd_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwadd_wv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwadd.w.mask.nxv4f64.nxv4f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwadd.w.mask.nxv4f64.nxv4f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfwadd_wv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -303,7 +303,7 @@ vfloat64m4_t test_vfwadd_wv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwadd_wf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwadd.w.mask.nxv4f64.f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwadd.w.mask.nxv4f64.f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfwadd_wf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -314,7 +314,7 @@ vfloat64m4_t test_vfwadd_wf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwadd_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwadd.mask.nxv8f64.nxv8f32.nxv8f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwadd.mask.nxv8f64.nxv8f32.nxv8f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfwadd_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -326,7 +326,7 @@ vfloat64m8_t test_vfwadd_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwadd_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwadd.mask.nxv8f64.nxv8f32.f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwadd.mask.nxv8f64.nxv8f32.f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfwadd_vf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -337,7 +337,7 @@ vfloat64m8_t test_vfwadd_vf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwadd_wv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwadd.w.mask.nxv8f64.nxv8f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwadd.w.mask.nxv8f64.nxv8f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfwadd_wv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -349,7 +349,7 @@ vfloat64m8_t test_vfwadd_wv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwadd_wf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwadd.w.mask.nxv8f64.f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwadd.w.mask.nxv8f64.f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfwadd_wf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwcvt.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwcvt.c
index 64ecbb9..bea98de 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwcvt.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwcvt.c
@@ -388,7 +388,7 @@ vfloat64m8_t test_vfwcvt_f_f_v_f64m8(vfloat32m4_t src, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_x_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfwcvt.f.x.v.mask.nxv1f32.nxv1i16.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfwcvt.f.x.v.mask.nxv1f32.nxv1i16.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfwcvt_f_x_v_f32mf2_m(vbool64_t mask,
@@ -400,7 +400,7 @@ vfloat32mf2_t test_vfwcvt_f_x_v_f32mf2_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_x_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfwcvt.f.x.v.mask.nxv2f32.nxv2i16.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfwcvt.f.x.v.mask.nxv2f32.nxv2i16.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfwcvt_f_x_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -411,7 +411,7 @@ vfloat32m1_t test_vfwcvt_f_x_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_x_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfwcvt.f.x.v.mask.nxv4f32.nxv4i16.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfwcvt.f.x.v.mask.nxv4f32.nxv4i16.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfwcvt_f_x_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -422,7 +422,7 @@ vfloat32m2_t test_vfwcvt_f_x_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_x_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfwcvt.f.x.v.mask.nxv8f32.nxv8i16.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfwcvt.f.x.v.mask.nxv8f32.nxv8i16.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfwcvt_f_x_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -433,7 +433,7 @@ vfloat32m4_t test_vfwcvt_f_x_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_x_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfwcvt.f.x.v.mask.nxv16f32.nxv16i16.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfwcvt.f.x.v.mask.nxv16f32.nxv16i16.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfwcvt_f_x_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -444,7 +444,7 @@ vfloat32m8_t test_vfwcvt_f_x_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_xu_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfwcvt.f.xu.v.mask.nxv1f32.nxv1i16.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfwcvt.f.xu.v.mask.nxv1f32.nxv1i16.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vfwcvt_f_xu_v_f32mf2_m(vbool64_t mask,
@@ -456,7 +456,7 @@ vfloat32mf2_t test_vfwcvt_f_xu_v_f32mf2_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_xu_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfwcvt.f.xu.v.mask.nxv2f32.nxv2i16.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfwcvt.f.xu.v.mask.nxv2f32.nxv2i16.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vfwcvt_f_xu_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -467,7 +467,7 @@ vfloat32m1_t test_vfwcvt_f_xu_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_xu_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfwcvt.f.xu.v.mask.nxv4f32.nxv4i16.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfwcvt.f.xu.v.mask.nxv4f32.nxv4i16.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vfwcvt_f_xu_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -478,7 +478,7 @@ vfloat32m2_t test_vfwcvt_f_xu_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_xu_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfwcvt.f.xu.v.mask.nxv8f32.nxv8i16.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfwcvt.f.xu.v.mask.nxv8f32.nxv8i16.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vfwcvt_f_xu_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -489,7 +489,7 @@ vfloat32m4_t test_vfwcvt_f_xu_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_xu_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfwcvt.f.xu.v.mask.nxv16f32.nxv16i16.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfwcvt.f.xu.v.mask.nxv16f32.nxv16i16.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[SRC:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vfwcvt_f_xu_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -500,7 +500,7 @@ vfloat32m8_t test_vfwcvt_f_xu_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_x_f_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vfwcvt.x.f.v.mask.nxv1i64.nxv1f32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vfwcvt.x.f.v.mask.nxv1i64.nxv1f32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vfwcvt_x_f_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -511,7 +511,7 @@ vint64m1_t test_vfwcvt_x_f_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_rtz_x_f_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vfwcvt.rtz.x.f.v.mask.nxv1i64.nxv1f32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vfwcvt.rtz.x.f.v.mask.nxv1i64.nxv1f32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vfwcvt_rtz_x_f_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -522,7 +522,7 @@ vint64m1_t test_vfwcvt_rtz_x_f_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_x_f_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vfwcvt.x.f.v.mask.nxv2i64.nxv2f32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vfwcvt.x.f.v.mask.nxv2i64.nxv2f32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vfwcvt_x_f_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -533,7 +533,7 @@ vint64m2_t test_vfwcvt_x_f_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_rtz_x_f_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vfwcvt.rtz.x.f.v.mask.nxv2i64.nxv2f32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vfwcvt.rtz.x.f.v.mask.nxv2i64.nxv2f32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vfwcvt_rtz_x_f_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -544,7 +544,7 @@ vint64m2_t test_vfwcvt_rtz_x_f_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_x_f_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vfwcvt.x.f.v.mask.nxv4i64.nxv4f32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vfwcvt.x.f.v.mask.nxv4i64.nxv4f32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vfwcvt_x_f_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -555,7 +555,7 @@ vint64m4_t test_vfwcvt_x_f_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_rtz_x_f_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vfwcvt.rtz.x.f.v.mask.nxv4i64.nxv4f32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vfwcvt.rtz.x.f.v.mask.nxv4i64.nxv4f32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vfwcvt_rtz_x_f_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -566,7 +566,7 @@ vint64m4_t test_vfwcvt_rtz_x_f_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_x_f_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vfwcvt.x.f.v.mask.nxv8i64.nxv8f32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vfwcvt.x.f.v.mask.nxv8i64.nxv8f32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vfwcvt_x_f_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -577,7 +577,7 @@ vint64m8_t test_vfwcvt_x_f_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_rtz_x_f_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vfwcvt.rtz.x.f.v.mask.nxv8i64.nxv8f32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vfwcvt.rtz.x.f.v.mask.nxv8i64.nxv8f32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vfwcvt_rtz_x_f_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -588,7 +588,7 @@ vint64m8_t test_vfwcvt_rtz_x_f_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_xu_f_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vfwcvt.xu.f.v.mask.nxv1i64.nxv1f32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vfwcvt.xu.f.v.mask.nxv1i64.nxv1f32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vfwcvt_xu_f_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -599,7 +599,7 @@ vuint64m1_t test_vfwcvt_xu_f_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_rtz_xu_f_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vfwcvt.rtz.xu.f.v.mask.nxv1i64.nxv1f32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vfwcvt.rtz.xu.f.v.mask.nxv1i64.nxv1f32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vfwcvt_rtz_xu_f_v_u64m1_m(vbool64_t mask,
@@ -611,7 +611,7 @@ vuint64m1_t test_vfwcvt_rtz_xu_f_v_u64m1_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_xu_f_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vfwcvt.xu.f.v.mask.nxv2i64.nxv2f32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vfwcvt.xu.f.v.mask.nxv2i64.nxv2f32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vfwcvt_xu_f_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -622,7 +622,7 @@ vuint64m2_t test_vfwcvt_xu_f_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_rtz_xu_f_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vfwcvt.rtz.xu.f.v.mask.nxv2i64.nxv2f32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vfwcvt.rtz.xu.f.v.mask.nxv2i64.nxv2f32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vfwcvt_rtz_xu_f_v_u64m2_m(vbool32_t mask,
@@ -634,7 +634,7 @@ vuint64m2_t test_vfwcvt_rtz_xu_f_v_u64m2_m(vbool32_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_xu_f_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vfwcvt.xu.f.v.mask.nxv4i64.nxv4f32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vfwcvt.xu.f.v.mask.nxv4i64.nxv4f32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vfwcvt_xu_f_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -645,7 +645,7 @@ vuint64m4_t test_vfwcvt_xu_f_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_rtz_xu_f_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vfwcvt.rtz.xu.f.v.mask.nxv4i64.nxv4f32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vfwcvt.rtz.xu.f.v.mask.nxv4i64.nxv4f32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vfwcvt_rtz_xu_f_v_u64m4_m(vbool16_t mask,
@@ -657,7 +657,7 @@ vuint64m4_t test_vfwcvt_rtz_xu_f_v_u64m4_m(vbool16_t mask,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_xu_f_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vfwcvt.xu.f.v.mask.nxv8i64.nxv8f32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vfwcvt.xu.f.v.mask.nxv8i64.nxv8f32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vfwcvt_xu_f_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
@@ -668,7 +668,7 @@ vuint64m8_t test_vfwcvt_xu_f_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_rtz_xu_f_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vfwcvt.rtz.xu.f.v.mask.nxv8i64.nxv8f32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vfwcvt.rtz.xu.f.v.mask.nxv8i64.nxv8f32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vfwcvt_rtz_xu_f_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
@@ -679,7 +679,7 @@ vuint64m8_t test_vfwcvt_rtz_xu_f_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_x_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwcvt.f.x.v.mask.nxv1f64.nxv1i32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwcvt.f.x.v.mask.nxv1f64.nxv1i32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwcvt_f_x_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -690,7 +690,7 @@ vfloat64m1_t test_vfwcvt_f_x_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_x_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwcvt.f.x.v.mask.nxv2f64.nxv2i32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwcvt.f.x.v.mask.nxv2f64.nxv2i32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfwcvt_f_x_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -701,7 +701,7 @@ vfloat64m2_t test_vfwcvt_f_x_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_x_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwcvt.f.x.v.mask.nxv4f64.nxv4i32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwcvt.f.x.v.mask.nxv4f64.nxv4i32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfwcvt_f_x_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -712,7 +712,7 @@ vfloat64m4_t test_vfwcvt_f_x_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_x_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwcvt.f.x.v.mask.nxv8f64.nxv8i32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwcvt.f.x.v.mask.nxv8f64.nxv8i32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfwcvt_f_x_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -723,7 +723,7 @@ vfloat64m8_t test_vfwcvt_f_x_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_xu_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwcvt.f.xu.v.mask.nxv1f64.nxv1i32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwcvt.f.xu.v.mask.nxv1f64.nxv1i32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwcvt_f_xu_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -734,7 +734,7 @@ vfloat64m1_t test_vfwcvt_f_xu_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_xu_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwcvt.f.xu.v.mask.nxv2f64.nxv2i32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwcvt.f.xu.v.mask.nxv2f64.nxv2i32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfwcvt_f_xu_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -745,7 +745,7 @@ vfloat64m2_t test_vfwcvt_f_xu_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_xu_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwcvt.f.xu.v.mask.nxv4f64.nxv4i32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwcvt.f.xu.v.mask.nxv4f64.nxv4i32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfwcvt_f_xu_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -756,7 +756,7 @@ vfloat64m4_t test_vfwcvt_f_xu_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_xu_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwcvt.f.xu.v.mask.nxv8f64.nxv8i32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwcvt.f.xu.v.mask.nxv8f64.nxv8i32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfwcvt_f_xu_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -767,7 +767,7 @@ vfloat64m8_t test_vfwcvt_f_xu_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_f_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwcvt.f.f.v.mask.nxv1f64.nxv1f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwcvt.f.f.v.mask.nxv1f64.nxv1f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[SRC:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwcvt_f_f_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -778,7 +778,7 @@ vfloat64m1_t test_vfwcvt_f_f_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_f_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwcvt.f.f.v.mask.nxv2f64.nxv2f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwcvt.f.f.v.mask.nxv2f64.nxv2f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[SRC:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfwcvt_f_f_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -789,7 +789,7 @@ vfloat64m2_t test_vfwcvt_f_f_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_f_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwcvt.f.f.v.mask.nxv4f64.nxv4f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwcvt.f.f.v.mask.nxv4f64.nxv4f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[SRC:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfwcvt_f_f_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -800,7 +800,7 @@ vfloat64m4_t test_vfwcvt_f_f_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwcvt_f_f_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwcvt.f.f.v.mask.nxv8f64.nxv8f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwcvt.f.f.v.mask.nxv8f64.nxv8f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[SRC:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfwcvt_f_f_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwmacc.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwmacc.c
index 208193d..6126db8 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwmacc.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwmacc.c
@@ -96,7 +96,7 @@ vfloat64m8_t test_vfwmacc_vf_f64m8(vfloat64m8_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfwmacc_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwmacc.mask.nxv1f64.nxv1f32.nxv1f32.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwmacc.mask.nxv1f64.nxv1f32.nxv1f32.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwmacc_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -108,7 +108,7 @@ vfloat64m1_t test_vfwmacc_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwmacc_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwmacc.mask.nxv1f64.f32.nxv1f32.i64(<vscale x 1 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwmacc.mask.nxv1f64.f32.nxv1f32.i64(<vscale x 1 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwmacc_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -119,7 +119,7 @@ vfloat64m1_t test_vfwmacc_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwmacc_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwmacc.mask.nxv2f64.nxv2f32.nxv2f32.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwmacc.mask.nxv2f64.nxv2f32.nxv2f32.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfwmacc_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -131,7 +131,7 @@ vfloat64m2_t test_vfwmacc_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwmacc_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwmacc.mask.nxv2f64.f32.nxv2f32.i64(<vscale x 2 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwmacc.mask.nxv2f64.f32.nxv2f32.i64(<vscale x 2 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfwmacc_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -142,7 +142,7 @@ vfloat64m2_t test_vfwmacc_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwmacc_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwmacc.mask.nxv4f64.nxv4f32.nxv4f32.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwmacc.mask.nxv4f64.nxv4f32.nxv4f32.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfwmacc_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -154,7 +154,7 @@ vfloat64m4_t test_vfwmacc_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwmacc_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwmacc.mask.nxv4f64.f32.nxv4f32.i64(<vscale x 4 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwmacc.mask.nxv4f64.f32.nxv4f32.i64(<vscale x 4 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfwmacc_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -165,7 +165,7 @@ vfloat64m4_t test_vfwmacc_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwmacc_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwmacc.mask.nxv8f64.nxv8f32.nxv8f32.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwmacc.mask.nxv8f64.nxv8f32.nxv8f32.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfwmacc_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
@@ -177,7 +177,7 @@ vfloat64m8_t test_vfwmacc_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwmacc_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwmacc.mask.nxv8f64.f32.nxv8f32.i64(<vscale x 8 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwmacc.mask.nxv8f64.f32.nxv8f32.i64(<vscale x 8 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfwmacc_vf_f64m8_m(vbool8_t mask, vfloat64m8_t acc, float op1,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwmsac.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwmsac.c
index b1bf8f5..d2e35dc 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwmsac.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwmsac.c
@@ -96,7 +96,7 @@ vfloat64m8_t test_vfwmsac_vf_f64m8(vfloat64m8_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfwmsac_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwmsac.mask.nxv1f64.nxv1f32.nxv1f32.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwmsac.mask.nxv1f64.nxv1f32.nxv1f32.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwmsac_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -108,7 +108,7 @@ vfloat64m1_t test_vfwmsac_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwmsac_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwmsac.mask.nxv1f64.f32.nxv1f32.i64(<vscale x 1 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwmsac.mask.nxv1f64.f32.nxv1f32.i64(<vscale x 1 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwmsac_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -119,7 +119,7 @@ vfloat64m1_t test_vfwmsac_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwmsac_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwmsac.mask.nxv2f64.nxv2f32.nxv2f32.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwmsac.mask.nxv2f64.nxv2f32.nxv2f32.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfwmsac_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -131,7 +131,7 @@ vfloat64m2_t test_vfwmsac_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwmsac_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwmsac.mask.nxv2f64.f32.nxv2f32.i64(<vscale x 2 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwmsac.mask.nxv2f64.f32.nxv2f32.i64(<vscale x 2 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfwmsac_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -142,7 +142,7 @@ vfloat64m2_t test_vfwmsac_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwmsac_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwmsac.mask.nxv4f64.nxv4f32.nxv4f32.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwmsac.mask.nxv4f64.nxv4f32.nxv4f32.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfwmsac_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -154,7 +154,7 @@ vfloat64m4_t test_vfwmsac_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwmsac_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwmsac.mask.nxv4f64.f32.nxv4f32.i64(<vscale x 4 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwmsac.mask.nxv4f64.f32.nxv4f32.i64(<vscale x 4 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfwmsac_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -165,7 +165,7 @@ vfloat64m4_t test_vfwmsac_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwmsac_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwmsac.mask.nxv8f64.nxv8f32.nxv8f32.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwmsac.mask.nxv8f64.nxv8f32.nxv8f32.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfwmsac_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
@@ -177,7 +177,7 @@ vfloat64m8_t test_vfwmsac_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwmsac_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwmsac.mask.nxv8f64.f32.nxv8f32.i64(<vscale x 8 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwmsac.mask.nxv8f64.f32.nxv8f32.i64(<vscale x 8 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfwmsac_vf_f64m8_m(vbool8_t mask, vfloat64m8_t acc, float op1,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwmul.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwmul.c
index b5ef0b1..eb1ba18 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwmul.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwmul.c
@@ -92,7 +92,7 @@ vfloat64m8_t test_vfwmul_vf_f64m8(vfloat32m4_t op1, float op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfwmul_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwmul.mask.nxv1f64.nxv1f32.nxv1f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwmul.mask.nxv1f64.nxv1f32.nxv1f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwmul_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -104,7 +104,7 @@ vfloat64m1_t test_vfwmul_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwmul_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwmul.mask.nxv1f64.nxv1f32.f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwmul.mask.nxv1f64.nxv1f32.f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwmul_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -115,7 +115,7 @@ vfloat64m1_t test_vfwmul_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwmul_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwmul.mask.nxv2f64.nxv2f32.nxv2f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwmul.mask.nxv2f64.nxv2f32.nxv2f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfwmul_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -127,7 +127,7 @@ vfloat64m2_t test_vfwmul_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwmul_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwmul.mask.nxv2f64.nxv2f32.f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwmul.mask.nxv2f64.nxv2f32.f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfwmul_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -138,7 +138,7 @@ vfloat64m2_t test_vfwmul_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwmul_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwmul.mask.nxv4f64.nxv4f32.nxv4f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwmul.mask.nxv4f64.nxv4f32.nxv4f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfwmul_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -150,7 +150,7 @@ vfloat64m4_t test_vfwmul_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwmul_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwmul.mask.nxv4f64.nxv4f32.f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwmul.mask.nxv4f64.nxv4f32.f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfwmul_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -161,7 +161,7 @@ vfloat64m4_t test_vfwmul_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwmul_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwmul.mask.nxv8f64.nxv8f32.nxv8f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwmul.mask.nxv8f64.nxv8f32.nxv8f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfwmul_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -173,7 +173,7 @@ vfloat64m8_t test_vfwmul_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwmul_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwmul.mask.nxv8f64.nxv8f32.f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwmul.mask.nxv8f64.nxv8f32.f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfwmul_vf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwnmacc.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwnmacc.c
index 7264c4c..f8a260a 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwnmacc.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwnmacc.c
@@ -96,7 +96,7 @@ vfloat64m8_t test_vfwnmacc_vf_f64m8(vfloat64m8_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfwnmacc_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwnmacc.mask.nxv1f64.nxv1f32.nxv1f32.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwnmacc.mask.nxv1f64.nxv1f32.nxv1f32.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwnmacc_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -108,7 +108,7 @@ vfloat64m1_t test_vfwnmacc_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwnmacc_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwnmacc.mask.nxv1f64.f32.nxv1f32.i64(<vscale x 1 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwnmacc.mask.nxv1f64.f32.nxv1f32.i64(<vscale x 1 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwnmacc_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -119,7 +119,7 @@ vfloat64m1_t test_vfwnmacc_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwnmacc_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwnmacc.mask.nxv2f64.nxv2f32.nxv2f32.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwnmacc.mask.nxv2f64.nxv2f32.nxv2f32.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfwnmacc_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -131,7 +131,7 @@ vfloat64m2_t test_vfwnmacc_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwnmacc_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwnmacc.mask.nxv2f64.f32.nxv2f32.i64(<vscale x 2 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwnmacc.mask.nxv2f64.f32.nxv2f32.i64(<vscale x 2 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfwnmacc_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -142,7 +142,7 @@ vfloat64m2_t test_vfwnmacc_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwnmacc_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwnmacc.mask.nxv4f64.nxv4f32.nxv4f32.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwnmacc.mask.nxv4f64.nxv4f32.nxv4f32.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfwnmacc_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -154,7 +154,7 @@ vfloat64m4_t test_vfwnmacc_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwnmacc_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwnmacc.mask.nxv4f64.f32.nxv4f32.i64(<vscale x 4 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwnmacc.mask.nxv4f64.f32.nxv4f32.i64(<vscale x 4 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfwnmacc_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -165,7 +165,7 @@ vfloat64m4_t test_vfwnmacc_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwnmacc_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwnmacc.mask.nxv8f64.nxv8f32.nxv8f32.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwnmacc.mask.nxv8f64.nxv8f32.nxv8f32.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfwnmacc_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
@@ -177,7 +177,7 @@ vfloat64m8_t test_vfwnmacc_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwnmacc_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwnmacc.mask.nxv8f64.f32.nxv8f32.i64(<vscale x 8 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwnmacc.mask.nxv8f64.f32.nxv8f32.i64(<vscale x 8 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfwnmacc_vf_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwnmsac.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwnmsac.c
index a006991..a5e5cc7 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwnmsac.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwnmsac.c
@@ -96,7 +96,7 @@ vfloat64m8_t test_vfwnmsac_vf_f64m8(vfloat64m8_t acc, float op1,
 //
 // CHECK-RV64-LABEL: @test_vfwnmsac_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwnmsac.mask.nxv1f64.nxv1f32.nxv1f32.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwnmsac.mask.nxv1f64.nxv1f32.nxv1f32.i64(<vscale x 1 x double> [[ACC:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwnmsac_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -108,7 +108,7 @@ vfloat64m1_t test_vfwnmsac_vv_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwnmsac_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwnmsac.mask.nxv1f64.f32.nxv1f32.i64(<vscale x 1 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwnmsac.mask.nxv1f64.f32.nxv1f32.i64(<vscale x 1 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwnmsac_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
@@ -119,7 +119,7 @@ vfloat64m1_t test_vfwnmsac_vf_f64m1_m(vbool64_t mask, vfloat64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwnmsac_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwnmsac.mask.nxv2f64.nxv2f32.nxv2f32.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwnmsac.mask.nxv2f64.nxv2f32.nxv2f32.i64(<vscale x 2 x double> [[ACC:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfwnmsac_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -131,7 +131,7 @@ vfloat64m2_t test_vfwnmsac_vv_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwnmsac_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwnmsac.mask.nxv2f64.f32.nxv2f32.i64(<vscale x 2 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwnmsac.mask.nxv2f64.f32.nxv2f32.i64(<vscale x 2 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfwnmsac_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
@@ -142,7 +142,7 @@ vfloat64m2_t test_vfwnmsac_vf_f64m2_m(vbool32_t mask, vfloat64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwnmsac_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwnmsac.mask.nxv4f64.nxv4f32.nxv4f32.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwnmsac.mask.nxv4f64.nxv4f32.nxv4f32.i64(<vscale x 4 x double> [[ACC:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfwnmsac_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -154,7 +154,7 @@ vfloat64m4_t test_vfwnmsac_vv_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwnmsac_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwnmsac.mask.nxv4f64.f32.nxv4f32.i64(<vscale x 4 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwnmsac.mask.nxv4f64.f32.nxv4f32.i64(<vscale x 4 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfwnmsac_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
@@ -165,7 +165,7 @@ vfloat64m4_t test_vfwnmsac_vf_f64m4_m(vbool16_t mask, vfloat64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwnmsac_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwnmsac.mask.nxv8f64.nxv8f32.nxv8f32.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwnmsac.mask.nxv8f64.nxv8f32.nxv8f32.i64(<vscale x 8 x double> [[ACC:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfwnmsac_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
@@ -177,7 +177,7 @@ vfloat64m8_t test_vfwnmsac_vv_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vfwnmsac_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwnmsac.mask.nxv8f64.f32.nxv8f32.i64(<vscale x 8 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwnmsac.mask.nxv8f64.f32.nxv8f32.i64(<vscale x 8 x double> [[ACC:%.*]], float [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfwnmsac_vf_f64m8_m(vbool8_t mask, vfloat64m8_t acc,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwredsum.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwredsum.c
index e70ef4e..be390c6 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwredsum.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwredsum.c
@@ -68,7 +68,7 @@ vfloat64m1_t test_vfwredsum_vs_f32m8_f64m1(vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfwredsum_vs_f32mf2_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwredsum.mask.nxv1f64.nxv1f32.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 1 x float> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwredsum.mask.nxv1f64.nxv1f32.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 1 x float> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwredsum_vs_f32mf2_f64m1_m(vbool64_t mask, vfloat64m1_t dst,
@@ -80,7 +80,7 @@ vfloat64m1_t test_vfwredsum_vs_f32mf2_f64m1_m(vbool64_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfwredsum_vs_f32m1_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwredsum.mask.nxv1f64.nxv2f32.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 2 x float> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwredsum.mask.nxv1f64.nxv2f32.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 2 x float> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwredsum_vs_f32m1_f64m1_m(vbool32_t mask, vfloat64m1_t dst,
@@ -92,7 +92,7 @@ vfloat64m1_t test_vfwredsum_vs_f32m1_f64m1_m(vbool32_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfwredsum_vs_f32m2_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwredsum.mask.nxv1f64.nxv4f32.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 4 x float> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwredsum.mask.nxv1f64.nxv4f32.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 4 x float> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwredsum_vs_f32m2_f64m1_m(vbool16_t mask, vfloat64m1_t dst,
@@ -104,7 +104,7 @@ vfloat64m1_t test_vfwredsum_vs_f32m2_f64m1_m(vbool16_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfwredsum_vs_f32m4_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwredsum.mask.nxv1f64.nxv8f32.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 8 x float> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwredsum.mask.nxv1f64.nxv8f32.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 8 x float> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwredsum_vs_f32m4_f64m1_m(vbool8_t mask, vfloat64m1_t dst,
@@ -116,7 +116,7 @@ vfloat64m1_t test_vfwredsum_vs_f32m4_f64m1_m(vbool8_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfwredsum_vs_f32m8_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwredsum.mask.nxv1f64.nxv16f32.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 16 x float> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwredsum.mask.nxv1f64.nxv16f32.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 16 x float> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwredsum_vs_f32m8_f64m1_m(vbool4_t mask, vfloat64m1_t dst,
@@ -188,7 +188,7 @@ vfloat64m1_t test_vfwredosum_vs_f32m8_f64m1(vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfwredosum_vs_f32mf2_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwredosum.mask.nxv1f64.nxv1f32.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 1 x float> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwredosum.mask.nxv1f64.nxv1f32.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 1 x float> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwredosum_vs_f32mf2_f64m1_m(vbool64_t mask, vfloat64m1_t dst,
@@ -200,7 +200,7 @@ vfloat64m1_t test_vfwredosum_vs_f32mf2_f64m1_m(vbool64_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfwredosum_vs_f32m1_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwredosum.mask.nxv1f64.nxv2f32.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 2 x float> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwredosum.mask.nxv1f64.nxv2f32.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 2 x float> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwredosum_vs_f32m1_f64m1_m(vbool32_t mask, vfloat64m1_t dst,
@@ -212,7 +212,7 @@ vfloat64m1_t test_vfwredosum_vs_f32m1_f64m1_m(vbool32_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfwredosum_vs_f32m2_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwredosum.mask.nxv1f64.nxv4f32.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 4 x float> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwredosum.mask.nxv1f64.nxv4f32.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 4 x float> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwredosum_vs_f32m2_f64m1_m(vbool16_t mask, vfloat64m1_t dst,
@@ -224,7 +224,7 @@ vfloat64m1_t test_vfwredosum_vs_f32m2_f64m1_m(vbool16_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfwredosum_vs_f32m4_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwredosum.mask.nxv1f64.nxv8f32.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 8 x float> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwredosum.mask.nxv1f64.nxv8f32.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 8 x float> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwredosum_vs_f32m4_f64m1_m(vbool8_t mask, vfloat64m1_t dst,
@@ -236,7 +236,7 @@ vfloat64m1_t test_vfwredosum_vs_f32m4_f64m1_m(vbool8_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vfwredosum_vs_f32m8_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwredosum.mask.nxv1f64.nxv16f32.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 16 x float> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwredosum.mask.nxv1f64.nxv16f32.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 16 x float> [[VECTOR:%.*]], <vscale x 1 x double> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwredosum_vs_f32m8_f64m1_m(vbool4_t mask, vfloat64m1_t dst,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwsub.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwsub.c
index 441d7a6..e20c244 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwsub.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vfwsub.c
@@ -176,7 +176,7 @@ vfloat64m8_t test_vfwsub_wf_f64m8(vfloat64m8_t op1, float op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vfwsub_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwsub.mask.nxv1f64.nxv1f32.nxv1f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwsub.mask.nxv1f64.nxv1f32.nxv1f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwsub_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -188,7 +188,7 @@ vfloat64m1_t test_vfwsub_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwsub_vf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwsub.mask.nxv1f64.nxv1f32.f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwsub.mask.nxv1f64.nxv1f32.f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwsub_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -199,7 +199,7 @@ vfloat64m1_t test_vfwsub_vf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwsub_wv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwsub.w.mask.nxv1f64.nxv1f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwsub.w.mask.nxv1f64.nxv1f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwsub_wv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -211,7 +211,7 @@ vfloat64m1_t test_vfwsub_wv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwsub_wf_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwsub.w.mask.nxv1f64.f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfwsub.w.mask.nxv1f64.f32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vfwsub_wf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -222,7 +222,7 @@ vfloat64m1_t test_vfwsub_wf_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwsub_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwsub.mask.nxv2f64.nxv2f32.nxv2f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwsub.mask.nxv2f64.nxv2f32.nxv2f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfwsub_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -234,7 +234,7 @@ vfloat64m2_t test_vfwsub_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwsub_vf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwsub.mask.nxv2f64.nxv2f32.f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwsub.mask.nxv2f64.nxv2f32.f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfwsub_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -245,7 +245,7 @@ vfloat64m2_t test_vfwsub_vf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwsub_wv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwsub.w.mask.nxv2f64.nxv2f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwsub.w.mask.nxv2f64.nxv2f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfwsub_wv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -257,7 +257,7 @@ vfloat64m2_t test_vfwsub_wv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwsub_wf_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwsub.w.mask.nxv2f64.f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfwsub.w.mask.nxv2f64.f32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vfwsub_wf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -268,7 +268,7 @@ vfloat64m2_t test_vfwsub_wf_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwsub_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwsub.mask.nxv4f64.nxv4f32.nxv4f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwsub.mask.nxv4f64.nxv4f32.nxv4f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfwsub_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -280,7 +280,7 @@ vfloat64m4_t test_vfwsub_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwsub_vf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwsub.mask.nxv4f64.nxv4f32.f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwsub.mask.nxv4f64.nxv4f32.f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfwsub_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -291,7 +291,7 @@ vfloat64m4_t test_vfwsub_vf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwsub_wv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwsub.w.mask.nxv4f64.nxv4f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwsub.w.mask.nxv4f64.nxv4f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfwsub_wv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -303,7 +303,7 @@ vfloat64m4_t test_vfwsub_wv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwsub_wf_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwsub.w.mask.nxv4f64.f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfwsub.w.mask.nxv4f64.f32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vfwsub_wf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -314,7 +314,7 @@ vfloat64m4_t test_vfwsub_wf_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwsub_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwsub.mask.nxv8f64.nxv8f32.nxv8f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwsub.mask.nxv8f64.nxv8f32.nxv8f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfwsub_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -326,7 +326,7 @@ vfloat64m8_t test_vfwsub_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwsub_vf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwsub.mask.nxv8f64.nxv8f32.f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwsub.mask.nxv8f64.nxv8f32.f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfwsub_vf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -337,7 +337,7 @@ vfloat64m8_t test_vfwsub_vf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwsub_wv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwsub.w.mask.nxv8f64.nxv8f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwsub.w.mask.nxv8f64.nxv8f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfwsub_wv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -349,7 +349,7 @@ vfloat64m8_t test_vfwsub_wv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vfwsub_wf_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwsub.w.mask.nxv8f64.f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfwsub.w.mask.nxv8f64.f32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vfwsub_wf_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vid.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vid.c
index 50b34a7..49b319e 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vid.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vid.c
@@ -7,7 +7,7 @@
 //
 // CHECK-RV64-LABEL: @test_vid_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vid.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vid.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vid_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
@@ -18,7 +18,7 @@ vuint8mf8_t test_vid_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vid_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vid.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vid.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vid_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
@@ -29,7 +29,7 @@ vuint8mf4_t test_vid_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vid_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vid.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vid.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vid_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
@@ -40,7 +40,7 @@ vuint8mf2_t test_vid_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vid_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vid.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vid.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vid_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, size_t vl) {
@@ -50,7 +50,7 @@ vuint8m1_t test_vid_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vid_v_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vid.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vid.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vid_v_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, size_t vl) {
@@ -60,7 +60,7 @@ vuint8m2_t test_vid_v_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vid_v_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vid.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vid.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vid_v_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, size_t vl) {
@@ -70,7 +70,7 @@ vuint8m4_t test_vid_v_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vid_v_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vid.mask.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vid.mask.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vid_v_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, size_t vl) {
@@ -80,7 +80,7 @@ vuint8m8_t test_vid_v_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vid_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vid.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vid.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vid_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -91,7 +91,7 @@ vuint16mf4_t test_vid_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vid_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vid.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vid.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vid_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -102,7 +102,7 @@ vuint16mf2_t test_vid_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vid_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vid.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vid.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vid_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -113,7 +113,7 @@ vuint16m1_t test_vid_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vid_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vid.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vid.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vid_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -124,7 +124,7 @@ vuint16m2_t test_vid_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vid_v_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vid.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vid.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vid_v_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -135,7 +135,7 @@ vuint16m4_t test_vid_v_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vid_v_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vid.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vid.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vid_v_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -146,7 +146,7 @@ vuint16m8_t test_vid_v_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vid_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vid.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vid.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vid_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -157,7 +157,7 @@ vuint32mf2_t test_vid_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vid_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vid.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vid.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vid_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -168,7 +168,7 @@ vuint32m1_t test_vid_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vid_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vid.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vid.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vid_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -179,7 +179,7 @@ vuint32m2_t test_vid_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vid_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vid.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vid.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vid_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -190,7 +190,7 @@ vuint32m4_t test_vid_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vid_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vid.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vid.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vid_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -201,7 +201,7 @@ vuint32m8_t test_vid_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vid_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vid.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vid.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vid_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -212,7 +212,7 @@ vuint64m1_t test_vid_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vid_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vid.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vid.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vid_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -223,7 +223,7 @@ vuint64m2_t test_vid_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vid_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vid.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vid.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vid_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -234,7 +234,7 @@ vuint64m4_t test_vid_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vid_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vid.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vid.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vid_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/viota.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/viota.c
index fbd199e..d90f8c0 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/viota.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/viota.c
@@ -7,7 +7,7 @@
 //
 // CHECK-RV64-LABEL: @test_viota_m_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.viota.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.viota.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_viota_m_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
@@ -18,7 +18,7 @@ vuint8mf8_t test_viota_m_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_viota_m_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.viota.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.viota.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_viota_m_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
@@ -29,7 +29,7 @@ vuint8mf4_t test_viota_m_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_viota_m_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.viota.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.viota.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_viota_m_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
@@ -40,7 +40,7 @@ vuint8mf2_t test_viota_m_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_viota_m_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.viota.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.viota.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_viota_m_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
@@ -51,7 +51,7 @@ vuint8m1_t test_viota_m_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_viota_m_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.viota.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i1> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.viota.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i1> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_viota_m_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
@@ -62,7 +62,7 @@ vuint8m2_t test_viota_m_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_viota_m_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.viota.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i1> [[OP1:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.viota.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i1> [[OP1:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_viota_m_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
@@ -73,7 +73,7 @@ vuint8m4_t test_viota_m_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_viota_m_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.viota.mask.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i1> [[OP1:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.viota.mask.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i1> [[OP1:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_viota_m_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
@@ -84,7 +84,7 @@ vuint8m8_t test_viota_m_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_viota_m_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.viota.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.viota.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_viota_m_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -95,7 +95,7 @@ vuint16mf4_t test_viota_m_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_viota_m_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.viota.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.viota.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_viota_m_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -106,7 +106,7 @@ vuint16mf2_t test_viota_m_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_viota_m_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.viota.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.viota.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_viota_m_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -117,7 +117,7 @@ vuint16m1_t test_viota_m_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_viota_m_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.viota.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.viota.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_viota_m_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -128,7 +128,7 @@ vuint16m2_t test_viota_m_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_viota_m_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.viota.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i1> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.viota.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i1> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_viota_m_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -139,7 +139,7 @@ vuint16m4_t test_viota_m_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_viota_m_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.viota.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i1> [[OP1:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.viota.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i1> [[OP1:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_viota_m_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -150,7 +150,7 @@ vuint16m8_t test_viota_m_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_viota_m_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.viota.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.viota.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_viota_m_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -161,7 +161,7 @@ vuint32mf2_t test_viota_m_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_viota_m_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.viota.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.viota.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_viota_m_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -172,7 +172,7 @@ vuint32m1_t test_viota_m_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_viota_m_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.viota.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.viota.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_viota_m_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -183,7 +183,7 @@ vuint32m2_t test_viota_m_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_viota_m_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.viota.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.viota.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_viota_m_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -194,7 +194,7 @@ vuint32m4_t test_viota_m_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_viota_m_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.viota.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i1> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.viota.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i1> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_viota_m_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -205,7 +205,7 @@ vuint32m8_t test_viota_m_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_viota_m_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.viota.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.viota.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_viota_m_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -216,7 +216,7 @@ vuint64m1_t test_viota_m_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_viota_m_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.viota.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.viota.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_viota_m_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -227,7 +227,7 @@ vuint64m2_t test_viota_m_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_viota_m_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.viota.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.viota.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_viota_m_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -238,7 +238,7 @@ vuint64m4_t test_viota_m_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_viota_m_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.viota.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.viota.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_viota_m_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vle.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vle.c
index 7122973..1a069c2 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vle.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vle.c
@@ -9,7 +9,7 @@
 // CHECK-RV64-LABEL: @test_vle8_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vle.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vle.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]
 //
 vint8mf8_t test_vle8_v_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, const int8_t *base, size_t vl) {
@@ -20,7 +20,7 @@ vint8mf8_t test_vle8_v_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, const int8_
 // CHECK-RV64-LABEL: @test_vle8_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vle.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vle.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP1]]
 //
 vint8mf4_t test_vle8_v_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, const int8_t *base, size_t vl) {
@@ -31,7 +31,7 @@ vint8mf4_t test_vle8_v_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, const int8_
 // CHECK-RV64-LABEL: @test_vle8_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vle.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vle.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP1]]
 //
 vint8mf2_t test_vle8_v_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, const int8_t *base, size_t vl) {
@@ -42,7 +42,7 @@ vint8mf2_t test_vle8_v_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, const int8_
 // CHECK-RV64-LABEL: @test_vle8_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vle.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vle.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP1]]
 //
 vint8m1_t test_vle8_v_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, const int8_t *base, size_t vl) {
@@ -53,7 +53,7 @@ vint8m1_t test_vle8_v_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, const int8_t *b
 // CHECK-RV64-LABEL: @test_vle8_v_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vle.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vle.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP1]]
 //
 vint8m2_t test_vle8_v_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, const int8_t *base, size_t vl) {
@@ -64,7 +64,7 @@ vint8m2_t test_vle8_v_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, const int8_t *b
 // CHECK-RV64-LABEL: @test_vle8_v_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vle.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vle.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP1]]
 //
 vint8m4_t test_vle8_v_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, const int8_t *base, size_t vl) {
@@ -75,7 +75,7 @@ vint8m4_t test_vle8_v_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, const int8_t *b
 // CHECK-RV64-LABEL: @test_vle8_v_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vle.mask.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vle.mask.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP1]]
 //
 vint8m8_t test_vle8_v_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, const int8_t *base, size_t vl) {
@@ -86,7 +86,7 @@ vint8m8_t test_vle8_v_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, const int8_t *b
 // CHECK-RV64-LABEL: @test_vle16_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vle.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vle.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP1]]
 //
 vint16mf4_t test_vle16_v_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, const int16_t *base, size_t vl) {
@@ -97,7 +97,7 @@ vint16mf4_t test_vle16_v_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vle16_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vle.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vle.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP1]]
 //
 vint16mf2_t test_vle16_v_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, const int16_t *base, size_t vl) {
@@ -108,7 +108,7 @@ vint16mf2_t test_vle16_v_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vle16_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vle.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vle.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP1]]
 //
 vint16m1_t test_vle16_v_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, const int16_t *base, size_t vl) {
@@ -119,7 +119,7 @@ vint16m1_t test_vle16_v_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, const int1
 // CHECK-RV64-LABEL: @test_vle16_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vle.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vle.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP1]]
 //
 vint16m2_t test_vle16_v_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, const int16_t *base, size_t vl) {
@@ -130,7 +130,7 @@ vint16m2_t test_vle16_v_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, const int16
 // CHECK-RV64-LABEL: @test_vle16_v_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vle.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vle.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP1]]
 //
 vint16m4_t test_vle16_v_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, const int16_t *base, size_t vl) {
@@ -141,7 +141,7 @@ vint16m4_t test_vle16_v_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, const int16
 // CHECK-RV64-LABEL: @test_vle16_v_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vle.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vle.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP1]]
 //
 vint16m8_t test_vle16_v_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, const int16_t *base, size_t vl) {
@@ -152,7 +152,7 @@ vint16m8_t test_vle16_v_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, const int16
 // CHECK-RV64-LABEL: @test_vle32_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vle.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vle.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vle32_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, const int32_t *base, size_t vl) {
@@ -163,7 +163,7 @@ vint32mf2_t test_vle32_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vle32_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vle.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vle.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vle32_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, const int32_t *base, size_t vl) {
@@ -174,7 +174,7 @@ vint32m1_t test_vle32_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, const int3
 // CHECK-RV64-LABEL: @test_vle32_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vle.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vle.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vle32_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, const int32_t *base, size_t vl) {
@@ -185,7 +185,7 @@ vint32m2_t test_vle32_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, const int3
 // CHECK-RV64-LABEL: @test_vle32_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vle.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vle.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vle32_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, const int32_t *base, size_t vl) {
@@ -196,7 +196,7 @@ vint32m4_t test_vle32_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, const int32
 // CHECK-RV64-LABEL: @test_vle32_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vle.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vle.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vle32_v_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, const int32_t *base, size_t vl) {
@@ -207,7 +207,7 @@ vint32m8_t test_vle32_v_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, const int32
 // CHECK-RV64-LABEL: @test_vle64_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vle.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vle.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vle64_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, const int64_t *base, size_t vl) {
@@ -218,7 +218,7 @@ vint64m1_t test_vle64_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, const int6
 // CHECK-RV64-LABEL: @test_vle64_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vle.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vle.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vle64_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, const int64_t *base, size_t vl) {
@@ -229,7 +229,7 @@ vint64m2_t test_vle64_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, const int6
 // CHECK-RV64-LABEL: @test_vle64_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vle.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vle.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vle64_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, const int64_t *base, size_t vl) {
@@ -240,7 +240,7 @@ vint64m4_t test_vle64_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, const int6
 // CHECK-RV64-LABEL: @test_vle64_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vle.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vle.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vle64_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, const int64_t *base, size_t vl) {
@@ -251,7 +251,7 @@ vint64m8_t test_vle64_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, const int64
 // CHECK-RV64-LABEL: @test_vle8_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vle.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vle.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]
 //
 vuint8mf8_t test_vle8_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, const uint8_t *base, size_t vl) {
@@ -262,7 +262,7 @@ vuint8mf8_t test_vle8_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, const uin
 // CHECK-RV64-LABEL: @test_vle8_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vle.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vle.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP1]]
 //
 vuint8mf4_t test_vle8_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, const uint8_t *base, size_t vl) {
@@ -273,7 +273,7 @@ vuint8mf4_t test_vle8_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, const uin
 // CHECK-RV64-LABEL: @test_vle8_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vle.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vle.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP1]]
 //
 vuint8mf2_t test_vle8_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, const uint8_t *base, size_t vl) {
@@ -284,7 +284,7 @@ vuint8mf2_t test_vle8_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, const uin
 // CHECK-RV64-LABEL: @test_vle8_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vle.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vle.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP1]]
 //
 vuint8m1_t test_vle8_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, const uint8_t *base, size_t vl) {
@@ -295,7 +295,7 @@ vuint8m1_t test_vle8_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, const uint8_t
 // CHECK-RV64-LABEL: @test_vle8_v_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vle.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vle.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP1]]
 //
 vuint8m2_t test_vle8_v_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, const uint8_t *base, size_t vl) {
@@ -306,7 +306,7 @@ vuint8m2_t test_vle8_v_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, const uint8_t
 // CHECK-RV64-LABEL: @test_vle8_v_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vle.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vle.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP1]]
 //
 vuint8m4_t test_vle8_v_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, const uint8_t *base, size_t vl) {
@@ -317,7 +317,7 @@ vuint8m4_t test_vle8_v_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, const uint8_t
 // CHECK-RV64-LABEL: @test_vle8_v_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vle.mask.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vle.mask.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP1]]
 //
 vuint8m8_t test_vle8_v_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, const uint8_t *base, size_t vl) {
@@ -328,7 +328,7 @@ vuint8m8_t test_vle8_v_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, const uint8_t
 // CHECK-RV64-LABEL: @test_vle16_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vle.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vle.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP1]]
 //
 vuint16mf4_t test_vle16_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, const uint16_t *base, size_t vl) {
@@ -339,7 +339,7 @@ vuint16mf4_t test_vle16_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vle16_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vle.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vle.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP1]]
 //
 vuint16mf2_t test_vle16_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, const uint16_t *base, size_t vl) {
@@ -350,7 +350,7 @@ vuint16mf2_t test_vle16_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vle16_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vle.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vle.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP1]]
 //
 vuint16m1_t test_vle16_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, const uint16_t *base, size_t vl) {
@@ -361,7 +361,7 @@ vuint16m1_t test_vle16_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, const ui
 // CHECK-RV64-LABEL: @test_vle16_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vle.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vle.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP1]]
 //
 vuint16m2_t test_vle16_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, const uint16_t *base, size_t vl) {
@@ -372,7 +372,7 @@ vuint16m2_t test_vle16_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, const uin
 // CHECK-RV64-LABEL: @test_vle16_v_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vle.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vle.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP1]]
 //
 vuint16m4_t test_vle16_v_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, const uint16_t *base, size_t vl) {
@@ -383,7 +383,7 @@ vuint16m4_t test_vle16_v_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, const uin
 // CHECK-RV64-LABEL: @test_vle16_v_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vle.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vle.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP1]]
 //
 vuint16m8_t test_vle16_v_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, const uint16_t *base, size_t vl) {
@@ -394,7 +394,7 @@ vuint16m8_t test_vle16_v_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, const uin
 // CHECK-RV64-LABEL: @test_vle32_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vle.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vle.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vle32_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, const uint32_t *base, size_t vl) {
@@ -405,7 +405,7 @@ vuint32mf2_t test_vle32_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vle32_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vle.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vle.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vle32_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, const uint32_t *base, size_t vl) {
@@ -416,7 +416,7 @@ vuint32m1_t test_vle32_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, const ui
 // CHECK-RV64-LABEL: @test_vle32_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vle.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vle.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vle32_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, const uint32_t *base, size_t vl) {
@@ -427,7 +427,7 @@ vuint32m2_t test_vle32_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, const ui
 // CHECK-RV64-LABEL: @test_vle32_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vle.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vle.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vle32_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, const uint32_t *base, size_t vl) {
@@ -438,7 +438,7 @@ vuint32m4_t test_vle32_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, const uin
 // CHECK-RV64-LABEL: @test_vle32_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vle.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vle.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vle32_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, const uint32_t *base, size_t vl) {
@@ -449,7 +449,7 @@ vuint32m8_t test_vle32_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, const uin
 // CHECK-RV64-LABEL: @test_vle64_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vle.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vle.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vle64_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, const uint64_t *base, size_t vl) {
@@ -460,7 +460,7 @@ vuint64m1_t test_vle64_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, const ui
 // CHECK-RV64-LABEL: @test_vle64_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vle.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vle.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vle64_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, const uint64_t *base, size_t vl) {
@@ -471,7 +471,7 @@ vuint64m2_t test_vle64_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, const ui
 // CHECK-RV64-LABEL: @test_vle64_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vle.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vle.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vle64_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, const uint64_t *base, size_t vl) {
@@ -482,7 +482,7 @@ vuint64m4_t test_vle64_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, const ui
 // CHECK-RV64-LABEL: @test_vle64_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vle.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vle.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vle64_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, const uint64_t *base, size_t vl) {
@@ -493,7 +493,7 @@ vuint64m8_t test_vle64_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, const uin
 // CHECK-RV64-LABEL: @test_vle32_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vle.mask.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vle.mask.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP1]]
 //
 vfloat32mf2_t test_vle32_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff, const float *base, size_t vl) {
@@ -504,7 +504,7 @@ vfloat32mf2_t test_vle32_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vle32_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vle.mask.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vle.mask.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP1]]
 //
 vfloat32m1_t test_vle32_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, const float *base, size_t vl) {
@@ -515,7 +515,7 @@ vfloat32m1_t test_vle32_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vle32_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vle.mask.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vle.mask.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 vfloat32m2_t test_vle32_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, const float *base, size_t vl) {
@@ -526,7 +526,7 @@ vfloat32m2_t test_vle32_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vle32_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vle.mask.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vle.mask.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP1]]
 //
 vfloat32m4_t test_vle32_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, const float *base, size_t vl) {
@@ -537,7 +537,7 @@ vfloat32m4_t test_vle32_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, const f
 // CHECK-RV64-LABEL: @test_vle32_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vle.mask.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vle.mask.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP1]]
 //
 vfloat32m8_t test_vle32_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff, const float *base, size_t vl) {
@@ -548,7 +548,7 @@ vfloat32m8_t test_vle32_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff, const f
 // CHECK-RV64-LABEL: @test_vle64_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vle.mask.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vle.mask.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP1]]
 //
 vfloat64m1_t test_vle64_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, const double *base, size_t vl) {
@@ -559,7 +559,7 @@ vfloat64m1_t test_vle64_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vle64_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vle.mask.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vle.mask.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 vfloat64m2_t test_vle64_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, const double *base, size_t vl) {
@@ -570,7 +570,7 @@ vfloat64m2_t test_vle64_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vle64_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vle.mask.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vle.mask.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP1]]
 //
 vfloat64m4_t test_vle64_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, const double *base, size_t vl) {
@@ -581,7 +581,7 @@ vfloat64m4_t test_vle64_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vle64_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vle.mask.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vle.mask.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP1]]
 //
 vfloat64m8_t test_vle64_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff, const double *base, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vloxei.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vloxei.c
index 167292d..e37ddb0 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vloxei.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vloxei.c
@@ -2111,7 +2111,7 @@ vfloat64m8_t test_vloxei64_v_f64m8(const double *base, vuint64m8_t bindex, size_
 // CHECK-RV64-LABEL: @test_vloxei8_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vloxei.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vloxei.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]
 //
 vint8mf8_t test_vloxei8_v_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, const int8_t *base, vuint8mf8_t bindex, size_t vl) {
@@ -2122,7 +2122,7 @@ vint8mf8_t test_vloxei8_v_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei8_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vloxei.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vloxei.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP1]]
 //
 vint8mf4_t test_vloxei8_v_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, const int8_t *base, vuint8mf4_t bindex, size_t vl) {
@@ -2133,7 +2133,7 @@ vint8mf4_t test_vloxei8_v_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei8_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vloxei.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vloxei.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP1]]
 //
 vint8mf2_t test_vloxei8_v_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, const int8_t *base, vuint8mf2_t bindex, size_t vl) {
@@ -2144,7 +2144,7 @@ vint8mf2_t test_vloxei8_v_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei8_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vloxei.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vloxei.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP1]]
 //
 vint8m1_t test_vloxei8_v_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, const int8_t *base, vuint8m1_t bindex, size_t vl) {
@@ -2155,7 +2155,7 @@ vint8m1_t test_vloxei8_v_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, const int8_t
 // CHECK-RV64-LABEL: @test_vloxei8_v_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vloxei.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vloxei.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP1]]
 //
 vint8m2_t test_vloxei8_v_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, const int8_t *base, vuint8m2_t bindex, size_t vl) {
@@ -2166,7 +2166,7 @@ vint8m2_t test_vloxei8_v_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, const int8_t
 // CHECK-RV64-LABEL: @test_vloxei8_v_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vloxei.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vloxei.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP1]]
 //
 vint8m4_t test_vloxei8_v_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, const int8_t *base, vuint8m4_t bindex, size_t vl) {
@@ -2177,7 +2177,7 @@ vint8m4_t test_vloxei8_v_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, const int8_t
 // CHECK-RV64-LABEL: @test_vloxei8_v_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vloxei.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vloxei.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP1]]
 //
 vint8m8_t test_vloxei8_v_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, const int8_t *base, vuint8m8_t bindex, size_t vl) {
@@ -2188,7 +2188,7 @@ vint8m8_t test_vloxei8_v_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, const int8_t
 // CHECK-RV64-LABEL: @test_vloxei16_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vloxei.mask.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vloxei.mask.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]
 //
 vint8mf8_t test_vloxei16_v_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, const int8_t *base, vuint16mf4_t bindex, size_t vl) {
@@ -2199,7 +2199,7 @@ vint8mf8_t test_vloxei16_v_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei16_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vloxei.mask.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vloxei.mask.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP1]]
 //
 vint8mf4_t test_vloxei16_v_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, const int8_t *base, vuint16mf2_t bindex, size_t vl) {
@@ -2210,7 +2210,7 @@ vint8mf4_t test_vloxei16_v_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei16_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vloxei.mask.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vloxei.mask.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP1]]
 //
 vint8mf2_t test_vloxei16_v_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, const int8_t *base, vuint16m1_t bindex, size_t vl) {
@@ -2221,7 +2221,7 @@ vint8mf2_t test_vloxei16_v_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei16_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vloxei.mask.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vloxei.mask.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP1]]
 //
 vint8m1_t test_vloxei16_v_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, const int8_t *base, vuint16m2_t bindex, size_t vl) {
@@ -2232,7 +2232,7 @@ vint8m1_t test_vloxei16_v_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, const int8_
 // CHECK-RV64-LABEL: @test_vloxei16_v_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vloxei.mask.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vloxei.mask.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP1]]
 //
 vint8m2_t test_vloxei16_v_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, const int8_t *base, vuint16m4_t bindex, size_t vl) {
@@ -2243,7 +2243,7 @@ vint8m2_t test_vloxei16_v_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, const int8_
 // CHECK-RV64-LABEL: @test_vloxei16_v_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vloxei.mask.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vloxei.mask.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP1]]
 //
 vint8m4_t test_vloxei16_v_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, const int8_t *base, vuint16m8_t bindex, size_t vl) {
@@ -2254,7 +2254,7 @@ vint8m4_t test_vloxei16_v_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, const int8_
 // CHECK-RV64-LABEL: @test_vloxei32_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vloxei.mask.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vloxei.mask.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]
 //
 vint8mf8_t test_vloxei32_v_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, const int8_t *base, vuint32mf2_t bindex, size_t vl) {
@@ -2265,7 +2265,7 @@ vint8mf8_t test_vloxei32_v_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei32_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vloxei.mask.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vloxei.mask.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP1]]
 //
 vint8mf4_t test_vloxei32_v_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, const int8_t *base, vuint32m1_t bindex, size_t vl) {
@@ -2276,7 +2276,7 @@ vint8mf4_t test_vloxei32_v_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei32_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vloxei.mask.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vloxei.mask.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP1]]
 //
 vint8mf2_t test_vloxei32_v_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, const int8_t *base, vuint32m2_t bindex, size_t vl) {
@@ -2287,7 +2287,7 @@ vint8mf2_t test_vloxei32_v_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei32_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vloxei.mask.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vloxei.mask.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP1]]
 //
 vint8m1_t test_vloxei32_v_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, const int8_t *base, vuint32m4_t bindex, size_t vl) {
@@ -2298,7 +2298,7 @@ vint8m1_t test_vloxei32_v_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, const int8_
 // CHECK-RV64-LABEL: @test_vloxei32_v_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vloxei.mask.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vloxei.mask.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP1]]
 //
 vint8m2_t test_vloxei32_v_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, const int8_t *base, vuint32m8_t bindex, size_t vl) {
@@ -2309,7 +2309,7 @@ vint8m2_t test_vloxei32_v_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, const int8_
 // CHECK-RV64-LABEL: @test_vloxei64_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vloxei.mask.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vloxei.mask.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]
 //
 vint8mf8_t test_vloxei64_v_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, const int8_t *base, vuint64m1_t bindex, size_t vl) {
@@ -2320,7 +2320,7 @@ vint8mf8_t test_vloxei64_v_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei64_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vloxei.mask.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vloxei.mask.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP1]]
 //
 vint8mf4_t test_vloxei64_v_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, const int8_t *base, vuint64m2_t bindex, size_t vl) {
@@ -2331,7 +2331,7 @@ vint8mf4_t test_vloxei64_v_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei64_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vloxei.mask.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vloxei.mask.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP1]]
 //
 vint8mf2_t test_vloxei64_v_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, const int8_t *base, vuint64m4_t bindex, size_t vl) {
@@ -2342,7 +2342,7 @@ vint8mf2_t test_vloxei64_v_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei64_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vloxei.mask.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vloxei.mask.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP1]]
 //
 vint8m1_t test_vloxei64_v_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, const int8_t *base, vuint64m8_t bindex, size_t vl) {
@@ -2353,7 +2353,7 @@ vint8m1_t test_vloxei64_v_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, const int8_
 // CHECK-RV64-LABEL: @test_vloxei8_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vloxei.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vloxei.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP1]]
 //
 vint16mf4_t test_vloxei8_v_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, const int16_t *base, vuint8mf8_t bindex, size_t vl) {
@@ -2364,7 +2364,7 @@ vint16mf4_t test_vloxei8_v_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei8_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vloxei.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vloxei.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP1]]
 //
 vint16mf2_t test_vloxei8_v_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, const int16_t *base, vuint8mf4_t bindex, size_t vl) {
@@ -2375,7 +2375,7 @@ vint16mf2_t test_vloxei8_v_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei8_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vloxei.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vloxei.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP1]]
 //
 vint16m1_t test_vloxei8_v_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, const int16_t *base, vuint8mf2_t bindex, size_t vl) {
@@ -2386,7 +2386,7 @@ vint16m1_t test_vloxei8_v_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei8_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vloxei.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vloxei.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP1]]
 //
 vint16m2_t test_vloxei8_v_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, const int16_t *base, vuint8m1_t bindex, size_t vl) {
@@ -2397,7 +2397,7 @@ vint16m2_t test_vloxei8_v_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, const int
 // CHECK-RV64-LABEL: @test_vloxei8_v_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vloxei.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vloxei.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP1]]
 //
 vint16m4_t test_vloxei8_v_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, const int16_t *base, vuint8m2_t bindex, size_t vl) {
@@ -2408,7 +2408,7 @@ vint16m4_t test_vloxei8_v_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, const int
 // CHECK-RV64-LABEL: @test_vloxei8_v_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vloxei.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vloxei.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP1]]
 //
 vint16m8_t test_vloxei8_v_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, const int16_t *base, vuint8m4_t bindex, size_t vl) {
@@ -2419,7 +2419,7 @@ vint16m8_t test_vloxei8_v_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, const int
 // CHECK-RV64-LABEL: @test_vloxei16_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vloxei.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vloxei.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP1]]
 //
 vint16mf4_t test_vloxei16_v_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, const int16_t *base, vuint16mf4_t bindex, size_t vl) {
@@ -2430,7 +2430,7 @@ vint16mf4_t test_vloxei16_v_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vloxei16_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vloxei.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vloxei.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP1]]
 //
 vint16mf2_t test_vloxei16_v_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, const int16_t *base, vuint16mf2_t bindex, size_t vl) {
@@ -2441,7 +2441,7 @@ vint16mf2_t test_vloxei16_v_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vloxei16_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vloxei.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vloxei.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP1]]
 //
 vint16m1_t test_vloxei16_v_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, const int16_t *base, vuint16m1_t bindex, size_t vl) {
@@ -2452,7 +2452,7 @@ vint16m1_t test_vloxei16_v_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei16_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vloxei.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vloxei.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP1]]
 //
 vint16m2_t test_vloxei16_v_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, const int16_t *base, vuint16m2_t bindex, size_t vl) {
@@ -2463,7 +2463,7 @@ vint16m2_t test_vloxei16_v_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei16_v_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vloxei.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vloxei.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP1]]
 //
 vint16m4_t test_vloxei16_v_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, const int16_t *base, vuint16m4_t bindex, size_t vl) {
@@ -2474,7 +2474,7 @@ vint16m4_t test_vloxei16_v_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei16_v_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vloxei.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vloxei.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP1]]
 //
 vint16m8_t test_vloxei16_v_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, const int16_t *base, vuint16m8_t bindex, size_t vl) {
@@ -2485,7 +2485,7 @@ vint16m8_t test_vloxei16_v_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei32_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vloxei.mask.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vloxei.mask.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP1]]
 //
 vint16mf4_t test_vloxei32_v_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, const int16_t *base, vuint32mf2_t bindex, size_t vl) {
@@ -2496,7 +2496,7 @@ vint16mf4_t test_vloxei32_v_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vloxei32_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vloxei.mask.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vloxei.mask.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP1]]
 //
 vint16mf2_t test_vloxei32_v_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, const int16_t *base, vuint32m1_t bindex, size_t vl) {
@@ -2507,7 +2507,7 @@ vint16mf2_t test_vloxei32_v_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vloxei32_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vloxei.mask.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vloxei.mask.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP1]]
 //
 vint16m1_t test_vloxei32_v_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, const int16_t *base, vuint32m2_t bindex, size_t vl) {
@@ -2518,7 +2518,7 @@ vint16m1_t test_vloxei32_v_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei32_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vloxei.mask.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vloxei.mask.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP1]]
 //
 vint16m2_t test_vloxei32_v_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, const int16_t *base, vuint32m4_t bindex, size_t vl) {
@@ -2529,7 +2529,7 @@ vint16m2_t test_vloxei32_v_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei32_v_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vloxei.mask.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vloxei.mask.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP1]]
 //
 vint16m4_t test_vloxei32_v_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, const int16_t *base, vuint32m8_t bindex, size_t vl) {
@@ -2540,7 +2540,7 @@ vint16m4_t test_vloxei32_v_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei64_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vloxei.mask.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vloxei.mask.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP1]]
 //
 vint16mf4_t test_vloxei64_v_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, const int16_t *base, vuint64m1_t bindex, size_t vl) {
@@ -2551,7 +2551,7 @@ vint16mf4_t test_vloxei64_v_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vloxei64_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vloxei.mask.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vloxei.mask.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP1]]
 //
 vint16mf2_t test_vloxei64_v_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, const int16_t *base, vuint64m2_t bindex, size_t vl) {
@@ -2562,7 +2562,7 @@ vint16mf2_t test_vloxei64_v_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vloxei64_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vloxei.mask.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vloxei.mask.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP1]]
 //
 vint16m1_t test_vloxei64_v_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, const int16_t *base, vuint64m4_t bindex, size_t vl) {
@@ -2573,7 +2573,7 @@ vint16m1_t test_vloxei64_v_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei64_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vloxei.mask.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vloxei.mask.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP1]]
 //
 vint16m2_t test_vloxei64_v_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, const int16_t *base, vuint64m8_t bindex, size_t vl) {
@@ -2584,7 +2584,7 @@ vint16m2_t test_vloxei64_v_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei8_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vloxei.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vloxei.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vloxei8_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, const int32_t *base, vuint8mf8_t bindex, size_t vl) {
@@ -2595,7 +2595,7 @@ vint32mf2_t test_vloxei8_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei8_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vloxei.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vloxei.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vloxei8_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, const int32_t *base, vuint8mf4_t bindex, size_t vl) {
@@ -2606,7 +2606,7 @@ vint32m1_t test_vloxei8_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei8_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vloxei.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vloxei.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vloxei8_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, const int32_t *base, vuint8mf2_t bindex, size_t vl) {
@@ -2617,7 +2617,7 @@ vint32m2_t test_vloxei8_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei8_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vloxei.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vloxei.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vloxei8_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, const int32_t *base, vuint8m1_t bindex, size_t vl) {
@@ -2628,7 +2628,7 @@ vint32m4_t test_vloxei8_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, const int
 // CHECK-RV64-LABEL: @test_vloxei8_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vloxei.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vloxei.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vloxei8_v_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, const int32_t *base, vuint8m2_t bindex, size_t vl) {
@@ -2639,7 +2639,7 @@ vint32m8_t test_vloxei8_v_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, const int
 // CHECK-RV64-LABEL: @test_vloxei16_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vloxei.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vloxei.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vloxei16_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, const int32_t *base, vuint16mf4_t bindex, size_t vl) {
@@ -2650,7 +2650,7 @@ vint32mf2_t test_vloxei16_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vloxei16_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vloxei.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vloxei.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vloxei16_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, const int32_t *base, vuint16mf2_t bindex, size_t vl) {
@@ -2661,7 +2661,7 @@ vint32m1_t test_vloxei16_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei16_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vloxei.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vloxei.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vloxei16_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, const int32_t *base, vuint16m1_t bindex, size_t vl) {
@@ -2672,7 +2672,7 @@ vint32m2_t test_vloxei16_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei16_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vloxei.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vloxei.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vloxei16_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, const int32_t *base, vuint16m2_t bindex, size_t vl) {
@@ -2683,7 +2683,7 @@ vint32m4_t test_vloxei16_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei16_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vloxei.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vloxei.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vloxei16_v_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, const int32_t *base, vuint16m4_t bindex, size_t vl) {
@@ -2694,7 +2694,7 @@ vint32m8_t test_vloxei16_v_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei32_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vloxei.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vloxei.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vloxei32_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, const int32_t *base, vuint32mf2_t bindex, size_t vl) {
@@ -2705,7 +2705,7 @@ vint32mf2_t test_vloxei32_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vloxei32_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vloxei.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vloxei.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vloxei32_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, const int32_t *base, vuint32m1_t bindex, size_t vl) {
@@ -2716,7 +2716,7 @@ vint32m1_t test_vloxei32_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei32_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vloxei.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vloxei.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vloxei32_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, const int32_t *base, vuint32m2_t bindex, size_t vl) {
@@ -2727,7 +2727,7 @@ vint32m2_t test_vloxei32_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei32_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vloxei.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vloxei.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vloxei32_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, const int32_t *base, vuint32m4_t bindex, size_t vl) {
@@ -2738,7 +2738,7 @@ vint32m4_t test_vloxei32_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei32_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vloxei.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vloxei.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vloxei32_v_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, const int32_t *base, vuint32m8_t bindex, size_t vl) {
@@ -2749,7 +2749,7 @@ vint32m8_t test_vloxei32_v_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei64_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vloxei.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vloxei.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vloxei64_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, const int32_t *base, vuint64m1_t bindex, size_t vl) {
@@ -2760,7 +2760,7 @@ vint32mf2_t test_vloxei64_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vloxei64_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vloxei.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vloxei.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vloxei64_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, const int32_t *base, vuint64m2_t bindex, size_t vl) {
@@ -2771,7 +2771,7 @@ vint32m1_t test_vloxei64_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei64_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vloxei.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vloxei.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vloxei64_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, const int32_t *base, vuint64m4_t bindex, size_t vl) {
@@ -2782,7 +2782,7 @@ vint32m2_t test_vloxei64_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei64_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vloxei.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vloxei.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vloxei64_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, const int32_t *base, vuint64m8_t bindex, size_t vl) {
@@ -2793,7 +2793,7 @@ vint32m4_t test_vloxei64_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei8_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vloxei.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vloxei.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vloxei8_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, const int64_t *base, vuint8mf8_t bindex, size_t vl) {
@@ -2804,7 +2804,7 @@ vint64m1_t test_vloxei8_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei8_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vloxei.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vloxei.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vloxei8_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, const int64_t *base, vuint8mf4_t bindex, size_t vl) {
@@ -2815,7 +2815,7 @@ vint64m2_t test_vloxei8_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei8_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vloxei.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vloxei.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vloxei8_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, const int64_t *base, vuint8mf2_t bindex, size_t vl) {
@@ -2826,7 +2826,7 @@ vint64m4_t test_vloxei8_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei8_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vloxei.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vloxei.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vloxei8_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, const int64_t *base, vuint8m1_t bindex, size_t vl) {
@@ -2837,7 +2837,7 @@ vint64m8_t test_vloxei8_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, const int
 // CHECK-RV64-LABEL: @test_vloxei16_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vloxei.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vloxei.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vloxei16_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, const int64_t *base, vuint16mf4_t bindex, size_t vl) {
@@ -2848,7 +2848,7 @@ vint64m1_t test_vloxei16_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei16_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vloxei.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vloxei.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vloxei16_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, const int64_t *base, vuint16mf2_t bindex, size_t vl) {
@@ -2859,7 +2859,7 @@ vint64m2_t test_vloxei16_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei16_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vloxei.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vloxei.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vloxei16_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, const int64_t *base, vuint16m1_t bindex, size_t vl) {
@@ -2870,7 +2870,7 @@ vint64m4_t test_vloxei16_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei16_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vloxei.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vloxei.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vloxei16_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, const int64_t *base, vuint16m2_t bindex, size_t vl) {
@@ -2881,7 +2881,7 @@ vint64m8_t test_vloxei16_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei32_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vloxei.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vloxei.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vloxei32_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, const int64_t *base, vuint32mf2_t bindex, size_t vl) {
@@ -2892,7 +2892,7 @@ vint64m1_t test_vloxei32_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei32_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vloxei.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vloxei.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vloxei32_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, const int64_t *base, vuint32m1_t bindex, size_t vl) {
@@ -2903,7 +2903,7 @@ vint64m2_t test_vloxei32_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei32_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vloxei.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vloxei.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vloxei32_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, const int64_t *base, vuint32m2_t bindex, size_t vl) {
@@ -2914,7 +2914,7 @@ vint64m4_t test_vloxei32_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei32_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vloxei.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vloxei.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vloxei32_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, const int64_t *base, vuint32m4_t bindex, size_t vl) {
@@ -2925,7 +2925,7 @@ vint64m8_t test_vloxei32_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei64_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vloxei.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vloxei.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vloxei64_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, const int64_t *base, vuint64m1_t bindex, size_t vl) {
@@ -2936,7 +2936,7 @@ vint64m1_t test_vloxei64_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei64_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vloxei.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vloxei.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vloxei64_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, const int64_t *base, vuint64m2_t bindex, size_t vl) {
@@ -2947,7 +2947,7 @@ vint64m2_t test_vloxei64_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei64_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vloxei.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vloxei.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vloxei64_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, const int64_t *base, vuint64m4_t bindex, size_t vl) {
@@ -2958,7 +2958,7 @@ vint64m4_t test_vloxei64_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vloxei64_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vloxei.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vloxei.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vloxei64_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, const int64_t *base, vuint64m8_t bindex, size_t vl) {
@@ -2969,7 +2969,7 @@ vint64m8_t test_vloxei64_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vloxei8_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vloxei.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vloxei.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]
 //
 vuint8mf8_t test_vloxei8_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, const uint8_t *base, vuint8mf8_t bindex, size_t vl) {
@@ -2980,7 +2980,7 @@ vuint8mf8_t test_vloxei8_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei8_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vloxei.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vloxei.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP1]]
 //
 vuint8mf4_t test_vloxei8_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, const uint8_t *base, vuint8mf4_t bindex, size_t vl) {
@@ -2991,7 +2991,7 @@ vuint8mf4_t test_vloxei8_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei8_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vloxei.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vloxei.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP1]]
 //
 vuint8mf2_t test_vloxei8_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, const uint8_t *base, vuint8mf2_t bindex, size_t vl) {
@@ -3002,7 +3002,7 @@ vuint8mf2_t test_vloxei8_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei8_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vloxei.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vloxei.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP1]]
 //
 vuint8m1_t test_vloxei8_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, const uint8_t *base, vuint8m1_t bindex, size_t vl) {
@@ -3013,7 +3013,7 @@ vuint8m1_t test_vloxei8_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, const uint
 // CHECK-RV64-LABEL: @test_vloxei8_v_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vloxei.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vloxei.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP1]]
 //
 vuint8m2_t test_vloxei8_v_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, const uint8_t *base, vuint8m2_t bindex, size_t vl) {
@@ -3024,7 +3024,7 @@ vuint8m2_t test_vloxei8_v_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, const uint
 // CHECK-RV64-LABEL: @test_vloxei8_v_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vloxei.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vloxei.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP1]]
 //
 vuint8m4_t test_vloxei8_v_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, const uint8_t *base, vuint8m4_t bindex, size_t vl) {
@@ -3035,7 +3035,7 @@ vuint8m4_t test_vloxei8_v_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, const uint
 // CHECK-RV64-LABEL: @test_vloxei8_v_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vloxei.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vloxei.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP1]]
 //
 vuint8m8_t test_vloxei8_v_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, const uint8_t *base, vuint8m8_t bindex, size_t vl) {
@@ -3046,7 +3046,7 @@ vuint8m8_t test_vloxei8_v_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, const uint
 // CHECK-RV64-LABEL: @test_vloxei16_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vloxei.mask.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vloxei.mask.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]
 //
 vuint8mf8_t test_vloxei16_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, const uint8_t *base, vuint16mf4_t bindex, size_t vl) {
@@ -3057,7 +3057,7 @@ vuint8mf8_t test_vloxei16_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei16_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vloxei.mask.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vloxei.mask.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP1]]
 //
 vuint8mf4_t test_vloxei16_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, const uint8_t *base, vuint16mf2_t bindex, size_t vl) {
@@ -3068,7 +3068,7 @@ vuint8mf4_t test_vloxei16_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei16_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vloxei.mask.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vloxei.mask.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP1]]
 //
 vuint8mf2_t test_vloxei16_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, const uint8_t *base, vuint16m1_t bindex, size_t vl) {
@@ -3079,7 +3079,7 @@ vuint8mf2_t test_vloxei16_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei16_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vloxei.mask.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vloxei.mask.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP1]]
 //
 vuint8m1_t test_vloxei16_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, const uint8_t *base, vuint16m2_t bindex, size_t vl) {
@@ -3090,7 +3090,7 @@ vuint8m1_t test_vloxei16_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, const uin
 // CHECK-RV64-LABEL: @test_vloxei16_v_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vloxei.mask.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vloxei.mask.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP1]]
 //
 vuint8m2_t test_vloxei16_v_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, const uint8_t *base, vuint16m4_t bindex, size_t vl) {
@@ -3101,7 +3101,7 @@ vuint8m2_t test_vloxei16_v_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, const uin
 // CHECK-RV64-LABEL: @test_vloxei16_v_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vloxei.mask.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vloxei.mask.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP1]]
 //
 vuint8m4_t test_vloxei16_v_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, const uint8_t *base, vuint16m8_t bindex, size_t vl) {
@@ -3112,7 +3112,7 @@ vuint8m4_t test_vloxei16_v_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, const uin
 // CHECK-RV64-LABEL: @test_vloxei32_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vloxei.mask.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vloxei.mask.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]
 //
 vuint8mf8_t test_vloxei32_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, const uint8_t *base, vuint32mf2_t bindex, size_t vl) {
@@ -3123,7 +3123,7 @@ vuint8mf8_t test_vloxei32_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei32_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vloxei.mask.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vloxei.mask.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP1]]
 //
 vuint8mf4_t test_vloxei32_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, const uint8_t *base, vuint32m1_t bindex, size_t vl) {
@@ -3134,7 +3134,7 @@ vuint8mf4_t test_vloxei32_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei32_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vloxei.mask.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vloxei.mask.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP1]]
 //
 vuint8mf2_t test_vloxei32_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, const uint8_t *base, vuint32m2_t bindex, size_t vl) {
@@ -3145,7 +3145,7 @@ vuint8mf2_t test_vloxei32_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei32_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vloxei.mask.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vloxei.mask.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP1]]
 //
 vuint8m1_t test_vloxei32_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, const uint8_t *base, vuint32m4_t bindex, size_t vl) {
@@ -3156,7 +3156,7 @@ vuint8m1_t test_vloxei32_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, const uin
 // CHECK-RV64-LABEL: @test_vloxei32_v_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vloxei.mask.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vloxei.mask.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP1]]
 //
 vuint8m2_t test_vloxei32_v_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, const uint8_t *base, vuint32m8_t bindex, size_t vl) {
@@ -3167,7 +3167,7 @@ vuint8m2_t test_vloxei32_v_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, const uin
 // CHECK-RV64-LABEL: @test_vloxei64_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vloxei.mask.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vloxei.mask.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]
 //
 vuint8mf8_t test_vloxei64_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, const uint8_t *base, vuint64m1_t bindex, size_t vl) {
@@ -3178,7 +3178,7 @@ vuint8mf8_t test_vloxei64_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei64_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vloxei.mask.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vloxei.mask.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP1]]
 //
 vuint8mf4_t test_vloxei64_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, const uint8_t *base, vuint64m2_t bindex, size_t vl) {
@@ -3189,7 +3189,7 @@ vuint8mf4_t test_vloxei64_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei64_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vloxei.mask.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vloxei.mask.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP1]]
 //
 vuint8mf2_t test_vloxei64_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, const uint8_t *base, vuint64m4_t bindex, size_t vl) {
@@ -3200,7 +3200,7 @@ vuint8mf2_t test_vloxei64_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei64_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vloxei.mask.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vloxei.mask.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP1]]
 //
 vuint8m1_t test_vloxei64_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, const uint8_t *base, vuint64m8_t bindex, size_t vl) {
@@ -3211,7 +3211,7 @@ vuint8m1_t test_vloxei64_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, const uin
 // CHECK-RV64-LABEL: @test_vloxei8_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vloxei.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vloxei.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP1]]
 //
 vuint16mf4_t test_vloxei8_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, const uint16_t *base, vuint8mf8_t bindex, size_t vl) {
@@ -3222,7 +3222,7 @@ vuint16mf4_t test_vloxei8_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vloxei8_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vloxei.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vloxei.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP1]]
 //
 vuint16mf2_t test_vloxei8_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, const uint16_t *base, vuint8mf4_t bindex, size_t vl) {
@@ -3233,7 +3233,7 @@ vuint16mf2_t test_vloxei8_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vloxei8_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vloxei.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vloxei.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP1]]
 //
 vuint16m1_t test_vloxei8_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, const uint16_t *base, vuint8mf2_t bindex, size_t vl) {
@@ -3244,7 +3244,7 @@ vuint16m1_t test_vloxei8_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei8_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vloxei.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vloxei.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP1]]
 //
 vuint16m2_t test_vloxei8_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, const uint16_t *base, vuint8m1_t bindex, size_t vl) {
@@ -3255,7 +3255,7 @@ vuint16m2_t test_vloxei8_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, const u
 // CHECK-RV64-LABEL: @test_vloxei8_v_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vloxei.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vloxei.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP1]]
 //
 vuint16m4_t test_vloxei8_v_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, const uint16_t *base, vuint8m2_t bindex, size_t vl) {
@@ -3266,7 +3266,7 @@ vuint16m4_t test_vloxei8_v_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, const u
 // CHECK-RV64-LABEL: @test_vloxei8_v_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vloxei.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vloxei.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP1]]
 //
 vuint16m8_t test_vloxei8_v_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, const uint16_t *base, vuint8m4_t bindex, size_t vl) {
@@ -3277,7 +3277,7 @@ vuint16m8_t test_vloxei8_v_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, const u
 // CHECK-RV64-LABEL: @test_vloxei16_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vloxei.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vloxei.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP1]]
 //
 vuint16mf4_t test_vloxei16_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, const uint16_t *base, vuint16mf4_t bindex, size_t vl) {
@@ -3288,7 +3288,7 @@ vuint16mf4_t test_vloxei16_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, co
 // CHECK-RV64-LABEL: @test_vloxei16_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vloxei.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vloxei.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP1]]
 //
 vuint16mf2_t test_vloxei16_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, const uint16_t *base, vuint16mf2_t bindex, size_t vl) {
@@ -3299,7 +3299,7 @@ vuint16mf2_t test_vloxei16_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, co
 // CHECK-RV64-LABEL: @test_vloxei16_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vloxei.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vloxei.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP1]]
 //
 vuint16m1_t test_vloxei16_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, const uint16_t *base, vuint16m1_t bindex, size_t vl) {
@@ -3310,7 +3310,7 @@ vuint16m1_t test_vloxei16_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei16_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vloxei.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vloxei.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP1]]
 //
 vuint16m2_t test_vloxei16_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, const uint16_t *base, vuint16m2_t bindex, size_t vl) {
@@ -3321,7 +3321,7 @@ vuint16m2_t test_vloxei16_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei16_v_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vloxei.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vloxei.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP1]]
 //
 vuint16m4_t test_vloxei16_v_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, const uint16_t *base, vuint16m4_t bindex, size_t vl) {
@@ -3332,7 +3332,7 @@ vuint16m4_t test_vloxei16_v_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei16_v_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vloxei.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vloxei.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP1]]
 //
 vuint16m8_t test_vloxei16_v_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, const uint16_t *base, vuint16m8_t bindex, size_t vl) {
@@ -3343,7 +3343,7 @@ vuint16m8_t test_vloxei16_v_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei32_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vloxei.mask.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vloxei.mask.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP1]]
 //
 vuint16mf4_t test_vloxei32_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, const uint16_t *base, vuint32mf2_t bindex, size_t vl) {
@@ -3354,7 +3354,7 @@ vuint16mf4_t test_vloxei32_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, co
 // CHECK-RV64-LABEL: @test_vloxei32_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vloxei.mask.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vloxei.mask.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP1]]
 //
 vuint16mf2_t test_vloxei32_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, const uint16_t *base, vuint32m1_t bindex, size_t vl) {
@@ -3365,7 +3365,7 @@ vuint16mf2_t test_vloxei32_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, co
 // CHECK-RV64-LABEL: @test_vloxei32_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vloxei.mask.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vloxei.mask.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP1]]
 //
 vuint16m1_t test_vloxei32_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, const uint16_t *base, vuint32m2_t bindex, size_t vl) {
@@ -3376,7 +3376,7 @@ vuint16m1_t test_vloxei32_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei32_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vloxei.mask.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vloxei.mask.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP1]]
 //
 vuint16m2_t test_vloxei32_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, const uint16_t *base, vuint32m4_t bindex, size_t vl) {
@@ -3387,7 +3387,7 @@ vuint16m2_t test_vloxei32_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei32_v_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vloxei.mask.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vloxei.mask.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP1]]
 //
 vuint16m4_t test_vloxei32_v_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, const uint16_t *base, vuint32m8_t bindex, size_t vl) {
@@ -3398,7 +3398,7 @@ vuint16m4_t test_vloxei32_v_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei64_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vloxei.mask.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vloxei.mask.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP1]]
 //
 vuint16mf4_t test_vloxei64_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, const uint16_t *base, vuint64m1_t bindex, size_t vl) {
@@ -3409,7 +3409,7 @@ vuint16mf4_t test_vloxei64_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, co
 // CHECK-RV64-LABEL: @test_vloxei64_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vloxei.mask.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vloxei.mask.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP1]]
 //
 vuint16mf2_t test_vloxei64_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, const uint16_t *base, vuint64m2_t bindex, size_t vl) {
@@ -3420,7 +3420,7 @@ vuint16mf2_t test_vloxei64_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, co
 // CHECK-RV64-LABEL: @test_vloxei64_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vloxei.mask.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vloxei.mask.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP1]]
 //
 vuint16m1_t test_vloxei64_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, const uint16_t *base, vuint64m4_t bindex, size_t vl) {
@@ -3431,7 +3431,7 @@ vuint16m1_t test_vloxei64_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei64_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vloxei.mask.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vloxei.mask.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP1]]
 //
 vuint16m2_t test_vloxei64_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, const uint16_t *base, vuint64m8_t bindex, size_t vl) {
@@ -3442,7 +3442,7 @@ vuint16m2_t test_vloxei64_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei8_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vloxei.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vloxei.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vloxei8_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, const uint32_t *base, vuint8mf8_t bindex, size_t vl) {
@@ -3453,7 +3453,7 @@ vuint32mf2_t test_vloxei8_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vloxei8_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vloxei.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vloxei.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vloxei8_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, const uint32_t *base, vuint8mf4_t bindex, size_t vl) {
@@ -3464,7 +3464,7 @@ vuint32m1_t test_vloxei8_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei8_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vloxei.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vloxei.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vloxei8_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, const uint32_t *base, vuint8mf2_t bindex, size_t vl) {
@@ -3475,7 +3475,7 @@ vuint32m2_t test_vloxei8_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei8_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vloxei.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vloxei.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vloxei8_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, const uint32_t *base, vuint8m1_t bindex, size_t vl) {
@@ -3486,7 +3486,7 @@ vuint32m4_t test_vloxei8_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, const u
 // CHECK-RV64-LABEL: @test_vloxei8_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vloxei.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vloxei.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vloxei8_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, const uint32_t *base, vuint8m2_t bindex, size_t vl) {
@@ -3497,7 +3497,7 @@ vuint32m8_t test_vloxei8_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, const u
 // CHECK-RV64-LABEL: @test_vloxei16_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vloxei.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vloxei.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vloxei16_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, const uint32_t *base, vuint16mf4_t bindex, size_t vl) {
@@ -3508,7 +3508,7 @@ vuint32mf2_t test_vloxei16_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, co
 // CHECK-RV64-LABEL: @test_vloxei16_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vloxei.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vloxei.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vloxei16_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, const uint32_t *base, vuint16mf2_t bindex, size_t vl) {
@@ -3519,7 +3519,7 @@ vuint32m1_t test_vloxei16_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei16_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vloxei.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vloxei.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vloxei16_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, const uint32_t *base, vuint16m1_t bindex, size_t vl) {
@@ -3530,7 +3530,7 @@ vuint32m2_t test_vloxei16_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei16_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vloxei.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vloxei.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vloxei16_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, const uint32_t *base, vuint16m2_t bindex, size_t vl) {
@@ -3541,7 +3541,7 @@ vuint32m4_t test_vloxei16_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei16_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vloxei.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vloxei.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vloxei16_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, const uint32_t *base, vuint16m4_t bindex, size_t vl) {
@@ -3552,7 +3552,7 @@ vuint32m8_t test_vloxei16_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei32_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vloxei.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vloxei.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vloxei32_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, const uint32_t *base, vuint32mf2_t bindex, size_t vl) {
@@ -3563,7 +3563,7 @@ vuint32mf2_t test_vloxei32_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, co
 // CHECK-RV64-LABEL: @test_vloxei32_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vloxei.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vloxei.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vloxei32_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, const uint32_t *base, vuint32m1_t bindex, size_t vl) {
@@ -3574,7 +3574,7 @@ vuint32m1_t test_vloxei32_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei32_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vloxei.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vloxei.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vloxei32_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, const uint32_t *base, vuint32m2_t bindex, size_t vl) {
@@ -3585,7 +3585,7 @@ vuint32m2_t test_vloxei32_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei32_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vloxei.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vloxei.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vloxei32_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, const uint32_t *base, vuint32m4_t bindex, size_t vl) {
@@ -3596,7 +3596,7 @@ vuint32m4_t test_vloxei32_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei32_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vloxei.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vloxei.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vloxei32_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, const uint32_t *base, vuint32m8_t bindex, size_t vl) {
@@ -3607,7 +3607,7 @@ vuint32m8_t test_vloxei32_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei64_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vloxei.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vloxei.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vloxei64_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, const uint32_t *base, vuint64m1_t bindex, size_t vl) {
@@ -3618,7 +3618,7 @@ vuint32mf2_t test_vloxei64_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, co
 // CHECK-RV64-LABEL: @test_vloxei64_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vloxei.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vloxei.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vloxei64_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, const uint32_t *base, vuint64m2_t bindex, size_t vl) {
@@ -3629,7 +3629,7 @@ vuint32m1_t test_vloxei64_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei64_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vloxei.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vloxei.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vloxei64_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, const uint32_t *base, vuint64m4_t bindex, size_t vl) {
@@ -3640,7 +3640,7 @@ vuint32m2_t test_vloxei64_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei64_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vloxei.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vloxei.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vloxei64_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, const uint32_t *base, vuint64m8_t bindex, size_t vl) {
@@ -3651,7 +3651,7 @@ vuint32m4_t test_vloxei64_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei8_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vloxei.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vloxei.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vloxei8_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, const uint64_t *base, vuint8mf8_t bindex, size_t vl) {
@@ -3662,7 +3662,7 @@ vuint64m1_t test_vloxei8_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei8_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vloxei.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vloxei.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vloxei8_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, const uint64_t *base, vuint8mf4_t bindex, size_t vl) {
@@ -3673,7 +3673,7 @@ vuint64m2_t test_vloxei8_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei8_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vloxei.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vloxei.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vloxei8_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, const uint64_t *base, vuint8mf2_t bindex, size_t vl) {
@@ -3684,7 +3684,7 @@ vuint64m4_t test_vloxei8_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei8_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vloxei.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vloxei.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vloxei8_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, const uint64_t *base, vuint8m1_t bindex, size_t vl) {
@@ -3695,7 +3695,7 @@ vuint64m8_t test_vloxei8_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, const u
 // CHECK-RV64-LABEL: @test_vloxei16_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vloxei.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vloxei.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vloxei16_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, const uint64_t *base, vuint16mf4_t bindex, size_t vl) {
@@ -3706,7 +3706,7 @@ vuint64m1_t test_vloxei16_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei16_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vloxei.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vloxei.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vloxei16_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, const uint64_t *base, vuint16mf2_t bindex, size_t vl) {
@@ -3717,7 +3717,7 @@ vuint64m2_t test_vloxei16_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei16_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vloxei.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vloxei.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vloxei16_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, const uint64_t *base, vuint16m1_t bindex, size_t vl) {
@@ -3728,7 +3728,7 @@ vuint64m4_t test_vloxei16_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei16_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vloxei.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vloxei.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vloxei16_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, const uint64_t *base, vuint16m2_t bindex, size_t vl) {
@@ -3739,7 +3739,7 @@ vuint64m8_t test_vloxei16_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei32_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vloxei.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vloxei.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vloxei32_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, const uint64_t *base, vuint32mf2_t bindex, size_t vl) {
@@ -3750,7 +3750,7 @@ vuint64m1_t test_vloxei32_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei32_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vloxei.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vloxei.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vloxei32_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, const uint64_t *base, vuint32m1_t bindex, size_t vl) {
@@ -3761,7 +3761,7 @@ vuint64m2_t test_vloxei32_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei32_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vloxei.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vloxei.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vloxei32_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, const uint64_t *base, vuint32m2_t bindex, size_t vl) {
@@ -3772,7 +3772,7 @@ vuint64m4_t test_vloxei32_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei32_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vloxei.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vloxei.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vloxei32_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, const uint64_t *base, vuint32m4_t bindex, size_t vl) {
@@ -3783,7 +3783,7 @@ vuint64m8_t test_vloxei32_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei64_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vloxei.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vloxei.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vloxei64_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, const uint64_t *base, vuint64m1_t bindex, size_t vl) {
@@ -3794,7 +3794,7 @@ vuint64m1_t test_vloxei64_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei64_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vloxei.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vloxei.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vloxei64_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, const uint64_t *base, vuint64m2_t bindex, size_t vl) {
@@ -3805,7 +3805,7 @@ vuint64m2_t test_vloxei64_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei64_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vloxei.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vloxei.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vloxei64_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, const uint64_t *base, vuint64m4_t bindex, size_t vl) {
@@ -3816,7 +3816,7 @@ vuint64m4_t test_vloxei64_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei64_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vloxei.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vloxei.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vloxei64_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, const uint64_t *base, vuint64m8_t bindex, size_t vl) {
@@ -3827,7 +3827,7 @@ vuint64m8_t test_vloxei64_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei8_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vloxei.mask.nxv1f32.nxv1i8.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vloxei.mask.nxv1f32.nxv1i8.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP1]]
 //
 vfloat32mf2_t test_vloxei8_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff, const float *base, vuint8mf8_t bindex, size_t vl) {
@@ -3838,7 +3838,7 @@ vfloat32mf2_t test_vloxei8_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff, c
 // CHECK-RV64-LABEL: @test_vloxei8_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vloxei.mask.nxv2f32.nxv2i8.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vloxei.mask.nxv2f32.nxv2i8.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP1]]
 //
 vfloat32m1_t test_vloxei8_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, const float *base, vuint8mf4_t bindex, size_t vl) {
@@ -3849,7 +3849,7 @@ vfloat32m1_t test_vloxei8_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vloxei8_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vloxei.mask.nxv4f32.nxv4i8.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vloxei.mask.nxv4f32.nxv4i8.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 vfloat32m2_t test_vloxei8_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, const float *base, vuint8mf2_t bindex, size_t vl) {
@@ -3860,7 +3860,7 @@ vfloat32m2_t test_vloxei8_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vloxei8_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vloxei.mask.nxv8f32.nxv8i8.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vloxei.mask.nxv8f32.nxv8i8.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP1]]
 //
 vfloat32m4_t test_vloxei8_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, const float *base, vuint8m1_t bindex, size_t vl) {
@@ -3871,7 +3871,7 @@ vfloat32m4_t test_vloxei8_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei8_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vloxei.mask.nxv16f32.nxv16i8.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vloxei.mask.nxv16f32.nxv16i8.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP1]]
 //
 vfloat32m8_t test_vloxei8_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff, const float *base, vuint8m2_t bindex, size_t vl) {
@@ -3882,7 +3882,7 @@ vfloat32m8_t test_vloxei8_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei16_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vloxei.mask.nxv1f32.nxv1i16.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vloxei.mask.nxv1f32.nxv1i16.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP1]]
 //
 vfloat32mf2_t test_vloxei16_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff, const float *base, vuint16mf4_t bindex, size_t vl) {
@@ -3893,7 +3893,7 @@ vfloat32mf2_t test_vloxei16_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vloxei16_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vloxei.mask.nxv2f32.nxv2i16.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vloxei.mask.nxv2f32.nxv2i16.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP1]]
 //
 vfloat32m1_t test_vloxei16_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, const float *base, vuint16mf2_t bindex, size_t vl) {
@@ -3904,7 +3904,7 @@ vfloat32m1_t test_vloxei16_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vloxei16_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vloxei.mask.nxv4f32.nxv4i16.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vloxei.mask.nxv4f32.nxv4i16.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 vfloat32m2_t test_vloxei16_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, const float *base, vuint16m1_t bindex, size_t vl) {
@@ -3915,7 +3915,7 @@ vfloat32m2_t test_vloxei16_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vloxei16_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vloxei.mask.nxv8f32.nxv8i16.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vloxei.mask.nxv8f32.nxv8i16.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP1]]
 //
 vfloat32m4_t test_vloxei16_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, const float *base, vuint16m2_t bindex, size_t vl) {
@@ -3926,7 +3926,7 @@ vfloat32m4_t test_vloxei16_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vloxei16_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vloxei.mask.nxv16f32.nxv16i16.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vloxei.mask.nxv16f32.nxv16i16.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP1]]
 //
 vfloat32m8_t test_vloxei16_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff, const float *base, vuint16m4_t bindex, size_t vl) {
@@ -3937,7 +3937,7 @@ vfloat32m8_t test_vloxei16_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vloxei32_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vloxei.mask.nxv1f32.nxv1i32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vloxei.mask.nxv1f32.nxv1i32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP1]]
 //
 vfloat32mf2_t test_vloxei32_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff, const float *base, vuint32mf2_t bindex, size_t vl) {
@@ -3948,7 +3948,7 @@ vfloat32mf2_t test_vloxei32_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vloxei32_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vloxei.mask.nxv2f32.nxv2i32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vloxei.mask.nxv2f32.nxv2i32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP1]]
 //
 vfloat32m1_t test_vloxei32_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, const float *base, vuint32m1_t bindex, size_t vl) {
@@ -3959,7 +3959,7 @@ vfloat32m1_t test_vloxei32_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vloxei32_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vloxei.mask.nxv4f32.nxv4i32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vloxei.mask.nxv4f32.nxv4i32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 vfloat32m2_t test_vloxei32_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, const float *base, vuint32m2_t bindex, size_t vl) {
@@ -3970,7 +3970,7 @@ vfloat32m2_t test_vloxei32_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vloxei32_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vloxei.mask.nxv8f32.nxv8i32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vloxei.mask.nxv8f32.nxv8i32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP1]]
 //
 vfloat32m4_t test_vloxei32_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, const float *base, vuint32m4_t bindex, size_t vl) {
@@ -3981,7 +3981,7 @@ vfloat32m4_t test_vloxei32_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vloxei32_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vloxei.mask.nxv16f32.nxv16i32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vloxei.mask.nxv16f32.nxv16i32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP1]]
 //
 vfloat32m8_t test_vloxei32_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff, const float *base, vuint32m8_t bindex, size_t vl) {
@@ -3992,7 +3992,7 @@ vfloat32m8_t test_vloxei32_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vloxei64_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vloxei.mask.nxv1f32.nxv1i64.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vloxei.mask.nxv1f32.nxv1i64.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP1]]
 //
 vfloat32mf2_t test_vloxei64_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff, const float *base, vuint64m1_t bindex, size_t vl) {
@@ -4003,7 +4003,7 @@ vfloat32mf2_t test_vloxei64_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vloxei64_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vloxei.mask.nxv2f32.nxv2i64.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vloxei.mask.nxv2f32.nxv2i64.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP1]]
 //
 vfloat32m1_t test_vloxei64_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, const float *base, vuint64m2_t bindex, size_t vl) {
@@ -4014,7 +4014,7 @@ vfloat32m1_t test_vloxei64_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vloxei64_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vloxei.mask.nxv4f32.nxv4i64.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vloxei.mask.nxv4f32.nxv4i64.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 vfloat32m2_t test_vloxei64_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, const float *base, vuint64m4_t bindex, size_t vl) {
@@ -4025,7 +4025,7 @@ vfloat32m2_t test_vloxei64_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vloxei64_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vloxei.mask.nxv8f32.nxv8i64.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vloxei.mask.nxv8f32.nxv8i64.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP1]]
 //
 vfloat32m4_t test_vloxei64_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, const float *base, vuint64m8_t bindex, size_t vl) {
@@ -4036,7 +4036,7 @@ vfloat32m4_t test_vloxei64_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vloxei8_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vloxei.mask.nxv1f64.nxv1i8.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vloxei.mask.nxv1f64.nxv1i8.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP1]]
 //
 vfloat64m1_t test_vloxei8_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, const double *base, vuint8mf8_t bindex, size_t vl) {
@@ -4047,7 +4047,7 @@ vfloat64m1_t test_vloxei8_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vloxei8_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vloxei.mask.nxv2f64.nxv2i8.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vloxei.mask.nxv2f64.nxv2i8.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 vfloat64m2_t test_vloxei8_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, const double *base, vuint8mf4_t bindex, size_t vl) {
@@ -4058,7 +4058,7 @@ vfloat64m2_t test_vloxei8_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vloxei8_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vloxei.mask.nxv4f64.nxv4i8.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vloxei.mask.nxv4f64.nxv4i8.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP1]]
 //
 vfloat64m4_t test_vloxei8_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, const double *base, vuint8mf2_t bindex, size_t vl) {
@@ -4069,7 +4069,7 @@ vfloat64m4_t test_vloxei8_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vloxei8_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vloxei.mask.nxv8f64.nxv8i8.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vloxei.mask.nxv8f64.nxv8i8.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP1]]
 //
 vfloat64m8_t test_vloxei8_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff, const double *base, vuint8m1_t bindex, size_t vl) {
@@ -4080,7 +4080,7 @@ vfloat64m8_t test_vloxei8_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vloxei16_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vloxei.mask.nxv1f64.nxv1i16.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vloxei.mask.nxv1f64.nxv1i16.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP1]]
 //
 vfloat64m1_t test_vloxei16_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, const double *base, vuint16mf4_t bindex, size_t vl) {
@@ -4091,7 +4091,7 @@ vfloat64m1_t test_vloxei16_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vloxei16_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vloxei.mask.nxv2f64.nxv2i16.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vloxei.mask.nxv2f64.nxv2i16.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 vfloat64m2_t test_vloxei16_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, const double *base, vuint16mf2_t bindex, size_t vl) {
@@ -4102,7 +4102,7 @@ vfloat64m2_t test_vloxei16_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vloxei16_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vloxei.mask.nxv4f64.nxv4i16.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vloxei.mask.nxv4f64.nxv4i16.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP1]]
 //
 vfloat64m4_t test_vloxei16_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, const double *base, vuint16m1_t bindex, size_t vl) {
@@ -4113,7 +4113,7 @@ vfloat64m4_t test_vloxei16_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vloxei16_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vloxei.mask.nxv8f64.nxv8i16.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vloxei.mask.nxv8f64.nxv8i16.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP1]]
 //
 vfloat64m8_t test_vloxei16_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff, const double *base, vuint16m2_t bindex, size_t vl) {
@@ -4124,7 +4124,7 @@ vfloat64m8_t test_vloxei16_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vloxei32_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vloxei.mask.nxv1f64.nxv1i32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vloxei.mask.nxv1f64.nxv1i32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP1]]
 //
 vfloat64m1_t test_vloxei32_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, const double *base, vuint32mf2_t bindex, size_t vl) {
@@ -4135,7 +4135,7 @@ vfloat64m1_t test_vloxei32_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vloxei32_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vloxei.mask.nxv2f64.nxv2i32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vloxei.mask.nxv2f64.nxv2i32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 vfloat64m2_t test_vloxei32_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, const double *base, vuint32m1_t bindex, size_t vl) {
@@ -4146,7 +4146,7 @@ vfloat64m2_t test_vloxei32_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vloxei32_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vloxei.mask.nxv4f64.nxv4i32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vloxei.mask.nxv4f64.nxv4i32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP1]]
 //
 vfloat64m4_t test_vloxei32_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, const double *base, vuint32m2_t bindex, size_t vl) {
@@ -4157,7 +4157,7 @@ vfloat64m4_t test_vloxei32_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vloxei32_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vloxei.mask.nxv8f64.nxv8i32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vloxei.mask.nxv8f64.nxv8i32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP1]]
 //
 vfloat64m8_t test_vloxei32_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff, const double *base, vuint32m4_t bindex, size_t vl) {
@@ -4168,7 +4168,7 @@ vfloat64m8_t test_vloxei32_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vloxei64_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vloxei.mask.nxv1f64.nxv1i64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vloxei.mask.nxv1f64.nxv1i64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP1]]
 //
 vfloat64m1_t test_vloxei64_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, const double *base, vuint64m1_t bindex, size_t vl) {
@@ -4179,7 +4179,7 @@ vfloat64m1_t test_vloxei64_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vloxei64_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vloxei.mask.nxv2f64.nxv2i64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vloxei.mask.nxv2f64.nxv2i64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 vfloat64m2_t test_vloxei64_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, const double *base, vuint64m2_t bindex, size_t vl) {
@@ -4190,7 +4190,7 @@ vfloat64m2_t test_vloxei64_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vloxei64_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vloxei.mask.nxv4f64.nxv4i64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vloxei.mask.nxv4f64.nxv4i64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP1]]
 //
 vfloat64m4_t test_vloxei64_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, const double *base, vuint64m4_t bindex, size_t vl) {
@@ -4201,7 +4201,7 @@ vfloat64m4_t test_vloxei64_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vloxei64_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vloxei.mask.nxv8f64.nxv8i64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vloxei.mask.nxv8f64.nxv8i64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP1]]
 //
 vfloat64m8_t test_vloxei64_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff, const double *base, vuint64m8_t bindex, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vlse.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vlse.c
index afab365..18b04a2 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vlse.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vlse.c
@@ -9,7 +9,7 @@
 // CHECK-RV64-LABEL: @test_vlse8_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vlse.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vlse.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]
 //
 vint8mf8_t test_vlse8_v_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
@@ -22,7 +22,7 @@ vint8mf8_t test_vlse8_v_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse8_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vlse.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vlse.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP1]]
 //
 vint8mf4_t test_vlse8_v_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
@@ -35,7 +35,7 @@ vint8mf4_t test_vlse8_v_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse8_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vlse.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vlse.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP1]]
 //
 vint8mf2_t test_vlse8_v_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
@@ -48,7 +48,7 @@ vint8mf2_t test_vlse8_v_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse8_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vlse.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vlse.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP1]]
 //
 vint8m1_t test_vlse8_v_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
@@ -61,7 +61,7 @@ vint8m1_t test_vlse8_v_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse8_v_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vlse.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vlse.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP1]]
 //
 vint8m2_t test_vlse8_v_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
@@ -74,7 +74,7 @@ vint8m2_t test_vlse8_v_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse8_v_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vlse.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vlse.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP1]]
 //
 vint8m4_t test_vlse8_v_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
@@ -87,7 +87,7 @@ vint8m4_t test_vlse8_v_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse8_v_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vlse.mask.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vlse.mask.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP1]]
 //
 vint8m8_t test_vlse8_v_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
@@ -100,7 +100,7 @@ vint8m8_t test_vlse8_v_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse16_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vlse.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vlse.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP1]]
 //
 vint16mf4_t test_vlse16_v_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -113,7 +113,7 @@ vint16mf4_t test_vlse16_v_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse16_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vlse.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vlse.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP1]]
 //
 vint16mf2_t test_vlse16_v_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -126,7 +126,7 @@ vint16mf2_t test_vlse16_v_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse16_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vlse.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vlse.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP1]]
 //
 vint16m1_t test_vlse16_v_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -139,7 +139,7 @@ vint16m1_t test_vlse16_v_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse16_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vlse.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vlse.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP1]]
 //
 vint16m2_t test_vlse16_v_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -152,7 +152,7 @@ vint16m2_t test_vlse16_v_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse16_v_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vlse.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vlse.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP1]]
 //
 vint16m4_t test_vlse16_v_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -165,7 +165,7 @@ vint16m4_t test_vlse16_v_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse16_v_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vlse.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vlse.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP1]]
 //
 vint16m8_t test_vlse16_v_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -178,7 +178,7 @@ vint16m8_t test_vlse16_v_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse32_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vlse.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vlse.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vlse32_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -191,7 +191,7 @@ vint32mf2_t test_vlse32_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse32_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vlse.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vlse.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vlse32_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -204,7 +204,7 @@ vint32m1_t test_vlse32_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse32_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vlse.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vlse.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vlse32_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -217,7 +217,7 @@ vint32m2_t test_vlse32_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse32_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vlse.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vlse.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vlse32_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -230,7 +230,7 @@ vint32m4_t test_vlse32_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse32_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vlse.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vlse.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vlse32_v_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -243,7 +243,7 @@ vint32m8_t test_vlse32_v_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse64_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vlse.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vlse.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vlse64_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -256,7 +256,7 @@ vint64m1_t test_vlse64_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse64_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vlse.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vlse.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vlse64_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -269,7 +269,7 @@ vint64m2_t test_vlse64_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse64_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vlse.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vlse.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vlse64_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -282,7 +282,7 @@ vint64m4_t test_vlse64_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse64_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vlse.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vlse.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vlse64_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -295,7 +295,7 @@ vint64m8_t test_vlse64_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse8_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vlse.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vlse.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]
 //
 vuint8mf8_t test_vlse8_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
@@ -308,7 +308,7 @@ vuint8mf8_t test_vlse8_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse8_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vlse.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vlse.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP1]]
 //
 vuint8mf4_t test_vlse8_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
@@ -321,7 +321,7 @@ vuint8mf4_t test_vlse8_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse8_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vlse.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vlse.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP1]]
 //
 vuint8mf2_t test_vlse8_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
@@ -334,7 +334,7 @@ vuint8mf2_t test_vlse8_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse8_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vlse.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vlse.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP1]]
 //
 vuint8m1_t test_vlse8_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
@@ -347,7 +347,7 @@ vuint8m1_t test_vlse8_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse8_v_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vlse.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vlse.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP1]]
 //
 vuint8m2_t test_vlse8_v_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
@@ -360,7 +360,7 @@ vuint8m2_t test_vlse8_v_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse8_v_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vlse.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vlse.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP1]]
 //
 vuint8m4_t test_vlse8_v_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
@@ -373,7 +373,7 @@ vuint8m4_t test_vlse8_v_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse8_v_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vlse.mask.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vlse.mask.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP1]]
 //
 vuint8m8_t test_vlse8_v_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
@@ -386,7 +386,7 @@ vuint8m8_t test_vlse8_v_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse16_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vlse.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vlse.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP1]]
 //
 vuint16mf4_t test_vlse16_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -399,7 +399,7 @@ vuint16mf4_t test_vlse16_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse16_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vlse.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vlse.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP1]]
 //
 vuint16mf2_t test_vlse16_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -412,7 +412,7 @@ vuint16mf2_t test_vlse16_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse16_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vlse.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vlse.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP1]]
 //
 vuint16m1_t test_vlse16_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -425,7 +425,7 @@ vuint16m1_t test_vlse16_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse16_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vlse.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vlse.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP1]]
 //
 vuint16m2_t test_vlse16_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -438,7 +438,7 @@ vuint16m2_t test_vlse16_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse16_v_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vlse.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vlse.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP1]]
 //
 vuint16m4_t test_vlse16_v_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -451,7 +451,7 @@ vuint16m4_t test_vlse16_v_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse16_v_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vlse.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vlse.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP1]]
 //
 vuint16m8_t test_vlse16_v_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -464,7 +464,7 @@ vuint16m8_t test_vlse16_v_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse32_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vlse.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vlse.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vlse32_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -477,7 +477,7 @@ vuint32mf2_t test_vlse32_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse32_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vlse.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vlse.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vlse32_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -490,7 +490,7 @@ vuint32m1_t test_vlse32_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse32_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vlse.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vlse.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vlse32_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -503,7 +503,7 @@ vuint32m2_t test_vlse32_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse32_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vlse.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vlse.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vlse32_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -516,7 +516,7 @@ vuint32m4_t test_vlse32_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse32_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vlse.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vlse.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vlse32_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -529,7 +529,7 @@ vuint32m8_t test_vlse32_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse64_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vlse.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vlse.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vlse64_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -542,7 +542,7 @@ vuint64m1_t test_vlse64_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse64_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vlse.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vlse.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vlse64_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -555,7 +555,7 @@ vuint64m2_t test_vlse64_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse64_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vlse.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vlse.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vlse64_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -568,7 +568,7 @@ vuint64m4_t test_vlse64_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse64_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vlse.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vlse.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vlse64_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
@@ -581,7 +581,7 @@ vuint64m8_t test_vlse64_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse32_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vlse.mask.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vlse.mask.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP1]]
 //
 vfloat32mf2_t test_vlse32_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -594,7 +594,7 @@ vfloat32mf2_t test_vlse32_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse32_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vlse.mask.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vlse.mask.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP1]]
 //
 vfloat32m1_t test_vlse32_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -607,7 +607,7 @@ vfloat32m1_t test_vlse32_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse32_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vlse.mask.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vlse.mask.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 vfloat32m2_t test_vlse32_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -620,7 +620,7 @@ vfloat32m2_t test_vlse32_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse32_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vlse.mask.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vlse.mask.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP1]]
 //
 vfloat32m4_t test_vlse32_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -633,7 +633,7 @@ vfloat32m4_t test_vlse32_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse32_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vlse.mask.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vlse.mask.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP1]]
 //
 vfloat32m8_t test_vlse32_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -646,7 +646,7 @@ vfloat32m8_t test_vlse32_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse64_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vlse.mask.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vlse.mask.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP1]]
 //
 vfloat64m1_t test_vlse64_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -659,7 +659,7 @@ vfloat64m1_t test_vlse64_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse64_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vlse.mask.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vlse.mask.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 vfloat64m2_t test_vlse64_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -672,7 +672,7 @@ vfloat64m2_t test_vlse64_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse64_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vlse.mask.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vlse.mask.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP1]]
 //
 vfloat64m4_t test_vlse64_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -685,7 +685,7 @@ vfloat64m4_t test_vlse64_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 // CHECK-RV64-LABEL: @test_vlse64_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vlse.mask.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vlse.mask.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP1]]
 //
 vfloat64m8_t test_vlse64_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vluxei.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vluxei.c
index d002c10..5363c07 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vluxei.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vluxei.c
@@ -2110,7 +2110,7 @@ vfloat64m8_t test_vluxei64_v_f64m8(const double *base, vuint64m8_t bindex, size_
 // CHECK-RV64-LABEL: @test_vluxei8_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vluxei.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vluxei.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]
 //
 vint8mf8_t test_vluxei8_v_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, const int8_t *base, vuint8mf8_t bindex, size_t vl) {
@@ -2121,7 +2121,7 @@ vint8mf8_t test_vluxei8_v_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei8_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vluxei.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vluxei.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP1]]
 //
 vint8mf4_t test_vluxei8_v_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, const int8_t *base, vuint8mf4_t bindex, size_t vl) {
@@ -2132,7 +2132,7 @@ vint8mf4_t test_vluxei8_v_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei8_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vluxei.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vluxei.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP1]]
 //
 vint8mf2_t test_vluxei8_v_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, const int8_t *base, vuint8mf2_t bindex, size_t vl) {
@@ -2143,7 +2143,7 @@ vint8mf2_t test_vluxei8_v_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei8_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vluxei.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vluxei.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP1]]
 //
 vint8m1_t test_vluxei8_v_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, const int8_t *base, vuint8m1_t bindex, size_t vl) {
@@ -2154,7 +2154,7 @@ vint8m1_t test_vluxei8_v_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, const int8_t
 // CHECK-RV64-LABEL: @test_vluxei8_v_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vluxei.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vluxei.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP1]]
 //
 vint8m2_t test_vluxei8_v_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, const int8_t *base, vuint8m2_t bindex, size_t vl) {
@@ -2165,7 +2165,7 @@ vint8m2_t test_vluxei8_v_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, const int8_t
 // CHECK-RV64-LABEL: @test_vluxei8_v_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vluxei.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vluxei.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP1]]
 //
 vint8m4_t test_vluxei8_v_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, const int8_t *base, vuint8m4_t bindex, size_t vl) {
@@ -2176,7 +2176,7 @@ vint8m4_t test_vluxei8_v_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, const int8_t
 // CHECK-RV64-LABEL: @test_vluxei8_v_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vluxei.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vluxei.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP1]]
 //
 vint8m8_t test_vluxei8_v_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, const int8_t *base, vuint8m8_t bindex, size_t vl) {
@@ -2187,7 +2187,7 @@ vint8m8_t test_vluxei8_v_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, const int8_t
 // CHECK-RV64-LABEL: @test_vluxei16_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vluxei.mask.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vluxei.mask.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]
 //
 vint8mf8_t test_vluxei16_v_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, const int8_t *base, vuint16mf4_t bindex, size_t vl) {
@@ -2198,7 +2198,7 @@ vint8mf8_t test_vluxei16_v_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei16_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vluxei.mask.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vluxei.mask.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP1]]
 //
 vint8mf4_t test_vluxei16_v_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, const int8_t *base, vuint16mf2_t bindex, size_t vl) {
@@ -2209,7 +2209,7 @@ vint8mf4_t test_vluxei16_v_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei16_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vluxei.mask.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vluxei.mask.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP1]]
 //
 vint8mf2_t test_vluxei16_v_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, const int8_t *base, vuint16m1_t bindex, size_t vl) {
@@ -2220,7 +2220,7 @@ vint8mf2_t test_vluxei16_v_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei16_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vluxei.mask.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vluxei.mask.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP1]]
 //
 vint8m1_t test_vluxei16_v_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, const int8_t *base, vuint16m2_t bindex, size_t vl) {
@@ -2231,7 +2231,7 @@ vint8m1_t test_vluxei16_v_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, const int8_
 // CHECK-RV64-LABEL: @test_vluxei16_v_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vluxei.mask.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vluxei.mask.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP1]]
 //
 vint8m2_t test_vluxei16_v_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, const int8_t *base, vuint16m4_t bindex, size_t vl) {
@@ -2242,7 +2242,7 @@ vint8m2_t test_vluxei16_v_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, const int8_
 // CHECK-RV64-LABEL: @test_vluxei16_v_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vluxei.mask.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vluxei.mask.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP1]]
 //
 vint8m4_t test_vluxei16_v_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, const int8_t *base, vuint16m8_t bindex, size_t vl) {
@@ -2253,7 +2253,7 @@ vint8m4_t test_vluxei16_v_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, const int8_
 // CHECK-RV64-LABEL: @test_vluxei32_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vluxei.mask.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vluxei.mask.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]
 //
 vint8mf8_t test_vluxei32_v_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, const int8_t *base, vuint32mf2_t bindex, size_t vl) {
@@ -2264,7 +2264,7 @@ vint8mf8_t test_vluxei32_v_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei32_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vluxei.mask.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vluxei.mask.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP1]]
 //
 vint8mf4_t test_vluxei32_v_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, const int8_t *base, vuint32m1_t bindex, size_t vl) {
@@ -2275,7 +2275,7 @@ vint8mf4_t test_vluxei32_v_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei32_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vluxei.mask.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vluxei.mask.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP1]]
 //
 vint8mf2_t test_vluxei32_v_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, const int8_t *base, vuint32m2_t bindex, size_t vl) {
@@ -2286,7 +2286,7 @@ vint8mf2_t test_vluxei32_v_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei32_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vluxei.mask.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vluxei.mask.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP1]]
 //
 vint8m1_t test_vluxei32_v_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, const int8_t *base, vuint32m4_t bindex, size_t vl) {
@@ -2297,7 +2297,7 @@ vint8m1_t test_vluxei32_v_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, const int8_
 // CHECK-RV64-LABEL: @test_vluxei32_v_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vluxei.mask.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vluxei.mask.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP1]]
 //
 vint8m2_t test_vluxei32_v_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, const int8_t *base, vuint32m8_t bindex, size_t vl) {
@@ -2308,7 +2308,7 @@ vint8m2_t test_vluxei32_v_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, const int8_
 // CHECK-RV64-LABEL: @test_vluxei64_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vluxei.mask.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vluxei.mask.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]
 //
 vint8mf8_t test_vluxei64_v_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, const int8_t *base, vuint64m1_t bindex, size_t vl) {
@@ -2319,7 +2319,7 @@ vint8mf8_t test_vluxei64_v_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei64_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vluxei.mask.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vluxei.mask.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP1]]
 //
 vint8mf4_t test_vluxei64_v_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, const int8_t *base, vuint64m2_t bindex, size_t vl) {
@@ -2330,7 +2330,7 @@ vint8mf4_t test_vluxei64_v_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei64_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vluxei.mask.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vluxei.mask.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP1]]
 //
 vint8mf2_t test_vluxei64_v_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, const int8_t *base, vuint64m4_t bindex, size_t vl) {
@@ -2341,7 +2341,7 @@ vint8mf2_t test_vluxei64_v_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei64_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vluxei.mask.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vluxei.mask.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP1]]
 //
 vint8m1_t test_vluxei64_v_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, const int8_t *base, vuint64m8_t bindex, size_t vl) {
@@ -2352,7 +2352,7 @@ vint8m1_t test_vluxei64_v_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, const int8_
 // CHECK-RV64-LABEL: @test_vluxei8_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vluxei.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vluxei.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP1]]
 //
 vint16mf4_t test_vluxei8_v_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, const int16_t *base, vuint8mf8_t bindex, size_t vl) {
@@ -2363,7 +2363,7 @@ vint16mf4_t test_vluxei8_v_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei8_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vluxei.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vluxei.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP1]]
 //
 vint16mf2_t test_vluxei8_v_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, const int16_t *base, vuint8mf4_t bindex, size_t vl) {
@@ -2374,7 +2374,7 @@ vint16mf2_t test_vluxei8_v_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei8_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vluxei.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vluxei.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP1]]
 //
 vint16m1_t test_vluxei8_v_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, const int16_t *base, vuint8mf2_t bindex, size_t vl) {
@@ -2385,7 +2385,7 @@ vint16m1_t test_vluxei8_v_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei8_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vluxei.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vluxei.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP1]]
 //
 vint16m2_t test_vluxei8_v_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, const int16_t *base, vuint8m1_t bindex, size_t vl) {
@@ -2396,7 +2396,7 @@ vint16m2_t test_vluxei8_v_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, const int
 // CHECK-RV64-LABEL: @test_vluxei8_v_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vluxei.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vluxei.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP1]]
 //
 vint16m4_t test_vluxei8_v_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, const int16_t *base, vuint8m2_t bindex, size_t vl) {
@@ -2407,7 +2407,7 @@ vint16m4_t test_vluxei8_v_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, const int
 // CHECK-RV64-LABEL: @test_vluxei8_v_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vluxei.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vluxei.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP1]]
 //
 vint16m8_t test_vluxei8_v_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, const int16_t *base, vuint8m4_t bindex, size_t vl) {
@@ -2418,7 +2418,7 @@ vint16m8_t test_vluxei8_v_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, const int
 // CHECK-RV64-LABEL: @test_vluxei16_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vluxei.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vluxei.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP1]]
 //
 vint16mf4_t test_vluxei16_v_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, const int16_t *base, vuint16mf4_t bindex, size_t vl) {
@@ -2429,7 +2429,7 @@ vint16mf4_t test_vluxei16_v_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vluxei16_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vluxei.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vluxei.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP1]]
 //
 vint16mf2_t test_vluxei16_v_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, const int16_t *base, vuint16mf2_t bindex, size_t vl) {
@@ -2440,7 +2440,7 @@ vint16mf2_t test_vluxei16_v_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vluxei16_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vluxei.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vluxei.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP1]]
 //
 vint16m1_t test_vluxei16_v_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, const int16_t *base, vuint16m1_t bindex, size_t vl) {
@@ -2451,7 +2451,7 @@ vint16m1_t test_vluxei16_v_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei16_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vluxei.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vluxei.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP1]]
 //
 vint16m2_t test_vluxei16_v_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, const int16_t *base, vuint16m2_t bindex, size_t vl) {
@@ -2462,7 +2462,7 @@ vint16m2_t test_vluxei16_v_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei16_v_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vluxei.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vluxei.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP1]]
 //
 vint16m4_t test_vluxei16_v_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, const int16_t *base, vuint16m4_t bindex, size_t vl) {
@@ -2473,7 +2473,7 @@ vint16m4_t test_vluxei16_v_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei16_v_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vluxei.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vluxei.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP1]]
 //
 vint16m8_t test_vluxei16_v_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, const int16_t *base, vuint16m8_t bindex, size_t vl) {
@@ -2484,7 +2484,7 @@ vint16m8_t test_vluxei16_v_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei32_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vluxei.mask.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vluxei.mask.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP1]]
 //
 vint16mf4_t test_vluxei32_v_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, const int16_t *base, vuint32mf2_t bindex, size_t vl) {
@@ -2495,7 +2495,7 @@ vint16mf4_t test_vluxei32_v_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vluxei32_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vluxei.mask.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vluxei.mask.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP1]]
 //
 vint16mf2_t test_vluxei32_v_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, const int16_t *base, vuint32m1_t bindex, size_t vl) {
@@ -2506,7 +2506,7 @@ vint16mf2_t test_vluxei32_v_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vluxei32_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vluxei.mask.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vluxei.mask.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP1]]
 //
 vint16m1_t test_vluxei32_v_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, const int16_t *base, vuint32m2_t bindex, size_t vl) {
@@ -2517,7 +2517,7 @@ vint16m1_t test_vluxei32_v_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei32_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vluxei.mask.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vluxei.mask.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP1]]
 //
 vint16m2_t test_vluxei32_v_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, const int16_t *base, vuint32m4_t bindex, size_t vl) {
@@ -2528,7 +2528,7 @@ vint16m2_t test_vluxei32_v_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei32_v_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vluxei.mask.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vluxei.mask.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP1]]
 //
 vint16m4_t test_vluxei32_v_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, const int16_t *base, vuint32m8_t bindex, size_t vl) {
@@ -2539,7 +2539,7 @@ vint16m4_t test_vluxei32_v_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei64_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vluxei.mask.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vluxei.mask.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP1]]
 //
 vint16mf4_t test_vluxei64_v_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, const int16_t *base, vuint64m1_t bindex, size_t vl) {
@@ -2550,7 +2550,7 @@ vint16mf4_t test_vluxei64_v_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vluxei64_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vluxei.mask.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vluxei.mask.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP1]]
 //
 vint16mf2_t test_vluxei64_v_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, const int16_t *base, vuint64m2_t bindex, size_t vl) {
@@ -2561,7 +2561,7 @@ vint16mf2_t test_vluxei64_v_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vluxei64_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vluxei.mask.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vluxei.mask.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP1]]
 //
 vint16m1_t test_vluxei64_v_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, const int16_t *base, vuint64m4_t bindex, size_t vl) {
@@ -2572,7 +2572,7 @@ vint16m1_t test_vluxei64_v_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei64_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vluxei.mask.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vluxei.mask.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP1]]
 //
 vint16m2_t test_vluxei64_v_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, const int16_t *base, vuint64m8_t bindex, size_t vl) {
@@ -2583,7 +2583,7 @@ vint16m2_t test_vluxei64_v_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei8_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vluxei.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vluxei.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vluxei8_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, const int32_t *base, vuint8mf8_t bindex, size_t vl) {
@@ -2594,7 +2594,7 @@ vint32mf2_t test_vluxei8_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei8_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vluxei.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vluxei.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vluxei8_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, const int32_t *base, vuint8mf4_t bindex, size_t vl) {
@@ -2605,7 +2605,7 @@ vint32m1_t test_vluxei8_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei8_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vluxei.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vluxei.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vluxei8_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, const int32_t *base, vuint8mf2_t bindex, size_t vl) {
@@ -2616,7 +2616,7 @@ vint32m2_t test_vluxei8_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei8_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vluxei.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vluxei.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vluxei8_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, const int32_t *base, vuint8m1_t bindex, size_t vl) {
@@ -2627,7 +2627,7 @@ vint32m4_t test_vluxei8_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, const int
 // CHECK-RV64-LABEL: @test_vluxei8_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vluxei.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vluxei.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vluxei8_v_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, const int32_t *base, vuint8m2_t bindex, size_t vl) {
@@ -2638,7 +2638,7 @@ vint32m8_t test_vluxei8_v_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, const int
 // CHECK-RV64-LABEL: @test_vluxei16_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vluxei.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vluxei.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vluxei16_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, const int32_t *base, vuint16mf4_t bindex, size_t vl) {
@@ -2649,7 +2649,7 @@ vint32mf2_t test_vluxei16_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vluxei16_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vluxei.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vluxei.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vluxei16_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, const int32_t *base, vuint16mf2_t bindex, size_t vl) {
@@ -2660,7 +2660,7 @@ vint32m1_t test_vluxei16_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei16_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vluxei.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vluxei.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vluxei16_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, const int32_t *base, vuint16m1_t bindex, size_t vl) {
@@ -2671,7 +2671,7 @@ vint32m2_t test_vluxei16_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei16_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vluxei.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vluxei.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vluxei16_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, const int32_t *base, vuint16m2_t bindex, size_t vl) {
@@ -2682,7 +2682,7 @@ vint32m4_t test_vluxei16_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei16_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vluxei.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vluxei.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vluxei16_v_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, const int32_t *base, vuint16m4_t bindex, size_t vl) {
@@ -2693,7 +2693,7 @@ vint32m8_t test_vluxei16_v_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei32_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vluxei.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vluxei.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vluxei32_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, const int32_t *base, vuint32mf2_t bindex, size_t vl) {
@@ -2704,7 +2704,7 @@ vint32mf2_t test_vluxei32_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vluxei32_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vluxei.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vluxei.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vluxei32_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, const int32_t *base, vuint32m1_t bindex, size_t vl) {
@@ -2715,7 +2715,7 @@ vint32m1_t test_vluxei32_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei32_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vluxei.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vluxei.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vluxei32_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, const int32_t *base, vuint32m2_t bindex, size_t vl) {
@@ -2726,7 +2726,7 @@ vint32m2_t test_vluxei32_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei32_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vluxei.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vluxei.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vluxei32_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, const int32_t *base, vuint32m4_t bindex, size_t vl) {
@@ -2737,7 +2737,7 @@ vint32m4_t test_vluxei32_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei32_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vluxei.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vluxei.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vint32m8_t test_vluxei32_v_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, const int32_t *base, vuint32m8_t bindex, size_t vl) {
@@ -2748,7 +2748,7 @@ vint32m8_t test_vluxei32_v_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei64_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vluxei.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vluxei.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vint32mf2_t test_vluxei64_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, const int32_t *base, vuint64m1_t bindex, size_t vl) {
@@ -2759,7 +2759,7 @@ vint32mf2_t test_vluxei64_v_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vluxei64_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vluxei.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vluxei.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vint32m1_t test_vluxei64_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, const int32_t *base, vuint64m2_t bindex, size_t vl) {
@@ -2770,7 +2770,7 @@ vint32m1_t test_vluxei64_v_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei64_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vluxei.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vluxei.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vint32m2_t test_vluxei64_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, const int32_t *base, vuint64m4_t bindex, size_t vl) {
@@ -2781,7 +2781,7 @@ vint32m2_t test_vluxei64_v_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei64_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vluxei.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vluxei.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vint32m4_t test_vluxei64_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, const int32_t *base, vuint64m8_t bindex, size_t vl) {
@@ -2792,7 +2792,7 @@ vint32m4_t test_vluxei64_v_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei8_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vluxei.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vluxei.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vluxei8_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, const int64_t *base, vuint8mf8_t bindex, size_t vl) {
@@ -2803,7 +2803,7 @@ vint64m1_t test_vluxei8_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei8_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vluxei.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vluxei.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vluxei8_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, const int64_t *base, vuint8mf4_t bindex, size_t vl) {
@@ -2814,7 +2814,7 @@ vint64m2_t test_vluxei8_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei8_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vluxei.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vluxei.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vluxei8_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, const int64_t *base, vuint8mf2_t bindex, size_t vl) {
@@ -2825,7 +2825,7 @@ vint64m4_t test_vluxei8_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei8_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vluxei.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vluxei.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vluxei8_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, const int64_t *base, vuint8m1_t bindex, size_t vl) {
@@ -2836,7 +2836,7 @@ vint64m8_t test_vluxei8_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, const int
 // CHECK-RV64-LABEL: @test_vluxei16_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vluxei.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vluxei.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vluxei16_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, const int64_t *base, vuint16mf4_t bindex, size_t vl) {
@@ -2847,7 +2847,7 @@ vint64m1_t test_vluxei16_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei16_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vluxei.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vluxei.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vluxei16_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, const int64_t *base, vuint16mf2_t bindex, size_t vl) {
@@ -2858,7 +2858,7 @@ vint64m2_t test_vluxei16_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei16_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vluxei.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vluxei.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vluxei16_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, const int64_t *base, vuint16m1_t bindex, size_t vl) {
@@ -2869,7 +2869,7 @@ vint64m4_t test_vluxei16_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei16_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vluxei.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vluxei.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vluxei16_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, const int64_t *base, vuint16m2_t bindex, size_t vl) {
@@ -2880,7 +2880,7 @@ vint64m8_t test_vluxei16_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei32_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vluxei.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vluxei.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vluxei32_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, const int64_t *base, vuint32mf2_t bindex, size_t vl) {
@@ -2891,7 +2891,7 @@ vint64m1_t test_vluxei32_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei32_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vluxei.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vluxei.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vluxei32_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, const int64_t *base, vuint32m1_t bindex, size_t vl) {
@@ -2902,7 +2902,7 @@ vint64m2_t test_vluxei32_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei32_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vluxei.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vluxei.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vluxei32_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, const int64_t *base, vuint32m2_t bindex, size_t vl) {
@@ -2913,7 +2913,7 @@ vint64m4_t test_vluxei32_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei32_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vluxei.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vluxei.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vluxei32_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, const int64_t *base, vuint32m4_t bindex, size_t vl) {
@@ -2924,7 +2924,7 @@ vint64m8_t test_vluxei32_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei64_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vluxei.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vluxei.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vint64m1_t test_vluxei64_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, const int64_t *base, vuint64m1_t bindex, size_t vl) {
@@ -2935,7 +2935,7 @@ vint64m1_t test_vluxei64_v_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei64_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vluxei.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vluxei.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vint64m2_t test_vluxei64_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, const int64_t *base, vuint64m2_t bindex, size_t vl) {
@@ -2946,7 +2946,7 @@ vint64m2_t test_vluxei64_v_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei64_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vluxei.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vluxei.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vint64m4_t test_vluxei64_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, const int64_t *base, vuint64m4_t bindex, size_t vl) {
@@ -2957,7 +2957,7 @@ vint64m4_t test_vluxei64_v_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, const i
 // CHECK-RV64-LABEL: @test_vluxei64_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vluxei.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vluxei.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vint64m8_t test_vluxei64_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, const int64_t *base, vuint64m8_t bindex, size_t vl) {
@@ -2968,7 +2968,7 @@ vint64m8_t test_vluxei64_v_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, const in
 // CHECK-RV64-LABEL: @test_vluxei8_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vluxei.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vluxei.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]
 //
 vuint8mf8_t test_vluxei8_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, const uint8_t *base, vuint8mf8_t bindex, size_t vl) {
@@ -2979,7 +2979,7 @@ vuint8mf8_t test_vluxei8_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei8_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vluxei.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vluxei.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP1]]
 //
 vuint8mf4_t test_vluxei8_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, const uint8_t *base, vuint8mf4_t bindex, size_t vl) {
@@ -2990,7 +2990,7 @@ vuint8mf4_t test_vluxei8_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei8_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vluxei.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vluxei.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP1]]
 //
 vuint8mf2_t test_vluxei8_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, const uint8_t *base, vuint8mf2_t bindex, size_t vl) {
@@ -3001,7 +3001,7 @@ vuint8mf2_t test_vluxei8_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei8_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vluxei.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vluxei.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP1]]
 //
 vuint8m1_t test_vluxei8_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, const uint8_t *base, vuint8m1_t bindex, size_t vl) {
@@ -3012,7 +3012,7 @@ vuint8m1_t test_vluxei8_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, const uint
 // CHECK-RV64-LABEL: @test_vluxei8_v_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vluxei.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vluxei.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP1]]
 //
 vuint8m2_t test_vluxei8_v_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, const uint8_t *base, vuint8m2_t bindex, size_t vl) {
@@ -3023,7 +3023,7 @@ vuint8m2_t test_vluxei8_v_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, const uint
 // CHECK-RV64-LABEL: @test_vluxei8_v_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vluxei.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vluxei.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP1]]
 //
 vuint8m4_t test_vluxei8_v_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, const uint8_t *base, vuint8m4_t bindex, size_t vl) {
@@ -3034,7 +3034,7 @@ vuint8m4_t test_vluxei8_v_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, const uint
 // CHECK-RV64-LABEL: @test_vluxei8_v_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vluxei.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vluxei.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP1]]
 //
 vuint8m8_t test_vluxei8_v_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, const uint8_t *base, vuint8m8_t bindex, size_t vl) {
@@ -3045,7 +3045,7 @@ vuint8m8_t test_vluxei8_v_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, const uint
 // CHECK-RV64-LABEL: @test_vluxei16_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vluxei.mask.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vluxei.mask.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]
 //
 vuint8mf8_t test_vluxei16_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, const uint8_t *base, vuint16mf4_t bindex, size_t vl) {
@@ -3056,7 +3056,7 @@ vuint8mf8_t test_vluxei16_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei16_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vluxei.mask.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vluxei.mask.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP1]]
 //
 vuint8mf4_t test_vluxei16_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, const uint8_t *base, vuint16mf2_t bindex, size_t vl) {
@@ -3067,7 +3067,7 @@ vuint8mf4_t test_vluxei16_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei16_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vluxei.mask.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vluxei.mask.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP1]]
 //
 vuint8mf2_t test_vluxei16_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, const uint8_t *base, vuint16m1_t bindex, size_t vl) {
@@ -3078,7 +3078,7 @@ vuint8mf2_t test_vluxei16_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei16_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vluxei.mask.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vluxei.mask.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP1]]
 //
 vuint8m1_t test_vluxei16_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, const uint8_t *base, vuint16m2_t bindex, size_t vl) {
@@ -3089,7 +3089,7 @@ vuint8m1_t test_vluxei16_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, const uin
 // CHECK-RV64-LABEL: @test_vluxei16_v_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vluxei.mask.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vluxei.mask.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP1]]
 //
 vuint8m2_t test_vluxei16_v_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, const uint8_t *base, vuint16m4_t bindex, size_t vl) {
@@ -3100,7 +3100,7 @@ vuint8m2_t test_vluxei16_v_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, const uin
 // CHECK-RV64-LABEL: @test_vluxei16_v_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vluxei.mask.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vluxei.mask.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP1]]
 //
 vuint8m4_t test_vluxei16_v_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, const uint8_t *base, vuint16m8_t bindex, size_t vl) {
@@ -3111,7 +3111,7 @@ vuint8m4_t test_vluxei16_v_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, const uin
 // CHECK-RV64-LABEL: @test_vluxei32_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vluxei.mask.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vluxei.mask.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]
 //
 vuint8mf8_t test_vluxei32_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, const uint8_t *base, vuint32mf2_t bindex, size_t vl) {
@@ -3122,7 +3122,7 @@ vuint8mf8_t test_vluxei32_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei32_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vluxei.mask.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vluxei.mask.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP1]]
 //
 vuint8mf4_t test_vluxei32_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, const uint8_t *base, vuint32m1_t bindex, size_t vl) {
@@ -3133,7 +3133,7 @@ vuint8mf4_t test_vluxei32_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei32_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vluxei.mask.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vluxei.mask.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP1]]
 //
 vuint8mf2_t test_vluxei32_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, const uint8_t *base, vuint32m2_t bindex, size_t vl) {
@@ -3144,7 +3144,7 @@ vuint8mf2_t test_vluxei32_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei32_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vluxei.mask.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vluxei.mask.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP1]]
 //
 vuint8m1_t test_vluxei32_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, const uint8_t *base, vuint32m4_t bindex, size_t vl) {
@@ -3155,7 +3155,7 @@ vuint8m1_t test_vluxei32_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, const uin
 // CHECK-RV64-LABEL: @test_vluxei32_v_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vluxei.mask.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vluxei.mask.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP1]]
 //
 vuint8m2_t test_vluxei32_v_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, const uint8_t *base, vuint32m8_t bindex, size_t vl) {
@@ -3166,7 +3166,7 @@ vuint8m2_t test_vluxei32_v_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, const uin
 // CHECK-RV64-LABEL: @test_vluxei64_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vluxei.mask.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vluxei.mask.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]
 //
 vuint8mf8_t test_vluxei64_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, const uint8_t *base, vuint64m1_t bindex, size_t vl) {
@@ -3177,7 +3177,7 @@ vuint8mf8_t test_vluxei64_v_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei64_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vluxei.mask.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vluxei.mask.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP1]]
 //
 vuint8mf4_t test_vluxei64_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, const uint8_t *base, vuint64m2_t bindex, size_t vl) {
@@ -3188,7 +3188,7 @@ vuint8mf4_t test_vluxei64_v_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei64_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vluxei.mask.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vluxei.mask.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP1]]
 //
 vuint8mf2_t test_vluxei64_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, const uint8_t *base, vuint64m4_t bindex, size_t vl) {
@@ -3199,7 +3199,7 @@ vuint8mf2_t test_vluxei64_v_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei64_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vluxei.mask.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vluxei.mask.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP1]]
 //
 vuint8m1_t test_vluxei64_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, const uint8_t *base, vuint64m8_t bindex, size_t vl) {
@@ -3210,7 +3210,7 @@ vuint8m1_t test_vluxei64_v_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, const uin
 // CHECK-RV64-LABEL: @test_vluxei8_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vluxei.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vluxei.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP1]]
 //
 vuint16mf4_t test_vluxei8_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, const uint16_t *base, vuint8mf8_t bindex, size_t vl) {
@@ -3221,7 +3221,7 @@ vuint16mf4_t test_vluxei8_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vluxei8_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vluxei.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vluxei.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP1]]
 //
 vuint16mf2_t test_vluxei8_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, const uint16_t *base, vuint8mf4_t bindex, size_t vl) {
@@ -3232,7 +3232,7 @@ vuint16mf2_t test_vluxei8_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vluxei8_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vluxei.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vluxei.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP1]]
 //
 vuint16m1_t test_vluxei8_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, const uint16_t *base, vuint8mf2_t bindex, size_t vl) {
@@ -3243,7 +3243,7 @@ vuint16m1_t test_vluxei8_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei8_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vluxei.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vluxei.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP1]]
 //
 vuint16m2_t test_vluxei8_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, const uint16_t *base, vuint8m1_t bindex, size_t vl) {
@@ -3254,7 +3254,7 @@ vuint16m2_t test_vluxei8_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, const u
 // CHECK-RV64-LABEL: @test_vluxei8_v_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vluxei.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vluxei.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP1]]
 //
 vuint16m4_t test_vluxei8_v_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, const uint16_t *base, vuint8m2_t bindex, size_t vl) {
@@ -3265,7 +3265,7 @@ vuint16m4_t test_vluxei8_v_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, const u
 // CHECK-RV64-LABEL: @test_vluxei8_v_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vluxei.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vluxei.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP1]]
 //
 vuint16m8_t test_vluxei8_v_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, const uint16_t *base, vuint8m4_t bindex, size_t vl) {
@@ -3276,7 +3276,7 @@ vuint16m8_t test_vluxei8_v_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, const u
 // CHECK-RV64-LABEL: @test_vluxei16_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vluxei.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vluxei.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP1]]
 //
 vuint16mf4_t test_vluxei16_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, const uint16_t *base, vuint16mf4_t bindex, size_t vl) {
@@ -3287,7 +3287,7 @@ vuint16mf4_t test_vluxei16_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, co
 // CHECK-RV64-LABEL: @test_vluxei16_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vluxei.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vluxei.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP1]]
 //
 vuint16mf2_t test_vluxei16_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, const uint16_t *base, vuint16mf2_t bindex, size_t vl) {
@@ -3298,7 +3298,7 @@ vuint16mf2_t test_vluxei16_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, co
 // CHECK-RV64-LABEL: @test_vluxei16_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vluxei.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vluxei.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP1]]
 //
 vuint16m1_t test_vluxei16_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, const uint16_t *base, vuint16m1_t bindex, size_t vl) {
@@ -3309,7 +3309,7 @@ vuint16m1_t test_vluxei16_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei16_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vluxei.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vluxei.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP1]]
 //
 vuint16m2_t test_vluxei16_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, const uint16_t *base, vuint16m2_t bindex, size_t vl) {
@@ -3320,7 +3320,7 @@ vuint16m2_t test_vluxei16_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei16_v_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vluxei.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vluxei.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP1]]
 //
 vuint16m4_t test_vluxei16_v_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, const uint16_t *base, vuint16m4_t bindex, size_t vl) {
@@ -3331,7 +3331,7 @@ vuint16m4_t test_vluxei16_v_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei16_v_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vluxei.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vluxei.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP1]]
 //
 vuint16m8_t test_vluxei16_v_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, const uint16_t *base, vuint16m8_t bindex, size_t vl) {
@@ -3342,7 +3342,7 @@ vuint16m8_t test_vluxei16_v_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei32_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vluxei.mask.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vluxei.mask.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP1]]
 //
 vuint16mf4_t test_vluxei32_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, const uint16_t *base, vuint32mf2_t bindex, size_t vl) {
@@ -3353,7 +3353,7 @@ vuint16mf4_t test_vluxei32_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, co
 // CHECK-RV64-LABEL: @test_vluxei32_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vluxei.mask.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vluxei.mask.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP1]]
 //
 vuint16mf2_t test_vluxei32_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, const uint16_t *base, vuint32m1_t bindex, size_t vl) {
@@ -3364,7 +3364,7 @@ vuint16mf2_t test_vluxei32_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, co
 // CHECK-RV64-LABEL: @test_vluxei32_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vluxei.mask.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vluxei.mask.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP1]]
 //
 vuint16m1_t test_vluxei32_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, const uint16_t *base, vuint32m2_t bindex, size_t vl) {
@@ -3375,7 +3375,7 @@ vuint16m1_t test_vluxei32_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei32_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vluxei.mask.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vluxei.mask.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP1]]
 //
 vuint16m2_t test_vluxei32_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, const uint16_t *base, vuint32m4_t bindex, size_t vl) {
@@ -3386,7 +3386,7 @@ vuint16m2_t test_vluxei32_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei32_v_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vluxei.mask.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vluxei.mask.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP1]]
 //
 vuint16m4_t test_vluxei32_v_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, const uint16_t *base, vuint32m8_t bindex, size_t vl) {
@@ -3397,7 +3397,7 @@ vuint16m4_t test_vluxei32_v_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei64_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vluxei.mask.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vluxei.mask.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP1]]
 //
 vuint16mf4_t test_vluxei64_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, const uint16_t *base, vuint64m1_t bindex, size_t vl) {
@@ -3408,7 +3408,7 @@ vuint16mf4_t test_vluxei64_v_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, co
 // CHECK-RV64-LABEL: @test_vluxei64_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vluxei.mask.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vluxei.mask.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP1]]
 //
 vuint16mf2_t test_vluxei64_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, const uint16_t *base, vuint64m2_t bindex, size_t vl) {
@@ -3419,7 +3419,7 @@ vuint16mf2_t test_vluxei64_v_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, co
 // CHECK-RV64-LABEL: @test_vluxei64_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vluxei.mask.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vluxei.mask.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP1]]
 //
 vuint16m1_t test_vluxei64_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, const uint16_t *base, vuint64m4_t bindex, size_t vl) {
@@ -3430,7 +3430,7 @@ vuint16m1_t test_vluxei64_v_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei64_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vluxei.mask.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vluxei.mask.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP1]]
 //
 vuint16m2_t test_vluxei64_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, const uint16_t *base, vuint64m8_t bindex, size_t vl) {
@@ -3441,7 +3441,7 @@ vuint16m2_t test_vluxei64_v_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei8_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vluxei.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vluxei.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vluxei8_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, const uint32_t *base, vuint8mf8_t bindex, size_t vl) {
@@ -3452,7 +3452,7 @@ vuint32mf2_t test_vluxei8_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vluxei8_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vluxei.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vluxei.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vluxei8_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, const uint32_t *base, vuint8mf4_t bindex, size_t vl) {
@@ -3463,7 +3463,7 @@ vuint32m1_t test_vluxei8_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei8_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vluxei.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vluxei.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vluxei8_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, const uint32_t *base, vuint8mf2_t bindex, size_t vl) {
@@ -3474,7 +3474,7 @@ vuint32m2_t test_vluxei8_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei8_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vluxei.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vluxei.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vluxei8_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, const uint32_t *base, vuint8m1_t bindex, size_t vl) {
@@ -3485,7 +3485,7 @@ vuint32m4_t test_vluxei8_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, const u
 // CHECK-RV64-LABEL: @test_vluxei8_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vluxei.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vluxei.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vluxei8_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, const uint32_t *base, vuint8m2_t bindex, size_t vl) {
@@ -3496,7 +3496,7 @@ vuint32m8_t test_vluxei8_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, const u
 // CHECK-RV64-LABEL: @test_vluxei16_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vluxei.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vluxei.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vluxei16_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, const uint32_t *base, vuint16mf4_t bindex, size_t vl) {
@@ -3507,7 +3507,7 @@ vuint32mf2_t test_vluxei16_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, co
 // CHECK-RV64-LABEL: @test_vluxei16_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vluxei.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vluxei.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vluxei16_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, const uint32_t *base, vuint16mf2_t bindex, size_t vl) {
@@ -3518,7 +3518,7 @@ vuint32m1_t test_vluxei16_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei16_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vluxei.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vluxei.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vluxei16_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, const uint32_t *base, vuint16m1_t bindex, size_t vl) {
@@ -3529,7 +3529,7 @@ vuint32m2_t test_vluxei16_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei16_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vluxei.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vluxei.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vluxei16_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, const uint32_t *base, vuint16m2_t bindex, size_t vl) {
@@ -3540,7 +3540,7 @@ vuint32m4_t test_vluxei16_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei16_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vluxei.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vluxei.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vluxei16_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, const uint32_t *base, vuint16m4_t bindex, size_t vl) {
@@ -3551,7 +3551,7 @@ vuint32m8_t test_vluxei16_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei32_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vluxei.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vluxei.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vluxei32_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, const uint32_t *base, vuint32mf2_t bindex, size_t vl) {
@@ -3562,7 +3562,7 @@ vuint32mf2_t test_vluxei32_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, co
 // CHECK-RV64-LABEL: @test_vluxei32_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vluxei.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vluxei.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vluxei32_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, const uint32_t *base, vuint32m1_t bindex, size_t vl) {
@@ -3573,7 +3573,7 @@ vuint32m1_t test_vluxei32_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei32_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vluxei.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vluxei.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vluxei32_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, const uint32_t *base, vuint32m2_t bindex, size_t vl) {
@@ -3584,7 +3584,7 @@ vuint32m2_t test_vluxei32_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei32_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vluxei.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vluxei.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vluxei32_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, const uint32_t *base, vuint32m4_t bindex, size_t vl) {
@@ -3595,7 +3595,7 @@ vuint32m4_t test_vluxei32_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei32_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vluxei.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vluxei.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP1]]
 //
 vuint32m8_t test_vluxei32_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, const uint32_t *base, vuint32m8_t bindex, size_t vl) {
@@ -3606,7 +3606,7 @@ vuint32m8_t test_vluxei32_v_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei64_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vluxei.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vluxei.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP1]]
 //
 vuint32mf2_t test_vluxei64_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, const uint32_t *base, vuint64m1_t bindex, size_t vl) {
@@ -3617,7 +3617,7 @@ vuint32mf2_t test_vluxei64_v_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, co
 // CHECK-RV64-LABEL: @test_vluxei64_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vluxei.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vluxei.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP1]]
 //
 vuint32m1_t test_vluxei64_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, const uint32_t *base, vuint64m2_t bindex, size_t vl) {
@@ -3628,7 +3628,7 @@ vuint32m1_t test_vluxei64_v_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei64_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vluxei.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vluxei.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
 //
 vuint32m2_t test_vluxei64_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, const uint32_t *base, vuint64m4_t bindex, size_t vl) {
@@ -3639,7 +3639,7 @@ vuint32m2_t test_vluxei64_v_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei64_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vluxei.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vluxei.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP1]]
 //
 vuint32m4_t test_vluxei64_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, const uint32_t *base, vuint64m8_t bindex, size_t vl) {
@@ -3650,7 +3650,7 @@ vuint32m4_t test_vluxei64_v_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei8_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vluxei.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vluxei.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vluxei8_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, const uint64_t *base, vuint8mf8_t bindex, size_t vl) {
@@ -3661,7 +3661,7 @@ vuint64m1_t test_vluxei8_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei8_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vluxei.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vluxei.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vluxei8_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, const uint64_t *base, vuint8mf4_t bindex, size_t vl) {
@@ -3672,7 +3672,7 @@ vuint64m2_t test_vluxei8_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei8_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vluxei.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vluxei.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vluxei8_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, const uint64_t *base, vuint8mf2_t bindex, size_t vl) {
@@ -3683,7 +3683,7 @@ vuint64m4_t test_vluxei8_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei8_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vluxei.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vluxei.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vluxei8_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, const uint64_t *base, vuint8m1_t bindex, size_t vl) {
@@ -3694,7 +3694,7 @@ vuint64m8_t test_vluxei8_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, const u
 // CHECK-RV64-LABEL: @test_vluxei16_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vluxei.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vluxei.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vluxei16_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, const uint64_t *base, vuint16mf4_t bindex, size_t vl) {
@@ -3705,7 +3705,7 @@ vuint64m1_t test_vluxei16_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei16_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vluxei.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vluxei.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vluxei16_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, const uint64_t *base, vuint16mf2_t bindex, size_t vl) {
@@ -3716,7 +3716,7 @@ vuint64m2_t test_vluxei16_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei16_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vluxei.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vluxei.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vluxei16_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, const uint64_t *base, vuint16m1_t bindex, size_t vl) {
@@ -3727,7 +3727,7 @@ vuint64m4_t test_vluxei16_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei16_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vluxei.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vluxei.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vluxei16_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, const uint64_t *base, vuint16m2_t bindex, size_t vl) {
@@ -3738,7 +3738,7 @@ vuint64m8_t test_vluxei16_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei32_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vluxei.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vluxei.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vluxei32_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, const uint64_t *base, vuint32mf2_t bindex, size_t vl) {
@@ -3749,7 +3749,7 @@ vuint64m1_t test_vluxei32_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei32_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vluxei.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vluxei.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vluxei32_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, const uint64_t *base, vuint32m1_t bindex, size_t vl) {
@@ -3760,7 +3760,7 @@ vuint64m2_t test_vluxei32_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei32_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vluxei.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vluxei.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vluxei32_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, const uint64_t *base, vuint32m2_t bindex, size_t vl) {
@@ -3771,7 +3771,7 @@ vuint64m4_t test_vluxei32_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei32_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vluxei.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vluxei.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vluxei32_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, const uint64_t *base, vuint32m4_t bindex, size_t vl) {
@@ -3782,7 +3782,7 @@ vuint64m8_t test_vluxei32_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei64_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vluxei.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vluxei.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP1]]
 //
 vuint64m1_t test_vluxei64_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, const uint64_t *base, vuint64m1_t bindex, size_t vl) {
@@ -3793,7 +3793,7 @@ vuint64m1_t test_vluxei64_v_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei64_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vluxei.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vluxei.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
 //
 vuint64m2_t test_vluxei64_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, const uint64_t *base, vuint64m2_t bindex, size_t vl) {
@@ -3804,7 +3804,7 @@ vuint64m2_t test_vluxei64_v_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei64_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vluxei.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vluxei.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP1]]
 //
 vuint64m4_t test_vluxei64_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, const uint64_t *base, vuint64m4_t bindex, size_t vl) {
@@ -3815,7 +3815,7 @@ vuint64m4_t test_vluxei64_v_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei64_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vluxei.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vluxei.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP1]]
 //
 vuint64m8_t test_vluxei64_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, const uint64_t *base, vuint64m8_t bindex, size_t vl) {
@@ -3826,7 +3826,7 @@ vuint64m8_t test_vluxei64_v_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei8_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vluxei.mask.nxv1f32.nxv1i8.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vluxei.mask.nxv1f32.nxv1i8.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP1]]
 //
 vfloat32mf2_t test_vluxei8_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff, const float *base, vuint8mf8_t bindex, size_t vl) {
@@ -3837,7 +3837,7 @@ vfloat32mf2_t test_vluxei8_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff, c
 // CHECK-RV64-LABEL: @test_vluxei8_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vluxei.mask.nxv2f32.nxv2i8.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vluxei.mask.nxv2f32.nxv2i8.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP1]]
 //
 vfloat32m1_t test_vluxei8_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, const float *base, vuint8mf4_t bindex, size_t vl) {
@@ -3848,7 +3848,7 @@ vfloat32m1_t test_vluxei8_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vluxei8_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vluxei.mask.nxv4f32.nxv4i8.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vluxei.mask.nxv4f32.nxv4i8.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 vfloat32m2_t test_vluxei8_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, const float *base, vuint8mf2_t bindex, size_t vl) {
@@ -3859,7 +3859,7 @@ vfloat32m2_t test_vluxei8_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vluxei8_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vluxei.mask.nxv8f32.nxv8i8.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vluxei.mask.nxv8f32.nxv8i8.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP1]]
 //
 vfloat32m4_t test_vluxei8_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, const float *base, vuint8m1_t bindex, size_t vl) {
@@ -3870,7 +3870,7 @@ vfloat32m4_t test_vluxei8_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei8_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vluxei.mask.nxv16f32.nxv16i8.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vluxei.mask.nxv16f32.nxv16i8.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP1]]
 //
 vfloat32m8_t test_vluxei8_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff, const float *base, vuint8m2_t bindex, size_t vl) {
@@ -3881,7 +3881,7 @@ vfloat32m8_t test_vluxei8_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei16_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vluxei.mask.nxv1f32.nxv1i16.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vluxei.mask.nxv1f32.nxv1i16.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP1]]
 //
 vfloat32mf2_t test_vluxei16_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff, const float *base, vuint16mf4_t bindex, size_t vl) {
@@ -3892,7 +3892,7 @@ vfloat32mf2_t test_vluxei16_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vluxei16_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vluxei.mask.nxv2f32.nxv2i16.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vluxei.mask.nxv2f32.nxv2i16.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP1]]
 //
 vfloat32m1_t test_vluxei16_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, const float *base, vuint16mf2_t bindex, size_t vl) {
@@ -3903,7 +3903,7 @@ vfloat32m1_t test_vluxei16_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vluxei16_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vluxei.mask.nxv4f32.nxv4i16.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vluxei.mask.nxv4f32.nxv4i16.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 vfloat32m2_t test_vluxei16_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, const float *base, vuint16m1_t bindex, size_t vl) {
@@ -3914,7 +3914,7 @@ vfloat32m2_t test_vluxei16_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vluxei16_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vluxei.mask.nxv8f32.nxv8i16.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vluxei.mask.nxv8f32.nxv8i16.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP1]]
 //
 vfloat32m4_t test_vluxei16_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, const float *base, vuint16m2_t bindex, size_t vl) {
@@ -3925,7 +3925,7 @@ vfloat32m4_t test_vluxei16_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vluxei16_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vluxei.mask.nxv16f32.nxv16i16.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vluxei.mask.nxv16f32.nxv16i16.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP1]]
 //
 vfloat32m8_t test_vluxei16_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff, const float *base, vuint16m4_t bindex, size_t vl) {
@@ -3936,7 +3936,7 @@ vfloat32m8_t test_vluxei16_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vluxei32_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vluxei.mask.nxv1f32.nxv1i32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vluxei.mask.nxv1f32.nxv1i32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP1]]
 //
 vfloat32mf2_t test_vluxei32_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff, const float *base, vuint32mf2_t bindex, size_t vl) {
@@ -3947,7 +3947,7 @@ vfloat32mf2_t test_vluxei32_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vluxei32_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vluxei.mask.nxv2f32.nxv2i32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vluxei.mask.nxv2f32.nxv2i32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP1]]
 //
 vfloat32m1_t test_vluxei32_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, const float *base, vuint32m1_t bindex, size_t vl) {
@@ -3958,7 +3958,7 @@ vfloat32m1_t test_vluxei32_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vluxei32_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vluxei.mask.nxv4f32.nxv4i32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vluxei.mask.nxv4f32.nxv4i32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 vfloat32m2_t test_vluxei32_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, const float *base, vuint32m2_t bindex, size_t vl) {
@@ -3969,7 +3969,7 @@ vfloat32m2_t test_vluxei32_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vluxei32_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vluxei.mask.nxv8f32.nxv8i32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vluxei.mask.nxv8f32.nxv8i32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP1]]
 //
 vfloat32m4_t test_vluxei32_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, const float *base, vuint32m4_t bindex, size_t vl) {
@@ -3980,7 +3980,7 @@ vfloat32m4_t test_vluxei32_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vluxei32_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vluxei.mask.nxv16f32.nxv16i32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x float> @llvm.riscv.vluxei.mask.nxv16f32.nxv16i32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP1]]
 //
 vfloat32m8_t test_vluxei32_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff, const float *base, vuint32m8_t bindex, size_t vl) {
@@ -3991,7 +3991,7 @@ vfloat32m8_t test_vluxei32_v_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vluxei64_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vluxei.mask.nxv1f32.nxv1i64.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x float> @llvm.riscv.vluxei.mask.nxv1f32.nxv1i64.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP1]]
 //
 vfloat32mf2_t test_vluxei64_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff, const float *base, vuint64m1_t bindex, size_t vl) {
@@ -4002,7 +4002,7 @@ vfloat32mf2_t test_vluxei64_v_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 // CHECK-RV64-LABEL: @test_vluxei64_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vluxei.mask.nxv2f32.nxv2i64.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x float> @llvm.riscv.vluxei.mask.nxv2f32.nxv2i64.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP1]]
 //
 vfloat32m1_t test_vluxei64_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, const float *base, vuint64m2_t bindex, size_t vl) {
@@ -4013,7 +4013,7 @@ vfloat32m1_t test_vluxei64_v_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vluxei64_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vluxei.mask.nxv4f32.nxv4i64.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x float> @llvm.riscv.vluxei.mask.nxv4f32.nxv4i64.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 vfloat32m2_t test_vluxei64_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, const float *base, vuint64m4_t bindex, size_t vl) {
@@ -4024,7 +4024,7 @@ vfloat32m2_t test_vluxei64_v_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vluxei64_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vluxei.mask.nxv8f32.nxv8i64.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x float> @llvm.riscv.vluxei.mask.nxv8f32.nxv8i64.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP1]]
 //
 vfloat32m4_t test_vluxei64_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, const float *base, vuint64m8_t bindex, size_t vl) {
@@ -4035,7 +4035,7 @@ vfloat32m4_t test_vluxei64_v_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vluxei8_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vluxei.mask.nxv1f64.nxv1i8.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vluxei.mask.nxv1f64.nxv1i8.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP1]]
 //
 vfloat64m1_t test_vluxei8_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, const double *base, vuint8mf8_t bindex, size_t vl) {
@@ -4046,7 +4046,7 @@ vfloat64m1_t test_vluxei8_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vluxei8_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vluxei.mask.nxv2f64.nxv2i8.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vluxei.mask.nxv2f64.nxv2i8.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 vfloat64m2_t test_vluxei8_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, const double *base, vuint8mf4_t bindex, size_t vl) {
@@ -4057,7 +4057,7 @@ vfloat64m2_t test_vluxei8_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vluxei8_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vluxei.mask.nxv4f64.nxv4i8.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vluxei.mask.nxv4f64.nxv4i8.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP1]]
 //
 vfloat64m4_t test_vluxei8_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, const double *base, vuint8mf2_t bindex, size_t vl) {
@@ -4068,7 +4068,7 @@ vfloat64m4_t test_vluxei8_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vluxei8_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vluxei.mask.nxv8f64.nxv8i8.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vluxei.mask.nxv8f64.nxv8i8.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP1]]
 //
 vfloat64m8_t test_vluxei8_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff, const double *base, vuint8m1_t bindex, size_t vl) {
@@ -4079,7 +4079,7 @@ vfloat64m8_t test_vluxei8_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff, const
 // CHECK-RV64-LABEL: @test_vluxei16_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vluxei.mask.nxv1f64.nxv1i16.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vluxei.mask.nxv1f64.nxv1i16.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP1]]
 //
 vfloat64m1_t test_vluxei16_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, const double *base, vuint16mf4_t bindex, size_t vl) {
@@ -4090,7 +4090,7 @@ vfloat64m1_t test_vluxei16_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vluxei16_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vluxei.mask.nxv2f64.nxv2i16.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vluxei.mask.nxv2f64.nxv2i16.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 vfloat64m2_t test_vluxei16_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, const double *base, vuint16mf2_t bindex, size_t vl) {
@@ -4101,7 +4101,7 @@ vfloat64m2_t test_vluxei16_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vluxei16_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vluxei.mask.nxv4f64.nxv4i16.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vluxei.mask.nxv4f64.nxv4i16.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP1]]
 //
 vfloat64m4_t test_vluxei16_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, const double *base, vuint16m1_t bindex, size_t vl) {
@@ -4112,7 +4112,7 @@ vfloat64m4_t test_vluxei16_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vluxei16_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vluxei.mask.nxv8f64.nxv8i16.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vluxei.mask.nxv8f64.nxv8i16.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP1]]
 //
 vfloat64m8_t test_vluxei16_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff, const double *base, vuint16m2_t bindex, size_t vl) {
@@ -4123,7 +4123,7 @@ vfloat64m8_t test_vluxei16_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vluxei32_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vluxei.mask.nxv1f64.nxv1i32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vluxei.mask.nxv1f64.nxv1i32.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP1]]
 //
 vfloat64m1_t test_vluxei32_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, const double *base, vuint32mf2_t bindex, size_t vl) {
@@ -4134,7 +4134,7 @@ vfloat64m1_t test_vluxei32_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vluxei32_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vluxei.mask.nxv2f64.nxv2i32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vluxei.mask.nxv2f64.nxv2i32.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 vfloat64m2_t test_vluxei32_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, const double *base, vuint32m1_t bindex, size_t vl) {
@@ -4145,7 +4145,7 @@ vfloat64m2_t test_vluxei32_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vluxei32_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vluxei.mask.nxv4f64.nxv4i32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vluxei.mask.nxv4f64.nxv4i32.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP1]]
 //
 vfloat64m4_t test_vluxei32_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, const double *base, vuint32m2_t bindex, size_t vl) {
@@ -4156,7 +4156,7 @@ vfloat64m4_t test_vluxei32_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vluxei32_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vluxei.mask.nxv8f64.nxv8i32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vluxei.mask.nxv8f64.nxv8i32.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP1]]
 //
 vfloat64m8_t test_vluxei32_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff, const double *base, vuint32m4_t bindex, size_t vl) {
@@ -4167,7 +4167,7 @@ vfloat64m8_t test_vluxei32_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff, cons
 // CHECK-RV64-LABEL: @test_vluxei64_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vluxei.mask.nxv1f64.nxv1i64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x double> @llvm.riscv.vluxei.mask.nxv1f64.nxv1i64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP1]]
 //
 vfloat64m1_t test_vluxei64_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, const double *base, vuint64m1_t bindex, size_t vl) {
@@ -4178,7 +4178,7 @@ vfloat64m1_t test_vluxei64_v_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vluxei64_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vluxei.mask.nxv2f64.nxv2i64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x double> @llvm.riscv.vluxei.mask.nxv2f64.nxv2i64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 vfloat64m2_t test_vluxei64_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, const double *base, vuint64m2_t bindex, size_t vl) {
@@ -4189,7 +4189,7 @@ vfloat64m2_t test_vluxei64_v_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vluxei64_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vluxei.mask.nxv4f64.nxv4i64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x double> @llvm.riscv.vluxei.mask.nxv4f64.nxv4i64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP1]]
 //
 vfloat64m4_t test_vluxei64_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, const double *base, vuint64m4_t bindex, size_t vl) {
@@ -4200,7 +4200,7 @@ vfloat64m4_t test_vluxei64_v_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff, con
 // CHECK-RV64-LABEL: @test_vluxei64_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vluxei.mask.nxv8f64.nxv8i64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vluxei.mask.nxv8f64.nxv8i64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP1]]
 //
 vfloat64m8_t test_vluxei64_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff, const double *base, vuint64m8_t bindex, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmacc.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmacc.c
index 66adf5b1..e902a63 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmacc.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmacc.c
@@ -887,7 +887,7 @@ vuint64m8_t test_vmacc_vx_u64m8(vuint64m8_t acc, uint64_t op1, vuint64m8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmacc.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmacc.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vmacc_vv_i8mf8_m(vbool64_t mask, vint8mf8_t acc, vint8mf8_t op1, vint8mf8_t op2, size_t vl) {
@@ -897,7 +897,7 @@ vint8mf8_t test_vmacc_vv_i8mf8_m(vbool64_t mask, vint8mf8_t acc, vint8mf8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmacc.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmacc.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vmacc_vx_i8mf8_m(vbool64_t mask, vint8mf8_t acc, int8_t op1, vint8mf8_t op2, size_t vl) {
@@ -907,7 +907,7 @@ vint8mf8_t test_vmacc_vx_i8mf8_m(vbool64_t mask, vint8mf8_t acc, int8_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmacc.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmacc.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vmacc_vv_i8mf4_m(vbool32_t mask, vint8mf4_t acc, vint8mf4_t op1, vint8mf4_t op2, size_t vl) {
@@ -917,7 +917,7 @@ vint8mf4_t test_vmacc_vv_i8mf4_m(vbool32_t mask, vint8mf4_t acc, vint8mf4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmacc.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmacc.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vmacc_vx_i8mf4_m(vbool32_t mask, vint8mf4_t acc, int8_t op1, vint8mf4_t op2, size_t vl) {
@@ -927,7 +927,7 @@ vint8mf4_t test_vmacc_vx_i8mf4_m(vbool32_t mask, vint8mf4_t acc, int8_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmacc.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmacc.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vmacc_vv_i8mf2_m(vbool16_t mask, vint8mf2_t acc, vint8mf2_t op1, vint8mf2_t op2, size_t vl) {
@@ -937,7 +937,7 @@ vint8mf2_t test_vmacc_vv_i8mf2_m(vbool16_t mask, vint8mf2_t acc, vint8mf2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmacc.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmacc.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vmacc_vx_i8mf2_m(vbool16_t mask, vint8mf2_t acc, int8_t op1, vint8mf2_t op2, size_t vl) {
@@ -947,7 +947,7 @@ vint8mf2_t test_vmacc_vx_i8mf2_m(vbool16_t mask, vint8mf2_t acc, int8_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmacc.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmacc.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vmacc_vv_i8m1_m(vbool8_t mask, vint8m1_t acc, vint8m1_t op1, vint8m1_t op2, size_t vl) {
@@ -957,7 +957,7 @@ vint8m1_t test_vmacc_vv_i8m1_m(vbool8_t mask, vint8m1_t acc, vint8m1_t op1, vint
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmacc.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmacc.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vmacc_vx_i8m1_m(vbool8_t mask, vint8m1_t acc, int8_t op1, vint8m1_t op2, size_t vl) {
@@ -967,7 +967,7 @@ vint8m1_t test_vmacc_vx_i8m1_m(vbool8_t mask, vint8m1_t acc, int8_t op1, vint8m1
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmacc.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmacc.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vmacc_vv_i8m2_m(vbool4_t mask, vint8m2_t acc, vint8m2_t op1, vint8m2_t op2, size_t vl) {
@@ -977,7 +977,7 @@ vint8m2_t test_vmacc_vv_i8m2_m(vbool4_t mask, vint8m2_t acc, vint8m2_t op1, vint
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmacc.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmacc.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vmacc_vx_i8m2_m(vbool4_t mask, vint8m2_t acc, int8_t op1, vint8m2_t op2, size_t vl) {
@@ -987,7 +987,7 @@ vint8m2_t test_vmacc_vx_i8m2_m(vbool4_t mask, vint8m2_t acc, int8_t op1, vint8m2
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmacc.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmacc.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vmacc_vv_i8m4_m(vbool2_t mask, vint8m4_t acc, vint8m4_t op1, vint8m4_t op2, size_t vl) {
@@ -997,7 +997,7 @@ vint8m4_t test_vmacc_vv_i8m4_m(vbool2_t mask, vint8m4_t acc, vint8m4_t op1, vint
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmacc.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmacc.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vmacc_vx_i8m4_m(vbool2_t mask, vint8m4_t acc, int8_t op1, vint8m4_t op2, size_t vl) {
@@ -1007,7 +1007,7 @@ vint8m4_t test_vmacc_vx_i8m4_m(vbool2_t mask, vint8m4_t acc, int8_t op1, vint8m4
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmacc.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[ACC:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmacc.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[ACC:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vmacc_vv_i8m8_m(vbool1_t mask, vint8m8_t acc, vint8m8_t op1, vint8m8_t op2, size_t vl) {
@@ -1017,7 +1017,7 @@ vint8m8_t test_vmacc_vv_i8m8_m(vbool1_t mask, vint8m8_t acc, vint8m8_t op1, vint
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmacc.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmacc.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vmacc_vx_i8m8_m(vbool1_t mask, vint8m8_t acc, int8_t op1, vint8m8_t op2, size_t vl) {
@@ -1027,7 +1027,7 @@ vint8m8_t test_vmacc_vx_i8m8_m(vbool1_t mask, vint8m8_t acc, int8_t op1, vint8m8
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmacc.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmacc.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vmacc_vv_i16mf4_m(vbool64_t mask, vint16mf4_t acc, vint16mf4_t op1, vint16mf4_t op2, size_t vl) {
@@ -1037,7 +1037,7 @@ vint16mf4_t test_vmacc_vv_i16mf4_m(vbool64_t mask, vint16mf4_t acc, vint16mf4_t
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmacc.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmacc.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vmacc_vx_i16mf4_m(vbool64_t mask, vint16mf4_t acc, int16_t op1, vint16mf4_t op2, size_t vl) {
@@ -1047,7 +1047,7 @@ vint16mf4_t test_vmacc_vx_i16mf4_m(vbool64_t mask, vint16mf4_t acc, int16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmacc.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmacc.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vmacc_vv_i16mf2_m(vbool32_t mask, vint16mf2_t acc, vint16mf2_t op1, vint16mf2_t op2, size_t vl) {
@@ -1057,7 +1057,7 @@ vint16mf2_t test_vmacc_vv_i16mf2_m(vbool32_t mask, vint16mf2_t acc, vint16mf2_t
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmacc.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmacc.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vmacc_vx_i16mf2_m(vbool32_t mask, vint16mf2_t acc, int16_t op1, vint16mf2_t op2, size_t vl) {
@@ -1067,7 +1067,7 @@ vint16mf2_t test_vmacc_vx_i16mf2_m(vbool32_t mask, vint16mf2_t acc, int16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmacc.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmacc.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vmacc_vv_i16m1_m(vbool16_t mask, vint16m1_t acc, vint16m1_t op1, vint16m1_t op2, size_t vl) {
@@ -1077,7 +1077,7 @@ vint16m1_t test_vmacc_vv_i16m1_m(vbool16_t mask, vint16m1_t acc, vint16m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmacc.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmacc.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vmacc_vx_i16m1_m(vbool16_t mask, vint16m1_t acc, int16_t op1, vint16m1_t op2, size_t vl) {
@@ -1087,7 +1087,7 @@ vint16m1_t test_vmacc_vx_i16m1_m(vbool16_t mask, vint16m1_t acc, int16_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmacc.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmacc.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vmacc_vv_i16m2_m(vbool8_t mask, vint16m2_t acc, vint16m2_t op1, vint16m2_t op2, size_t vl) {
@@ -1097,7 +1097,7 @@ vint16m2_t test_vmacc_vv_i16m2_m(vbool8_t mask, vint16m2_t acc, vint16m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmacc.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmacc.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vmacc_vx_i16m2_m(vbool8_t mask, vint16m2_t acc, int16_t op1, vint16m2_t op2, size_t vl) {
@@ -1107,7 +1107,7 @@ vint16m2_t test_vmacc_vx_i16m2_m(vbool8_t mask, vint16m2_t acc, int16_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmacc.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmacc.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vmacc_vv_i16m4_m(vbool4_t mask, vint16m4_t acc, vint16m4_t op1, vint16m4_t op2, size_t vl) {
@@ -1117,7 +1117,7 @@ vint16m4_t test_vmacc_vv_i16m4_m(vbool4_t mask, vint16m4_t acc, vint16m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmacc.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmacc.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vmacc_vx_i16m4_m(vbool4_t mask, vint16m4_t acc, int16_t op1, vint16m4_t op2, size_t vl) {
@@ -1127,7 +1127,7 @@ vint16m4_t test_vmacc_vx_i16m4_m(vbool4_t mask, vint16m4_t acc, int16_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmacc.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmacc.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vmacc_vv_i16m8_m(vbool2_t mask, vint16m8_t acc, vint16m8_t op1, vint16m8_t op2, size_t vl) {
@@ -1137,7 +1137,7 @@ vint16m8_t test_vmacc_vv_i16m8_m(vbool2_t mask, vint16m8_t acc, vint16m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmacc.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmacc.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vmacc_vx_i16m8_m(vbool2_t mask, vint16m8_t acc, int16_t op1, vint16m8_t op2, size_t vl) {
@@ -1147,7 +1147,7 @@ vint16m8_t test_vmacc_vx_i16m8_m(vbool2_t mask, vint16m8_t acc, int16_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmacc.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmacc.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vmacc_vv_i32mf2_m(vbool64_t mask, vint32mf2_t acc, vint32mf2_t op1, vint32mf2_t op2, size_t vl) {
@@ -1157,7 +1157,7 @@ vint32mf2_t test_vmacc_vv_i32mf2_m(vbool64_t mask, vint32mf2_t acc, vint32mf2_t
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmacc.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmacc.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vmacc_vx_i32mf2_m(vbool64_t mask, vint32mf2_t acc, int32_t op1, vint32mf2_t op2, size_t vl) {
@@ -1167,7 +1167,7 @@ vint32mf2_t test_vmacc_vx_i32mf2_m(vbool64_t mask, vint32mf2_t acc, int32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmacc.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmacc.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vmacc_vv_i32m1_m(vbool32_t mask, vint32m1_t acc, vint32m1_t op1, vint32m1_t op2, size_t vl) {
@@ -1177,7 +1177,7 @@ vint32m1_t test_vmacc_vv_i32m1_m(vbool32_t mask, vint32m1_t acc, vint32m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmacc.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmacc.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vmacc_vx_i32m1_m(vbool32_t mask, vint32m1_t acc, int32_t op1, vint32m1_t op2, size_t vl) {
@@ -1187,7 +1187,7 @@ vint32m1_t test_vmacc_vx_i32m1_m(vbool32_t mask, vint32m1_t acc, int32_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmacc.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmacc.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vmacc_vv_i32m2_m(vbool16_t mask, vint32m2_t acc, vint32m2_t op1, vint32m2_t op2, size_t vl) {
@@ -1197,7 +1197,7 @@ vint32m2_t test_vmacc_vv_i32m2_m(vbool16_t mask, vint32m2_t acc, vint32m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmacc.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmacc.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vmacc_vx_i32m2_m(vbool16_t mask, vint32m2_t acc, int32_t op1, vint32m2_t op2, size_t vl) {
@@ -1207,7 +1207,7 @@ vint32m2_t test_vmacc_vx_i32m2_m(vbool16_t mask, vint32m2_t acc, int32_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmacc.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmacc.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vmacc_vv_i32m4_m(vbool8_t mask, vint32m4_t acc, vint32m4_t op1, vint32m4_t op2, size_t vl) {
@@ -1217,7 +1217,7 @@ vint32m4_t test_vmacc_vv_i32m4_m(vbool8_t mask, vint32m4_t acc, vint32m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmacc.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmacc.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vmacc_vx_i32m4_m(vbool8_t mask, vint32m4_t acc, int32_t op1, vint32m4_t op2, size_t vl) {
@@ -1227,7 +1227,7 @@ vint32m4_t test_vmacc_vx_i32m4_m(vbool8_t mask, vint32m4_t acc, int32_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmacc.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmacc.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vmacc_vv_i32m8_m(vbool4_t mask, vint32m8_t acc, vint32m8_t op1, vint32m8_t op2, size_t vl) {
@@ -1237,7 +1237,7 @@ vint32m8_t test_vmacc_vv_i32m8_m(vbool4_t mask, vint32m8_t acc, vint32m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmacc.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmacc.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vmacc_vx_i32m8_m(vbool4_t mask, vint32m8_t acc, int32_t op1, vint32m8_t op2, size_t vl) {
@@ -1247,7 +1247,7 @@ vint32m8_t test_vmacc_vx_i32m8_m(vbool4_t mask, vint32m8_t acc, int32_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmacc.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmacc.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vmacc_vv_i64m1_m(vbool64_t mask, vint64m1_t acc, vint64m1_t op1, vint64m1_t op2, size_t vl) {
@@ -1257,7 +1257,7 @@ vint64m1_t test_vmacc_vv_i64m1_m(vbool64_t mask, vint64m1_t acc, vint64m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmacc.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmacc.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vmacc_vx_i64m1_m(vbool64_t mask, vint64m1_t acc, int64_t op1, vint64m1_t op2, size_t vl) {
@@ -1267,7 +1267,7 @@ vint64m1_t test_vmacc_vx_i64m1_m(vbool64_t mask, vint64m1_t acc, int64_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmacc.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmacc.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vmacc_vv_i64m2_m(vbool32_t mask, vint64m2_t acc, vint64m2_t op1, vint64m2_t op2, size_t vl) {
@@ -1277,7 +1277,7 @@ vint64m2_t test_vmacc_vv_i64m2_m(vbool32_t mask, vint64m2_t acc, vint64m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmacc.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmacc.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vmacc_vx_i64m2_m(vbool32_t mask, vint64m2_t acc, int64_t op1, vint64m2_t op2, size_t vl) {
@@ -1287,7 +1287,7 @@ vint64m2_t test_vmacc_vx_i64m2_m(vbool32_t mask, vint64m2_t acc, int64_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmacc.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmacc.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vmacc_vv_i64m4_m(vbool16_t mask, vint64m4_t acc, vint64m4_t op1, vint64m4_t op2, size_t vl) {
@@ -1297,7 +1297,7 @@ vint64m4_t test_vmacc_vv_i64m4_m(vbool16_t mask, vint64m4_t acc, vint64m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmacc.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmacc.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vmacc_vx_i64m4_m(vbool16_t mask, vint64m4_t acc, int64_t op1, vint64m4_t op2, size_t vl) {
@@ -1307,7 +1307,7 @@ vint64m4_t test_vmacc_vx_i64m4_m(vbool16_t mask, vint64m4_t acc, int64_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmacc.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmacc.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vmacc_vv_i64m8_m(vbool8_t mask, vint64m8_t acc, vint64m8_t op1, vint64m8_t op2, size_t vl) {
@@ -1317,7 +1317,7 @@ vint64m8_t test_vmacc_vv_i64m8_m(vbool8_t mask, vint64m8_t acc, vint64m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmacc.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmacc.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vmacc_vx_i64m8_m(vbool8_t mask, vint64m8_t acc, int64_t op1, vint64m8_t op2, size_t vl) {
@@ -1327,7 +1327,7 @@ vint64m8_t test_vmacc_vx_i64m8_m(vbool8_t mask, vint64m8_t acc, int64_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmacc.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmacc.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vmacc_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t acc, vuint8mf8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -1337,7 +1337,7 @@ vuint8mf8_t test_vmacc_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t acc, vuint8mf8_t o
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmacc.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmacc.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vmacc_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t acc, uint8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -1347,7 +1347,7 @@ vuint8mf8_t test_vmacc_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t acc, uint8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmacc.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmacc.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vmacc_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t acc, vuint8mf4_t op1, vuint8mf4_t op2, size_t vl) {
@@ -1357,7 +1357,7 @@ vuint8mf4_t test_vmacc_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t acc, vuint8mf4_t o
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmacc.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmacc.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vmacc_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t acc, uint8_t op1, vuint8mf4_t op2, size_t vl) {
@@ -1367,7 +1367,7 @@ vuint8mf4_t test_vmacc_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t acc, uint8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmacc.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmacc.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vmacc_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t acc, vuint8mf2_t op1, vuint8mf2_t op2, size_t vl) {
@@ -1377,7 +1377,7 @@ vuint8mf2_t test_vmacc_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t acc, vuint8mf2_t o
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmacc.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmacc.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vmacc_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t acc, uint8_t op1, vuint8mf2_t op2, size_t vl) {
@@ -1387,7 +1387,7 @@ vuint8mf2_t test_vmacc_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t acc, uint8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmacc.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmacc.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vmacc_vv_u8m1_m(vbool8_t mask, vuint8m1_t acc, vuint8m1_t op1, vuint8m1_t op2, size_t vl) {
@@ -1397,7 +1397,7 @@ vuint8m1_t test_vmacc_vv_u8m1_m(vbool8_t mask, vuint8m1_t acc, vuint8m1_t op1, v
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmacc.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmacc.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vmacc_vx_u8m1_m(vbool8_t mask, vuint8m1_t acc, uint8_t op1, vuint8m1_t op2, size_t vl) {
@@ -1407,7 +1407,7 @@ vuint8m1_t test_vmacc_vx_u8m1_m(vbool8_t mask, vuint8m1_t acc, uint8_t op1, vuin
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmacc.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmacc.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vmacc_vv_u8m2_m(vbool4_t mask, vuint8m2_t acc, vuint8m2_t op1, vuint8m2_t op2, size_t vl) {
@@ -1417,7 +1417,7 @@ vuint8m2_t test_vmacc_vv_u8m2_m(vbool4_t mask, vuint8m2_t acc, vuint8m2_t op1, v
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmacc.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmacc.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vmacc_vx_u8m2_m(vbool4_t mask, vuint8m2_t acc, uint8_t op1, vuint8m2_t op2, size_t vl) {
@@ -1427,7 +1427,7 @@ vuint8m2_t test_vmacc_vx_u8m2_m(vbool4_t mask, vuint8m2_t acc, uint8_t op1, vuin
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmacc.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmacc.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vmacc_vv_u8m4_m(vbool2_t mask, vuint8m4_t acc, vuint8m4_t op1, vuint8m4_t op2, size_t vl) {
@@ -1437,7 +1437,7 @@ vuint8m4_t test_vmacc_vv_u8m4_m(vbool2_t mask, vuint8m4_t acc, vuint8m4_t op1, v
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmacc.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmacc.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vmacc_vx_u8m4_m(vbool2_t mask, vuint8m4_t acc, uint8_t op1, vuint8m4_t op2, size_t vl) {
@@ -1447,7 +1447,7 @@ vuint8m4_t test_vmacc_vx_u8m4_m(vbool2_t mask, vuint8m4_t acc, uint8_t op1, vuin
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmacc.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[ACC:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmacc.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[ACC:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vmacc_vv_u8m8_m(vbool1_t mask, vuint8m8_t acc, vuint8m8_t op1, vuint8m8_t op2, size_t vl) {
@@ -1457,7 +1457,7 @@ vuint8m8_t test_vmacc_vv_u8m8_m(vbool1_t mask, vuint8m8_t acc, vuint8m8_t op1, v
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmacc.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmacc.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vmacc_vx_u8m8_m(vbool1_t mask, vuint8m8_t acc, uint8_t op1, vuint8m8_t op2, size_t vl) {
@@ -1467,7 +1467,7 @@ vuint8m8_t test_vmacc_vx_u8m8_m(vbool1_t mask, vuint8m8_t acc, uint8_t op1, vuin
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmacc.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmacc.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vmacc_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t acc, vuint16mf4_t op1, vuint16mf4_t op2, size_t vl) {
@@ -1477,7 +1477,7 @@ vuint16mf4_t test_vmacc_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t acc, vuint16mf4
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmacc.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmacc.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vmacc_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t acc, uint16_t op1, vuint16mf4_t op2, size_t vl) {
@@ -1487,7 +1487,7 @@ vuint16mf4_t test_vmacc_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t acc, uint16_t o
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmacc.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmacc.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vmacc_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t acc, vuint16mf2_t op1, vuint16mf2_t op2, size_t vl) {
@@ -1497,7 +1497,7 @@ vuint16mf2_t test_vmacc_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t acc, vuint16mf2
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmacc.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmacc.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vmacc_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t acc, uint16_t op1, vuint16mf2_t op2, size_t vl) {
@@ -1507,7 +1507,7 @@ vuint16mf2_t test_vmacc_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t acc, uint16_t o
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmacc.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmacc.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vmacc_vv_u16m1_m(vbool16_t mask, vuint16m1_t acc, vuint16m1_t op1, vuint16m1_t op2, size_t vl) {
@@ -1517,7 +1517,7 @@ vuint16m1_t test_vmacc_vv_u16m1_m(vbool16_t mask, vuint16m1_t acc, vuint16m1_t o
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmacc.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmacc.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vmacc_vx_u16m1_m(vbool16_t mask, vuint16m1_t acc, uint16_t op1, vuint16m1_t op2, size_t vl) {
@@ -1527,7 +1527,7 @@ vuint16m1_t test_vmacc_vx_u16m1_m(vbool16_t mask, vuint16m1_t acc, uint16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmacc.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmacc.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vmacc_vv_u16m2_m(vbool8_t mask, vuint16m2_t acc, vuint16m2_t op1, vuint16m2_t op2, size_t vl) {
@@ -1537,7 +1537,7 @@ vuint16m2_t test_vmacc_vv_u16m2_m(vbool8_t mask, vuint16m2_t acc, vuint16m2_t op
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmacc.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmacc.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vmacc_vx_u16m2_m(vbool8_t mask, vuint16m2_t acc, uint16_t op1, vuint16m2_t op2, size_t vl) {
@@ -1547,7 +1547,7 @@ vuint16m2_t test_vmacc_vx_u16m2_m(vbool8_t mask, vuint16m2_t acc, uint16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmacc.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmacc.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vmacc_vv_u16m4_m(vbool4_t mask, vuint16m4_t acc, vuint16m4_t op1, vuint16m4_t op2, size_t vl) {
@@ -1557,7 +1557,7 @@ vuint16m4_t test_vmacc_vv_u16m4_m(vbool4_t mask, vuint16m4_t acc, vuint16m4_t op
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmacc.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmacc.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vmacc_vx_u16m4_m(vbool4_t mask, vuint16m4_t acc, uint16_t op1, vuint16m4_t op2, size_t vl) {
@@ -1567,7 +1567,7 @@ vuint16m4_t test_vmacc_vx_u16m4_m(vbool4_t mask, vuint16m4_t acc, uint16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmacc.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmacc.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vmacc_vv_u16m8_m(vbool2_t mask, vuint16m8_t acc, vuint16m8_t op1, vuint16m8_t op2, size_t vl) {
@@ -1577,7 +1577,7 @@ vuint16m8_t test_vmacc_vv_u16m8_m(vbool2_t mask, vuint16m8_t acc, vuint16m8_t op
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmacc.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmacc.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vmacc_vx_u16m8_m(vbool2_t mask, vuint16m8_t acc, uint16_t op1, vuint16m8_t op2, size_t vl) {
@@ -1587,7 +1587,7 @@ vuint16m8_t test_vmacc_vx_u16m8_m(vbool2_t mask, vuint16m8_t acc, uint16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmacc.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmacc.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vmacc_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t acc, vuint32mf2_t op1, vuint32mf2_t op2, size_t vl) {
@@ -1597,7 +1597,7 @@ vuint32mf2_t test_vmacc_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t acc, vuint32mf2
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmacc.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmacc.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vmacc_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t acc, uint32_t op1, vuint32mf2_t op2, size_t vl) {
@@ -1607,7 +1607,7 @@ vuint32mf2_t test_vmacc_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t acc, uint32_t o
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmacc.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmacc.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vmacc_vv_u32m1_m(vbool32_t mask, vuint32m1_t acc, vuint32m1_t op1, vuint32m1_t op2, size_t vl) {
@@ -1617,7 +1617,7 @@ vuint32m1_t test_vmacc_vv_u32m1_m(vbool32_t mask, vuint32m1_t acc, vuint32m1_t o
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmacc.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmacc.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vmacc_vx_u32m1_m(vbool32_t mask, vuint32m1_t acc, uint32_t op1, vuint32m1_t op2, size_t vl) {
@@ -1627,7 +1627,7 @@ vuint32m1_t test_vmacc_vx_u32m1_m(vbool32_t mask, vuint32m1_t acc, uint32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmacc.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmacc.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vmacc_vv_u32m2_m(vbool16_t mask, vuint32m2_t acc, vuint32m2_t op1, vuint32m2_t op2, size_t vl) {
@@ -1637,7 +1637,7 @@ vuint32m2_t test_vmacc_vv_u32m2_m(vbool16_t mask, vuint32m2_t acc, vuint32m2_t o
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmacc.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmacc.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vmacc_vx_u32m2_m(vbool16_t mask, vuint32m2_t acc, uint32_t op1, vuint32m2_t op2, size_t vl) {
@@ -1647,7 +1647,7 @@ vuint32m2_t test_vmacc_vx_u32m2_m(vbool16_t mask, vuint32m2_t acc, uint32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmacc.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmacc.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vmacc_vv_u32m4_m(vbool8_t mask, vuint32m4_t acc, vuint32m4_t op1, vuint32m4_t op2, size_t vl) {
@@ -1657,7 +1657,7 @@ vuint32m4_t test_vmacc_vv_u32m4_m(vbool8_t mask, vuint32m4_t acc, vuint32m4_t op
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmacc.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmacc.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vmacc_vx_u32m4_m(vbool8_t mask, vuint32m4_t acc, uint32_t op1, vuint32m4_t op2, size_t vl) {
@@ -1667,7 +1667,7 @@ vuint32m4_t test_vmacc_vx_u32m4_m(vbool8_t mask, vuint32m4_t acc, uint32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmacc.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmacc.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vmacc_vv_u32m8_m(vbool4_t mask, vuint32m8_t acc, vuint32m8_t op1, vuint32m8_t op2, size_t vl) {
@@ -1677,7 +1677,7 @@ vuint32m8_t test_vmacc_vv_u32m8_m(vbool4_t mask, vuint32m8_t acc, vuint32m8_t op
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmacc.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmacc.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vmacc_vx_u32m8_m(vbool4_t mask, vuint32m8_t acc, uint32_t op1, vuint32m8_t op2, size_t vl) {
@@ -1687,7 +1687,7 @@ vuint32m8_t test_vmacc_vx_u32m8_m(vbool4_t mask, vuint32m8_t acc, uint32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmacc.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmacc.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vmacc_vv_u64m1_m(vbool64_t mask, vuint64m1_t acc, vuint64m1_t op1, vuint64m1_t op2, size_t vl) {
@@ -1697,7 +1697,7 @@ vuint64m1_t test_vmacc_vv_u64m1_m(vbool64_t mask, vuint64m1_t acc, vuint64m1_t o
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmacc.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmacc.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vmacc_vx_u64m1_m(vbool64_t mask, vuint64m1_t acc, uint64_t op1, vuint64m1_t op2, size_t vl) {
@@ -1707,7 +1707,7 @@ vuint64m1_t test_vmacc_vx_u64m1_m(vbool64_t mask, vuint64m1_t acc, uint64_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmacc.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmacc.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vmacc_vv_u64m2_m(vbool32_t mask, vuint64m2_t acc, vuint64m2_t op1, vuint64m2_t op2, size_t vl) {
@@ -1717,7 +1717,7 @@ vuint64m2_t test_vmacc_vv_u64m2_m(vbool32_t mask, vuint64m2_t acc, vuint64m2_t o
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmacc.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmacc.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vmacc_vx_u64m2_m(vbool32_t mask, vuint64m2_t acc, uint64_t op1, vuint64m2_t op2, size_t vl) {
@@ -1727,7 +1727,7 @@ vuint64m2_t test_vmacc_vx_u64m2_m(vbool32_t mask, vuint64m2_t acc, uint64_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmacc.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmacc.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vmacc_vv_u64m4_m(vbool16_t mask, vuint64m4_t acc, vuint64m4_t op1, vuint64m4_t op2, size_t vl) {
@@ -1737,7 +1737,7 @@ vuint64m4_t test_vmacc_vv_u64m4_m(vbool16_t mask, vuint64m4_t acc, vuint64m4_t o
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmacc.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmacc.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vmacc_vx_u64m4_m(vbool16_t mask, vuint64m4_t acc, uint64_t op1, vuint64m4_t op2, size_t vl) {
@@ -1747,7 +1747,7 @@ vuint64m4_t test_vmacc_vx_u64m4_m(vbool16_t mask, vuint64m4_t acc, uint64_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmacc_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmacc.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmacc.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vmacc_vv_u64m8_m(vbool8_t mask, vuint64m8_t acc, vuint64m8_t op1, vuint64m8_t op2, size_t vl) {
@@ -1757,7 +1757,7 @@ vuint64m8_t test_vmacc_vv_u64m8_m(vbool8_t mask, vuint64m8_t acc, vuint64m8_t op
 //
 // CHECK-RV64-LABEL: @test_vmacc_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmacc.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmacc.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vmacc_vx_u64m8_m(vbool8_t mask, vuint64m8_t acc, uint64_t op1, vuint64m8_t op2, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmadd.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmadd.c
index e1f4812..4bd51db 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmadd.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmadd.c
@@ -887,7 +887,7 @@ vuint64m8_t test_vmadd_vx_u64m8(vuint64m8_t acc, uint64_t op1, vuint64m8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmadd.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmadd.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vmadd_vv_i8mf8_m(vbool64_t mask, vint8mf8_t acc, vint8mf8_t op1, vint8mf8_t op2, size_t vl) {
@@ -897,7 +897,7 @@ vint8mf8_t test_vmadd_vv_i8mf8_m(vbool64_t mask, vint8mf8_t acc, vint8mf8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmadd.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmadd.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vmadd_vx_i8mf8_m(vbool64_t mask, vint8mf8_t acc, int8_t op1, vint8mf8_t op2, size_t vl) {
@@ -907,7 +907,7 @@ vint8mf8_t test_vmadd_vx_i8mf8_m(vbool64_t mask, vint8mf8_t acc, int8_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmadd.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmadd.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vmadd_vv_i8mf4_m(vbool32_t mask, vint8mf4_t acc, vint8mf4_t op1, vint8mf4_t op2, size_t vl) {
@@ -917,7 +917,7 @@ vint8mf4_t test_vmadd_vv_i8mf4_m(vbool32_t mask, vint8mf4_t acc, vint8mf4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmadd.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmadd.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vmadd_vx_i8mf4_m(vbool32_t mask, vint8mf4_t acc, int8_t op1, vint8mf4_t op2, size_t vl) {
@@ -927,7 +927,7 @@ vint8mf4_t test_vmadd_vx_i8mf4_m(vbool32_t mask, vint8mf4_t acc, int8_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmadd.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmadd.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vmadd_vv_i8mf2_m(vbool16_t mask, vint8mf2_t acc, vint8mf2_t op1, vint8mf2_t op2, size_t vl) {
@@ -937,7 +937,7 @@ vint8mf2_t test_vmadd_vv_i8mf2_m(vbool16_t mask, vint8mf2_t acc, vint8mf2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmadd.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmadd.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vmadd_vx_i8mf2_m(vbool16_t mask, vint8mf2_t acc, int8_t op1, vint8mf2_t op2, size_t vl) {
@@ -947,7 +947,7 @@ vint8mf2_t test_vmadd_vx_i8mf2_m(vbool16_t mask, vint8mf2_t acc, int8_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmadd.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmadd.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vmadd_vv_i8m1_m(vbool8_t mask, vint8m1_t acc, vint8m1_t op1, vint8m1_t op2, size_t vl) {
@@ -957,7 +957,7 @@ vint8m1_t test_vmadd_vv_i8m1_m(vbool8_t mask, vint8m1_t acc, vint8m1_t op1, vint
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmadd.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmadd.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vmadd_vx_i8m1_m(vbool8_t mask, vint8m1_t acc, int8_t op1, vint8m1_t op2, size_t vl) {
@@ -967,7 +967,7 @@ vint8m1_t test_vmadd_vx_i8m1_m(vbool8_t mask, vint8m1_t acc, int8_t op1, vint8m1
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmadd.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmadd.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vmadd_vv_i8m2_m(vbool4_t mask, vint8m2_t acc, vint8m2_t op1, vint8m2_t op2, size_t vl) {
@@ -977,7 +977,7 @@ vint8m2_t test_vmadd_vv_i8m2_m(vbool4_t mask, vint8m2_t acc, vint8m2_t op1, vint
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmadd.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmadd.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vmadd_vx_i8m2_m(vbool4_t mask, vint8m2_t acc, int8_t op1, vint8m2_t op2, size_t vl) {
@@ -987,7 +987,7 @@ vint8m2_t test_vmadd_vx_i8m2_m(vbool4_t mask, vint8m2_t acc, int8_t op1, vint8m2
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmadd.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmadd.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vmadd_vv_i8m4_m(vbool2_t mask, vint8m4_t acc, vint8m4_t op1, vint8m4_t op2, size_t vl) {
@@ -997,7 +997,7 @@ vint8m4_t test_vmadd_vv_i8m4_m(vbool2_t mask, vint8m4_t acc, vint8m4_t op1, vint
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmadd.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmadd.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vmadd_vx_i8m4_m(vbool2_t mask, vint8m4_t acc, int8_t op1, vint8m4_t op2, size_t vl) {
@@ -1007,7 +1007,7 @@ vint8m4_t test_vmadd_vx_i8m4_m(vbool2_t mask, vint8m4_t acc, int8_t op1, vint8m4
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmadd.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[ACC:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmadd.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[ACC:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vmadd_vv_i8m8_m(vbool1_t mask, vint8m8_t acc, vint8m8_t op1, vint8m8_t op2, size_t vl) {
@@ -1017,7 +1017,7 @@ vint8m8_t test_vmadd_vv_i8m8_m(vbool1_t mask, vint8m8_t acc, vint8m8_t op1, vint
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmadd.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmadd.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vmadd_vx_i8m8_m(vbool1_t mask, vint8m8_t acc, int8_t op1, vint8m8_t op2, size_t vl) {
@@ -1027,7 +1027,7 @@ vint8m8_t test_vmadd_vx_i8m8_m(vbool1_t mask, vint8m8_t acc, int8_t op1, vint8m8
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmadd.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmadd.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vmadd_vv_i16mf4_m(vbool64_t mask, vint16mf4_t acc, vint16mf4_t op1, vint16mf4_t op2, size_t vl) {
@@ -1037,7 +1037,7 @@ vint16mf4_t test_vmadd_vv_i16mf4_m(vbool64_t mask, vint16mf4_t acc, vint16mf4_t
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmadd.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmadd.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vmadd_vx_i16mf4_m(vbool64_t mask, vint16mf4_t acc, int16_t op1, vint16mf4_t op2, size_t vl) {
@@ -1047,7 +1047,7 @@ vint16mf4_t test_vmadd_vx_i16mf4_m(vbool64_t mask, vint16mf4_t acc, int16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmadd.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmadd.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vmadd_vv_i16mf2_m(vbool32_t mask, vint16mf2_t acc, vint16mf2_t op1, vint16mf2_t op2, size_t vl) {
@@ -1057,7 +1057,7 @@ vint16mf2_t test_vmadd_vv_i16mf2_m(vbool32_t mask, vint16mf2_t acc, vint16mf2_t
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmadd.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmadd.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vmadd_vx_i16mf2_m(vbool32_t mask, vint16mf2_t acc, int16_t op1, vint16mf2_t op2, size_t vl) {
@@ -1067,7 +1067,7 @@ vint16mf2_t test_vmadd_vx_i16mf2_m(vbool32_t mask, vint16mf2_t acc, int16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmadd.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmadd.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vmadd_vv_i16m1_m(vbool16_t mask, vint16m1_t acc, vint16m1_t op1, vint16m1_t op2, size_t vl) {
@@ -1077,7 +1077,7 @@ vint16m1_t test_vmadd_vv_i16m1_m(vbool16_t mask, vint16m1_t acc, vint16m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmadd.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmadd.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vmadd_vx_i16m1_m(vbool16_t mask, vint16m1_t acc, int16_t op1, vint16m1_t op2, size_t vl) {
@@ -1087,7 +1087,7 @@ vint16m1_t test_vmadd_vx_i16m1_m(vbool16_t mask, vint16m1_t acc, int16_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmadd.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmadd.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vmadd_vv_i16m2_m(vbool8_t mask, vint16m2_t acc, vint16m2_t op1, vint16m2_t op2, size_t vl) {
@@ -1097,7 +1097,7 @@ vint16m2_t test_vmadd_vv_i16m2_m(vbool8_t mask, vint16m2_t acc, vint16m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmadd.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmadd.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vmadd_vx_i16m2_m(vbool8_t mask, vint16m2_t acc, int16_t op1, vint16m2_t op2, size_t vl) {
@@ -1107,7 +1107,7 @@ vint16m2_t test_vmadd_vx_i16m2_m(vbool8_t mask, vint16m2_t acc, int16_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmadd.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmadd.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vmadd_vv_i16m4_m(vbool4_t mask, vint16m4_t acc, vint16m4_t op1, vint16m4_t op2, size_t vl) {
@@ -1117,7 +1117,7 @@ vint16m4_t test_vmadd_vv_i16m4_m(vbool4_t mask, vint16m4_t acc, vint16m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmadd.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmadd.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vmadd_vx_i16m4_m(vbool4_t mask, vint16m4_t acc, int16_t op1, vint16m4_t op2, size_t vl) {
@@ -1127,7 +1127,7 @@ vint16m4_t test_vmadd_vx_i16m4_m(vbool4_t mask, vint16m4_t acc, int16_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmadd.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmadd.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vmadd_vv_i16m8_m(vbool2_t mask, vint16m8_t acc, vint16m8_t op1, vint16m8_t op2, size_t vl) {
@@ -1137,7 +1137,7 @@ vint16m8_t test_vmadd_vv_i16m8_m(vbool2_t mask, vint16m8_t acc, vint16m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmadd.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmadd.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vmadd_vx_i16m8_m(vbool2_t mask, vint16m8_t acc, int16_t op1, vint16m8_t op2, size_t vl) {
@@ -1147,7 +1147,7 @@ vint16m8_t test_vmadd_vx_i16m8_m(vbool2_t mask, vint16m8_t acc, int16_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmadd.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmadd.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vmadd_vv_i32mf2_m(vbool64_t mask, vint32mf2_t acc, vint32mf2_t op1, vint32mf2_t op2, size_t vl) {
@@ -1157,7 +1157,7 @@ vint32mf2_t test_vmadd_vv_i32mf2_m(vbool64_t mask, vint32mf2_t acc, vint32mf2_t
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmadd.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmadd.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vmadd_vx_i32mf2_m(vbool64_t mask, vint32mf2_t acc, int32_t op1, vint32mf2_t op2, size_t vl) {
@@ -1167,7 +1167,7 @@ vint32mf2_t test_vmadd_vx_i32mf2_m(vbool64_t mask, vint32mf2_t acc, int32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmadd.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmadd.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vmadd_vv_i32m1_m(vbool32_t mask, vint32m1_t acc, vint32m1_t op1, vint32m1_t op2, size_t vl) {
@@ -1177,7 +1177,7 @@ vint32m1_t test_vmadd_vv_i32m1_m(vbool32_t mask, vint32m1_t acc, vint32m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmadd.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmadd.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vmadd_vx_i32m1_m(vbool32_t mask, vint32m1_t acc, int32_t op1, vint32m1_t op2, size_t vl) {
@@ -1187,7 +1187,7 @@ vint32m1_t test_vmadd_vx_i32m1_m(vbool32_t mask, vint32m1_t acc, int32_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmadd.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmadd.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vmadd_vv_i32m2_m(vbool16_t mask, vint32m2_t acc, vint32m2_t op1, vint32m2_t op2, size_t vl) {
@@ -1197,7 +1197,7 @@ vint32m2_t test_vmadd_vv_i32m2_m(vbool16_t mask, vint32m2_t acc, vint32m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmadd.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmadd.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vmadd_vx_i32m2_m(vbool16_t mask, vint32m2_t acc, int32_t op1, vint32m2_t op2, size_t vl) {
@@ -1207,7 +1207,7 @@ vint32m2_t test_vmadd_vx_i32m2_m(vbool16_t mask, vint32m2_t acc, int32_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmadd.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmadd.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vmadd_vv_i32m4_m(vbool8_t mask, vint32m4_t acc, vint32m4_t op1, vint32m4_t op2, size_t vl) {
@@ -1217,7 +1217,7 @@ vint32m4_t test_vmadd_vv_i32m4_m(vbool8_t mask, vint32m4_t acc, vint32m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmadd.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmadd.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vmadd_vx_i32m4_m(vbool8_t mask, vint32m4_t acc, int32_t op1, vint32m4_t op2, size_t vl) {
@@ -1227,7 +1227,7 @@ vint32m4_t test_vmadd_vx_i32m4_m(vbool8_t mask, vint32m4_t acc, int32_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmadd.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmadd.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vmadd_vv_i32m8_m(vbool4_t mask, vint32m8_t acc, vint32m8_t op1, vint32m8_t op2, size_t vl) {
@@ -1237,7 +1237,7 @@ vint32m8_t test_vmadd_vv_i32m8_m(vbool4_t mask, vint32m8_t acc, vint32m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmadd.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmadd.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vmadd_vx_i32m8_m(vbool4_t mask, vint32m8_t acc, int32_t op1, vint32m8_t op2, size_t vl) {
@@ -1247,7 +1247,7 @@ vint32m8_t test_vmadd_vx_i32m8_m(vbool4_t mask, vint32m8_t acc, int32_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmadd.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmadd.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vmadd_vv_i64m1_m(vbool64_t mask, vint64m1_t acc, vint64m1_t op1, vint64m1_t op2, size_t vl) {
@@ -1257,7 +1257,7 @@ vint64m1_t test_vmadd_vv_i64m1_m(vbool64_t mask, vint64m1_t acc, vint64m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmadd.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmadd.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vmadd_vx_i64m1_m(vbool64_t mask, vint64m1_t acc, int64_t op1, vint64m1_t op2, size_t vl) {
@@ -1267,7 +1267,7 @@ vint64m1_t test_vmadd_vx_i64m1_m(vbool64_t mask, vint64m1_t acc, int64_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmadd.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmadd.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vmadd_vv_i64m2_m(vbool32_t mask, vint64m2_t acc, vint64m2_t op1, vint64m2_t op2, size_t vl) {
@@ -1277,7 +1277,7 @@ vint64m2_t test_vmadd_vv_i64m2_m(vbool32_t mask, vint64m2_t acc, vint64m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmadd.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmadd.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vmadd_vx_i64m2_m(vbool32_t mask, vint64m2_t acc, int64_t op1, vint64m2_t op2, size_t vl) {
@@ -1287,7 +1287,7 @@ vint64m2_t test_vmadd_vx_i64m2_m(vbool32_t mask, vint64m2_t acc, int64_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmadd.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmadd.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vmadd_vv_i64m4_m(vbool16_t mask, vint64m4_t acc, vint64m4_t op1, vint64m4_t op2, size_t vl) {
@@ -1297,7 +1297,7 @@ vint64m4_t test_vmadd_vv_i64m4_m(vbool16_t mask, vint64m4_t acc, vint64m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmadd.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmadd.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vmadd_vx_i64m4_m(vbool16_t mask, vint64m4_t acc, int64_t op1, vint64m4_t op2, size_t vl) {
@@ -1307,7 +1307,7 @@ vint64m4_t test_vmadd_vx_i64m4_m(vbool16_t mask, vint64m4_t acc, int64_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmadd.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmadd.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vmadd_vv_i64m8_m(vbool8_t mask, vint64m8_t acc, vint64m8_t op1, vint64m8_t op2, size_t vl) {
@@ -1317,7 +1317,7 @@ vint64m8_t test_vmadd_vv_i64m8_m(vbool8_t mask, vint64m8_t acc, vint64m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmadd.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmadd.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vmadd_vx_i64m8_m(vbool8_t mask, vint64m8_t acc, int64_t op1, vint64m8_t op2, size_t vl) {
@@ -1327,7 +1327,7 @@ vint64m8_t test_vmadd_vx_i64m8_m(vbool8_t mask, vint64m8_t acc, int64_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmadd.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmadd.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vmadd_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t acc, vuint8mf8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -1337,7 +1337,7 @@ vuint8mf8_t test_vmadd_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t acc, vuint8mf8_t o
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmadd.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmadd.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vmadd_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t acc, uint8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -1347,7 +1347,7 @@ vuint8mf8_t test_vmadd_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t acc, uint8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmadd.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmadd.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vmadd_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t acc, vuint8mf4_t op1, vuint8mf4_t op2, size_t vl) {
@@ -1357,7 +1357,7 @@ vuint8mf4_t test_vmadd_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t acc, vuint8mf4_t o
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmadd.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmadd.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vmadd_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t acc, uint8_t op1, vuint8mf4_t op2, size_t vl) {
@@ -1367,7 +1367,7 @@ vuint8mf4_t test_vmadd_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t acc, uint8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmadd.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmadd.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vmadd_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t acc, vuint8mf2_t op1, vuint8mf2_t op2, size_t vl) {
@@ -1377,7 +1377,7 @@ vuint8mf2_t test_vmadd_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t acc, vuint8mf2_t o
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmadd.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmadd.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vmadd_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t acc, uint8_t op1, vuint8mf2_t op2, size_t vl) {
@@ -1387,7 +1387,7 @@ vuint8mf2_t test_vmadd_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t acc, uint8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmadd.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmadd.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vmadd_vv_u8m1_m(vbool8_t mask, vuint8m1_t acc, vuint8m1_t op1, vuint8m1_t op2, size_t vl) {
@@ -1397,7 +1397,7 @@ vuint8m1_t test_vmadd_vv_u8m1_m(vbool8_t mask, vuint8m1_t acc, vuint8m1_t op1, v
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmadd.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmadd.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vmadd_vx_u8m1_m(vbool8_t mask, vuint8m1_t acc, uint8_t op1, vuint8m1_t op2, size_t vl) {
@@ -1407,7 +1407,7 @@ vuint8m1_t test_vmadd_vx_u8m1_m(vbool8_t mask, vuint8m1_t acc, uint8_t op1, vuin
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmadd.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmadd.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vmadd_vv_u8m2_m(vbool4_t mask, vuint8m2_t acc, vuint8m2_t op1, vuint8m2_t op2, size_t vl) {
@@ -1417,7 +1417,7 @@ vuint8m2_t test_vmadd_vv_u8m2_m(vbool4_t mask, vuint8m2_t acc, vuint8m2_t op1, v
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmadd.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmadd.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vmadd_vx_u8m2_m(vbool4_t mask, vuint8m2_t acc, uint8_t op1, vuint8m2_t op2, size_t vl) {
@@ -1427,7 +1427,7 @@ vuint8m2_t test_vmadd_vx_u8m2_m(vbool4_t mask, vuint8m2_t acc, uint8_t op1, vuin
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmadd.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmadd.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vmadd_vv_u8m4_m(vbool2_t mask, vuint8m4_t acc, vuint8m4_t op1, vuint8m4_t op2, size_t vl) {
@@ -1437,7 +1437,7 @@ vuint8m4_t test_vmadd_vv_u8m4_m(vbool2_t mask, vuint8m4_t acc, vuint8m4_t op1, v
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmadd.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmadd.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vmadd_vx_u8m4_m(vbool2_t mask, vuint8m4_t acc, uint8_t op1, vuint8m4_t op2, size_t vl) {
@@ -1447,7 +1447,7 @@ vuint8m4_t test_vmadd_vx_u8m4_m(vbool2_t mask, vuint8m4_t acc, uint8_t op1, vuin
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmadd.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[ACC:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmadd.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[ACC:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vmadd_vv_u8m8_m(vbool1_t mask, vuint8m8_t acc, vuint8m8_t op1, vuint8m8_t op2, size_t vl) {
@@ -1457,7 +1457,7 @@ vuint8m8_t test_vmadd_vv_u8m8_m(vbool1_t mask, vuint8m8_t acc, vuint8m8_t op1, v
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmadd.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmadd.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vmadd_vx_u8m8_m(vbool1_t mask, vuint8m8_t acc, uint8_t op1, vuint8m8_t op2, size_t vl) {
@@ -1467,7 +1467,7 @@ vuint8m8_t test_vmadd_vx_u8m8_m(vbool1_t mask, vuint8m8_t acc, uint8_t op1, vuin
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmadd.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmadd.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vmadd_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t acc, vuint16mf4_t op1, vuint16mf4_t op2, size_t vl) {
@@ -1477,7 +1477,7 @@ vuint16mf4_t test_vmadd_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t acc, vuint16mf4
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmadd.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmadd.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vmadd_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t acc, uint16_t op1, vuint16mf4_t op2, size_t vl) {
@@ -1487,7 +1487,7 @@ vuint16mf4_t test_vmadd_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t acc, uint16_t o
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmadd.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmadd.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vmadd_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t acc, vuint16mf2_t op1, vuint16mf2_t op2, size_t vl) {
@@ -1497,7 +1497,7 @@ vuint16mf2_t test_vmadd_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t acc, vuint16mf2
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmadd.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmadd.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vmadd_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t acc, uint16_t op1, vuint16mf2_t op2, size_t vl) {
@@ -1507,7 +1507,7 @@ vuint16mf2_t test_vmadd_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t acc, uint16_t o
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmadd.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmadd.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vmadd_vv_u16m1_m(vbool16_t mask, vuint16m1_t acc, vuint16m1_t op1, vuint16m1_t op2, size_t vl) {
@@ -1517,7 +1517,7 @@ vuint16m1_t test_vmadd_vv_u16m1_m(vbool16_t mask, vuint16m1_t acc, vuint16m1_t o
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmadd.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmadd.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vmadd_vx_u16m1_m(vbool16_t mask, vuint16m1_t acc, uint16_t op1, vuint16m1_t op2, size_t vl) {
@@ -1527,7 +1527,7 @@ vuint16m1_t test_vmadd_vx_u16m1_m(vbool16_t mask, vuint16m1_t acc, uint16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmadd.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmadd.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vmadd_vv_u16m2_m(vbool8_t mask, vuint16m2_t acc, vuint16m2_t op1, vuint16m2_t op2, size_t vl) {
@@ -1537,7 +1537,7 @@ vuint16m2_t test_vmadd_vv_u16m2_m(vbool8_t mask, vuint16m2_t acc, vuint16m2_t op
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmadd.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmadd.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vmadd_vx_u16m2_m(vbool8_t mask, vuint16m2_t acc, uint16_t op1, vuint16m2_t op2, size_t vl) {
@@ -1547,7 +1547,7 @@ vuint16m2_t test_vmadd_vx_u16m2_m(vbool8_t mask, vuint16m2_t acc, uint16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmadd.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmadd.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vmadd_vv_u16m4_m(vbool4_t mask, vuint16m4_t acc, vuint16m4_t op1, vuint16m4_t op2, size_t vl) {
@@ -1557,7 +1557,7 @@ vuint16m4_t test_vmadd_vv_u16m4_m(vbool4_t mask, vuint16m4_t acc, vuint16m4_t op
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmadd.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmadd.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vmadd_vx_u16m4_m(vbool4_t mask, vuint16m4_t acc, uint16_t op1, vuint16m4_t op2, size_t vl) {
@@ -1567,7 +1567,7 @@ vuint16m4_t test_vmadd_vx_u16m4_m(vbool4_t mask, vuint16m4_t acc, uint16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmadd.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmadd.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vmadd_vv_u16m8_m(vbool2_t mask, vuint16m8_t acc, vuint16m8_t op1, vuint16m8_t op2, size_t vl) {
@@ -1577,7 +1577,7 @@ vuint16m8_t test_vmadd_vv_u16m8_m(vbool2_t mask, vuint16m8_t acc, vuint16m8_t op
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmadd.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmadd.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vmadd_vx_u16m8_m(vbool2_t mask, vuint16m8_t acc, uint16_t op1, vuint16m8_t op2, size_t vl) {
@@ -1587,7 +1587,7 @@ vuint16m8_t test_vmadd_vx_u16m8_m(vbool2_t mask, vuint16m8_t acc, uint16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmadd.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmadd.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vmadd_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t acc, vuint32mf2_t op1, vuint32mf2_t op2, size_t vl) {
@@ -1597,7 +1597,7 @@ vuint32mf2_t test_vmadd_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t acc, vuint32mf2
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmadd.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmadd.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vmadd_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t acc, uint32_t op1, vuint32mf2_t op2, size_t vl) {
@@ -1607,7 +1607,7 @@ vuint32mf2_t test_vmadd_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t acc, uint32_t o
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmadd.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmadd.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vmadd_vv_u32m1_m(vbool32_t mask, vuint32m1_t acc, vuint32m1_t op1, vuint32m1_t op2, size_t vl) {
@@ -1617,7 +1617,7 @@ vuint32m1_t test_vmadd_vv_u32m1_m(vbool32_t mask, vuint32m1_t acc, vuint32m1_t o
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmadd.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmadd.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vmadd_vx_u32m1_m(vbool32_t mask, vuint32m1_t acc, uint32_t op1, vuint32m1_t op2, size_t vl) {
@@ -1627,7 +1627,7 @@ vuint32m1_t test_vmadd_vx_u32m1_m(vbool32_t mask, vuint32m1_t acc, uint32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmadd.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmadd.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vmadd_vv_u32m2_m(vbool16_t mask, vuint32m2_t acc, vuint32m2_t op1, vuint32m2_t op2, size_t vl) {
@@ -1637,7 +1637,7 @@ vuint32m2_t test_vmadd_vv_u32m2_m(vbool16_t mask, vuint32m2_t acc, vuint32m2_t o
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmadd.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmadd.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vmadd_vx_u32m2_m(vbool16_t mask, vuint32m2_t acc, uint32_t op1, vuint32m2_t op2, size_t vl) {
@@ -1647,7 +1647,7 @@ vuint32m2_t test_vmadd_vx_u32m2_m(vbool16_t mask, vuint32m2_t acc, uint32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmadd.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmadd.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vmadd_vv_u32m4_m(vbool8_t mask, vuint32m4_t acc, vuint32m4_t op1, vuint32m4_t op2, size_t vl) {
@@ -1657,7 +1657,7 @@ vuint32m4_t test_vmadd_vv_u32m4_m(vbool8_t mask, vuint32m4_t acc, vuint32m4_t op
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmadd.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmadd.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vmadd_vx_u32m4_m(vbool8_t mask, vuint32m4_t acc, uint32_t op1, vuint32m4_t op2, size_t vl) {
@@ -1667,7 +1667,7 @@ vuint32m4_t test_vmadd_vx_u32m4_m(vbool8_t mask, vuint32m4_t acc, uint32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmadd.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmadd.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vmadd_vv_u32m8_m(vbool4_t mask, vuint32m8_t acc, vuint32m8_t op1, vuint32m8_t op2, size_t vl) {
@@ -1677,7 +1677,7 @@ vuint32m8_t test_vmadd_vv_u32m8_m(vbool4_t mask, vuint32m8_t acc, vuint32m8_t op
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmadd.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmadd.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vmadd_vx_u32m8_m(vbool4_t mask, vuint32m8_t acc, uint32_t op1, vuint32m8_t op2, size_t vl) {
@@ -1687,7 +1687,7 @@ vuint32m8_t test_vmadd_vx_u32m8_m(vbool4_t mask, vuint32m8_t acc, uint32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmadd.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmadd.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vmadd_vv_u64m1_m(vbool64_t mask, vuint64m1_t acc, vuint64m1_t op1, vuint64m1_t op2, size_t vl) {
@@ -1697,7 +1697,7 @@ vuint64m1_t test_vmadd_vv_u64m1_m(vbool64_t mask, vuint64m1_t acc, vuint64m1_t o
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmadd.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmadd.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vmadd_vx_u64m1_m(vbool64_t mask, vuint64m1_t acc, uint64_t op1, vuint64m1_t op2, size_t vl) {
@@ -1707,7 +1707,7 @@ vuint64m1_t test_vmadd_vx_u64m1_m(vbool64_t mask, vuint64m1_t acc, uint64_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmadd.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmadd.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vmadd_vv_u64m2_m(vbool32_t mask, vuint64m2_t acc, vuint64m2_t op1, vuint64m2_t op2, size_t vl) {
@@ -1717,7 +1717,7 @@ vuint64m2_t test_vmadd_vv_u64m2_m(vbool32_t mask, vuint64m2_t acc, vuint64m2_t o
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmadd.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmadd.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vmadd_vx_u64m2_m(vbool32_t mask, vuint64m2_t acc, uint64_t op1, vuint64m2_t op2, size_t vl) {
@@ -1727,7 +1727,7 @@ vuint64m2_t test_vmadd_vx_u64m2_m(vbool32_t mask, vuint64m2_t acc, uint64_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmadd.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmadd.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vmadd_vv_u64m4_m(vbool16_t mask, vuint64m4_t acc, vuint64m4_t op1, vuint64m4_t op2, size_t vl) {
@@ -1737,7 +1737,7 @@ vuint64m4_t test_vmadd_vv_u64m4_m(vbool16_t mask, vuint64m4_t acc, vuint64m4_t o
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmadd.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmadd.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vmadd_vx_u64m4_m(vbool16_t mask, vuint64m4_t acc, uint64_t op1, vuint64m4_t op2, size_t vl) {
@@ -1747,7 +1747,7 @@ vuint64m4_t test_vmadd_vx_u64m4_m(vbool16_t mask, vuint64m4_t acc, uint64_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmadd_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmadd.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmadd.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vmadd_vv_u64m8_m(vbool8_t mask, vuint64m8_t acc, vuint64m8_t op1, vuint64m8_t op2, size_t vl) {
@@ -1757,7 +1757,7 @@ vuint64m8_t test_vmadd_vv_u64m8_m(vbool8_t mask, vuint64m8_t acc, vuint64m8_t op
 //
 // CHECK-RV64-LABEL: @test_vmadd_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmadd.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmadd.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vmadd_vx_u64m8_m(vbool8_t mask, vuint64m8_t acc, uint64_t op1, vuint64m8_t op2, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmax.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmax.c
index 83f3d6b..06a6f51 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmax.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmax.c
@@ -887,7 +887,7 @@ vuint64m8_t test_vmaxu_vx_u64m8(vuint64m8_t op1, uint64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmax.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmax.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vmax_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, vint8mf8_t op2, size_t vl) {
@@ -897,7 +897,7 @@ vint8mf8_t test_vmax_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmax.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmax.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vmax_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, int8_t op2, size_t vl) {
@@ -907,7 +907,7 @@ vint8mf8_t test_vmax_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmax.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmax.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vmax_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, vint8mf4_t op2, size_t vl) {
@@ -917,7 +917,7 @@ vint8mf4_t test_vmax_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmax.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmax.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vmax_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, int8_t op2, size_t vl) {
@@ -927,7 +927,7 @@ vint8mf4_t test_vmax_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmax.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmax.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vmax_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, vint8mf2_t op2, size_t vl) {
@@ -937,7 +937,7 @@ vint8mf2_t test_vmax_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmax.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmax.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vmax_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, int8_t op2, size_t vl) {
@@ -947,7 +947,7 @@ vint8mf2_t test_vmax_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmax.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmax.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vmax_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, vint8m1_t op2, size_t vl) {
@@ -957,7 +957,7 @@ vint8m1_t test_vmax_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmax.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmax.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vmax_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, int8_t op2, size_t vl) {
@@ -967,7 +967,7 @@ vint8m1_t test_vmax_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmax.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmax.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vmax_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, vint8m2_t op2, size_t vl) {
@@ -977,7 +977,7 @@ vint8m2_t test_vmax_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmax.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmax.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vmax_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, int8_t op2, size_t vl) {
@@ -987,7 +987,7 @@ vint8m2_t test_vmax_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmax.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmax.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vmax_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, vint8m4_t op2, size_t vl) {
@@ -997,7 +997,7 @@ vint8m4_t test_vmax_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmax.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmax.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vmax_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, int8_t op2, size_t vl) {
@@ -1007,7 +1007,7 @@ vint8m4_t test_vmax_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmax.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmax.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vmax_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, vint8m8_t op2, size_t vl) {
@@ -1017,7 +1017,7 @@ vint8m8_t test_vmax_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmax.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmax.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vmax_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, int8_t op2, size_t vl) {
@@ -1027,7 +1027,7 @@ vint8m8_t test_vmax_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmax.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmax.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vmax_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, vint16mf4_t op2, size_t vl) {
@@ -1037,7 +1037,7 @@ vint16mf4_t test_vmax_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmax.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmax.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vmax_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, int16_t op2, size_t vl) {
@@ -1047,7 +1047,7 @@ vint16mf4_t test_vmax_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmax.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmax.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vmax_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, vint16mf2_t op2, size_t vl) {
@@ -1057,7 +1057,7 @@ vint16mf2_t test_vmax_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmax.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmax.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vmax_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, int16_t op2, size_t vl) {
@@ -1067,7 +1067,7 @@ vint16mf2_t test_vmax_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmax.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmax.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vmax_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, vint16m1_t op2, size_t vl) {
@@ -1077,7 +1077,7 @@ vint16m1_t test_vmax_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmax.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmax.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vmax_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, int16_t op2, size_t vl) {
@@ -1087,7 +1087,7 @@ vint16m1_t test_vmax_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmax.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmax.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vmax_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, vint16m2_t op2, size_t vl) {
@@ -1097,7 +1097,7 @@ vint16m2_t test_vmax_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmax.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmax.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vmax_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, int16_t op2, size_t vl) {
@@ -1107,7 +1107,7 @@ vint16m2_t test_vmax_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmax.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmax.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vmax_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, vint16m4_t op2, size_t vl) {
@@ -1117,7 +1117,7 @@ vint16m4_t test_vmax_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmax.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmax.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vmax_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, int16_t op2, size_t vl) {
@@ -1127,7 +1127,7 @@ vint16m4_t test_vmax_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmax.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmax.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vmax_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, vint16m8_t op2, size_t vl) {
@@ -1137,7 +1137,7 @@ vint16m8_t test_vmax_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmax.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmax.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vmax_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, int16_t op2, size_t vl) {
@@ -1147,7 +1147,7 @@ vint16m8_t test_vmax_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmax.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmax.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vmax_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, vint32mf2_t op2, size_t vl) {
@@ -1157,7 +1157,7 @@ vint32mf2_t test_vmax_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmax.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmax.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vmax_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, int32_t op2, size_t vl) {
@@ -1167,7 +1167,7 @@ vint32mf2_t test_vmax_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmax.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmax.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vmax_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, vint32m1_t op2, size_t vl) {
@@ -1177,7 +1177,7 @@ vint32m1_t test_vmax_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmax.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmax.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vmax_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, int32_t op2, size_t vl) {
@@ -1187,7 +1187,7 @@ vint32m1_t test_vmax_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmax.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmax.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vmax_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, vint32m2_t op2, size_t vl) {
@@ -1197,7 +1197,7 @@ vint32m2_t test_vmax_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmax.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmax.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vmax_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, int32_t op2, size_t vl) {
@@ -1207,7 +1207,7 @@ vint32m2_t test_vmax_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmax.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmax.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vmax_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, vint32m4_t op2, size_t vl) {
@@ -1217,7 +1217,7 @@ vint32m4_t test_vmax_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmax.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmax.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vmax_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, int32_t op2, size_t vl) {
@@ -1227,7 +1227,7 @@ vint32m4_t test_vmax_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmax.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmax.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vmax_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, vint32m8_t op2, size_t vl) {
@@ -1237,7 +1237,7 @@ vint32m8_t test_vmax_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmax.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmax.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vmax_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, int32_t op2, size_t vl) {
@@ -1247,7 +1247,7 @@ vint32m8_t test_vmax_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmax.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmax.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vmax_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, vint64m1_t op2, size_t vl) {
@@ -1257,7 +1257,7 @@ vint64m1_t test_vmax_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmax.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmax.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vmax_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, int64_t op2, size_t vl) {
@@ -1267,7 +1267,7 @@ vint64m1_t test_vmax_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmax.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmax.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vmax_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, vint64m2_t op2, size_t vl) {
@@ -1277,7 +1277,7 @@ vint64m2_t test_vmax_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmax.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmax.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vmax_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, int64_t op2, size_t vl) {
@@ -1287,7 +1287,7 @@ vint64m2_t test_vmax_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmax.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmax.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vmax_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, vint64m4_t op2, size_t vl) {
@@ -1297,7 +1297,7 @@ vint64m4_t test_vmax_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmax.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmax.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vmax_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, int64_t op2, size_t vl) {
@@ -1307,7 +1307,7 @@ vint64m4_t test_vmax_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmax.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmax.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vmax_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, vint64m8_t op2, size_t vl) {
@@ -1317,7 +1317,7 @@ vint64m8_t test_vmax_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vmax_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmax.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmax.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vmax_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, int64_t op2, size_t vl) {
@@ -1327,7 +1327,7 @@ vint64m8_t test_vmax_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmaxu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmaxu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vmaxu_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -1337,7 +1337,7 @@ vuint8mf8_t test_vmaxu_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmaxu.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmaxu.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vmaxu_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, uint8_t op2, size_t vl) {
@@ -1347,7 +1347,7 @@ vuint8mf8_t test_vmaxu_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmaxu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmaxu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vmaxu_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, vuint8mf4_t op2, size_t vl) {
@@ -1357,7 +1357,7 @@ vuint8mf4_t test_vmaxu_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmaxu.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmaxu.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vmaxu_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, uint8_t op2, size_t vl) {
@@ -1367,7 +1367,7 @@ vuint8mf4_t test_vmaxu_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmaxu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmaxu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vmaxu_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, vuint8mf2_t op2, size_t vl) {
@@ -1377,7 +1377,7 @@ vuint8mf2_t test_vmaxu_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmaxu.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmaxu.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vmaxu_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, uint8_t op2, size_t vl) {
@@ -1387,7 +1387,7 @@ vuint8mf2_t test_vmaxu_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmaxu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmaxu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vmaxu_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, vuint8m1_t op2, size_t vl) {
@@ -1397,7 +1397,7 @@ vuint8m1_t test_vmaxu_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmaxu.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmaxu.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vmaxu_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, uint8_t op2, size_t vl) {
@@ -1407,7 +1407,7 @@ vuint8m1_t test_vmaxu_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmaxu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmaxu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vmaxu_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, vuint8m2_t op2, size_t vl) {
@@ -1417,7 +1417,7 @@ vuint8m2_t test_vmaxu_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmaxu.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmaxu.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vmaxu_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, uint8_t op2, size_t vl) {
@@ -1427,7 +1427,7 @@ vuint8m2_t test_vmaxu_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmaxu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmaxu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vmaxu_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, vuint8m4_t op2, size_t vl) {
@@ -1437,7 +1437,7 @@ vuint8m4_t test_vmaxu_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmaxu.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmaxu.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vmaxu_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, uint8_t op2, size_t vl) {
@@ -1447,7 +1447,7 @@ vuint8m4_t test_vmaxu_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmaxu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmaxu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vmaxu_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, vuint8m8_t op2, size_t vl) {
@@ -1457,7 +1457,7 @@ vuint8m8_t test_vmaxu_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmaxu.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmaxu.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vmaxu_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, uint8_t op2, size_t vl) {
@@ -1467,7 +1467,7 @@ vuint8m8_t test_vmaxu_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmaxu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmaxu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vmaxu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, vuint16mf4_t op2, size_t vl) {
@@ -1477,7 +1477,7 @@ vuint16mf4_t test_vmaxu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmaxu.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmaxu.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vmaxu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, uint16_t op2, size_t vl) {
@@ -1487,7 +1487,7 @@ vuint16mf4_t test_vmaxu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmaxu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmaxu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vmaxu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, vuint16mf2_t op2, size_t vl) {
@@ -1497,7 +1497,7 @@ vuint16mf2_t test_vmaxu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmaxu.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmaxu.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vmaxu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, uint16_t op2, size_t vl) {
@@ -1507,7 +1507,7 @@ vuint16mf2_t test_vmaxu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmaxu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmaxu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vmaxu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, vuint16m1_t op2, size_t vl) {
@@ -1517,7 +1517,7 @@ vuint16m1_t test_vmaxu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmaxu.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmaxu.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vmaxu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, uint16_t op2, size_t vl) {
@@ -1527,7 +1527,7 @@ vuint16m1_t test_vmaxu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmaxu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmaxu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vmaxu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, vuint16m2_t op2, size_t vl) {
@@ -1537,7 +1537,7 @@ vuint16m2_t test_vmaxu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmaxu.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmaxu.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vmaxu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, uint16_t op2, size_t vl) {
@@ -1547,7 +1547,7 @@ vuint16m2_t test_vmaxu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmaxu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmaxu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vmaxu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, vuint16m4_t op2, size_t vl) {
@@ -1557,7 +1557,7 @@ vuint16m4_t test_vmaxu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmaxu.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmaxu.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vmaxu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, uint16_t op2, size_t vl) {
@@ -1567,7 +1567,7 @@ vuint16m4_t test_vmaxu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmaxu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmaxu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vmaxu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, vuint16m8_t op2, size_t vl) {
@@ -1577,7 +1577,7 @@ vuint16m8_t test_vmaxu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmaxu.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmaxu.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vmaxu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, uint16_t op2, size_t vl) {
@@ -1587,7 +1587,7 @@ vuint16m8_t test_vmaxu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmaxu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmaxu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vmaxu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, vuint32mf2_t op2, size_t vl) {
@@ -1597,7 +1597,7 @@ vuint32mf2_t test_vmaxu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmaxu.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmaxu.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vmaxu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, uint32_t op2, size_t vl) {
@@ -1607,7 +1607,7 @@ vuint32mf2_t test_vmaxu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmaxu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmaxu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vmaxu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, vuint32m1_t op2, size_t vl) {
@@ -1617,7 +1617,7 @@ vuint32m1_t test_vmaxu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmaxu.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmaxu.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vmaxu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, uint32_t op2, size_t vl) {
@@ -1627,7 +1627,7 @@ vuint32m1_t test_vmaxu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmaxu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmaxu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vmaxu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, vuint32m2_t op2, size_t vl) {
@@ -1637,7 +1637,7 @@ vuint32m2_t test_vmaxu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmaxu.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmaxu.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vmaxu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, uint32_t op2, size_t vl) {
@@ -1647,7 +1647,7 @@ vuint32m2_t test_vmaxu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmaxu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmaxu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vmaxu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, vuint32m4_t op2, size_t vl) {
@@ -1657,7 +1657,7 @@ vuint32m4_t test_vmaxu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmaxu.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmaxu.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vmaxu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, uint32_t op2, size_t vl) {
@@ -1667,7 +1667,7 @@ vuint32m4_t test_vmaxu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmaxu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmaxu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vmaxu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, vuint32m8_t op2, size_t vl) {
@@ -1677,7 +1677,7 @@ vuint32m8_t test_vmaxu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmaxu.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmaxu.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vmaxu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, uint32_t op2, size_t vl) {
@@ -1687,7 +1687,7 @@ vuint32m8_t test_vmaxu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmaxu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmaxu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vmaxu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, vuint64m1_t op2, size_t vl) {
@@ -1697,7 +1697,7 @@ vuint64m1_t test_vmaxu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmaxu.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmaxu.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vmaxu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, uint64_t op2, size_t vl) {
@@ -1707,7 +1707,7 @@ vuint64m1_t test_vmaxu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmaxu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmaxu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vmaxu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, vuint64m2_t op2, size_t vl) {
@@ -1717,7 +1717,7 @@ vuint64m2_t test_vmaxu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmaxu.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmaxu.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vmaxu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, uint64_t op2, size_t vl) {
@@ -1727,7 +1727,7 @@ vuint64m2_t test_vmaxu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmaxu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmaxu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vmaxu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, vuint64m4_t op2, size_t vl) {
@@ -1737,7 +1737,7 @@ vuint64m4_t test_vmaxu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmaxu.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmaxu.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vmaxu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, uint64_t op2, size_t vl) {
@@ -1747,7 +1747,7 @@ vuint64m4_t test_vmaxu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmaxu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmaxu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vmaxu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, vuint64m8_t op2, size_t vl) {
@@ -1757,7 +1757,7 @@ vuint64m8_t test_vmaxu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vmaxu_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmaxu.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmaxu.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vmaxu_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, uint64_t op2, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmerge.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmerge.c
index 6e336d4..ecbe041 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmerge.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmerge.c
@@ -8,7 +8,7 @@
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i8mf8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmerge.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmerge.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vmerge_vvm_i8mf8(vbool64_t mask, vint8mf8_t op1, vint8mf8_t op2,
@@ -19,7 +19,7 @@ vint8mf8_t test_vmerge_vvm_i8mf8(vbool64_t mask, vint8mf8_t op1, vint8mf8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i8mf8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmerge.nxv1i8.i8.i64(<vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmerge.nxv1i8.i8.i64(<vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vmerge_vxm_i8mf8(vbool64_t mask, vint8mf8_t op1, int8_t op2,
@@ -30,7 +30,7 @@ vint8mf8_t test_vmerge_vxm_i8mf8(vbool64_t mask, vint8mf8_t op1, int8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i8mf4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmerge.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmerge.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vmerge_vvm_i8mf4(vbool32_t mask, vint8mf4_t op1, vint8mf4_t op2,
@@ -41,7 +41,7 @@ vint8mf4_t test_vmerge_vvm_i8mf4(vbool32_t mask, vint8mf4_t op1, vint8mf4_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i8mf4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmerge.nxv2i8.i8.i64(<vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmerge.nxv2i8.i8.i64(<vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vmerge_vxm_i8mf4(vbool32_t mask, vint8mf4_t op1, int8_t op2,
@@ -52,7 +52,7 @@ vint8mf4_t test_vmerge_vxm_i8mf4(vbool32_t mask, vint8mf4_t op1, int8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i8mf2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmerge.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmerge.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vmerge_vvm_i8mf2(vbool16_t mask, vint8mf2_t op1, vint8mf2_t op2,
@@ -63,7 +63,7 @@ vint8mf2_t test_vmerge_vvm_i8mf2(vbool16_t mask, vint8mf2_t op1, vint8mf2_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i8mf2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmerge.nxv4i8.i8.i64(<vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmerge.nxv4i8.i8.i64(<vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vmerge_vxm_i8mf2(vbool16_t mask, vint8mf2_t op1, int8_t op2,
@@ -74,7 +74,7 @@ vint8mf2_t test_vmerge_vxm_i8mf2(vbool16_t mask, vint8mf2_t op1, int8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i8m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmerge.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmerge.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vmerge_vvm_i8m1(vbool8_t mask, vint8m1_t op1, vint8m1_t op2,
@@ -85,7 +85,7 @@ vint8m1_t test_vmerge_vvm_i8m1(vbool8_t mask, vint8m1_t op1, vint8m1_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i8m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmerge.nxv8i8.i8.i64(<vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmerge.nxv8i8.i8.i64(<vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vmerge_vxm_i8m1(vbool8_t mask, vint8m1_t op1, int8_t op2,
@@ -96,7 +96,7 @@ vint8m1_t test_vmerge_vxm_i8m1(vbool8_t mask, vint8m1_t op1, int8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i8m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmerge.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmerge.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vmerge_vvm_i8m2(vbool4_t mask, vint8m2_t op1, vint8m2_t op2,
@@ -107,7 +107,7 @@ vint8m2_t test_vmerge_vvm_i8m2(vbool4_t mask, vint8m2_t op1, vint8m2_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i8m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmerge.nxv16i8.i8.i64(<vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmerge.nxv16i8.i8.i64(<vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vmerge_vxm_i8m2(vbool4_t mask, vint8m2_t op1, int8_t op2,
@@ -118,7 +118,7 @@ vint8m2_t test_vmerge_vxm_i8m2(vbool4_t mask, vint8m2_t op1, int8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i8m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmerge.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmerge.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vmerge_vvm_i8m4(vbool2_t mask, vint8m4_t op1, vint8m4_t op2,
@@ -129,7 +129,7 @@ vint8m4_t test_vmerge_vvm_i8m4(vbool2_t mask, vint8m4_t op1, vint8m4_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i8m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmerge.nxv32i8.i8.i64(<vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmerge.nxv32i8.i8.i64(<vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vmerge_vxm_i8m4(vbool2_t mask, vint8m4_t op1, int8_t op2,
@@ -140,7 +140,7 @@ vint8m4_t test_vmerge_vxm_i8m4(vbool2_t mask, vint8m4_t op1, int8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i8m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmerge.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmerge.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vmerge_vvm_i8m8(vbool1_t mask, vint8m8_t op1, vint8m8_t op2,
@@ -151,7 +151,7 @@ vint8m8_t test_vmerge_vvm_i8m8(vbool1_t mask, vint8m8_t op1, vint8m8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i8m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmerge.nxv64i8.i8.i64(<vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmerge.nxv64i8.i8.i64(<vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vmerge_vxm_i8m8(vbool1_t mask, vint8m8_t op1, int8_t op2,
@@ -162,7 +162,7 @@ vint8m8_t test_vmerge_vxm_i8m8(vbool1_t mask, vint8m8_t op1, int8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i16mf4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmerge.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmerge.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vmerge_vvm_i16mf4(vbool64_t mask, vint16mf4_t op1,
@@ -173,7 +173,7 @@ vint16mf4_t test_vmerge_vvm_i16mf4(vbool64_t mask, vint16mf4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i16mf4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmerge.nxv1i16.i16.i64(<vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmerge.nxv1i16.i16.i64(<vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vmerge_vxm_i16mf4(vbool64_t mask, vint16mf4_t op1, int16_t op2,
@@ -184,7 +184,7 @@ vint16mf4_t test_vmerge_vxm_i16mf4(vbool64_t mask, vint16mf4_t op1, int16_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i16mf2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmerge.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmerge.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vmerge_vvm_i16mf2(vbool32_t mask, vint16mf2_t op1,
@@ -195,7 +195,7 @@ vint16mf2_t test_vmerge_vvm_i16mf2(vbool32_t mask, vint16mf2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i16mf2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmerge.nxv2i16.i16.i64(<vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmerge.nxv2i16.i16.i64(<vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vmerge_vxm_i16mf2(vbool32_t mask, vint16mf2_t op1, int16_t op2,
@@ -206,7 +206,7 @@ vint16mf2_t test_vmerge_vxm_i16mf2(vbool32_t mask, vint16mf2_t op1, int16_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i16m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmerge.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmerge.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vmerge_vvm_i16m1(vbool16_t mask, vint16m1_t op1, vint16m1_t op2,
@@ -217,7 +217,7 @@ vint16m1_t test_vmerge_vvm_i16m1(vbool16_t mask, vint16m1_t op1, vint16m1_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i16m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmerge.nxv4i16.i16.i64(<vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmerge.nxv4i16.i16.i64(<vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vmerge_vxm_i16m1(vbool16_t mask, vint16m1_t op1, int16_t op2,
@@ -228,7 +228,7 @@ vint16m1_t test_vmerge_vxm_i16m1(vbool16_t mask, vint16m1_t op1, int16_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i16m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmerge.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmerge.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vmerge_vvm_i16m2(vbool8_t mask, vint16m2_t op1, vint16m2_t op2,
@@ -239,7 +239,7 @@ vint16m2_t test_vmerge_vvm_i16m2(vbool8_t mask, vint16m2_t op1, vint16m2_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i16m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmerge.nxv8i16.i16.i64(<vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmerge.nxv8i16.i16.i64(<vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vmerge_vxm_i16m2(vbool8_t mask, vint16m2_t op1, int16_t op2,
@@ -250,7 +250,7 @@ vint16m2_t test_vmerge_vxm_i16m2(vbool8_t mask, vint16m2_t op1, int16_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i16m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmerge.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmerge.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vmerge_vvm_i16m4(vbool4_t mask, vint16m4_t op1, vint16m4_t op2,
@@ -261,7 +261,7 @@ vint16m4_t test_vmerge_vvm_i16m4(vbool4_t mask, vint16m4_t op1, vint16m4_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i16m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmerge.nxv16i16.i16.i64(<vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmerge.nxv16i16.i16.i64(<vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vmerge_vxm_i16m4(vbool4_t mask, vint16m4_t op1, int16_t op2,
@@ -272,7 +272,7 @@ vint16m4_t test_vmerge_vxm_i16m4(vbool4_t mask, vint16m4_t op1, int16_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i16m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmerge.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmerge.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vmerge_vvm_i16m8(vbool2_t mask, vint16m8_t op1, vint16m8_t op2,
@@ -283,7 +283,7 @@ vint16m8_t test_vmerge_vvm_i16m8(vbool2_t mask, vint16m8_t op1, vint16m8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i16m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmerge.nxv32i16.i16.i64(<vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmerge.nxv32i16.i16.i64(<vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vmerge_vxm_i16m8(vbool2_t mask, vint16m8_t op1, int16_t op2,
@@ -294,7 +294,7 @@ vint16m8_t test_vmerge_vxm_i16m8(vbool2_t mask, vint16m8_t op1, int16_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i32mf2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmerge.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmerge.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vmerge_vvm_i32mf2(vbool64_t mask, vint32mf2_t op1,
@@ -305,7 +305,7 @@ vint32mf2_t test_vmerge_vvm_i32mf2(vbool64_t mask, vint32mf2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i32mf2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmerge.nxv1i32.i32.i64(<vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmerge.nxv1i32.i32.i64(<vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vmerge_vxm_i32mf2(vbool64_t mask, vint32mf2_t op1, int32_t op2,
@@ -316,7 +316,7 @@ vint32mf2_t test_vmerge_vxm_i32mf2(vbool64_t mask, vint32mf2_t op1, int32_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i32m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmerge.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmerge.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vmerge_vvm_i32m1(vbool32_t mask, vint32m1_t op1, vint32m1_t op2,
@@ -327,7 +327,7 @@ vint32m1_t test_vmerge_vvm_i32m1(vbool32_t mask, vint32m1_t op1, vint32m1_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i32m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmerge.nxv2i32.i32.i64(<vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmerge.nxv2i32.i32.i64(<vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vmerge_vxm_i32m1(vbool32_t mask, vint32m1_t op1, int32_t op2,
@@ -338,7 +338,7 @@ vint32m1_t test_vmerge_vxm_i32m1(vbool32_t mask, vint32m1_t op1, int32_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i32m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmerge.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmerge.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vmerge_vvm_i32m2(vbool16_t mask, vint32m2_t op1, vint32m2_t op2,
@@ -349,7 +349,7 @@ vint32m2_t test_vmerge_vvm_i32m2(vbool16_t mask, vint32m2_t op1, vint32m2_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i32m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmerge.nxv4i32.i32.i64(<vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmerge.nxv4i32.i32.i64(<vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vmerge_vxm_i32m2(vbool16_t mask, vint32m2_t op1, int32_t op2,
@@ -360,7 +360,7 @@ vint32m2_t test_vmerge_vxm_i32m2(vbool16_t mask, vint32m2_t op1, int32_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i32m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmerge.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmerge.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vmerge_vvm_i32m4(vbool8_t mask, vint32m4_t op1, vint32m4_t op2,
@@ -371,7 +371,7 @@ vint32m4_t test_vmerge_vvm_i32m4(vbool8_t mask, vint32m4_t op1, vint32m4_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i32m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmerge.nxv8i32.i32.i64(<vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmerge.nxv8i32.i32.i64(<vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vmerge_vxm_i32m4(vbool8_t mask, vint32m4_t op1, int32_t op2,
@@ -382,7 +382,7 @@ vint32m4_t test_vmerge_vxm_i32m4(vbool8_t mask, vint32m4_t op1, int32_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i32m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmerge.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmerge.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vmerge_vvm_i32m8(vbool4_t mask, vint32m8_t op1, vint32m8_t op2,
@@ -393,7 +393,7 @@ vint32m8_t test_vmerge_vvm_i32m8(vbool4_t mask, vint32m8_t op1, vint32m8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i32m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmerge.nxv16i32.i32.i64(<vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmerge.nxv16i32.i32.i64(<vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vmerge_vxm_i32m8(vbool4_t mask, vint32m8_t op1, int32_t op2,
@@ -404,7 +404,7 @@ vint32m8_t test_vmerge_vxm_i32m8(vbool4_t mask, vint32m8_t op1, int32_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i64m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmerge.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmerge.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vmerge_vvm_i64m1(vbool64_t mask, vint64m1_t op1, vint64m1_t op2,
@@ -415,7 +415,7 @@ vint64m1_t test_vmerge_vvm_i64m1(vbool64_t mask, vint64m1_t op1, vint64m1_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i64m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmerge.nxv1i64.i64.i64(<vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmerge.nxv1i64.i64.i64(<vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vmerge_vxm_i64m1(vbool64_t mask, vint64m1_t op1, int64_t op2,
@@ -426,7 +426,7 @@ vint64m1_t test_vmerge_vxm_i64m1(vbool64_t mask, vint64m1_t op1, int64_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i64m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmerge.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmerge.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vmerge_vvm_i64m2(vbool32_t mask, vint64m2_t op1, vint64m2_t op2,
@@ -437,7 +437,7 @@ vint64m2_t test_vmerge_vvm_i64m2(vbool32_t mask, vint64m2_t op1, vint64m2_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i64m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmerge.nxv2i64.i64.i64(<vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmerge.nxv2i64.i64.i64(<vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vmerge_vxm_i64m2(vbool32_t mask, vint64m2_t op1, int64_t op2,
@@ -448,7 +448,7 @@ vint64m2_t test_vmerge_vxm_i64m2(vbool32_t mask, vint64m2_t op1, int64_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i64m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmerge.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmerge.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vmerge_vvm_i64m4(vbool16_t mask, vint64m4_t op1, vint64m4_t op2,
@@ -459,7 +459,7 @@ vint64m4_t test_vmerge_vvm_i64m4(vbool16_t mask, vint64m4_t op1, vint64m4_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i64m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmerge.nxv4i64.i64.i64(<vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmerge.nxv4i64.i64.i64(<vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vmerge_vxm_i64m4(vbool16_t mask, vint64m4_t op1, int64_t op2,
@@ -470,7 +470,7 @@ vint64m4_t test_vmerge_vxm_i64m4(vbool16_t mask, vint64m4_t op1, int64_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_i64m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmerge.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmerge.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vmerge_vvm_i64m8(vbool8_t mask, vint64m8_t op1, vint64m8_t op2,
@@ -481,7 +481,7 @@ vint64m8_t test_vmerge_vvm_i64m8(vbool8_t mask, vint64m8_t op1, vint64m8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_i64m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmerge.nxv8i64.i64.i64(<vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmerge.nxv8i64.i64.i64(<vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vmerge_vxm_i64m8(vbool8_t mask, vint64m8_t op1, int64_t op2,
@@ -492,7 +492,7 @@ vint64m8_t test_vmerge_vxm_i64m8(vbool8_t mask, vint64m8_t op1, int64_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u8mf8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmerge.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmerge.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vmerge_vvm_u8mf8(vbool64_t mask, vuint8mf8_t op1,
@@ -503,7 +503,7 @@ vuint8mf8_t test_vmerge_vvm_u8mf8(vbool64_t mask, vuint8mf8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u8mf8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmerge.nxv1i8.i8.i64(<vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmerge.nxv1i8.i8.i64(<vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vmerge_vxm_u8mf8(vbool64_t mask, vuint8mf8_t op1, uint8_t op2,
@@ -514,7 +514,7 @@ vuint8mf8_t test_vmerge_vxm_u8mf8(vbool64_t mask, vuint8mf8_t op1, uint8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u8mf4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmerge.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmerge.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vmerge_vvm_u8mf4(vbool32_t mask, vuint8mf4_t op1,
@@ -525,7 +525,7 @@ vuint8mf4_t test_vmerge_vvm_u8mf4(vbool32_t mask, vuint8mf4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u8mf4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmerge.nxv2i8.i8.i64(<vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmerge.nxv2i8.i8.i64(<vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vmerge_vxm_u8mf4(vbool32_t mask, vuint8mf4_t op1, uint8_t op2,
@@ -536,7 +536,7 @@ vuint8mf4_t test_vmerge_vxm_u8mf4(vbool32_t mask, vuint8mf4_t op1, uint8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u8mf2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmerge.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmerge.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vmerge_vvm_u8mf2(vbool16_t mask, vuint8mf2_t op1,
@@ -547,7 +547,7 @@ vuint8mf2_t test_vmerge_vvm_u8mf2(vbool16_t mask, vuint8mf2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u8mf2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmerge.nxv4i8.i8.i64(<vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmerge.nxv4i8.i8.i64(<vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vmerge_vxm_u8mf2(vbool16_t mask, vuint8mf2_t op1, uint8_t op2,
@@ -558,7 +558,7 @@ vuint8mf2_t test_vmerge_vxm_u8mf2(vbool16_t mask, vuint8mf2_t op1, uint8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u8m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmerge.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmerge.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vmerge_vvm_u8m1(vbool8_t mask, vuint8m1_t op1, vuint8m1_t op2,
@@ -569,7 +569,7 @@ vuint8m1_t test_vmerge_vvm_u8m1(vbool8_t mask, vuint8m1_t op1, vuint8m1_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u8m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmerge.nxv8i8.i8.i64(<vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmerge.nxv8i8.i8.i64(<vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vmerge_vxm_u8m1(vbool8_t mask, vuint8m1_t op1, uint8_t op2,
@@ -580,7 +580,7 @@ vuint8m1_t test_vmerge_vxm_u8m1(vbool8_t mask, vuint8m1_t op1, uint8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u8m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmerge.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmerge.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vmerge_vvm_u8m2(vbool4_t mask, vuint8m2_t op1, vuint8m2_t op2,
@@ -591,7 +591,7 @@ vuint8m2_t test_vmerge_vvm_u8m2(vbool4_t mask, vuint8m2_t op1, vuint8m2_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u8m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmerge.nxv16i8.i8.i64(<vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmerge.nxv16i8.i8.i64(<vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vmerge_vxm_u8m2(vbool4_t mask, vuint8m2_t op1, uint8_t op2,
@@ -602,7 +602,7 @@ vuint8m2_t test_vmerge_vxm_u8m2(vbool4_t mask, vuint8m2_t op1, uint8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u8m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmerge.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmerge.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vmerge_vvm_u8m4(vbool2_t mask, vuint8m4_t op1, vuint8m4_t op2,
@@ -613,7 +613,7 @@ vuint8m4_t test_vmerge_vvm_u8m4(vbool2_t mask, vuint8m4_t op1, vuint8m4_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u8m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmerge.nxv32i8.i8.i64(<vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmerge.nxv32i8.i8.i64(<vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vmerge_vxm_u8m4(vbool2_t mask, vuint8m4_t op1, uint8_t op2,
@@ -624,7 +624,7 @@ vuint8m4_t test_vmerge_vxm_u8m4(vbool2_t mask, vuint8m4_t op1, uint8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u8m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmerge.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmerge.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vmerge_vvm_u8m8(vbool1_t mask, vuint8m8_t op1, vuint8m8_t op2,
@@ -635,7 +635,7 @@ vuint8m8_t test_vmerge_vvm_u8m8(vbool1_t mask, vuint8m8_t op1, vuint8m8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u8m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmerge.nxv64i8.i8.i64(<vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmerge.nxv64i8.i8.i64(<vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vmerge_vxm_u8m8(vbool1_t mask, vuint8m8_t op1, uint8_t op2,
@@ -646,7 +646,7 @@ vuint8m8_t test_vmerge_vxm_u8m8(vbool1_t mask, vuint8m8_t op1, uint8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u16mf4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmerge.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmerge.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vmerge_vvm_u16mf4(vbool64_t mask, vuint16mf4_t op1,
@@ -657,7 +657,7 @@ vuint16mf4_t test_vmerge_vvm_u16mf4(vbool64_t mask, vuint16mf4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u16mf4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmerge.nxv1i16.i16.i64(<vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmerge.nxv1i16.i16.i64(<vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vmerge_vxm_u16mf4(vbool64_t mask, vuint16mf4_t op1,
@@ -668,7 +668,7 @@ vuint16mf4_t test_vmerge_vxm_u16mf4(vbool64_t mask, vuint16mf4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u16mf2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmerge.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmerge.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vmerge_vvm_u16mf2(vbool32_t mask, vuint16mf2_t op1,
@@ -679,7 +679,7 @@ vuint16mf2_t test_vmerge_vvm_u16mf2(vbool32_t mask, vuint16mf2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u16mf2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmerge.nxv2i16.i16.i64(<vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmerge.nxv2i16.i16.i64(<vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vmerge_vxm_u16mf2(vbool32_t mask, vuint16mf2_t op1,
@@ -690,7 +690,7 @@ vuint16mf2_t test_vmerge_vxm_u16mf2(vbool32_t mask, vuint16mf2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u16m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmerge.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmerge.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vmerge_vvm_u16m1(vbool16_t mask, vuint16m1_t op1,
@@ -701,7 +701,7 @@ vuint16m1_t test_vmerge_vvm_u16m1(vbool16_t mask, vuint16m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u16m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmerge.nxv4i16.i16.i64(<vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmerge.nxv4i16.i16.i64(<vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vmerge_vxm_u16m1(vbool16_t mask, vuint16m1_t op1, uint16_t op2,
@@ -712,7 +712,7 @@ vuint16m1_t test_vmerge_vxm_u16m1(vbool16_t mask, vuint16m1_t op1, uint16_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u16m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmerge.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmerge.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vmerge_vvm_u16m2(vbool8_t mask, vuint16m2_t op1,
@@ -723,7 +723,7 @@ vuint16m2_t test_vmerge_vvm_u16m2(vbool8_t mask, vuint16m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u16m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmerge.nxv8i16.i16.i64(<vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmerge.nxv8i16.i16.i64(<vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vmerge_vxm_u16m2(vbool8_t mask, vuint16m2_t op1, uint16_t op2,
@@ -734,7 +734,7 @@ vuint16m2_t test_vmerge_vxm_u16m2(vbool8_t mask, vuint16m2_t op1, uint16_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u16m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmerge.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmerge.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vmerge_vvm_u16m4(vbool4_t mask, vuint16m4_t op1,
@@ -745,7 +745,7 @@ vuint16m4_t test_vmerge_vvm_u16m4(vbool4_t mask, vuint16m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u16m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmerge.nxv16i16.i16.i64(<vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmerge.nxv16i16.i16.i64(<vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vmerge_vxm_u16m4(vbool4_t mask, vuint16m4_t op1, uint16_t op2,
@@ -756,7 +756,7 @@ vuint16m4_t test_vmerge_vxm_u16m4(vbool4_t mask, vuint16m4_t op1, uint16_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u16m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmerge.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmerge.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vmerge_vvm_u16m8(vbool2_t mask, vuint16m8_t op1,
@@ -767,7 +767,7 @@ vuint16m8_t test_vmerge_vvm_u16m8(vbool2_t mask, vuint16m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u16m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmerge.nxv32i16.i16.i64(<vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmerge.nxv32i16.i16.i64(<vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vmerge_vxm_u16m8(vbool2_t mask, vuint16m8_t op1, uint16_t op2,
@@ -778,7 +778,7 @@ vuint16m8_t test_vmerge_vxm_u16m8(vbool2_t mask, vuint16m8_t op1, uint16_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u32mf2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmerge.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmerge.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vmerge_vvm_u32mf2(vbool64_t mask, vuint32mf2_t op1,
@@ -789,7 +789,7 @@ vuint32mf2_t test_vmerge_vvm_u32mf2(vbool64_t mask, vuint32mf2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u32mf2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmerge.nxv1i32.i32.i64(<vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmerge.nxv1i32.i32.i64(<vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vmerge_vxm_u32mf2(vbool64_t mask, vuint32mf2_t op1,
@@ -800,7 +800,7 @@ vuint32mf2_t test_vmerge_vxm_u32mf2(vbool64_t mask, vuint32mf2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u32m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmerge.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmerge.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vmerge_vvm_u32m1(vbool32_t mask, vuint32m1_t op1,
@@ -811,7 +811,7 @@ vuint32m1_t test_vmerge_vvm_u32m1(vbool32_t mask, vuint32m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u32m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmerge.nxv2i32.i32.i64(<vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmerge.nxv2i32.i32.i64(<vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vmerge_vxm_u32m1(vbool32_t mask, vuint32m1_t op1, uint32_t op2,
@@ -822,7 +822,7 @@ vuint32m1_t test_vmerge_vxm_u32m1(vbool32_t mask, vuint32m1_t op1, uint32_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u32m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmerge.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmerge.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vmerge_vvm_u32m2(vbool16_t mask, vuint32m2_t op1,
@@ -833,7 +833,7 @@ vuint32m2_t test_vmerge_vvm_u32m2(vbool16_t mask, vuint32m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u32m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmerge.nxv4i32.i32.i64(<vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmerge.nxv4i32.i32.i64(<vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vmerge_vxm_u32m2(vbool16_t mask, vuint32m2_t op1, uint32_t op2,
@@ -844,7 +844,7 @@ vuint32m2_t test_vmerge_vxm_u32m2(vbool16_t mask, vuint32m2_t op1, uint32_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u32m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmerge.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmerge.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vmerge_vvm_u32m4(vbool8_t mask, vuint32m4_t op1,
@@ -855,7 +855,7 @@ vuint32m4_t test_vmerge_vvm_u32m4(vbool8_t mask, vuint32m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u32m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmerge.nxv8i32.i32.i64(<vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmerge.nxv8i32.i32.i64(<vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vmerge_vxm_u32m4(vbool8_t mask, vuint32m4_t op1, uint32_t op2,
@@ -866,7 +866,7 @@ vuint32m4_t test_vmerge_vxm_u32m4(vbool8_t mask, vuint32m4_t op1, uint32_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u32m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmerge.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmerge.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vmerge_vvm_u32m8(vbool4_t mask, vuint32m8_t op1,
@@ -877,7 +877,7 @@ vuint32m8_t test_vmerge_vvm_u32m8(vbool4_t mask, vuint32m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u32m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmerge.nxv16i32.i32.i64(<vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmerge.nxv16i32.i32.i64(<vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vmerge_vxm_u32m8(vbool4_t mask, vuint32m8_t op1, uint32_t op2,
@@ -888,7 +888,7 @@ vuint32m8_t test_vmerge_vxm_u32m8(vbool4_t mask, vuint32m8_t op1, uint32_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u64m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmerge.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmerge.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vmerge_vvm_u64m1(vbool64_t mask, vuint64m1_t op1,
@@ -899,7 +899,7 @@ vuint64m1_t test_vmerge_vvm_u64m1(vbool64_t mask, vuint64m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u64m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmerge.nxv1i64.i64.i64(<vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmerge.nxv1i64.i64.i64(<vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vmerge_vxm_u64m1(vbool64_t mask, vuint64m1_t op1, uint64_t op2,
@@ -910,7 +910,7 @@ vuint64m1_t test_vmerge_vxm_u64m1(vbool64_t mask, vuint64m1_t op1, uint64_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u64m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmerge.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmerge.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vmerge_vvm_u64m2(vbool32_t mask, vuint64m2_t op1,
@@ -921,7 +921,7 @@ vuint64m2_t test_vmerge_vvm_u64m2(vbool32_t mask, vuint64m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u64m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmerge.nxv2i64.i64.i64(<vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmerge.nxv2i64.i64.i64(<vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vmerge_vxm_u64m2(vbool32_t mask, vuint64m2_t op1, uint64_t op2,
@@ -932,7 +932,7 @@ vuint64m2_t test_vmerge_vxm_u64m2(vbool32_t mask, vuint64m2_t op1, uint64_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u64m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmerge.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmerge.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vmerge_vvm_u64m4(vbool16_t mask, vuint64m4_t op1,
@@ -943,7 +943,7 @@ vuint64m4_t test_vmerge_vvm_u64m4(vbool16_t mask, vuint64m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u64m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmerge.nxv4i64.i64.i64(<vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmerge.nxv4i64.i64.i64(<vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vmerge_vxm_u64m4(vbool16_t mask, vuint64m4_t op1, uint64_t op2,
@@ -954,7 +954,7 @@ vuint64m4_t test_vmerge_vxm_u64m4(vbool16_t mask, vuint64m4_t op1, uint64_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_u64m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmerge.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmerge.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vmerge_vvm_u64m8(vbool8_t mask, vuint64m8_t op1,
@@ -965,7 +965,7 @@ vuint64m8_t test_vmerge_vvm_u64m8(vbool8_t mask, vuint64m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vxm_u64m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmerge.nxv8i64.i64.i64(<vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmerge.nxv8i64.i64.i64(<vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vmerge_vxm_u64m8(vbool8_t mask, vuint64m8_t op1, uint64_t op2,
@@ -976,7 +976,7 @@ vuint64m8_t test_vmerge_vxm_u64m8(vbool8_t mask, vuint64m8_t op1, uint64_t op2,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_f32mf2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmerge.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vfmerge.nxv1f32.nxv1f32.i64(<vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vmerge_vvm_f32mf2(vbool64_t mask, vfloat32mf2_t op1,
@@ -987,7 +987,7 @@ vfloat32mf2_t test_vmerge_vvm_f32mf2(vbool64_t mask, vfloat32mf2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_f32m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmerge.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vfmerge.nxv2f32.nxv2f32.i64(<vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vmerge_vvm_f32m1(vbool32_t mask, vfloat32m1_t op1,
@@ -998,7 +998,7 @@ vfloat32m1_t test_vmerge_vvm_f32m1(vbool32_t mask, vfloat32m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_f32m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmerge.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vfmerge.nxv4f32.nxv4f32.i64(<vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vmerge_vvm_f32m2(vbool16_t mask, vfloat32m2_t op1,
@@ -1009,7 +1009,7 @@ vfloat32m2_t test_vmerge_vvm_f32m2(vbool16_t mask, vfloat32m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_f32m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmerge.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vfmerge.nxv8f32.nxv8f32.i64(<vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vmerge_vvm_f32m4(vbool8_t mask, vfloat32m4_t op1,
@@ -1020,7 +1020,7 @@ vfloat32m4_t test_vmerge_vvm_f32m4(vbool8_t mask, vfloat32m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_f32m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmerge.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vfmerge.nxv16f32.nxv16f32.i64(<vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vmerge_vvm_f32m8(vbool4_t mask, vfloat32m8_t op1,
@@ -1031,7 +1031,7 @@ vfloat32m8_t test_vmerge_vvm_f32m8(vbool4_t mask, vfloat32m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_f64m1(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmerge.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vfmerge.nxv1f64.nxv1f64.i64(<vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vmerge_vvm_f64m1(vbool64_t mask, vfloat64m1_t op1,
@@ -1042,7 +1042,7 @@ vfloat64m1_t test_vmerge_vvm_f64m1(vbool64_t mask, vfloat64m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_f64m2(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmerge.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vfmerge.nxv2f64.nxv2f64.i64(<vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vmerge_vvm_f64m2(vbool32_t mask, vfloat64m2_t op1,
@@ -1053,7 +1053,7 @@ vfloat64m2_t test_vmerge_vvm_f64m2(vbool32_t mask, vfloat64m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_f64m4(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmerge.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vfmerge.nxv4f64.nxv4f64.i64(<vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vmerge_vvm_f64m4(vbool16_t mask, vfloat64m4_t op1,
@@ -1064,7 +1064,7 @@ vfloat64m4_t test_vmerge_vvm_f64m4(vbool16_t mask, vfloat64m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmerge_vvm_f64m8(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmerge.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vfmerge.nxv8f64.nxv8f64.i64(<vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vmerge_vvm_f64m8(vbool8_t mask, vfloat64m8_t op1,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmfeq.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmfeq.c
index 3415ba7..d7f5b59 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmfeq.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmfeq.c
@@ -194,7 +194,7 @@ vbool8_t test_vmfeq_vf_f64m8_b8(vfloat64m8_t op1, double op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vmfeq_vv_f32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfeq.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfeq.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmfeq_vv_f32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -206,7 +206,7 @@ vbool64_t test_vmfeq_vv_f32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfeq_vf_f32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfeq.mask.nxv1f32.f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfeq.mask.nxv1f32.f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmfeq_vf_f32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -217,7 +217,7 @@ vbool64_t test_vmfeq_vf_f32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfeq_vv_f32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfeq.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfeq.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmfeq_vv_f32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -229,7 +229,7 @@ vbool32_t test_vmfeq_vv_f32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfeq_vf_f32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfeq.mask.nxv2f32.f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfeq.mask.nxv2f32.f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmfeq_vf_f32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -240,7 +240,7 @@ vbool32_t test_vmfeq_vf_f32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfeq_vv_f32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfeq.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfeq.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmfeq_vv_f32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -252,7 +252,7 @@ vbool16_t test_vmfeq_vv_f32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfeq_vf_f32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfeq.mask.nxv4f32.f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfeq.mask.nxv4f32.f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmfeq_vf_f32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -263,7 +263,7 @@ vbool16_t test_vmfeq_vf_f32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfeq_vv_f32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfeq.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfeq.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmfeq_vv_f32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -275,7 +275,7 @@ vbool8_t test_vmfeq_vv_f32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfeq_vf_f32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfeq.mask.nxv8f32.f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfeq.mask.nxv8f32.f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmfeq_vf_f32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -286,7 +286,7 @@ vbool8_t test_vmfeq_vf_f32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfeq_vv_f32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmfeq.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmfeq.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmfeq_vv_f32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -298,7 +298,7 @@ vbool4_t test_vmfeq_vv_f32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfeq_vf_f32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmfeq.mask.nxv16f32.f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmfeq.mask.nxv16f32.f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmfeq_vf_f32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -309,7 +309,7 @@ vbool4_t test_vmfeq_vf_f32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfeq_vv_f64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfeq.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfeq.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmfeq_vv_f64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -321,7 +321,7 @@ vbool64_t test_vmfeq_vv_f64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfeq_vf_f64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfeq.mask.nxv1f64.f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfeq.mask.nxv1f64.f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmfeq_vf_f64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -332,7 +332,7 @@ vbool64_t test_vmfeq_vf_f64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfeq_vv_f64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfeq.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfeq.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmfeq_vv_f64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -344,7 +344,7 @@ vbool32_t test_vmfeq_vv_f64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfeq_vf_f64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfeq.mask.nxv2f64.f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfeq.mask.nxv2f64.f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmfeq_vf_f64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -355,7 +355,7 @@ vbool32_t test_vmfeq_vf_f64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfeq_vv_f64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfeq.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfeq.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmfeq_vv_f64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -367,7 +367,7 @@ vbool16_t test_vmfeq_vv_f64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfeq_vf_f64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfeq.mask.nxv4f64.f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfeq.mask.nxv4f64.f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmfeq_vf_f64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -378,7 +378,7 @@ vbool16_t test_vmfeq_vf_f64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfeq_vv_f64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfeq.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfeq.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmfeq_vv_f64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -390,7 +390,7 @@ vbool8_t test_vmfeq_vv_f64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfeq_vf_f64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfeq.mask.nxv8f64.f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfeq.mask.nxv8f64.f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmfeq_vf_f64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmfge.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmfge.c
index cb0b456..e4624ab 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmfge.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmfge.c
@@ -188,7 +188,7 @@ vbool8_t test_vmfge_vf_f64m8_b8 (vfloat64m8_t op1, double op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vmfge_vv_f32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfge.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfge.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmfge_vv_f32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vfloat32mf2_t op1, vfloat32mf2_t op2, size_t vl) {
@@ -198,7 +198,7 @@ vbool64_t test_vmfge_vv_f32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vmfge_vf_f32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfge.mask.nxv1f32.f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfge.mask.nxv1f32.f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmfge_vf_f32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vfloat32mf2_t op1, float op2, size_t vl) {
@@ -208,7 +208,7 @@ vbool64_t test_vmfge_vf_f32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vmfge_vv_f32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfge.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfge.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmfge_vv_f32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vfloat32m1_t op1, vfloat32m1_t op2, size_t vl) {
@@ -218,7 +218,7 @@ vbool32_t test_vmfge_vv_f32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vmfge_vf_f32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfge.mask.nxv2f32.f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfge.mask.nxv2f32.f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmfge_vf_f32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vfloat32m1_t op1, float op2, size_t vl) {
@@ -228,7 +228,7 @@ vbool32_t test_vmfge_vf_f32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vmfge_vv_f32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfge.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfge.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmfge_vv_f32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vfloat32m2_t op1, vfloat32m2_t op2, size_t vl) {
@@ -238,7 +238,7 @@ vbool16_t test_vmfge_vv_f32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vmfge_vf_f32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfge.mask.nxv4f32.f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfge.mask.nxv4f32.f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmfge_vf_f32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vfloat32m2_t op1, float op2, size_t vl) {
@@ -248,7 +248,7 @@ vbool16_t test_vmfge_vf_f32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vmfge_vv_f32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfge.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfge.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmfge_vv_f32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vfloat32m4_t op1, vfloat32m4_t op2, size_t vl) {
@@ -258,7 +258,7 @@ vbool8_t test_vmfge_vv_f32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vfloat32m4
 //
 // CHECK-RV64-LABEL: @test_vmfge_vf_f32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfge.mask.nxv8f32.f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfge.mask.nxv8f32.f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmfge_vf_f32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vfloat32m4_t op1, float op2, size_t vl) {
@@ -268,7 +268,7 @@ vbool8_t test_vmfge_vf_f32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vfloat32m4
 //
 // CHECK-RV64-LABEL: @test_vmfge_vv_f32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmfge.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmfge.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmfge_vv_f32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vfloat32m8_t op1, vfloat32m8_t op2, size_t vl) {
@@ -278,7 +278,7 @@ vbool4_t test_vmfge_vv_f32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vfloat32m8
 //
 // CHECK-RV64-LABEL: @test_vmfge_vf_f32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmfge.mask.nxv16f32.f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmfge.mask.nxv16f32.f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmfge_vf_f32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vfloat32m8_t op1, float op2, size_t vl) {
@@ -288,7 +288,7 @@ vbool4_t test_vmfge_vf_f32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vfloat32m8
 //
 // CHECK-RV64-LABEL: @test_vmfge_vv_f64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfge.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfge.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmfge_vv_f64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vfloat64m1_t op1, vfloat64m1_t op2, size_t vl) {
@@ -298,7 +298,7 @@ vbool64_t test_vmfge_vv_f64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vmfge_vf_f64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfge.mask.nxv1f64.f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfge.mask.nxv1f64.f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmfge_vf_f64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vfloat64m1_t op1, double op2, size_t vl) {
@@ -308,7 +308,7 @@ vbool64_t test_vmfge_vf_f64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vmfge_vv_f64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfge.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfge.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmfge_vv_f64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vfloat64m2_t op1, vfloat64m2_t op2, size_t vl) {
@@ -318,7 +318,7 @@ vbool32_t test_vmfge_vv_f64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vmfge_vf_f64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfge.mask.nxv2f64.f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfge.mask.nxv2f64.f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmfge_vf_f64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vfloat64m2_t op1, double op2, size_t vl) {
@@ -328,7 +328,7 @@ vbool32_t test_vmfge_vf_f64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vmfge_vv_f64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfge.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfge.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmfge_vv_f64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vfloat64m4_t op1, vfloat64m4_t op2, size_t vl) {
@@ -338,7 +338,7 @@ vbool16_t test_vmfge_vv_f64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vmfge_vf_f64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfge.mask.nxv4f64.f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfge.mask.nxv4f64.f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmfge_vf_f64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vfloat64m4_t op1, double op2, size_t vl) {
@@ -348,7 +348,7 @@ vbool16_t test_vmfge_vf_f64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vmfge_vv_f64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfge.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfge.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmfge_vv_f64m8_b8_m (vbool8_t mask, vbool8_t maskedoff, vfloat64m8_t op1, vfloat64m8_t op2, size_t vl) {
@@ -358,7 +358,7 @@ vbool8_t test_vmfge_vv_f64m8_b8_m (vbool8_t mask, vbool8_t maskedoff, vfloat64m8
 //
 // CHECK-RV64-LABEL: @test_vmfge_vf_f64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfge.mask.nxv8f64.f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfge.mask.nxv8f64.f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmfge_vf_f64m8_b8_m (vbool8_t mask, vbool8_t maskedoff, vfloat64m8_t op1, double op2, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmfgt.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmfgt.c
index 889df32..2cf1abcc 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmfgt.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmfgt.c
@@ -188,7 +188,7 @@ vbool8_t test_vmfgt_vf_f64m8_b8 (vfloat64m8_t op1, double op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vmfgt_vv_f32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfgt.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfgt.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmfgt_vv_f32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vfloat32mf2_t op1, vfloat32mf2_t op2, size_t vl) {
@@ -198,7 +198,7 @@ vbool64_t test_vmfgt_vv_f32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vmfgt_vf_f32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfgt.mask.nxv1f32.f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfgt.mask.nxv1f32.f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmfgt_vf_f32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vfloat32mf2_t op1, float op2, size_t vl) {
@@ -208,7 +208,7 @@ vbool64_t test_vmfgt_vf_f32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vfloa
 //
 // CHECK-RV64-LABEL: @test_vmfgt_vv_f32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfgt.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfgt.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmfgt_vv_f32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vfloat32m1_t op1, vfloat32m1_t op2, size_t vl) {
@@ -218,7 +218,7 @@ vbool32_t test_vmfgt_vv_f32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vmfgt_vf_f32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfgt.mask.nxv2f32.f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfgt.mask.nxv2f32.f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmfgt_vf_f32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vfloat32m1_t op1, float op2, size_t vl) {
@@ -228,7 +228,7 @@ vbool32_t test_vmfgt_vf_f32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vmfgt_vv_f32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfgt.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfgt.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmfgt_vv_f32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vfloat32m2_t op1, vfloat32m2_t op2, size_t vl) {
@@ -238,7 +238,7 @@ vbool16_t test_vmfgt_vv_f32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vmfgt_vf_f32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfgt.mask.nxv4f32.f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfgt.mask.nxv4f32.f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmfgt_vf_f32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vfloat32m2_t op1, float op2, size_t vl) {
@@ -248,7 +248,7 @@ vbool16_t test_vmfgt_vf_f32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vmfgt_vv_f32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfgt.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfgt.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmfgt_vv_f32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vfloat32m4_t op1, vfloat32m4_t op2, size_t vl) {
@@ -258,7 +258,7 @@ vbool8_t test_vmfgt_vv_f32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vfloat32m4
 //
 // CHECK-RV64-LABEL: @test_vmfgt_vf_f32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfgt.mask.nxv8f32.f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfgt.mask.nxv8f32.f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmfgt_vf_f32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vfloat32m4_t op1, float op2, size_t vl) {
@@ -268,7 +268,7 @@ vbool8_t test_vmfgt_vf_f32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vfloat32m4
 //
 // CHECK-RV64-LABEL: @test_vmfgt_vv_f32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmfgt.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmfgt.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmfgt_vv_f32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vfloat32m8_t op1, vfloat32m8_t op2, size_t vl) {
@@ -278,7 +278,7 @@ vbool4_t test_vmfgt_vv_f32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vfloat32m8
 //
 // CHECK-RV64-LABEL: @test_vmfgt_vf_f32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmfgt.mask.nxv16f32.f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmfgt.mask.nxv16f32.f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmfgt_vf_f32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vfloat32m8_t op1, float op2, size_t vl) {
@@ -288,7 +288,7 @@ vbool4_t test_vmfgt_vf_f32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vfloat32m8
 //
 // CHECK-RV64-LABEL: @test_vmfgt_vv_f64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfgt.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfgt.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmfgt_vv_f64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vfloat64m1_t op1, vfloat64m1_t op2, size_t vl) {
@@ -298,7 +298,7 @@ vbool64_t test_vmfgt_vv_f64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vmfgt_vf_f64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfgt.mask.nxv1f64.f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfgt.mask.nxv1f64.f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmfgt_vf_f64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vfloat64m1_t op1, double op2, size_t vl) {
@@ -308,7 +308,7 @@ vbool64_t test_vmfgt_vf_f64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vmfgt_vv_f64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfgt.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfgt.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmfgt_vv_f64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vfloat64m2_t op1, vfloat64m2_t op2, size_t vl) {
@@ -318,7 +318,7 @@ vbool32_t test_vmfgt_vv_f64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vmfgt_vf_f64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfgt.mask.nxv2f64.f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfgt.mask.nxv2f64.f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmfgt_vf_f64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vfloat64m2_t op1, double op2, size_t vl) {
@@ -328,7 +328,7 @@ vbool32_t test_vmfgt_vf_f64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vmfgt_vv_f64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfgt.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfgt.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmfgt_vv_f64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vfloat64m4_t op1, vfloat64m4_t op2, size_t vl) {
@@ -338,7 +338,7 @@ vbool16_t test_vmfgt_vv_f64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vmfgt_vf_f64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfgt.mask.nxv4f64.f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfgt.mask.nxv4f64.f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmfgt_vf_f64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vfloat64m4_t op1, double op2, size_t vl) {
@@ -348,7 +348,7 @@ vbool16_t test_vmfgt_vf_f64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vfloat
 //
 // CHECK-RV64-LABEL: @test_vmfgt_vv_f64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfgt.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfgt.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmfgt_vv_f64m8_b8_m (vbool8_t mask, vbool8_t maskedoff, vfloat64m8_t op1, vfloat64m8_t op2, size_t vl) {
@@ -358,7 +358,7 @@ vbool8_t test_vmfgt_vv_f64m8_b8_m (vbool8_t mask, vbool8_t maskedoff, vfloat64m8
 //
 // CHECK-RV64-LABEL: @test_vmfgt_vf_f64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfgt.mask.nxv8f64.f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfgt.mask.nxv8f64.f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmfgt_vf_f64m8_b8_m (vbool8_t mask, vbool8_t maskedoff, vfloat64m8_t op1, double op2, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmfle.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmfle.c
index 3cfeeaa..fe16113 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmfle.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmfle.c
@@ -194,7 +194,7 @@ vbool8_t test_vmfle_vf_f64m8_b8(vfloat64m8_t op1, double op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vmfle_vv_f32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfle.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfle.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmfle_vv_f32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -206,7 +206,7 @@ vbool64_t test_vmfle_vv_f32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfle_vf_f32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfle.mask.nxv1f32.f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfle.mask.nxv1f32.f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmfle_vf_f32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -217,7 +217,7 @@ vbool64_t test_vmfle_vf_f32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfle_vv_f32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfle.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfle.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmfle_vv_f32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -229,7 +229,7 @@ vbool32_t test_vmfle_vv_f32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfle_vf_f32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfle.mask.nxv2f32.f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfle.mask.nxv2f32.f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmfle_vf_f32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -240,7 +240,7 @@ vbool32_t test_vmfle_vf_f32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfle_vv_f32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfle.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfle.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmfle_vv_f32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -252,7 +252,7 @@ vbool16_t test_vmfle_vv_f32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfle_vf_f32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfle.mask.nxv4f32.f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfle.mask.nxv4f32.f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmfle_vf_f32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -263,7 +263,7 @@ vbool16_t test_vmfle_vf_f32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfle_vv_f32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfle.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfle.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmfle_vv_f32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -275,7 +275,7 @@ vbool8_t test_vmfle_vv_f32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfle_vf_f32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfle.mask.nxv8f32.f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfle.mask.nxv8f32.f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmfle_vf_f32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -286,7 +286,7 @@ vbool8_t test_vmfle_vf_f32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfle_vv_f32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmfle.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmfle.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmfle_vv_f32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -298,7 +298,7 @@ vbool4_t test_vmfle_vv_f32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfle_vf_f32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmfle.mask.nxv16f32.f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmfle.mask.nxv16f32.f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmfle_vf_f32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -309,7 +309,7 @@ vbool4_t test_vmfle_vf_f32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfle_vv_f64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfle.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfle.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmfle_vv_f64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -321,7 +321,7 @@ vbool64_t test_vmfle_vv_f64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfle_vf_f64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfle.mask.nxv1f64.f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfle.mask.nxv1f64.f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmfle_vf_f64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -332,7 +332,7 @@ vbool64_t test_vmfle_vf_f64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfle_vv_f64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfle.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfle.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmfle_vv_f64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -344,7 +344,7 @@ vbool32_t test_vmfle_vv_f64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfle_vf_f64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfle.mask.nxv2f64.f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfle.mask.nxv2f64.f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmfle_vf_f64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -355,7 +355,7 @@ vbool32_t test_vmfle_vf_f64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfle_vv_f64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfle.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfle.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmfle_vv_f64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -367,7 +367,7 @@ vbool16_t test_vmfle_vv_f64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfle_vf_f64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfle.mask.nxv4f64.f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfle.mask.nxv4f64.f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmfle_vf_f64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -378,7 +378,7 @@ vbool16_t test_vmfle_vf_f64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfle_vv_f64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfle.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfle.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmfle_vv_f64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -390,7 +390,7 @@ vbool8_t test_vmfle_vv_f64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfle_vf_f64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfle.mask.nxv8f64.f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfle.mask.nxv8f64.f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmfle_vf_f64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmflt.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmflt.c
index d1c5fd5..231433d 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmflt.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmflt.c
@@ -194,7 +194,7 @@ vbool8_t test_vmflt_vf_f64m8_b8(vfloat64m8_t op1, double op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vmflt_vv_f32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmflt.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmflt.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmflt_vv_f32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -206,7 +206,7 @@ vbool64_t test_vmflt_vv_f32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmflt_vf_f32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmflt.mask.nxv1f32.f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmflt.mask.nxv1f32.f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmflt_vf_f32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -217,7 +217,7 @@ vbool64_t test_vmflt_vf_f32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmflt_vv_f32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmflt.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmflt.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmflt_vv_f32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -229,7 +229,7 @@ vbool32_t test_vmflt_vv_f32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmflt_vf_f32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmflt.mask.nxv2f32.f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmflt.mask.nxv2f32.f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmflt_vf_f32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -240,7 +240,7 @@ vbool32_t test_vmflt_vf_f32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmflt_vv_f32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmflt.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmflt.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmflt_vv_f32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -252,7 +252,7 @@ vbool16_t test_vmflt_vv_f32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmflt_vf_f32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmflt.mask.nxv4f32.f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmflt.mask.nxv4f32.f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmflt_vf_f32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -263,7 +263,7 @@ vbool16_t test_vmflt_vf_f32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmflt_vv_f32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmflt.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmflt.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmflt_vv_f32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -275,7 +275,7 @@ vbool8_t test_vmflt_vv_f32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmflt_vf_f32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmflt.mask.nxv8f32.f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmflt.mask.nxv8f32.f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmflt_vf_f32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -286,7 +286,7 @@ vbool8_t test_vmflt_vf_f32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmflt_vv_f32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmflt.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmflt.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmflt_vv_f32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -298,7 +298,7 @@ vbool4_t test_vmflt_vv_f32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmflt_vf_f32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmflt.mask.nxv16f32.f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmflt.mask.nxv16f32.f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmflt_vf_f32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -309,7 +309,7 @@ vbool4_t test_vmflt_vf_f32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmflt_vv_f64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmflt.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmflt.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmflt_vv_f64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -321,7 +321,7 @@ vbool64_t test_vmflt_vv_f64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmflt_vf_f64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmflt.mask.nxv1f64.f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmflt.mask.nxv1f64.f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmflt_vf_f64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -332,7 +332,7 @@ vbool64_t test_vmflt_vf_f64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmflt_vv_f64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmflt.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmflt.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmflt_vv_f64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -344,7 +344,7 @@ vbool32_t test_vmflt_vv_f64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmflt_vf_f64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmflt.mask.nxv2f64.f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmflt.mask.nxv2f64.f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmflt_vf_f64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -355,7 +355,7 @@ vbool32_t test_vmflt_vf_f64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmflt_vv_f64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmflt.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmflt.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmflt_vv_f64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -367,7 +367,7 @@ vbool16_t test_vmflt_vv_f64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmflt_vf_f64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmflt.mask.nxv4f64.f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmflt.mask.nxv4f64.f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmflt_vf_f64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -378,7 +378,7 @@ vbool16_t test_vmflt_vf_f64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmflt_vv_f64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmflt.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmflt.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmflt_vv_f64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -390,7 +390,7 @@ vbool8_t test_vmflt_vv_f64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmflt_vf_f64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmflt.mask.nxv8f64.f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmflt.mask.nxv8f64.f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmflt_vf_f64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmfne.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmfne.c
index 5426e25..7a14a8e 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmfne.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmfne.c
@@ -194,7 +194,7 @@ vbool8_t test_vmfne_vf_f64m8_b8(vfloat64m8_t op1, double op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vmfne_vv_f32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfne.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfne.mask.nxv1f32.nxv1f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x float> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmfne_vv_f32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -206,7 +206,7 @@ vbool64_t test_vmfne_vv_f32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfne_vf_f32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfne.mask.nxv1f32.f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfne.mask.nxv1f32.f32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmfne_vf_f32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -217,7 +217,7 @@ vbool64_t test_vmfne_vf_f32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfne_vv_f32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfne.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfne.mask.nxv2f32.nxv2f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x float> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmfne_vv_f32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -229,7 +229,7 @@ vbool32_t test_vmfne_vv_f32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfne_vf_f32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfne.mask.nxv2f32.f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfne.mask.nxv2f32.f32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmfne_vf_f32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -240,7 +240,7 @@ vbool32_t test_vmfne_vf_f32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfne_vv_f32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfne.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfne.mask.nxv4f32.nxv4f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmfne_vv_f32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -252,7 +252,7 @@ vbool16_t test_vmfne_vv_f32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfne_vf_f32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfne.mask.nxv4f32.f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfne.mask.nxv4f32.f32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmfne_vf_f32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -263,7 +263,7 @@ vbool16_t test_vmfne_vf_f32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfne_vv_f32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfne.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfne.mask.nxv8f32.nxv8f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x float> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmfne_vv_f32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -275,7 +275,7 @@ vbool8_t test_vmfne_vv_f32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfne_vf_f32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfne.mask.nxv8f32.f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfne.mask.nxv8f32.f32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmfne_vf_f32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -286,7 +286,7 @@ vbool8_t test_vmfne_vf_f32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfne_vv_f32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmfne.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmfne.mask.nxv16f32.nxv16f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x float> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmfne_vv_f32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -298,7 +298,7 @@ vbool4_t test_vmfne_vv_f32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfne_vf_f32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmfne.mask.nxv16f32.f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmfne.mask.nxv16f32.f32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], float [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmfne_vf_f32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -309,7 +309,7 @@ vbool4_t test_vmfne_vf_f32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfne_vv_f64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfne.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfne.mask.nxv1f64.nxv1f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x double> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmfne_vv_f64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -321,7 +321,7 @@ vbool64_t test_vmfne_vv_f64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfne_vf_f64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfne.mask.nxv1f64.f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmfne.mask.nxv1f64.f64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmfne_vf_f64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -332,7 +332,7 @@ vbool64_t test_vmfne_vf_f64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfne_vv_f64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfne.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfne.mask.nxv2f64.nxv2f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmfne_vv_f64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -344,7 +344,7 @@ vbool32_t test_vmfne_vv_f64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfne_vf_f64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfne.mask.nxv2f64.f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmfne.mask.nxv2f64.f64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmfne_vf_f64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -355,7 +355,7 @@ vbool32_t test_vmfne_vf_f64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfne_vv_f64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfne.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfne.mask.nxv4f64.nxv4f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x double> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmfne_vv_f64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -367,7 +367,7 @@ vbool16_t test_vmfne_vv_f64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfne_vf_f64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfne.mask.nxv4f64.f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmfne.mask.nxv4f64.f64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmfne_vf_f64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -378,7 +378,7 @@ vbool16_t test_vmfne_vf_f64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfne_vv_f64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfne.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfne.mask.nxv8f64.nxv8f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x double> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmfne_vv_f64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -390,7 +390,7 @@ vbool8_t test_vmfne_vv_f64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmfne_vf_f64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfne.mask.nxv8f64.f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR6]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmfne.mask.nxv8f64.f64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], double [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmfne_vf_f64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmin.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmin.c
index c8d26b4..7b78825 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmin.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmin.c
@@ -887,7 +887,7 @@ vuint64m8_t test_vminu_vx_u64m8(vuint64m8_t op1, uint64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmin.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmin.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vmin_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, vint8mf8_t op2, size_t vl) {
@@ -897,7 +897,7 @@ vint8mf8_t test_vmin_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmin.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmin.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vmin_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, int8_t op2, size_t vl) {
@@ -907,7 +907,7 @@ vint8mf8_t test_vmin_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmin.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmin.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vmin_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, vint8mf4_t op2, size_t vl) {
@@ -917,7 +917,7 @@ vint8mf4_t test_vmin_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmin.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmin.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vmin_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, int8_t op2, size_t vl) {
@@ -927,7 +927,7 @@ vint8mf4_t test_vmin_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmin.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmin.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vmin_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, vint8mf2_t op2, size_t vl) {
@@ -937,7 +937,7 @@ vint8mf2_t test_vmin_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmin.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmin.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vmin_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, int8_t op2, size_t vl) {
@@ -947,7 +947,7 @@ vint8mf2_t test_vmin_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmin.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmin.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vmin_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, vint8m1_t op2, size_t vl) {
@@ -957,7 +957,7 @@ vint8m1_t test_vmin_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmin.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmin.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vmin_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, int8_t op2, size_t vl) {
@@ -967,7 +967,7 @@ vint8m1_t test_vmin_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmin.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmin.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vmin_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, vint8m2_t op2, size_t vl) {
@@ -977,7 +977,7 @@ vint8m2_t test_vmin_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmin.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmin.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vmin_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, int8_t op2, size_t vl) {
@@ -987,7 +987,7 @@ vint8m2_t test_vmin_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmin.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmin.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vmin_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, vint8m4_t op2, size_t vl) {
@@ -997,7 +997,7 @@ vint8m4_t test_vmin_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmin.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmin.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vmin_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, int8_t op2, size_t vl) {
@@ -1007,7 +1007,7 @@ vint8m4_t test_vmin_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmin.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmin.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vmin_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, vint8m8_t op2, size_t vl) {
@@ -1017,7 +1017,7 @@ vint8m8_t test_vmin_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmin.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmin.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vmin_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, int8_t op2, size_t vl) {
@@ -1027,7 +1027,7 @@ vint8m8_t test_vmin_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmin.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmin.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vmin_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, vint16mf4_t op2, size_t vl) {
@@ -1037,7 +1037,7 @@ vint16mf4_t test_vmin_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmin.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmin.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vmin_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, int16_t op2, size_t vl) {
@@ -1047,7 +1047,7 @@ vint16mf4_t test_vmin_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmin.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmin.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vmin_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, vint16mf2_t op2, size_t vl) {
@@ -1057,7 +1057,7 @@ vint16mf2_t test_vmin_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmin.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmin.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vmin_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, int16_t op2, size_t vl) {
@@ -1067,7 +1067,7 @@ vint16mf2_t test_vmin_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmin.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmin.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vmin_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, vint16m1_t op2, size_t vl) {
@@ -1077,7 +1077,7 @@ vint16m1_t test_vmin_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmin.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmin.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vmin_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, int16_t op2, size_t vl) {
@@ -1087,7 +1087,7 @@ vint16m1_t test_vmin_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmin.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmin.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vmin_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, vint16m2_t op2, size_t vl) {
@@ -1097,7 +1097,7 @@ vint16m2_t test_vmin_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmin.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmin.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vmin_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, int16_t op2, size_t vl) {
@@ -1107,7 +1107,7 @@ vint16m2_t test_vmin_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmin.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmin.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vmin_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, vint16m4_t op2, size_t vl) {
@@ -1117,7 +1117,7 @@ vint16m4_t test_vmin_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmin.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmin.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vmin_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, int16_t op2, size_t vl) {
@@ -1127,7 +1127,7 @@ vint16m4_t test_vmin_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmin.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmin.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vmin_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, vint16m8_t op2, size_t vl) {
@@ -1137,7 +1137,7 @@ vint16m8_t test_vmin_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmin.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmin.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vmin_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, int16_t op2, size_t vl) {
@@ -1147,7 +1147,7 @@ vint16m8_t test_vmin_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmin.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmin.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vmin_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, vint32mf2_t op2, size_t vl) {
@@ -1157,7 +1157,7 @@ vint32mf2_t test_vmin_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmin.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmin.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vmin_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, int32_t op2, size_t vl) {
@@ -1167,7 +1167,7 @@ vint32mf2_t test_vmin_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmin.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmin.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vmin_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, vint32m1_t op2, size_t vl) {
@@ -1177,7 +1177,7 @@ vint32m1_t test_vmin_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmin.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmin.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vmin_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, int32_t op2, size_t vl) {
@@ -1187,7 +1187,7 @@ vint32m1_t test_vmin_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmin.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmin.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vmin_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, vint32m2_t op2, size_t vl) {
@@ -1197,7 +1197,7 @@ vint32m2_t test_vmin_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmin.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmin.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vmin_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, int32_t op2, size_t vl) {
@@ -1207,7 +1207,7 @@ vint32m2_t test_vmin_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmin.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmin.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vmin_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, vint32m4_t op2, size_t vl) {
@@ -1217,7 +1217,7 @@ vint32m4_t test_vmin_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmin.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmin.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vmin_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, int32_t op2, size_t vl) {
@@ -1227,7 +1227,7 @@ vint32m4_t test_vmin_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmin.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmin.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vmin_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, vint32m8_t op2, size_t vl) {
@@ -1237,7 +1237,7 @@ vint32m8_t test_vmin_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmin.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmin.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vmin_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, int32_t op2, size_t vl) {
@@ -1247,7 +1247,7 @@ vint32m8_t test_vmin_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmin.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmin.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vmin_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, vint64m1_t op2, size_t vl) {
@@ -1257,7 +1257,7 @@ vint64m1_t test_vmin_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmin.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmin.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vmin_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, int64_t op2, size_t vl) {
@@ -1267,7 +1267,7 @@ vint64m1_t test_vmin_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmin.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmin.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vmin_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, vint64m2_t op2, size_t vl) {
@@ -1277,7 +1277,7 @@ vint64m2_t test_vmin_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmin.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmin.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vmin_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, int64_t op2, size_t vl) {
@@ -1287,7 +1287,7 @@ vint64m2_t test_vmin_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmin.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmin.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vmin_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, vint64m4_t op2, size_t vl) {
@@ -1297,7 +1297,7 @@ vint64m4_t test_vmin_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmin.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmin.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vmin_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, int64_t op2, size_t vl) {
@@ -1307,7 +1307,7 @@ vint64m4_t test_vmin_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmin.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmin.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vmin_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, vint64m8_t op2, size_t vl) {
@@ -1317,7 +1317,7 @@ vint64m8_t test_vmin_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vmin_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmin.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmin.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vmin_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, int64_t op2, size_t vl) {
@@ -1327,7 +1327,7 @@ vint64m8_t test_vmin_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vminu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vminu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vminu_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -1337,7 +1337,7 @@ vuint8mf8_t test_vminu_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vminu.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vminu.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vminu_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, uint8_t op2, size_t vl) {
@@ -1347,7 +1347,7 @@ vuint8mf8_t test_vminu_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vminu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vminu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vminu_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, vuint8mf4_t op2, size_t vl) {
@@ -1357,7 +1357,7 @@ vuint8mf4_t test_vminu_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vminu.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vminu.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vminu_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, uint8_t op2, size_t vl) {
@@ -1367,7 +1367,7 @@ vuint8mf4_t test_vminu_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vminu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vminu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vminu_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, vuint8mf2_t op2, size_t vl) {
@@ -1377,7 +1377,7 @@ vuint8mf2_t test_vminu_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vminu.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vminu.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vminu_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, uint8_t op2, size_t vl) {
@@ -1387,7 +1387,7 @@ vuint8mf2_t test_vminu_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vminu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vminu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vminu_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, vuint8m1_t op2, size_t vl) {
@@ -1397,7 +1397,7 @@ vuint8m1_t test_vminu_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vminu.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vminu.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vminu_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, uint8_t op2, size_t vl) {
@@ -1407,7 +1407,7 @@ vuint8m1_t test_vminu_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vminu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vminu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vminu_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, vuint8m2_t op2, size_t vl) {
@@ -1417,7 +1417,7 @@ vuint8m2_t test_vminu_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vminu.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vminu.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vminu_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, uint8_t op2, size_t vl) {
@@ -1427,7 +1427,7 @@ vuint8m2_t test_vminu_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vminu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vminu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vminu_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, vuint8m4_t op2, size_t vl) {
@@ -1437,7 +1437,7 @@ vuint8m4_t test_vminu_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vminu.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vminu.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vminu_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, uint8_t op2, size_t vl) {
@@ -1447,7 +1447,7 @@ vuint8m4_t test_vminu_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vminu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vminu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vminu_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, vuint8m8_t op2, size_t vl) {
@@ -1457,7 +1457,7 @@ vuint8m8_t test_vminu_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vminu.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vminu.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vminu_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, uint8_t op2, size_t vl) {
@@ -1467,7 +1467,7 @@ vuint8m8_t test_vminu_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vminu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vminu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vminu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, vuint16mf4_t op2, size_t vl) {
@@ -1477,7 +1477,7 @@ vuint16mf4_t test_vminu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vminu.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vminu.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vminu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, uint16_t op2, size_t vl) {
@@ -1487,7 +1487,7 @@ vuint16mf4_t test_vminu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vminu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vminu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vminu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, vuint16mf2_t op2, size_t vl) {
@@ -1497,7 +1497,7 @@ vuint16mf2_t test_vminu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vminu.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vminu.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vminu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, uint16_t op2, size_t vl) {
@@ -1507,7 +1507,7 @@ vuint16mf2_t test_vminu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vminu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vminu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vminu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, vuint16m1_t op2, size_t vl) {
@@ -1517,7 +1517,7 @@ vuint16m1_t test_vminu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vminu.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vminu.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vminu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, uint16_t op2, size_t vl) {
@@ -1527,7 +1527,7 @@ vuint16m1_t test_vminu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vminu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vminu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vminu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, vuint16m2_t op2, size_t vl) {
@@ -1537,7 +1537,7 @@ vuint16m2_t test_vminu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vminu.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vminu.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vminu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, uint16_t op2, size_t vl) {
@@ -1547,7 +1547,7 @@ vuint16m2_t test_vminu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vminu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vminu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vminu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, vuint16m4_t op2, size_t vl) {
@@ -1557,7 +1557,7 @@ vuint16m4_t test_vminu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vminu.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vminu.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vminu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, uint16_t op2, size_t vl) {
@@ -1567,7 +1567,7 @@ vuint16m4_t test_vminu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vminu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vminu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vminu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, vuint16m8_t op2, size_t vl) {
@@ -1577,7 +1577,7 @@ vuint16m8_t test_vminu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vminu.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vminu.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vminu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, uint16_t op2, size_t vl) {
@@ -1587,7 +1587,7 @@ vuint16m8_t test_vminu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vminu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vminu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vminu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, vuint32mf2_t op2, size_t vl) {
@@ -1597,7 +1597,7 @@ vuint32mf2_t test_vminu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vminu.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vminu.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vminu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, uint32_t op2, size_t vl) {
@@ -1607,7 +1607,7 @@ vuint32mf2_t test_vminu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vminu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vminu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vminu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, vuint32m1_t op2, size_t vl) {
@@ -1617,7 +1617,7 @@ vuint32m1_t test_vminu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vminu.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vminu.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vminu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, uint32_t op2, size_t vl) {
@@ -1627,7 +1627,7 @@ vuint32m1_t test_vminu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vminu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vminu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vminu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, vuint32m2_t op2, size_t vl) {
@@ -1637,7 +1637,7 @@ vuint32m2_t test_vminu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vminu.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vminu.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vminu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, uint32_t op2, size_t vl) {
@@ -1647,7 +1647,7 @@ vuint32m2_t test_vminu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vminu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vminu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vminu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, vuint32m4_t op2, size_t vl) {
@@ -1657,7 +1657,7 @@ vuint32m4_t test_vminu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vminu.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vminu.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vminu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, uint32_t op2, size_t vl) {
@@ -1667,7 +1667,7 @@ vuint32m4_t test_vminu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vminu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vminu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vminu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, vuint32m8_t op2, size_t vl) {
@@ -1677,7 +1677,7 @@ vuint32m8_t test_vminu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vminu.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vminu.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vminu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, uint32_t op2, size_t vl) {
@@ -1687,7 +1687,7 @@ vuint32m8_t test_vminu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vminu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vminu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vminu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, vuint64m1_t op2, size_t vl) {
@@ -1697,7 +1697,7 @@ vuint64m1_t test_vminu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vminu.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vminu.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vminu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, uint64_t op2, size_t vl) {
@@ -1707,7 +1707,7 @@ vuint64m1_t test_vminu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vminu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vminu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vminu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, vuint64m2_t op2, size_t vl) {
@@ -1717,7 +1717,7 @@ vuint64m2_t test_vminu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vminu.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vminu.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vminu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, uint64_t op2, size_t vl) {
@@ -1727,7 +1727,7 @@ vuint64m2_t test_vminu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vminu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vminu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vminu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, vuint64m4_t op2, size_t vl) {
@@ -1737,7 +1737,7 @@ vuint64m4_t test_vminu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vminu.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vminu.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vminu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, uint64_t op2, size_t vl) {
@@ -1747,7 +1747,7 @@ vuint64m4_t test_vminu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vminu_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vminu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vminu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vminu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, vuint64m8_t op2, size_t vl) {
@@ -1757,7 +1757,7 @@ vuint64m8_t test_vminu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vminu_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vminu.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vminu.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vminu_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, uint64_t op2, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsbf.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsbf.c
index be8576c8..551b0d8 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsbf.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsbf.c
@@ -63,7 +63,7 @@ vbool64_t test_vmsbf_m_b64(vbool64_t op1, size_t vl) { return vmsbf(op1, vl); }
 //
 // CHECK-RV64-LABEL: @test_vmsbf_m_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsbf.mask.nxv64i1.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i1> [[OP1:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsbf.mask.nxv64i1.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i1> [[OP1:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmsbf_m_b1_m(vbool1_t mask, vbool1_t maskedoff, vbool1_t op1,
@@ -74,7 +74,7 @@ vbool1_t test_vmsbf_m_b1_m(vbool1_t mask, vbool1_t maskedoff, vbool1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmsbf_m_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsbf.mask.nxv32i1.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i1> [[OP1:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsbf.mask.nxv32i1.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i1> [[OP1:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsbf_m_b2_m(vbool2_t mask, vbool2_t maskedoff, vbool2_t op1,
@@ -85,7 +85,7 @@ vbool2_t test_vmsbf_m_b2_m(vbool2_t mask, vbool2_t maskedoff, vbool2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmsbf_m_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsbf.mask.nxv16i1.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i1> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsbf.mask.nxv16i1.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i1> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsbf_m_b4_m(vbool4_t mask, vbool4_t maskedoff, vbool4_t op1,
@@ -96,7 +96,7 @@ vbool4_t test_vmsbf_m_b4_m(vbool4_t mask, vbool4_t maskedoff, vbool4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmsbf_m_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsbf.mask.nxv8i1.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsbf.mask.nxv8i1.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsbf_m_b8_m(vbool8_t mask, vbool8_t maskedoff, vbool8_t op1,
@@ -107,7 +107,7 @@ vbool8_t test_vmsbf_m_b8_m(vbool8_t mask, vbool8_t maskedoff, vbool8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmsbf_m_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsbf.mask.nxv4i1.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsbf.mask.nxv4i1.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsbf_m_b16_m(vbool16_t mask, vbool16_t maskedoff, vbool16_t op1,
@@ -118,7 +118,7 @@ vbool16_t test_vmsbf_m_b16_m(vbool16_t mask, vbool16_t maskedoff, vbool16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmsbf_m_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsbf.mask.nxv2i1.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsbf.mask.nxv2i1.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsbf_m_b32_m(vbool32_t mask, vbool32_t maskedoff, vbool32_t op1,
@@ -129,7 +129,7 @@ vbool32_t test_vmsbf_m_b32_m(vbool32_t mask, vbool32_t maskedoff, vbool32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmsbf_m_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsbf.mask.nxv1i1.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsbf.mask.nxv1i1.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsbf_m_b64_m(vbool64_t mask, vbool64_t maskedoff, vbool64_t op1,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmseq.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmseq.c
index 8a3e495..f0b4841 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmseq.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmseq.c
@@ -893,7 +893,7 @@ vbool8_t test_vmseq_vx_u64m8_b8(vuint64m8_t op1, uint64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmseq_vv_i8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -904,7 +904,7 @@ vbool64_t test_vmseq_vv_i8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmseq_vx_i8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -915,7 +915,7 @@ vbool64_t test_vmseq_vx_i8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmseq_vv_i8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -926,7 +926,7 @@ vbool32_t test_vmseq_vv_i8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmseq_vx_i8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -937,7 +937,7 @@ vbool32_t test_vmseq_vx_i8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmseq_vv_i8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -948,7 +948,7 @@ vbool16_t test_vmseq_vv_i8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmseq_vx_i8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -959,7 +959,7 @@ vbool16_t test_vmseq_vx_i8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmseq_vv_i8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -970,7 +970,7 @@ vbool8_t test_vmseq_vv_i8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmseq_vx_i8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -981,7 +981,7 @@ vbool8_t test_vmseq_vx_i8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmseq_vv_i8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -992,7 +992,7 @@ vbool4_t test_vmseq_vv_i8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmseq_vx_i8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1003,7 +1003,7 @@ vbool4_t test_vmseq_vx_i8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmseq.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmseq.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmseq_vv_i8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1014,7 +1014,7 @@ vbool2_t test_vmseq_vv_i8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmseq.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmseq.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmseq_vx_i8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1025,7 +1025,7 @@ vbool2_t test_vmseq_vx_i8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmseq.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmseq.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmseq_vv_i8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
@@ -1036,7 +1036,7 @@ vbool1_t test_vmseq_vv_i8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmseq.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmseq.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmseq_vx_i8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
@@ -1047,7 +1047,7 @@ vbool1_t test_vmseq_vx_i8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmseq_vv_i16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1059,7 +1059,7 @@ vbool64_t test_vmseq_vv_i16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmseq_vx_i16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1070,7 +1070,7 @@ vbool64_t test_vmseq_vx_i16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmseq_vv_i16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1082,7 +1082,7 @@ vbool32_t test_vmseq_vv_i16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmseq_vx_i16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1093,7 +1093,7 @@ vbool32_t test_vmseq_vx_i16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmseq_vv_i16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1104,7 +1104,7 @@ vbool16_t test_vmseq_vv_i16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmseq_vx_i16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1115,7 +1115,7 @@ vbool16_t test_vmseq_vx_i16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmseq_vv_i16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1126,7 +1126,7 @@ vbool8_t test_vmseq_vv_i16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmseq_vx_i16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1137,7 +1137,7 @@ vbool8_t test_vmseq_vx_i16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmseq_vv_i16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1148,7 +1148,7 @@ vbool4_t test_vmseq_vv_i16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmseq_vx_i16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1159,7 +1159,7 @@ vbool4_t test_vmseq_vx_i16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmseq.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmseq.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmseq_vv_i16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1170,7 +1170,7 @@ vbool2_t test_vmseq_vv_i16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmseq.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmseq.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmseq_vx_i16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1181,7 +1181,7 @@ vbool2_t test_vmseq_vx_i16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmseq_vv_i32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1193,7 +1193,7 @@ vbool64_t test_vmseq_vv_i32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmseq_vx_i32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1204,7 +1204,7 @@ vbool64_t test_vmseq_vx_i32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmseq_vv_i32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1215,7 +1215,7 @@ vbool32_t test_vmseq_vv_i32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmseq_vx_i32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1226,7 +1226,7 @@ vbool32_t test_vmseq_vx_i32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmseq_vv_i32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1237,7 +1237,7 @@ vbool16_t test_vmseq_vv_i32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmseq_vx_i32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1248,7 +1248,7 @@ vbool16_t test_vmseq_vx_i32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmseq_vv_i32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1259,7 +1259,7 @@ vbool8_t test_vmseq_vv_i32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmseq_vx_i32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1270,7 +1270,7 @@ vbool8_t test_vmseq_vx_i32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmseq_vv_i32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1281,7 +1281,7 @@ vbool4_t test_vmseq_vv_i32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmseq_vx_i32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1292,7 +1292,7 @@ vbool4_t test_vmseq_vx_i32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmseq_vv_i64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1303,7 +1303,7 @@ vbool64_t test_vmseq_vv_i64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmseq_vx_i64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1314,7 +1314,7 @@ vbool64_t test_vmseq_vx_i64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmseq_vv_i64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1325,7 +1325,7 @@ vbool32_t test_vmseq_vv_i64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmseq_vx_i64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1336,7 +1336,7 @@ vbool32_t test_vmseq_vx_i64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmseq_vv_i64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1347,7 +1347,7 @@ vbool16_t test_vmseq_vv_i64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmseq_vx_i64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1358,7 +1358,7 @@ vbool16_t test_vmseq_vx_i64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_i64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmseq_vv_i64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1369,7 +1369,7 @@ vbool8_t test_vmseq_vv_i64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_i64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmseq_vx_i64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1380,7 +1380,7 @@ vbool8_t test_vmseq_vx_i64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmseq_vv_u8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1392,7 +1392,7 @@ vbool64_t test_vmseq_vv_u8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmseq_vx_u8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1403,7 +1403,7 @@ vbool64_t test_vmseq_vx_u8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmseq_vv_u8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1415,7 +1415,7 @@ vbool32_t test_vmseq_vv_u8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmseq_vx_u8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1426,7 +1426,7 @@ vbool32_t test_vmseq_vx_u8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmseq_vv_u8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1438,7 +1438,7 @@ vbool16_t test_vmseq_vv_u8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmseq_vx_u8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1449,7 +1449,7 @@ vbool16_t test_vmseq_vx_u8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmseq_vv_u8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1460,7 +1460,7 @@ vbool8_t test_vmseq_vv_u8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmseq_vx_u8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1471,7 +1471,7 @@ vbool8_t test_vmseq_vx_u8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmseq_vv_u8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1482,7 +1482,7 @@ vbool4_t test_vmseq_vv_u8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmseq_vx_u8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1493,7 +1493,7 @@ vbool4_t test_vmseq_vx_u8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmseq.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmseq.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmseq_vv_u8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1504,7 +1504,7 @@ vbool2_t test_vmseq_vv_u8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmseq.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmseq.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmseq_vx_u8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1515,7 +1515,7 @@ vbool2_t test_vmseq_vx_u8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmseq.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmseq.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmseq_vv_u8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
@@ -1526,7 +1526,7 @@ vbool1_t test_vmseq_vv_u8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmseq.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmseq.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmseq_vx_u8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
@@ -1537,7 +1537,7 @@ vbool1_t test_vmseq_vx_u8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmseq_vv_u16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1549,7 +1549,7 @@ vbool64_t test_vmseq_vv_u16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmseq_vx_u16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1561,7 +1561,7 @@ vbool64_t test_vmseq_vx_u16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmseq_vv_u16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1573,7 +1573,7 @@ vbool32_t test_vmseq_vv_u16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmseq_vx_u16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1585,7 +1585,7 @@ vbool32_t test_vmseq_vx_u16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmseq_vv_u16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1597,7 +1597,7 @@ vbool16_t test_vmseq_vv_u16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmseq_vx_u16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1608,7 +1608,7 @@ vbool16_t test_vmseq_vx_u16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmseq_vv_u16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1619,7 +1619,7 @@ vbool8_t test_vmseq_vv_u16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmseq_vx_u16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1630,7 +1630,7 @@ vbool8_t test_vmseq_vx_u16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmseq_vv_u16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1641,7 +1641,7 @@ vbool4_t test_vmseq_vv_u16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmseq_vx_u16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1652,7 +1652,7 @@ vbool4_t test_vmseq_vx_u16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmseq.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmseq.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmseq_vv_u16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1663,7 +1663,7 @@ vbool2_t test_vmseq_vv_u16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmseq.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmseq.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmseq_vx_u16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1674,7 +1674,7 @@ vbool2_t test_vmseq_vx_u16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmseq_vv_u32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1686,7 +1686,7 @@ vbool64_t test_vmseq_vv_u32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmseq_vx_u32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1698,7 +1698,7 @@ vbool64_t test_vmseq_vx_u32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmseq_vv_u32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1710,7 +1710,7 @@ vbool32_t test_vmseq_vv_u32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmseq_vx_u32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1721,7 +1721,7 @@ vbool32_t test_vmseq_vx_u32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmseq_vv_u32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1733,7 +1733,7 @@ vbool16_t test_vmseq_vv_u32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmseq_vx_u32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1744,7 +1744,7 @@ vbool16_t test_vmseq_vx_u32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmseq_vv_u32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1755,7 +1755,7 @@ vbool8_t test_vmseq_vv_u32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmseq_vx_u32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1766,7 +1766,7 @@ vbool8_t test_vmseq_vx_u32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmseq_vv_u32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1777,7 +1777,7 @@ vbool4_t test_vmseq_vv_u32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmseq_vx_u32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1788,7 +1788,7 @@ vbool4_t test_vmseq_vx_u32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmseq_vv_u64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1800,7 +1800,7 @@ vbool64_t test_vmseq_vv_u64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmseq_vx_u64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1811,7 +1811,7 @@ vbool64_t test_vmseq_vx_u64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmseq_vv_u64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1823,7 +1823,7 @@ vbool32_t test_vmseq_vv_u64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmseq_vx_u64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1834,7 +1834,7 @@ vbool32_t test_vmseq_vx_u64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmseq_vv_u64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1846,7 +1846,7 @@ vbool16_t test_vmseq_vv_u64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmseq_vx_u64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1857,7 +1857,7 @@ vbool16_t test_vmseq_vx_u64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vv_u64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmseq_vv_u64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1868,7 +1868,7 @@ vbool8_t test_vmseq_vv_u64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmseq_vx_u64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmseq_vx_u64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsge.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsge.c
index d01222e..4a990e8 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsge.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsge.c
@@ -887,7 +887,7 @@ vbool8_t test_vmsgeu_vx_u64m8_b8 (vuint64m8_t op1, uint64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsge_vv_i8mf8_b64_m (vbool64_t mask, vbool64_t maskedoff, vint8mf8_t op1, vint8mf8_t op2, size_t vl) {
@@ -897,7 +897,7 @@ vbool64_t test_vmsge_vv_i8mf8_b64_m (vbool64_t mask, vbool64_t maskedoff, vint8m
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsge_vx_i8mf8_b64_m (vbool64_t mask, vbool64_t maskedoff, vint8mf8_t op1, int8_t op2, size_t vl) {
@@ -907,7 +907,7 @@ vbool64_t test_vmsge_vx_i8mf8_b64_m (vbool64_t mask, vbool64_t maskedoff, vint8m
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsge_vv_i8mf4_b32_m (vbool32_t mask, vbool32_t maskedoff, vint8mf4_t op1, vint8mf4_t op2, size_t vl) {
@@ -917,7 +917,7 @@ vbool32_t test_vmsge_vv_i8mf4_b32_m (vbool32_t mask, vbool32_t maskedoff, vint8m
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsge_vx_i8mf4_b32_m (vbool32_t mask, vbool32_t maskedoff, vint8mf4_t op1, int8_t op2, size_t vl) {
@@ -927,7 +927,7 @@ vbool32_t test_vmsge_vx_i8mf4_b32_m (vbool32_t mask, vbool32_t maskedoff, vint8m
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsge_vv_i8mf2_b16_m (vbool16_t mask, vbool16_t maskedoff, vint8mf2_t op1, vint8mf2_t op2, size_t vl) {
@@ -937,7 +937,7 @@ vbool16_t test_vmsge_vv_i8mf2_b16_m (vbool16_t mask, vbool16_t maskedoff, vint8m
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsge_vx_i8mf2_b16_m (vbool16_t mask, vbool16_t maskedoff, vint8mf2_t op1, int8_t op2, size_t vl) {
@@ -947,7 +947,7 @@ vbool16_t test_vmsge_vx_i8mf2_b16_m (vbool16_t mask, vbool16_t maskedoff, vint8m
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsge_vv_i8m1_b8_m (vbool8_t mask, vbool8_t maskedoff, vint8m1_t op1, vint8m1_t op2, size_t vl) {
@@ -957,7 +957,7 @@ vbool8_t test_vmsge_vv_i8m1_b8_m (vbool8_t mask, vbool8_t maskedoff, vint8m1_t o
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsge_vx_i8m1_b8_m (vbool8_t mask, vbool8_t maskedoff, vint8m1_t op1, int8_t op2, size_t vl) {
@@ -967,7 +967,7 @@ vbool8_t test_vmsge_vx_i8m1_b8_m (vbool8_t mask, vbool8_t maskedoff, vint8m1_t o
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsge.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsge.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsge_vv_i8m2_b4_m (vbool4_t mask, vbool4_t maskedoff, vint8m2_t op1, vint8m2_t op2, size_t vl) {
@@ -977,7 +977,7 @@ vbool4_t test_vmsge_vv_i8m2_b4_m (vbool4_t mask, vbool4_t maskedoff, vint8m2_t o
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsge.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsge.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsge_vx_i8m2_b4_m (vbool4_t mask, vbool4_t maskedoff, vint8m2_t op1, int8_t op2, size_t vl) {
@@ -987,7 +987,7 @@ vbool4_t test_vmsge_vx_i8m2_b4_m (vbool4_t mask, vbool4_t maskedoff, vint8m2_t o
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsge.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsge.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsge_vv_i8m4_b2_m (vbool2_t mask, vbool2_t maskedoff, vint8m4_t op1, vint8m4_t op2, size_t vl) {
@@ -997,7 +997,7 @@ vbool2_t test_vmsge_vv_i8m4_b2_m (vbool2_t mask, vbool2_t maskedoff, vint8m4_t o
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsge.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsge.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsge_vx_i8m4_b2_m (vbool2_t mask, vbool2_t maskedoff, vint8m4_t op1, int8_t op2, size_t vl) {
@@ -1007,7 +1007,7 @@ vbool2_t test_vmsge_vx_i8m4_b2_m (vbool2_t mask, vbool2_t maskedoff, vint8m4_t o
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsge.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsge.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmsge_vv_i8m8_b1_m (vbool1_t mask, vbool1_t maskedoff, vint8m8_t op1, vint8m8_t op2, size_t vl) {
@@ -1017,7 +1017,7 @@ vbool1_t test_vmsge_vv_i8m8_b1_m (vbool1_t mask, vbool1_t maskedoff, vint8m8_t o
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsge.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsge.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmsge_vx_i8m8_b1_m (vbool1_t mask, vbool1_t maskedoff, vint8m8_t op1, int8_t op2, size_t vl) {
@@ -1027,7 +1027,7 @@ vbool1_t test_vmsge_vx_i8m8_b1_m (vbool1_t mask, vbool1_t maskedoff, vint8m8_t o
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsge_vv_i16mf4_b64_m (vbool64_t mask, vbool64_t maskedoff, vint16mf4_t op1, vint16mf4_t op2, size_t vl) {
@@ -1037,7 +1037,7 @@ vbool64_t test_vmsge_vv_i16mf4_b64_m (vbool64_t mask, vbool64_t maskedoff, vint1
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsge_vx_i16mf4_b64_m (vbool64_t mask, vbool64_t maskedoff, vint16mf4_t op1, int16_t op2, size_t vl) {
@@ -1047,7 +1047,7 @@ vbool64_t test_vmsge_vx_i16mf4_b64_m (vbool64_t mask, vbool64_t maskedoff, vint1
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsge_vv_i16mf2_b32_m (vbool32_t mask, vbool32_t maskedoff, vint16mf2_t op1, vint16mf2_t op2, size_t vl) {
@@ -1057,7 +1057,7 @@ vbool32_t test_vmsge_vv_i16mf2_b32_m (vbool32_t mask, vbool32_t maskedoff, vint1
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsge_vx_i16mf2_b32_m (vbool32_t mask, vbool32_t maskedoff, vint16mf2_t op1, int16_t op2, size_t vl) {
@@ -1067,7 +1067,7 @@ vbool32_t test_vmsge_vx_i16mf2_b32_m (vbool32_t mask, vbool32_t maskedoff, vint1
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsge_vv_i16m1_b16_m (vbool16_t mask, vbool16_t maskedoff, vint16m1_t op1, vint16m1_t op2, size_t vl) {
@@ -1077,7 +1077,7 @@ vbool16_t test_vmsge_vv_i16m1_b16_m (vbool16_t mask, vbool16_t maskedoff, vint16
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsge_vx_i16m1_b16_m (vbool16_t mask, vbool16_t maskedoff, vint16m1_t op1, int16_t op2, size_t vl) {
@@ -1087,7 +1087,7 @@ vbool16_t test_vmsge_vx_i16m1_b16_m (vbool16_t mask, vbool16_t maskedoff, vint16
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsge_vv_i16m2_b8_m (vbool8_t mask, vbool8_t maskedoff, vint16m2_t op1, vint16m2_t op2, size_t vl) {
@@ -1097,7 +1097,7 @@ vbool8_t test_vmsge_vv_i16m2_b8_m (vbool8_t mask, vbool8_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsge_vx_i16m2_b8_m (vbool8_t mask, vbool8_t maskedoff, vint16m2_t op1, int16_t op2, size_t vl) {
@@ -1107,7 +1107,7 @@ vbool8_t test_vmsge_vx_i16m2_b8_m (vbool8_t mask, vbool8_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsge.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsge.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsge_vv_i16m4_b4_m (vbool4_t mask, vbool4_t maskedoff, vint16m4_t op1, vint16m4_t op2, size_t vl) {
@@ -1117,7 +1117,7 @@ vbool4_t test_vmsge_vv_i16m4_b4_m (vbool4_t mask, vbool4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsge.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsge.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsge_vx_i16m4_b4_m (vbool4_t mask, vbool4_t maskedoff, vint16m4_t op1, int16_t op2, size_t vl) {
@@ -1127,7 +1127,7 @@ vbool4_t test_vmsge_vx_i16m4_b4_m (vbool4_t mask, vbool4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsge.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsge.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsge_vv_i16m8_b2_m (vbool2_t mask, vbool2_t maskedoff, vint16m8_t op1, vint16m8_t op2, size_t vl) {
@@ -1137,7 +1137,7 @@ vbool2_t test_vmsge_vv_i16m8_b2_m (vbool2_t mask, vbool2_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsge.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsge.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsge_vx_i16m8_b2_m (vbool2_t mask, vbool2_t maskedoff, vint16m8_t op1, int16_t op2, size_t vl) {
@@ -1147,7 +1147,7 @@ vbool2_t test_vmsge_vx_i16m8_b2_m (vbool2_t mask, vbool2_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsge_vv_i32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vint32mf2_t op1, vint32mf2_t op2, size_t vl) {
@@ -1157,7 +1157,7 @@ vbool64_t test_vmsge_vv_i32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vint3
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsge_vx_i32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vint32mf2_t op1, int32_t op2, size_t vl) {
@@ -1167,7 +1167,7 @@ vbool64_t test_vmsge_vx_i32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vint3
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsge_vv_i32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vint32m1_t op1, vint32m1_t op2, size_t vl) {
@@ -1177,7 +1177,7 @@ vbool32_t test_vmsge_vv_i32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vint32
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsge_vx_i32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vint32m1_t op1, int32_t op2, size_t vl) {
@@ -1187,7 +1187,7 @@ vbool32_t test_vmsge_vx_i32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vint32
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsge_vv_i32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vint32m2_t op1, vint32m2_t op2, size_t vl) {
@@ -1197,7 +1197,7 @@ vbool16_t test_vmsge_vv_i32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vint32
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsge_vx_i32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vint32m2_t op1, int32_t op2, size_t vl) {
@@ -1207,7 +1207,7 @@ vbool16_t test_vmsge_vx_i32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vint32
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsge_vv_i32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vint32m4_t op1, vint32m4_t op2, size_t vl) {
@@ -1217,7 +1217,7 @@ vbool8_t test_vmsge_vv_i32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsge_vx_i32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vint32m4_t op1, int32_t op2, size_t vl) {
@@ -1227,7 +1227,7 @@ vbool8_t test_vmsge_vx_i32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsge.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsge.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsge_vv_i32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vint32m8_t op1, vint32m8_t op2, size_t vl) {
@@ -1237,7 +1237,7 @@ vbool4_t test_vmsge_vv_i32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsge.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsge.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsge_vx_i32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vint32m8_t op1, int32_t op2, size_t vl) {
@@ -1247,7 +1247,7 @@ vbool4_t test_vmsge_vx_i32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsge_vv_i64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vint64m1_t op1, vint64m1_t op2, size_t vl) {
@@ -1257,7 +1257,7 @@ vbool64_t test_vmsge_vv_i64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vint64
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsge_vx_i64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vint64m1_t op1, int64_t op2, size_t vl) {
@@ -1267,7 +1267,7 @@ vbool64_t test_vmsge_vx_i64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vint64
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsge_vv_i64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vint64m2_t op1, vint64m2_t op2, size_t vl) {
@@ -1277,7 +1277,7 @@ vbool32_t test_vmsge_vv_i64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vint64
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsge_vx_i64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vint64m2_t op1, int64_t op2, size_t vl) {
@@ -1287,7 +1287,7 @@ vbool32_t test_vmsge_vx_i64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vint64
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsge_vv_i64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vint64m4_t op1, vint64m4_t op2, size_t vl) {
@@ -1297,7 +1297,7 @@ vbool16_t test_vmsge_vv_i64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vint64
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsge_vx_i64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vint64m4_t op1, int64_t op2, size_t vl) {
@@ -1307,7 +1307,7 @@ vbool16_t test_vmsge_vx_i64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vint64
 //
 // CHECK-RV64-LABEL: @test_vmsge_vv_i64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsge_vv_i64m8_b8_m (vbool8_t mask, vbool8_t maskedoff, vint64m8_t op1, vint64m8_t op2, size_t vl) {
@@ -1317,7 +1317,7 @@ vbool8_t test_vmsge_vv_i64m8_b8_m (vbool8_t mask, vbool8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vmsge_vx_i64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsge_vx_i64m8_b8_m (vbool8_t mask, vbool8_t maskedoff, vint64m8_t op1, int64_t op2, size_t vl) {
@@ -1327,7 +1327,7 @@ vbool8_t test_vmsge_vx_i64m8_b8_m (vbool8_t mask, vbool8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgeu_vv_u8mf8_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint8mf8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -1337,7 +1337,7 @@ vbool64_t test_vmsgeu_vv_u8mf8_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgeu_vx_u8mf8_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint8mf8_t op1, uint8_t op2, size_t vl) {
@@ -1347,7 +1347,7 @@ vbool64_t test_vmsgeu_vx_u8mf8_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgeu_vv_u8mf4_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint8mf4_t op1, vuint8mf4_t op2, size_t vl) {
@@ -1357,7 +1357,7 @@ vbool32_t test_vmsgeu_vv_u8mf4_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgeu_vx_u8mf4_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint8mf4_t op1, uint8_t op2, size_t vl) {
@@ -1367,7 +1367,7 @@ vbool32_t test_vmsgeu_vx_u8mf4_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgeu_vv_u8mf2_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint8mf2_t op1, vuint8mf2_t op2, size_t vl) {
@@ -1377,7 +1377,7 @@ vbool16_t test_vmsgeu_vv_u8mf2_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgeu_vx_u8mf2_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint8mf2_t op1, uint8_t op2, size_t vl) {
@@ -1387,7 +1387,7 @@ vbool16_t test_vmsgeu_vx_u8mf2_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgeu_vv_u8m1_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint8m1_t op1, vuint8m1_t op2, size_t vl) {
@@ -1397,7 +1397,7 @@ vbool8_t test_vmsgeu_vv_u8m1_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint8m1_t
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgeu_vx_u8m1_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint8m1_t op1, uint8_t op2, size_t vl) {
@@ -1407,7 +1407,7 @@ vbool8_t test_vmsgeu_vx_u8m1_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint8m1_t
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgeu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgeu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsgeu_vv_u8m2_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint8m2_t op1, vuint8m2_t op2, size_t vl) {
@@ -1417,7 +1417,7 @@ vbool4_t test_vmsgeu_vv_u8m2_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint8m2_t
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgeu.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgeu.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsgeu_vx_u8m2_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint8m2_t op1, uint8_t op2, size_t vl) {
@@ -1427,7 +1427,7 @@ vbool4_t test_vmsgeu_vx_u8m2_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint8m2_t
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgeu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgeu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsgeu_vv_u8m4_b2_m (vbool2_t mask, vbool2_t maskedoff, vuint8m4_t op1, vuint8m4_t op2, size_t vl) {
@@ -1437,7 +1437,7 @@ vbool2_t test_vmsgeu_vv_u8m4_b2_m (vbool2_t mask, vbool2_t maskedoff, vuint8m4_t
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgeu.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgeu.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsgeu_vx_u8m4_b2_m (vbool2_t mask, vbool2_t maskedoff, vuint8m4_t op1, uint8_t op2, size_t vl) {
@@ -1447,7 +1447,7 @@ vbool2_t test_vmsgeu_vx_u8m4_b2_m (vbool2_t mask, vbool2_t maskedoff, vuint8m4_t
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsgeu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsgeu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmsgeu_vv_u8m8_b1_m (vbool1_t mask, vbool1_t maskedoff, vuint8m8_t op1, vuint8m8_t op2, size_t vl) {
@@ -1457,7 +1457,7 @@ vbool1_t test_vmsgeu_vv_u8m8_b1_m (vbool1_t mask, vbool1_t maskedoff, vuint8m8_t
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsgeu.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsgeu.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmsgeu_vx_u8m8_b1_m (vbool1_t mask, vbool1_t maskedoff, vuint8m8_t op1, uint8_t op2, size_t vl) {
@@ -1467,7 +1467,7 @@ vbool1_t test_vmsgeu_vx_u8m8_b1_m (vbool1_t mask, vbool1_t maskedoff, vuint8m8_t
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgeu_vv_u16mf4_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint16mf4_t op1, vuint16mf4_t op2, size_t vl) {
@@ -1477,7 +1477,7 @@ vbool64_t test_vmsgeu_vv_u16mf4_b64_m (vbool64_t mask, vbool64_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgeu_vx_u16mf4_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint16mf4_t op1, uint16_t op2, size_t vl) {
@@ -1487,7 +1487,7 @@ vbool64_t test_vmsgeu_vx_u16mf4_b64_m (vbool64_t mask, vbool64_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgeu_vv_u16mf2_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint16mf2_t op1, vuint16mf2_t op2, size_t vl) {
@@ -1497,7 +1497,7 @@ vbool32_t test_vmsgeu_vv_u16mf2_b32_m (vbool32_t mask, vbool32_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgeu_vx_u16mf2_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint16mf2_t op1, uint16_t op2, size_t vl) {
@@ -1507,7 +1507,7 @@ vbool32_t test_vmsgeu_vx_u16mf2_b32_m (vbool32_t mask, vbool32_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgeu_vv_u16m1_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint16m1_t op1, vuint16m1_t op2, size_t vl) {
@@ -1517,7 +1517,7 @@ vbool16_t test_vmsgeu_vv_u16m1_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgeu_vx_u16m1_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint16m1_t op1, uint16_t op2, size_t vl) {
@@ -1527,7 +1527,7 @@ vbool16_t test_vmsgeu_vx_u16m1_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgeu_vv_u16m2_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint16m2_t op1, vuint16m2_t op2, size_t vl) {
@@ -1537,7 +1537,7 @@ vbool8_t test_vmsgeu_vv_u16m2_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint16m2
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgeu_vx_u16m2_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint16m2_t op1, uint16_t op2, size_t vl) {
@@ -1547,7 +1547,7 @@ vbool8_t test_vmsgeu_vx_u16m2_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint16m2
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgeu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgeu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsgeu_vv_u16m4_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint16m4_t op1, vuint16m4_t op2, size_t vl) {
@@ -1557,7 +1557,7 @@ vbool4_t test_vmsgeu_vv_u16m4_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint16m4
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgeu.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgeu.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsgeu_vx_u16m4_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint16m4_t op1, uint16_t op2, size_t vl) {
@@ -1567,7 +1567,7 @@ vbool4_t test_vmsgeu_vx_u16m4_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint16m4
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgeu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgeu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsgeu_vv_u16m8_b2_m (vbool2_t mask, vbool2_t maskedoff, vuint16m8_t op1, vuint16m8_t op2, size_t vl) {
@@ -1577,7 +1577,7 @@ vbool2_t test_vmsgeu_vv_u16m8_b2_m (vbool2_t mask, vbool2_t maskedoff, vuint16m8
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgeu.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgeu.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsgeu_vx_u16m8_b2_m (vbool2_t mask, vbool2_t maskedoff, vuint16m8_t op1, uint16_t op2, size_t vl) {
@@ -1587,7 +1587,7 @@ vbool2_t test_vmsgeu_vx_u16m8_b2_m (vbool2_t mask, vbool2_t maskedoff, vuint16m8
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgeu_vv_u32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint32mf2_t op1, vuint32mf2_t op2, size_t vl) {
@@ -1597,7 +1597,7 @@ vbool64_t test_vmsgeu_vv_u32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgeu_vx_u32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint32mf2_t op1, uint32_t op2, size_t vl) {
@@ -1607,7 +1607,7 @@ vbool64_t test_vmsgeu_vx_u32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgeu_vv_u32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint32m1_t op1, vuint32m1_t op2, size_t vl) {
@@ -1617,7 +1617,7 @@ vbool32_t test_vmsgeu_vv_u32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgeu_vx_u32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint32m1_t op1, uint32_t op2, size_t vl) {
@@ -1627,7 +1627,7 @@ vbool32_t test_vmsgeu_vx_u32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgeu_vv_u32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint32m2_t op1, vuint32m2_t op2, size_t vl) {
@@ -1637,7 +1637,7 @@ vbool16_t test_vmsgeu_vv_u32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgeu_vx_u32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint32m2_t op1, uint32_t op2, size_t vl) {
@@ -1647,7 +1647,7 @@ vbool16_t test_vmsgeu_vx_u32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgeu_vv_u32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint32m4_t op1, vuint32m4_t op2, size_t vl) {
@@ -1657,7 +1657,7 @@ vbool8_t test_vmsgeu_vv_u32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint32m4
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgeu_vx_u32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint32m4_t op1, uint32_t op2, size_t vl) {
@@ -1667,7 +1667,7 @@ vbool8_t test_vmsgeu_vx_u32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint32m4
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgeu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgeu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsgeu_vv_u32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint32m8_t op1, vuint32m8_t op2, size_t vl) {
@@ -1677,7 +1677,7 @@ vbool4_t test_vmsgeu_vv_u32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint32m8
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgeu.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgeu.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsgeu_vx_u32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint32m8_t op1, uint32_t op2, size_t vl) {
@@ -1687,7 +1687,7 @@ vbool4_t test_vmsgeu_vx_u32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint32m8
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgeu_vv_u64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint64m1_t op1, vuint64m1_t op2, size_t vl) {
@@ -1697,7 +1697,7 @@ vbool64_t test_vmsgeu_vv_u64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgeu_vx_u64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint64m1_t op1, uint64_t op2, size_t vl) {
@@ -1707,7 +1707,7 @@ vbool64_t test_vmsgeu_vx_u64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgeu_vv_u64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint64m2_t op1, vuint64m2_t op2, size_t vl) {
@@ -1717,7 +1717,7 @@ vbool32_t test_vmsgeu_vv_u64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgeu_vx_u64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint64m2_t op1, uint64_t op2, size_t vl) {
@@ -1727,7 +1727,7 @@ vbool32_t test_vmsgeu_vx_u64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgeu_vv_u64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint64m4_t op1, vuint64m4_t op2, size_t vl) {
@@ -1737,7 +1737,7 @@ vbool16_t test_vmsgeu_vv_u64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgeu_vx_u64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint64m4_t op1, uint64_t op2, size_t vl) {
@@ -1747,7 +1747,7 @@ vbool16_t test_vmsgeu_vx_u64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vv_u64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgeu_vv_u64m8_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint64m8_t op1, vuint64m8_t op2, size_t vl) {
@@ -1757,7 +1757,7 @@ vbool8_t test_vmsgeu_vv_u64m8_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint64m8
 //
 // CHECK-RV64-LABEL: @test_vmsgeu_vx_u64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgeu_vx_u64m8_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint64m8_t op1, uint64_t op2, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsgt.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsgt.c
index 6b95f0b..634db12 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsgt.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsgt.c
@@ -887,7 +887,7 @@ vbool8_t test_vmsgtu_vx_u64m8_b8 (vuint64m8_t op1, uint64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgt_vv_i8mf8_b64_m (vbool64_t mask, vbool64_t maskedoff, vint8mf8_t op1, vint8mf8_t op2, size_t vl) {
@@ -897,7 +897,7 @@ vbool64_t test_vmsgt_vv_i8mf8_b64_m (vbool64_t mask, vbool64_t maskedoff, vint8m
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgt_vx_i8mf8_b64_m (vbool64_t mask, vbool64_t maskedoff, vint8mf8_t op1, int8_t op2, size_t vl) {
@@ -907,7 +907,7 @@ vbool64_t test_vmsgt_vx_i8mf8_b64_m (vbool64_t mask, vbool64_t maskedoff, vint8m
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgt_vv_i8mf4_b32_m (vbool32_t mask, vbool32_t maskedoff, vint8mf4_t op1, vint8mf4_t op2, size_t vl) {
@@ -917,7 +917,7 @@ vbool32_t test_vmsgt_vv_i8mf4_b32_m (vbool32_t mask, vbool32_t maskedoff, vint8m
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgt_vx_i8mf4_b32_m (vbool32_t mask, vbool32_t maskedoff, vint8mf4_t op1, int8_t op2, size_t vl) {
@@ -927,7 +927,7 @@ vbool32_t test_vmsgt_vx_i8mf4_b32_m (vbool32_t mask, vbool32_t maskedoff, vint8m
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgt_vv_i8mf2_b16_m (vbool16_t mask, vbool16_t maskedoff, vint8mf2_t op1, vint8mf2_t op2, size_t vl) {
@@ -937,7 +937,7 @@ vbool16_t test_vmsgt_vv_i8mf2_b16_m (vbool16_t mask, vbool16_t maskedoff, vint8m
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgt_vx_i8mf2_b16_m (vbool16_t mask, vbool16_t maskedoff, vint8mf2_t op1, int8_t op2, size_t vl) {
@@ -947,7 +947,7 @@ vbool16_t test_vmsgt_vx_i8mf2_b16_m (vbool16_t mask, vbool16_t maskedoff, vint8m
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgt_vv_i8m1_b8_m (vbool8_t mask, vbool8_t maskedoff, vint8m1_t op1, vint8m1_t op2, size_t vl) {
@@ -957,7 +957,7 @@ vbool8_t test_vmsgt_vv_i8m1_b8_m (vbool8_t mask, vbool8_t maskedoff, vint8m1_t o
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgt_vx_i8m1_b8_m (vbool8_t mask, vbool8_t maskedoff, vint8m1_t op1, int8_t op2, size_t vl) {
@@ -967,7 +967,7 @@ vbool8_t test_vmsgt_vx_i8m1_b8_m (vbool8_t mask, vbool8_t maskedoff, vint8m1_t o
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgt.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgt.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsgt_vv_i8m2_b4_m (vbool4_t mask, vbool4_t maskedoff, vint8m2_t op1, vint8m2_t op2, size_t vl) {
@@ -977,7 +977,7 @@ vbool4_t test_vmsgt_vv_i8m2_b4_m (vbool4_t mask, vbool4_t maskedoff, vint8m2_t o
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgt.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgt.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsgt_vx_i8m2_b4_m (vbool4_t mask, vbool4_t maskedoff, vint8m2_t op1, int8_t op2, size_t vl) {
@@ -987,7 +987,7 @@ vbool4_t test_vmsgt_vx_i8m2_b4_m (vbool4_t mask, vbool4_t maskedoff, vint8m2_t o
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgt.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgt.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsgt_vv_i8m4_b2_m (vbool2_t mask, vbool2_t maskedoff, vint8m4_t op1, vint8m4_t op2, size_t vl) {
@@ -997,7 +997,7 @@ vbool2_t test_vmsgt_vv_i8m4_b2_m (vbool2_t mask, vbool2_t maskedoff, vint8m4_t o
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgt.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgt.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsgt_vx_i8m4_b2_m (vbool2_t mask, vbool2_t maskedoff, vint8m4_t op1, int8_t op2, size_t vl) {
@@ -1007,7 +1007,7 @@ vbool2_t test_vmsgt_vx_i8m4_b2_m (vbool2_t mask, vbool2_t maskedoff, vint8m4_t o
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsgt.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsgt.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmsgt_vv_i8m8_b1_m (vbool1_t mask, vbool1_t maskedoff, vint8m8_t op1, vint8m8_t op2, size_t vl) {
@@ -1017,7 +1017,7 @@ vbool1_t test_vmsgt_vv_i8m8_b1_m (vbool1_t mask, vbool1_t maskedoff, vint8m8_t o
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsgt.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsgt.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmsgt_vx_i8m8_b1_m (vbool1_t mask, vbool1_t maskedoff, vint8m8_t op1, int8_t op2, size_t vl) {
@@ -1027,7 +1027,7 @@ vbool1_t test_vmsgt_vx_i8m8_b1_m (vbool1_t mask, vbool1_t maskedoff, vint8m8_t o
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgt_vv_i16mf4_b64_m (vbool64_t mask, vbool64_t maskedoff, vint16mf4_t op1, vint16mf4_t op2, size_t vl) {
@@ -1037,7 +1037,7 @@ vbool64_t test_vmsgt_vv_i16mf4_b64_m (vbool64_t mask, vbool64_t maskedoff, vint1
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgt_vx_i16mf4_b64_m (vbool64_t mask, vbool64_t maskedoff, vint16mf4_t op1, int16_t op2, size_t vl) {
@@ -1047,7 +1047,7 @@ vbool64_t test_vmsgt_vx_i16mf4_b64_m (vbool64_t mask, vbool64_t maskedoff, vint1
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgt_vv_i16mf2_b32_m (vbool32_t mask, vbool32_t maskedoff, vint16mf2_t op1, vint16mf2_t op2, size_t vl) {
@@ -1057,7 +1057,7 @@ vbool32_t test_vmsgt_vv_i16mf2_b32_m (vbool32_t mask, vbool32_t maskedoff, vint1
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgt_vx_i16mf2_b32_m (vbool32_t mask, vbool32_t maskedoff, vint16mf2_t op1, int16_t op2, size_t vl) {
@@ -1067,7 +1067,7 @@ vbool32_t test_vmsgt_vx_i16mf2_b32_m (vbool32_t mask, vbool32_t maskedoff, vint1
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgt_vv_i16m1_b16_m (vbool16_t mask, vbool16_t maskedoff, vint16m1_t op1, vint16m1_t op2, size_t vl) {
@@ -1077,7 +1077,7 @@ vbool16_t test_vmsgt_vv_i16m1_b16_m (vbool16_t mask, vbool16_t maskedoff, vint16
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgt_vx_i16m1_b16_m (vbool16_t mask, vbool16_t maskedoff, vint16m1_t op1, int16_t op2, size_t vl) {
@@ -1087,7 +1087,7 @@ vbool16_t test_vmsgt_vx_i16m1_b16_m (vbool16_t mask, vbool16_t maskedoff, vint16
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgt_vv_i16m2_b8_m (vbool8_t mask, vbool8_t maskedoff, vint16m2_t op1, vint16m2_t op2, size_t vl) {
@@ -1097,7 +1097,7 @@ vbool8_t test_vmsgt_vv_i16m2_b8_m (vbool8_t mask, vbool8_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgt_vx_i16m2_b8_m (vbool8_t mask, vbool8_t maskedoff, vint16m2_t op1, int16_t op2, size_t vl) {
@@ -1107,7 +1107,7 @@ vbool8_t test_vmsgt_vx_i16m2_b8_m (vbool8_t mask, vbool8_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgt.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgt.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsgt_vv_i16m4_b4_m (vbool4_t mask, vbool4_t maskedoff, vint16m4_t op1, vint16m4_t op2, size_t vl) {
@@ -1117,7 +1117,7 @@ vbool4_t test_vmsgt_vv_i16m4_b4_m (vbool4_t mask, vbool4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgt.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgt.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsgt_vx_i16m4_b4_m (vbool4_t mask, vbool4_t maskedoff, vint16m4_t op1, int16_t op2, size_t vl) {
@@ -1127,7 +1127,7 @@ vbool4_t test_vmsgt_vx_i16m4_b4_m (vbool4_t mask, vbool4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgt.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgt.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsgt_vv_i16m8_b2_m (vbool2_t mask, vbool2_t maskedoff, vint16m8_t op1, vint16m8_t op2, size_t vl) {
@@ -1137,7 +1137,7 @@ vbool2_t test_vmsgt_vv_i16m8_b2_m (vbool2_t mask, vbool2_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgt.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgt.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsgt_vx_i16m8_b2_m (vbool2_t mask, vbool2_t maskedoff, vint16m8_t op1, int16_t op2, size_t vl) {
@@ -1147,7 +1147,7 @@ vbool2_t test_vmsgt_vx_i16m8_b2_m (vbool2_t mask, vbool2_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgt_vv_i32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vint32mf2_t op1, vint32mf2_t op2, size_t vl) {
@@ -1157,7 +1157,7 @@ vbool64_t test_vmsgt_vv_i32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vint3
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgt_vx_i32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vint32mf2_t op1, int32_t op2, size_t vl) {
@@ -1167,7 +1167,7 @@ vbool64_t test_vmsgt_vx_i32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vint3
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgt_vv_i32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vint32m1_t op1, vint32m1_t op2, size_t vl) {
@@ -1177,7 +1177,7 @@ vbool32_t test_vmsgt_vv_i32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vint32
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgt_vx_i32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vint32m1_t op1, int32_t op2, size_t vl) {
@@ -1187,7 +1187,7 @@ vbool32_t test_vmsgt_vx_i32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vint32
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgt_vv_i32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vint32m2_t op1, vint32m2_t op2, size_t vl) {
@@ -1197,7 +1197,7 @@ vbool16_t test_vmsgt_vv_i32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vint32
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgt_vx_i32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vint32m2_t op1, int32_t op2, size_t vl) {
@@ -1207,7 +1207,7 @@ vbool16_t test_vmsgt_vx_i32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vint32
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgt_vv_i32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vint32m4_t op1, vint32m4_t op2, size_t vl) {
@@ -1217,7 +1217,7 @@ vbool8_t test_vmsgt_vv_i32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgt_vx_i32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vint32m4_t op1, int32_t op2, size_t vl) {
@@ -1227,7 +1227,7 @@ vbool8_t test_vmsgt_vx_i32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgt.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgt.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsgt_vv_i32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vint32m8_t op1, vint32m8_t op2, size_t vl) {
@@ -1237,7 +1237,7 @@ vbool4_t test_vmsgt_vv_i32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgt.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgt.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsgt_vx_i32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vint32m8_t op1, int32_t op2, size_t vl) {
@@ -1247,7 +1247,7 @@ vbool4_t test_vmsgt_vx_i32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgt_vv_i64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vint64m1_t op1, vint64m1_t op2, size_t vl) {
@@ -1257,7 +1257,7 @@ vbool64_t test_vmsgt_vv_i64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vint64
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgt_vx_i64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vint64m1_t op1, int64_t op2, size_t vl) {
@@ -1267,7 +1267,7 @@ vbool64_t test_vmsgt_vx_i64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vint64
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgt_vv_i64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vint64m2_t op1, vint64m2_t op2, size_t vl) {
@@ -1277,7 +1277,7 @@ vbool32_t test_vmsgt_vv_i64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vint64
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgt_vx_i64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vint64m2_t op1, int64_t op2, size_t vl) {
@@ -1287,7 +1287,7 @@ vbool32_t test_vmsgt_vx_i64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vint64
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgt_vv_i64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vint64m4_t op1, vint64m4_t op2, size_t vl) {
@@ -1297,7 +1297,7 @@ vbool16_t test_vmsgt_vv_i64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vint64
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgt_vx_i64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vint64m4_t op1, int64_t op2, size_t vl) {
@@ -1307,7 +1307,7 @@ vbool16_t test_vmsgt_vx_i64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vint64
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vv_i64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgt_vv_i64m8_b8_m (vbool8_t mask, vbool8_t maskedoff, vint64m8_t op1, vint64m8_t op2, size_t vl) {
@@ -1317,7 +1317,7 @@ vbool8_t test_vmsgt_vv_i64m8_b8_m (vbool8_t mask, vbool8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vmsgt_vx_i64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgt_vx_i64m8_b8_m (vbool8_t mask, vbool8_t maskedoff, vint64m8_t op1, int64_t op2, size_t vl) {
@@ -1327,7 +1327,7 @@ vbool8_t test_vmsgt_vx_i64m8_b8_m (vbool8_t mask, vbool8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgtu_vv_u8mf8_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint8mf8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -1337,7 +1337,7 @@ vbool64_t test_vmsgtu_vv_u8mf8_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgtu_vx_u8mf8_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint8mf8_t op1, uint8_t op2, size_t vl) {
@@ -1347,7 +1347,7 @@ vbool64_t test_vmsgtu_vx_u8mf8_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgtu_vv_u8mf4_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint8mf4_t op1, vuint8mf4_t op2, size_t vl) {
@@ -1357,7 +1357,7 @@ vbool32_t test_vmsgtu_vv_u8mf4_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgtu_vx_u8mf4_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint8mf4_t op1, uint8_t op2, size_t vl) {
@@ -1367,7 +1367,7 @@ vbool32_t test_vmsgtu_vx_u8mf4_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgtu_vv_u8mf2_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint8mf2_t op1, vuint8mf2_t op2, size_t vl) {
@@ -1377,7 +1377,7 @@ vbool16_t test_vmsgtu_vv_u8mf2_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgtu_vx_u8mf2_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint8mf2_t op1, uint8_t op2, size_t vl) {
@@ -1387,7 +1387,7 @@ vbool16_t test_vmsgtu_vx_u8mf2_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgtu_vv_u8m1_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint8m1_t op1, vuint8m1_t op2, size_t vl) {
@@ -1397,7 +1397,7 @@ vbool8_t test_vmsgtu_vv_u8m1_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint8m1_t
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgtu_vx_u8m1_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint8m1_t op1, uint8_t op2, size_t vl) {
@@ -1407,7 +1407,7 @@ vbool8_t test_vmsgtu_vx_u8m1_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint8m1_t
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgtu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgtu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsgtu_vv_u8m2_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint8m2_t op1, vuint8m2_t op2, size_t vl) {
@@ -1417,7 +1417,7 @@ vbool4_t test_vmsgtu_vv_u8m2_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint8m2_t
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgtu.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgtu.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsgtu_vx_u8m2_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint8m2_t op1, uint8_t op2, size_t vl) {
@@ -1427,7 +1427,7 @@ vbool4_t test_vmsgtu_vx_u8m2_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint8m2_t
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgtu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgtu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsgtu_vv_u8m4_b2_m (vbool2_t mask, vbool2_t maskedoff, vuint8m4_t op1, vuint8m4_t op2, size_t vl) {
@@ -1437,7 +1437,7 @@ vbool2_t test_vmsgtu_vv_u8m4_b2_m (vbool2_t mask, vbool2_t maskedoff, vuint8m4_t
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgtu.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgtu.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsgtu_vx_u8m4_b2_m (vbool2_t mask, vbool2_t maskedoff, vuint8m4_t op1, uint8_t op2, size_t vl) {
@@ -1447,7 +1447,7 @@ vbool2_t test_vmsgtu_vx_u8m4_b2_m (vbool2_t mask, vbool2_t maskedoff, vuint8m4_t
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsgtu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsgtu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmsgtu_vv_u8m8_b1_m (vbool1_t mask, vbool1_t maskedoff, vuint8m8_t op1, vuint8m8_t op2, size_t vl) {
@@ -1457,7 +1457,7 @@ vbool1_t test_vmsgtu_vv_u8m8_b1_m (vbool1_t mask, vbool1_t maskedoff, vuint8m8_t
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsgtu.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsgtu.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmsgtu_vx_u8m8_b1_m (vbool1_t mask, vbool1_t maskedoff, vuint8m8_t op1, uint8_t op2, size_t vl) {
@@ -1467,7 +1467,7 @@ vbool1_t test_vmsgtu_vx_u8m8_b1_m (vbool1_t mask, vbool1_t maskedoff, vuint8m8_t
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgtu_vv_u16mf4_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint16mf4_t op1, vuint16mf4_t op2, size_t vl) {
@@ -1477,7 +1477,7 @@ vbool64_t test_vmsgtu_vv_u16mf4_b64_m (vbool64_t mask, vbool64_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgtu_vx_u16mf4_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint16mf4_t op1, uint16_t op2, size_t vl) {
@@ -1487,7 +1487,7 @@ vbool64_t test_vmsgtu_vx_u16mf4_b64_m (vbool64_t mask, vbool64_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgtu_vv_u16mf2_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint16mf2_t op1, vuint16mf2_t op2, size_t vl) {
@@ -1497,7 +1497,7 @@ vbool32_t test_vmsgtu_vv_u16mf2_b32_m (vbool32_t mask, vbool32_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgtu_vx_u16mf2_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint16mf2_t op1, uint16_t op2, size_t vl) {
@@ -1507,7 +1507,7 @@ vbool32_t test_vmsgtu_vx_u16mf2_b32_m (vbool32_t mask, vbool32_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgtu_vv_u16m1_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint16m1_t op1, vuint16m1_t op2, size_t vl) {
@@ -1517,7 +1517,7 @@ vbool16_t test_vmsgtu_vv_u16m1_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgtu_vx_u16m1_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint16m1_t op1, uint16_t op2, size_t vl) {
@@ -1527,7 +1527,7 @@ vbool16_t test_vmsgtu_vx_u16m1_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgtu_vv_u16m2_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint16m2_t op1, vuint16m2_t op2, size_t vl) {
@@ -1537,7 +1537,7 @@ vbool8_t test_vmsgtu_vv_u16m2_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint16m2
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgtu_vx_u16m2_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint16m2_t op1, uint16_t op2, size_t vl) {
@@ -1547,7 +1547,7 @@ vbool8_t test_vmsgtu_vx_u16m2_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint16m2
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgtu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgtu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsgtu_vv_u16m4_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint16m4_t op1, vuint16m4_t op2, size_t vl) {
@@ -1557,7 +1557,7 @@ vbool4_t test_vmsgtu_vv_u16m4_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint16m4
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgtu.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgtu.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsgtu_vx_u16m4_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint16m4_t op1, uint16_t op2, size_t vl) {
@@ -1567,7 +1567,7 @@ vbool4_t test_vmsgtu_vx_u16m4_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint16m4
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgtu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgtu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsgtu_vv_u16m8_b2_m (vbool2_t mask, vbool2_t maskedoff, vuint16m8_t op1, vuint16m8_t op2, size_t vl) {
@@ -1577,7 +1577,7 @@ vbool2_t test_vmsgtu_vv_u16m8_b2_m (vbool2_t mask, vbool2_t maskedoff, vuint16m8
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgtu.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsgtu.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsgtu_vx_u16m8_b2_m (vbool2_t mask, vbool2_t maskedoff, vuint16m8_t op1, uint16_t op2, size_t vl) {
@@ -1587,7 +1587,7 @@ vbool2_t test_vmsgtu_vx_u16m8_b2_m (vbool2_t mask, vbool2_t maskedoff, vuint16m8
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgtu_vv_u32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint32mf2_t op1, vuint32mf2_t op2, size_t vl) {
@@ -1597,7 +1597,7 @@ vbool64_t test_vmsgtu_vv_u32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgtu_vx_u32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint32mf2_t op1, uint32_t op2, size_t vl) {
@@ -1607,7 +1607,7 @@ vbool64_t test_vmsgtu_vx_u32mf2_b64_m (vbool64_t mask, vbool64_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgtu_vv_u32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint32m1_t op1, vuint32m1_t op2, size_t vl) {
@@ -1617,7 +1617,7 @@ vbool32_t test_vmsgtu_vv_u32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgtu_vx_u32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint32m1_t op1, uint32_t op2, size_t vl) {
@@ -1627,7 +1627,7 @@ vbool32_t test_vmsgtu_vx_u32m1_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgtu_vv_u32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint32m2_t op1, vuint32m2_t op2, size_t vl) {
@@ -1637,7 +1637,7 @@ vbool16_t test_vmsgtu_vv_u32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgtu_vx_u32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint32m2_t op1, uint32_t op2, size_t vl) {
@@ -1647,7 +1647,7 @@ vbool16_t test_vmsgtu_vx_u32m2_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgtu_vv_u32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint32m4_t op1, vuint32m4_t op2, size_t vl) {
@@ -1657,7 +1657,7 @@ vbool8_t test_vmsgtu_vv_u32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint32m4
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgtu_vx_u32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint32m4_t op1, uint32_t op2, size_t vl) {
@@ -1667,7 +1667,7 @@ vbool8_t test_vmsgtu_vx_u32m4_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint32m4
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgtu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgtu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsgtu_vv_u32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint32m8_t op1, vuint32m8_t op2, size_t vl) {
@@ -1677,7 +1677,7 @@ vbool4_t test_vmsgtu_vv_u32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint32m8
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgtu.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsgtu.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsgtu_vx_u32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint32m8_t op1, uint32_t op2, size_t vl) {
@@ -1687,7 +1687,7 @@ vbool4_t test_vmsgtu_vx_u32m8_b4_m (vbool4_t mask, vbool4_t maskedoff, vuint32m8
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgtu_vv_u64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint64m1_t op1, vuint64m1_t op2, size_t vl) {
@@ -1697,7 +1697,7 @@ vbool64_t test_vmsgtu_vv_u64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsgtu_vx_u64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint64m1_t op1, uint64_t op2, size_t vl) {
@@ -1707,7 +1707,7 @@ vbool64_t test_vmsgtu_vx_u64m1_b64_m (vbool64_t mask, vbool64_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgtu_vv_u64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint64m2_t op1, vuint64m2_t op2, size_t vl) {
@@ -1717,7 +1717,7 @@ vbool32_t test_vmsgtu_vv_u64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsgtu_vx_u64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint64m2_t op1, uint64_t op2, size_t vl) {
@@ -1727,7 +1727,7 @@ vbool32_t test_vmsgtu_vx_u64m2_b32_m (vbool32_t mask, vbool32_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgtu_vv_u64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint64m4_t op1, vuint64m4_t op2, size_t vl) {
@@ -1737,7 +1737,7 @@ vbool16_t test_vmsgtu_vv_u64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsgtu_vx_u64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint64m4_t op1, uint64_t op2, size_t vl) {
@@ -1747,7 +1747,7 @@ vbool16_t test_vmsgtu_vx_u64m4_b16_m (vbool16_t mask, vbool16_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vv_u64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgtu_vv_u64m8_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint64m8_t op1, vuint64m8_t op2, size_t vl) {
@@ -1757,7 +1757,7 @@ vbool8_t test_vmsgtu_vv_u64m8_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint64m8
 //
 // CHECK-RV64-LABEL: @test_vmsgtu_vx_u64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsgtu_vx_u64m8_b8_m (vbool8_t mask, vbool8_t maskedoff, vuint64m8_t op1, uint64_t op2, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsif.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsif.c
index d337d63a..d655d42 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsif.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsif.c
@@ -63,7 +63,7 @@ vbool64_t test_vmsif_m_b64(vbool64_t op1, size_t vl) { return vmsif(op1, vl); }
 //
 // CHECK-RV64-LABEL: @test_vmsif_m_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsif.mask.nxv64i1.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i1> [[OP1:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsif.mask.nxv64i1.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i1> [[OP1:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmsif_m_b1_m(vbool1_t mask, vbool1_t maskedoff, vbool1_t op1,
@@ -74,7 +74,7 @@ vbool1_t test_vmsif_m_b1_m(vbool1_t mask, vbool1_t maskedoff, vbool1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmsif_m_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsif.mask.nxv32i1.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i1> [[OP1:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsif.mask.nxv32i1.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i1> [[OP1:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsif_m_b2_m(vbool2_t mask, vbool2_t maskedoff, vbool2_t op1,
@@ -85,7 +85,7 @@ vbool2_t test_vmsif_m_b2_m(vbool2_t mask, vbool2_t maskedoff, vbool2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmsif_m_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsif.mask.nxv16i1.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i1> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsif.mask.nxv16i1.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i1> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsif_m_b4_m(vbool4_t mask, vbool4_t maskedoff, vbool4_t op1,
@@ -96,7 +96,7 @@ vbool4_t test_vmsif_m_b4_m(vbool4_t mask, vbool4_t maskedoff, vbool4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmsif_m_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsif.mask.nxv8i1.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsif.mask.nxv8i1.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsif_m_b8_m(vbool8_t mask, vbool8_t maskedoff, vbool8_t op1,
@@ -107,7 +107,7 @@ vbool8_t test_vmsif_m_b8_m(vbool8_t mask, vbool8_t maskedoff, vbool8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmsif_m_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsif.mask.nxv4i1.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsif.mask.nxv4i1.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsif_m_b16_m(vbool16_t mask, vbool16_t maskedoff, vbool16_t op1,
@@ -118,7 +118,7 @@ vbool16_t test_vmsif_m_b16_m(vbool16_t mask, vbool16_t maskedoff, vbool16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmsif_m_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsif.mask.nxv2i1.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsif.mask.nxv2i1.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsif_m_b32_m(vbool32_t mask, vbool32_t maskedoff, vbool32_t op1,
@@ -129,7 +129,7 @@ vbool32_t test_vmsif_m_b32_m(vbool32_t mask, vbool32_t maskedoff, vbool32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmsif_m_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsif.mask.nxv1i1.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsif.mask.nxv1i1.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsif_m_b64_m(vbool64_t mask, vbool64_t maskedoff, vbool64_t op1,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsle.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsle.c
index d185473..12b6637 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsle.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsle.c
@@ -902,7 +902,7 @@ vbool8_t test_vmsleu_vx_u64m8_b8(vuint64m8_t op1, uint64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsle_vv_i8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -913,7 +913,7 @@ vbool64_t test_vmsle_vv_i8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsle_vx_i8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -924,7 +924,7 @@ vbool64_t test_vmsle_vx_i8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsle_vv_i8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -935,7 +935,7 @@ vbool32_t test_vmsle_vv_i8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsle_vx_i8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -946,7 +946,7 @@ vbool32_t test_vmsle_vx_i8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsle_vv_i8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -957,7 +957,7 @@ vbool16_t test_vmsle_vv_i8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsle_vx_i8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -968,7 +968,7 @@ vbool16_t test_vmsle_vx_i8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsle_vv_i8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -979,7 +979,7 @@ vbool8_t test_vmsle_vv_i8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsle_vx_i8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -990,7 +990,7 @@ vbool8_t test_vmsle_vx_i8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsle.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsle.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsle_vv_i8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1001,7 +1001,7 @@ vbool4_t test_vmsle_vv_i8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsle.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsle.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsle_vx_i8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1012,7 +1012,7 @@ vbool4_t test_vmsle_vx_i8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsle.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsle.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsle_vv_i8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1023,7 +1023,7 @@ vbool2_t test_vmsle_vv_i8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsle.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsle.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsle_vx_i8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1034,7 +1034,7 @@ vbool2_t test_vmsle_vx_i8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsle.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsle.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmsle_vv_i8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
@@ -1045,7 +1045,7 @@ vbool1_t test_vmsle_vv_i8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsle.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsle.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmsle_vx_i8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
@@ -1056,7 +1056,7 @@ vbool1_t test_vmsle_vx_i8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsle_vv_i16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1068,7 +1068,7 @@ vbool64_t test_vmsle_vv_i16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsle_vx_i16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1079,7 +1079,7 @@ vbool64_t test_vmsle_vx_i16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsle_vv_i16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1091,7 +1091,7 @@ vbool32_t test_vmsle_vv_i16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsle_vx_i16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1102,7 +1102,7 @@ vbool32_t test_vmsle_vx_i16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsle_vv_i16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1113,7 +1113,7 @@ vbool16_t test_vmsle_vv_i16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsle_vx_i16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1124,7 +1124,7 @@ vbool16_t test_vmsle_vx_i16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsle_vv_i16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1135,7 +1135,7 @@ vbool8_t test_vmsle_vv_i16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsle_vx_i16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1146,7 +1146,7 @@ vbool8_t test_vmsle_vx_i16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsle.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsle.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsle_vv_i16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1157,7 +1157,7 @@ vbool4_t test_vmsle_vv_i16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsle.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsle.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsle_vx_i16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1168,7 +1168,7 @@ vbool4_t test_vmsle_vx_i16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsle.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsle.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsle_vv_i16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1179,7 +1179,7 @@ vbool2_t test_vmsle_vv_i16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsle.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsle.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsle_vx_i16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1190,7 +1190,7 @@ vbool2_t test_vmsle_vx_i16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsle_vv_i32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1202,7 +1202,7 @@ vbool64_t test_vmsle_vv_i32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsle_vx_i32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1213,7 +1213,7 @@ vbool64_t test_vmsle_vx_i32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsle_vv_i32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1224,7 +1224,7 @@ vbool32_t test_vmsle_vv_i32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsle_vx_i32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1235,7 +1235,7 @@ vbool32_t test_vmsle_vx_i32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsle_vv_i32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1246,7 +1246,7 @@ vbool16_t test_vmsle_vv_i32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsle_vx_i32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1257,7 +1257,7 @@ vbool16_t test_vmsle_vx_i32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsle_vv_i32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1268,7 +1268,7 @@ vbool8_t test_vmsle_vv_i32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsle_vx_i32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1279,7 +1279,7 @@ vbool8_t test_vmsle_vx_i32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsle.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsle.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsle_vv_i32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1290,7 +1290,7 @@ vbool4_t test_vmsle_vv_i32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsle.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsle.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsle_vx_i32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1301,7 +1301,7 @@ vbool4_t test_vmsle_vx_i32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsle_vv_i64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1312,7 +1312,7 @@ vbool64_t test_vmsle_vv_i64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsle_vx_i64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1323,7 +1323,7 @@ vbool64_t test_vmsle_vx_i64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsle_vv_i64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1334,7 +1334,7 @@ vbool32_t test_vmsle_vv_i64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsle_vx_i64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1345,7 +1345,7 @@ vbool32_t test_vmsle_vx_i64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsle_vv_i64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1356,7 +1356,7 @@ vbool16_t test_vmsle_vv_i64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsle_vx_i64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1367,7 +1367,7 @@ vbool16_t test_vmsle_vx_i64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vv_i64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsle_vv_i64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1378,7 +1378,7 @@ vbool8_t test_vmsle_vv_i64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsle_vx_i64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsle_vx_i64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1389,7 +1389,7 @@ vbool8_t test_vmsle_vx_i64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsleu_vv_u8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1401,7 +1401,7 @@ vbool64_t test_vmsleu_vv_u8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsleu_vx_u8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1412,7 +1412,7 @@ vbool64_t test_vmsleu_vx_u8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsleu_vv_u8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1424,7 +1424,7 @@ vbool32_t test_vmsleu_vv_u8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsleu_vx_u8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1435,7 +1435,7 @@ vbool32_t test_vmsleu_vx_u8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsleu_vv_u8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1447,7 +1447,7 @@ vbool16_t test_vmsleu_vv_u8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsleu_vx_u8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1458,7 +1458,7 @@ vbool16_t test_vmsleu_vx_u8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsleu_vv_u8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1469,7 +1469,7 @@ vbool8_t test_vmsleu_vv_u8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsleu_vx_u8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1480,7 +1480,7 @@ vbool8_t test_vmsleu_vx_u8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsleu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsleu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsleu_vv_u8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1491,7 +1491,7 @@ vbool4_t test_vmsleu_vv_u8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsleu.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsleu.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsleu_vx_u8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1502,7 +1502,7 @@ vbool4_t test_vmsleu_vx_u8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsleu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsleu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsleu_vv_u8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1513,7 +1513,7 @@ vbool2_t test_vmsleu_vv_u8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsleu.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsleu.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsleu_vx_u8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1524,7 +1524,7 @@ vbool2_t test_vmsleu_vx_u8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsleu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsleu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmsleu_vv_u8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
@@ -1535,7 +1535,7 @@ vbool1_t test_vmsleu_vv_u8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsleu.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsleu.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmsleu_vx_u8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
@@ -1546,7 +1546,7 @@ vbool1_t test_vmsleu_vx_u8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsleu_vv_u16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1558,7 +1558,7 @@ vbool64_t test_vmsleu_vv_u16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsleu_vx_u16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1570,7 +1570,7 @@ vbool64_t test_vmsleu_vx_u16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsleu_vv_u16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1582,7 +1582,7 @@ vbool32_t test_vmsleu_vv_u16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsleu_vx_u16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1594,7 +1594,7 @@ vbool32_t test_vmsleu_vx_u16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsleu_vv_u16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1606,7 +1606,7 @@ vbool16_t test_vmsleu_vv_u16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsleu_vx_u16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1617,7 +1617,7 @@ vbool16_t test_vmsleu_vx_u16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsleu_vv_u16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1629,7 +1629,7 @@ vbool8_t test_vmsleu_vv_u16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsleu_vx_u16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1640,7 +1640,7 @@ vbool8_t test_vmsleu_vx_u16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsleu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsleu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsleu_vv_u16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1652,7 +1652,7 @@ vbool4_t test_vmsleu_vv_u16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsleu.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsleu.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsleu_vx_u16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1663,7 +1663,7 @@ vbool4_t test_vmsleu_vx_u16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsleu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsleu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsleu_vv_u16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1675,7 +1675,7 @@ vbool2_t test_vmsleu_vv_u16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsleu.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsleu.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsleu_vx_u16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1686,7 +1686,7 @@ vbool2_t test_vmsleu_vx_u16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsleu_vv_u32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1698,7 +1698,7 @@ vbool64_t test_vmsleu_vv_u32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsleu_vx_u32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1710,7 +1710,7 @@ vbool64_t test_vmsleu_vx_u32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsleu_vv_u32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1722,7 +1722,7 @@ vbool32_t test_vmsleu_vv_u32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsleu_vx_u32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1733,7 +1733,7 @@ vbool32_t test_vmsleu_vx_u32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsleu_vv_u32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1745,7 +1745,7 @@ vbool16_t test_vmsleu_vv_u32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsleu_vx_u32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1756,7 +1756,7 @@ vbool16_t test_vmsleu_vx_u32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsleu_vv_u32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1768,7 +1768,7 @@ vbool8_t test_vmsleu_vv_u32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsleu_vx_u32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1779,7 +1779,7 @@ vbool8_t test_vmsleu_vx_u32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsleu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsleu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsleu_vv_u32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1791,7 +1791,7 @@ vbool4_t test_vmsleu_vv_u32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsleu.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsleu.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsleu_vx_u32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1802,7 +1802,7 @@ vbool4_t test_vmsleu_vx_u32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsleu_vv_u64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1814,7 +1814,7 @@ vbool64_t test_vmsleu_vv_u64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsleu_vx_u64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1825,7 +1825,7 @@ vbool64_t test_vmsleu_vx_u64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsleu_vv_u64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1837,7 +1837,7 @@ vbool32_t test_vmsleu_vv_u64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsleu_vx_u64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1848,7 +1848,7 @@ vbool32_t test_vmsleu_vx_u64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsleu_vv_u64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1860,7 +1860,7 @@ vbool16_t test_vmsleu_vv_u64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsleu_vx_u64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1871,7 +1871,7 @@ vbool16_t test_vmsleu_vx_u64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vv_u64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsleu_vv_u64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1883,7 +1883,7 @@ vbool8_t test_vmsleu_vv_u64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsleu_vx_u64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsleu_vx_u64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmslt.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmslt.c
index 24a5b00..07b864c 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmslt.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmslt.c
@@ -902,7 +902,7 @@ vbool8_t test_vmsltu_vx_u64m8_b8(vuint64m8_t op1, uint64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmslt_vv_i8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -913,7 +913,7 @@ vbool64_t test_vmslt_vv_i8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmslt_vx_i8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -924,7 +924,7 @@ vbool64_t test_vmslt_vx_i8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmslt_vv_i8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -935,7 +935,7 @@ vbool32_t test_vmslt_vv_i8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmslt_vx_i8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -946,7 +946,7 @@ vbool32_t test_vmslt_vx_i8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmslt_vv_i8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -957,7 +957,7 @@ vbool16_t test_vmslt_vv_i8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmslt_vx_i8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -968,7 +968,7 @@ vbool16_t test_vmslt_vx_i8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmslt_vv_i8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -979,7 +979,7 @@ vbool8_t test_vmslt_vv_i8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmslt_vx_i8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -990,7 +990,7 @@ vbool8_t test_vmslt_vx_i8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmslt.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmslt.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmslt_vv_i8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1001,7 +1001,7 @@ vbool4_t test_vmslt_vv_i8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmslt.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmslt.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmslt_vx_i8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1012,7 +1012,7 @@ vbool4_t test_vmslt_vx_i8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmslt.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmslt.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmslt_vv_i8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1023,7 +1023,7 @@ vbool2_t test_vmslt_vv_i8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmslt.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmslt.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmslt_vx_i8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1034,7 +1034,7 @@ vbool2_t test_vmslt_vx_i8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmslt.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmslt.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmslt_vv_i8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
@@ -1045,7 +1045,7 @@ vbool1_t test_vmslt_vv_i8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmslt.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmslt.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmslt_vx_i8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
@@ -1056,7 +1056,7 @@ vbool1_t test_vmslt_vx_i8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmslt_vv_i16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1068,7 +1068,7 @@ vbool64_t test_vmslt_vv_i16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmslt_vx_i16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1079,7 +1079,7 @@ vbool64_t test_vmslt_vx_i16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmslt_vv_i16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1091,7 +1091,7 @@ vbool32_t test_vmslt_vv_i16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmslt_vx_i16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1102,7 +1102,7 @@ vbool32_t test_vmslt_vx_i16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmslt_vv_i16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1113,7 +1113,7 @@ vbool16_t test_vmslt_vv_i16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmslt_vx_i16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1124,7 +1124,7 @@ vbool16_t test_vmslt_vx_i16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmslt_vv_i16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1135,7 +1135,7 @@ vbool8_t test_vmslt_vv_i16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmslt_vx_i16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1146,7 +1146,7 @@ vbool8_t test_vmslt_vx_i16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmslt.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmslt.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmslt_vv_i16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1157,7 +1157,7 @@ vbool4_t test_vmslt_vv_i16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmslt.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmslt.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmslt_vx_i16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1168,7 +1168,7 @@ vbool4_t test_vmslt_vx_i16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmslt.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmslt.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmslt_vv_i16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1179,7 +1179,7 @@ vbool2_t test_vmslt_vv_i16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmslt.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmslt.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmslt_vx_i16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1190,7 +1190,7 @@ vbool2_t test_vmslt_vx_i16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmslt_vv_i32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1202,7 +1202,7 @@ vbool64_t test_vmslt_vv_i32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmslt_vx_i32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1213,7 +1213,7 @@ vbool64_t test_vmslt_vx_i32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmslt_vv_i32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1224,7 +1224,7 @@ vbool32_t test_vmslt_vv_i32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmslt_vx_i32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1235,7 +1235,7 @@ vbool32_t test_vmslt_vx_i32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmslt_vv_i32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1246,7 +1246,7 @@ vbool16_t test_vmslt_vv_i32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmslt_vx_i32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1257,7 +1257,7 @@ vbool16_t test_vmslt_vx_i32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmslt_vv_i32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1268,7 +1268,7 @@ vbool8_t test_vmslt_vv_i32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmslt_vx_i32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1279,7 +1279,7 @@ vbool8_t test_vmslt_vx_i32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmslt.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmslt.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmslt_vv_i32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1290,7 +1290,7 @@ vbool4_t test_vmslt_vv_i32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmslt.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmslt.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmslt_vx_i32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1301,7 +1301,7 @@ vbool4_t test_vmslt_vx_i32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmslt_vv_i64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1312,7 +1312,7 @@ vbool64_t test_vmslt_vv_i64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmslt_vx_i64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1323,7 +1323,7 @@ vbool64_t test_vmslt_vx_i64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmslt_vv_i64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1334,7 +1334,7 @@ vbool32_t test_vmslt_vv_i64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmslt_vx_i64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1345,7 +1345,7 @@ vbool32_t test_vmslt_vx_i64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmslt_vv_i64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1356,7 +1356,7 @@ vbool16_t test_vmslt_vv_i64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmslt_vx_i64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1367,7 +1367,7 @@ vbool16_t test_vmslt_vx_i64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vv_i64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmslt_vv_i64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1378,7 +1378,7 @@ vbool8_t test_vmslt_vv_i64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmslt_vx_i64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmslt_vx_i64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1389,7 +1389,7 @@ vbool8_t test_vmslt_vx_i64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsltu_vv_u8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1401,7 +1401,7 @@ vbool64_t test_vmsltu_vv_u8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsltu_vx_u8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1412,7 +1412,7 @@ vbool64_t test_vmsltu_vx_u8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsltu_vv_u8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1424,7 +1424,7 @@ vbool32_t test_vmsltu_vv_u8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsltu_vx_u8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1435,7 +1435,7 @@ vbool32_t test_vmsltu_vx_u8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsltu_vv_u8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1447,7 +1447,7 @@ vbool16_t test_vmsltu_vv_u8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsltu_vx_u8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1458,7 +1458,7 @@ vbool16_t test_vmsltu_vx_u8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsltu_vv_u8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1469,7 +1469,7 @@ vbool8_t test_vmsltu_vv_u8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsltu_vx_u8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1480,7 +1480,7 @@ vbool8_t test_vmsltu_vx_u8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsltu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsltu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsltu_vv_u8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1491,7 +1491,7 @@ vbool4_t test_vmsltu_vv_u8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsltu.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsltu.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsltu_vx_u8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1502,7 +1502,7 @@ vbool4_t test_vmsltu_vx_u8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsltu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsltu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsltu_vv_u8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1513,7 +1513,7 @@ vbool2_t test_vmsltu_vv_u8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsltu.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsltu.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsltu_vx_u8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1524,7 +1524,7 @@ vbool2_t test_vmsltu_vx_u8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsltu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsltu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmsltu_vv_u8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
@@ -1535,7 +1535,7 @@ vbool1_t test_vmsltu_vv_u8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsltu.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsltu.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmsltu_vx_u8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
@@ -1546,7 +1546,7 @@ vbool1_t test_vmsltu_vx_u8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsltu_vv_u16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1558,7 +1558,7 @@ vbool64_t test_vmsltu_vv_u16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsltu_vx_u16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1570,7 +1570,7 @@ vbool64_t test_vmsltu_vx_u16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsltu_vv_u16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1582,7 +1582,7 @@ vbool32_t test_vmsltu_vv_u16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsltu_vx_u16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1594,7 +1594,7 @@ vbool32_t test_vmsltu_vx_u16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsltu_vv_u16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1606,7 +1606,7 @@ vbool16_t test_vmsltu_vv_u16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsltu_vx_u16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1617,7 +1617,7 @@ vbool16_t test_vmsltu_vx_u16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsltu_vv_u16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1629,7 +1629,7 @@ vbool8_t test_vmsltu_vv_u16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsltu_vx_u16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1640,7 +1640,7 @@ vbool8_t test_vmsltu_vx_u16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsltu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsltu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsltu_vv_u16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1652,7 +1652,7 @@ vbool4_t test_vmsltu_vv_u16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsltu.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsltu.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsltu_vx_u16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1663,7 +1663,7 @@ vbool4_t test_vmsltu_vx_u16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsltu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsltu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsltu_vv_u16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1675,7 +1675,7 @@ vbool2_t test_vmsltu_vv_u16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsltu.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsltu.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsltu_vx_u16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1686,7 +1686,7 @@ vbool2_t test_vmsltu_vx_u16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsltu_vv_u32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1698,7 +1698,7 @@ vbool64_t test_vmsltu_vv_u32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsltu_vx_u32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1710,7 +1710,7 @@ vbool64_t test_vmsltu_vx_u32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsltu_vv_u32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1722,7 +1722,7 @@ vbool32_t test_vmsltu_vv_u32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsltu_vx_u32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1733,7 +1733,7 @@ vbool32_t test_vmsltu_vx_u32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsltu_vv_u32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1745,7 +1745,7 @@ vbool16_t test_vmsltu_vv_u32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsltu_vx_u32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1756,7 +1756,7 @@ vbool16_t test_vmsltu_vx_u32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsltu_vv_u32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1768,7 +1768,7 @@ vbool8_t test_vmsltu_vv_u32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsltu_vx_u32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1779,7 +1779,7 @@ vbool8_t test_vmsltu_vx_u32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsltu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsltu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsltu_vv_u32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1791,7 +1791,7 @@ vbool4_t test_vmsltu_vv_u32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsltu.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsltu.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsltu_vx_u32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1802,7 +1802,7 @@ vbool4_t test_vmsltu_vx_u32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsltu_vv_u64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1814,7 +1814,7 @@ vbool64_t test_vmsltu_vv_u64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsltu_vx_u64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1825,7 +1825,7 @@ vbool64_t test_vmsltu_vx_u64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsltu_vv_u64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1837,7 +1837,7 @@ vbool32_t test_vmsltu_vv_u64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsltu_vx_u64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1848,7 +1848,7 @@ vbool32_t test_vmsltu_vx_u64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsltu_vv_u64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1860,7 +1860,7 @@ vbool16_t test_vmsltu_vv_u64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsltu_vx_u64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1871,7 +1871,7 @@ vbool16_t test_vmsltu_vx_u64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vv_u64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsltu_vv_u64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1883,7 +1883,7 @@ vbool8_t test_vmsltu_vv_u64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsltu_vx_u64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsltu_vx_u64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsne.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsne.c
index a3fb1cd..61b2d97 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsne.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsne.c
@@ -893,7 +893,7 @@ vbool8_t test_vmsne_vx_u64m8_b8(vuint64m8_t op1, uint64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsne_vv_i8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -904,7 +904,7 @@ vbool64_t test_vmsne_vv_i8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsne_vx_i8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -915,7 +915,7 @@ vbool64_t test_vmsne_vx_i8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsne_vv_i8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -926,7 +926,7 @@ vbool32_t test_vmsne_vv_i8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsne_vx_i8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -937,7 +937,7 @@ vbool32_t test_vmsne_vx_i8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsne_vv_i8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -948,7 +948,7 @@ vbool16_t test_vmsne_vv_i8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsne_vx_i8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -959,7 +959,7 @@ vbool16_t test_vmsne_vx_i8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsne_vv_i8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -970,7 +970,7 @@ vbool8_t test_vmsne_vv_i8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsne_vx_i8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -981,7 +981,7 @@ vbool8_t test_vmsne_vx_i8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsne_vv_i8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -992,7 +992,7 @@ vbool4_t test_vmsne_vv_i8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsne_vx_i8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1003,7 +1003,7 @@ vbool4_t test_vmsne_vx_i8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsne.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsne.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsne_vv_i8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1014,7 +1014,7 @@ vbool2_t test_vmsne_vv_i8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsne.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsne.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsne_vx_i8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1025,7 +1025,7 @@ vbool2_t test_vmsne_vx_i8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsne.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsne.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmsne_vv_i8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
@@ -1036,7 +1036,7 @@ vbool1_t test_vmsne_vv_i8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsne.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsne.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmsne_vx_i8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
@@ -1047,7 +1047,7 @@ vbool1_t test_vmsne_vx_i8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsne_vv_i16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1059,7 +1059,7 @@ vbool64_t test_vmsne_vv_i16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsne_vx_i16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1070,7 +1070,7 @@ vbool64_t test_vmsne_vx_i16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsne_vv_i16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1082,7 +1082,7 @@ vbool32_t test_vmsne_vv_i16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsne_vx_i16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1093,7 +1093,7 @@ vbool32_t test_vmsne_vx_i16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsne_vv_i16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1104,7 +1104,7 @@ vbool16_t test_vmsne_vv_i16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsne_vx_i16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1115,7 +1115,7 @@ vbool16_t test_vmsne_vx_i16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsne_vv_i16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1126,7 +1126,7 @@ vbool8_t test_vmsne_vv_i16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsne_vx_i16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1137,7 +1137,7 @@ vbool8_t test_vmsne_vx_i16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsne_vv_i16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1148,7 +1148,7 @@ vbool4_t test_vmsne_vv_i16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsne_vx_i16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1159,7 +1159,7 @@ vbool4_t test_vmsne_vx_i16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsne.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsne.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsne_vv_i16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1170,7 +1170,7 @@ vbool2_t test_vmsne_vv_i16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsne.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsne.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsne_vx_i16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1181,7 +1181,7 @@ vbool2_t test_vmsne_vx_i16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsne_vv_i32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1193,7 +1193,7 @@ vbool64_t test_vmsne_vv_i32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsne_vx_i32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1204,7 +1204,7 @@ vbool64_t test_vmsne_vx_i32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsne_vv_i32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1215,7 +1215,7 @@ vbool32_t test_vmsne_vv_i32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsne_vx_i32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1226,7 +1226,7 @@ vbool32_t test_vmsne_vx_i32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsne_vv_i32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1237,7 +1237,7 @@ vbool16_t test_vmsne_vv_i32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsne_vx_i32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1248,7 +1248,7 @@ vbool16_t test_vmsne_vx_i32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsne_vv_i32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1259,7 +1259,7 @@ vbool8_t test_vmsne_vv_i32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsne_vx_i32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1270,7 +1270,7 @@ vbool8_t test_vmsne_vx_i32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsne_vv_i32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1281,7 +1281,7 @@ vbool4_t test_vmsne_vv_i32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsne_vx_i32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1292,7 +1292,7 @@ vbool4_t test_vmsne_vx_i32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsne_vv_i64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1303,7 +1303,7 @@ vbool64_t test_vmsne_vv_i64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsne_vx_i64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1314,7 +1314,7 @@ vbool64_t test_vmsne_vx_i64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsne_vv_i64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1325,7 +1325,7 @@ vbool32_t test_vmsne_vv_i64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsne_vx_i64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1336,7 +1336,7 @@ vbool32_t test_vmsne_vx_i64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsne_vv_i64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1347,7 +1347,7 @@ vbool16_t test_vmsne_vv_i64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsne_vx_i64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1358,7 +1358,7 @@ vbool16_t test_vmsne_vx_i64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_i64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsne_vv_i64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1369,7 +1369,7 @@ vbool8_t test_vmsne_vv_i64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_i64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsne_vx_i64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1380,7 +1380,7 @@ vbool8_t test_vmsne_vx_i64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsne_vv_u8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1392,7 +1392,7 @@ vbool64_t test_vmsne_vv_u8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u8mf8_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i8.i8.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsne_vx_u8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1403,7 +1403,7 @@ vbool64_t test_vmsne_vx_u8mf8_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsne_vv_u8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1415,7 +1415,7 @@ vbool32_t test_vmsne_vv_u8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u8mf4_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i8.i8.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsne_vx_u8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1426,7 +1426,7 @@ vbool32_t test_vmsne_vx_u8mf4_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsne_vv_u8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1438,7 +1438,7 @@ vbool16_t test_vmsne_vv_u8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u8mf2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i8.i8.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsne_vx_u8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1449,7 +1449,7 @@ vbool16_t test_vmsne_vx_u8mf2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsne_vv_u8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1460,7 +1460,7 @@ vbool8_t test_vmsne_vv_u8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u8m1_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i8.i8.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsne_vx_u8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1471,7 +1471,7 @@ vbool8_t test_vmsne_vx_u8m1_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsne_vv_u8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1482,7 +1482,7 @@ vbool4_t test_vmsne_vv_u8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u8m2_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i8.i8.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsne_vx_u8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1493,7 +1493,7 @@ vbool4_t test_vmsne_vx_u8m2_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsne.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsne.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsne_vv_u8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1504,7 +1504,7 @@ vbool2_t test_vmsne_vv_u8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u8m4_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsne.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsne.mask.nxv32i8.i8.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsne_vx_u8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1515,7 +1515,7 @@ vbool2_t test_vmsne_vx_u8m4_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsne.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsne.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmsne_vv_u8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
@@ -1526,7 +1526,7 @@ vbool1_t test_vmsne_vv_u8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u8m8_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsne.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsne.mask.nxv64i8.i8.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmsne_vx_u8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
@@ -1537,7 +1537,7 @@ vbool1_t test_vmsne_vx_u8m8_b1_m(vbool1_t mask, vbool1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsne_vv_u16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1549,7 +1549,7 @@ vbool64_t test_vmsne_vv_u16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u16mf4_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i16.i16.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsne_vx_u16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1561,7 +1561,7 @@ vbool64_t test_vmsne_vx_u16mf4_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsne_vv_u16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1573,7 +1573,7 @@ vbool32_t test_vmsne_vv_u16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u16mf2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i16.i16.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsne_vx_u16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1585,7 +1585,7 @@ vbool32_t test_vmsne_vx_u16mf2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsne_vv_u16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1597,7 +1597,7 @@ vbool16_t test_vmsne_vv_u16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u16m1_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i16.i16.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsne_vx_u16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1608,7 +1608,7 @@ vbool16_t test_vmsne_vx_u16m1_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsne_vv_u16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1619,7 +1619,7 @@ vbool8_t test_vmsne_vv_u16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u16m2_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i16.i16.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsne_vx_u16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1630,7 +1630,7 @@ vbool8_t test_vmsne_vx_u16m2_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsne_vv_u16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1641,7 +1641,7 @@ vbool4_t test_vmsne_vv_u16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u16m4_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i16.i16.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsne_vx_u16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1652,7 +1652,7 @@ vbool4_t test_vmsne_vx_u16m4_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsne.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsne.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsne_vv_u16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1663,7 +1663,7 @@ vbool2_t test_vmsne_vv_u16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u16m8_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsne.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsne.mask.nxv32i16.i16.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsne_vx_u16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
@@ -1674,7 +1674,7 @@ vbool2_t test_vmsne_vx_u16m8_b2_m(vbool2_t mask, vbool2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsne_vv_u32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1686,7 +1686,7 @@ vbool64_t test_vmsne_vv_u32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u32mf2_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i32.i32.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsne_vx_u32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1698,7 +1698,7 @@ vbool64_t test_vmsne_vx_u32mf2_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsne_vv_u32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1710,7 +1710,7 @@ vbool32_t test_vmsne_vv_u32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u32m1_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i32.i32.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsne_vx_u32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1721,7 +1721,7 @@ vbool32_t test_vmsne_vx_u32m1_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsne_vv_u32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1733,7 +1733,7 @@ vbool16_t test_vmsne_vv_u32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u32m2_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i32.i32.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsne_vx_u32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1744,7 +1744,7 @@ vbool16_t test_vmsne_vx_u32m2_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsne_vv_u32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1755,7 +1755,7 @@ vbool8_t test_vmsne_vv_u32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u32m4_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i32.i32.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsne_vx_u32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1766,7 +1766,7 @@ vbool8_t test_vmsne_vx_u32m4_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsne_vv_u32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1777,7 +1777,7 @@ vbool4_t test_vmsne_vv_u32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u32m8_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i32.i32.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsne_vx_u32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
@@ -1788,7 +1788,7 @@ vbool4_t test_vmsne_vx_u32m8_b4_m(vbool4_t mask, vbool4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsne_vv_u64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1800,7 +1800,7 @@ vbool64_t test_vmsne_vv_u64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u64m1_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i64.i64.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsne_vx_u64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
@@ -1811,7 +1811,7 @@ vbool64_t test_vmsne_vx_u64m1_b64_m(vbool64_t mask, vbool64_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsne_vv_u64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1823,7 +1823,7 @@ vbool32_t test_vmsne_vv_u64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u64m2_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i64.i64.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsne_vx_u64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
@@ -1834,7 +1834,7 @@ vbool32_t test_vmsne_vx_u64m2_b32_m(vbool32_t mask, vbool32_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsne_vv_u64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1846,7 +1846,7 @@ vbool16_t test_vmsne_vv_u64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u64m4_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i64.i64.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsne_vx_u64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
@@ -1857,7 +1857,7 @@ vbool16_t test_vmsne_vx_u64m4_b16_m(vbool16_t mask, vbool16_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vv_u64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsne_vv_u64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
@@ -1868,7 +1868,7 @@ vbool8_t test_vmsne_vv_u64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vmsne_vx_u64m8_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i64.i64.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsne_vx_u64m8_b8_m(vbool8_t mask, vbool8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsof.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsof.c
index 52f5f97..03cec0e 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsof.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmsof.c
@@ -63,7 +63,7 @@ vbool64_t test_vmsof_m_b64(vbool64_t op1, size_t vl) { return vmsof(op1, vl); }
 //
 // CHECK-RV64-LABEL: @test_vmsof_m_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsof.mask.nxv64i1.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i1> [[OP1:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i1> @llvm.riscv.vmsof.mask.nxv64i1.i64(<vscale x 64 x i1> [[MASKEDOFF:%.*]], <vscale x 64 x i1> [[OP1:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i1> [[TMP0]]
 //
 vbool1_t test_vmsof_m_b1_m(vbool1_t mask, vbool1_t maskedoff, vbool1_t op1,
@@ -74,7 +74,7 @@ vbool1_t test_vmsof_m_b1_m(vbool1_t mask, vbool1_t maskedoff, vbool1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmsof_m_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsof.mask.nxv32i1.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i1> [[OP1:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i1> @llvm.riscv.vmsof.mask.nxv32i1.i64(<vscale x 32 x i1> [[MASKEDOFF:%.*]], <vscale x 32 x i1> [[OP1:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i1> [[TMP0]]
 //
 vbool2_t test_vmsof_m_b2_m(vbool2_t mask, vbool2_t maskedoff, vbool2_t op1,
@@ -85,7 +85,7 @@ vbool2_t test_vmsof_m_b2_m(vbool2_t mask, vbool2_t maskedoff, vbool2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmsof_m_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsof.mask.nxv16i1.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i1> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i1> @llvm.riscv.vmsof.mask.nxv16i1.i64(<vscale x 16 x i1> [[MASKEDOFF:%.*]], <vscale x 16 x i1> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
 //
 vbool4_t test_vmsof_m_b4_m(vbool4_t mask, vbool4_t maskedoff, vbool4_t op1,
@@ -96,7 +96,7 @@ vbool4_t test_vmsof_m_b4_m(vbool4_t mask, vbool4_t maskedoff, vbool4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmsof_m_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsof.mask.nxv8i1.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i1> @llvm.riscv.vmsof.mask.nxv8i1.i64(<vscale x 8 x i1> [[MASKEDOFF:%.*]], <vscale x 8 x i1> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i1> [[TMP0]]
 //
 vbool8_t test_vmsof_m_b8_m(vbool8_t mask, vbool8_t maskedoff, vbool8_t op1,
@@ -107,7 +107,7 @@ vbool8_t test_vmsof_m_b8_m(vbool8_t mask, vbool8_t maskedoff, vbool8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmsof_m_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsof.mask.nxv4i1.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i1> @llvm.riscv.vmsof.mask.nxv4i1.i64(<vscale x 4 x i1> [[MASKEDOFF:%.*]], <vscale x 4 x i1> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i1> [[TMP0]]
 //
 vbool16_t test_vmsof_m_b16_m(vbool16_t mask, vbool16_t maskedoff, vbool16_t op1,
@@ -118,7 +118,7 @@ vbool16_t test_vmsof_m_b16_m(vbool16_t mask, vbool16_t maskedoff, vbool16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmsof_m_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsof.mask.nxv2i1.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i1> @llvm.riscv.vmsof.mask.nxv2i1.i64(<vscale x 2 x i1> [[MASKEDOFF:%.*]], <vscale x 2 x i1> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i1> [[TMP0]]
 //
 vbool32_t test_vmsof_m_b32_m(vbool32_t mask, vbool32_t maskedoff, vbool32_t op1,
@@ -129,7 +129,7 @@ vbool32_t test_vmsof_m_b32_m(vbool32_t mask, vbool32_t maskedoff, vbool32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmsof_m_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsof.mask.nxv1i1.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i1> @llvm.riscv.vmsof.mask.nxv1i1.i64(<vscale x 1 x i1> [[MASKEDOFF:%.*]], <vscale x 1 x i1> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i1> [[TMP0]]
 //
 vbool64_t test_vmsof_m_b64_m(vbool64_t mask, vbool64_t maskedoff, vbool64_t op1,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmul.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmul.c
index 024050f..435ffae 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmul.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vmul.c
@@ -2207,7 +2207,7 @@ vint64m8_t test_vmulhsu_vx_i64m8(vint64m8_t op1, uint64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmul.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmul.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vmul_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, vint8mf8_t op2, size_t vl) {
@@ -2217,7 +2217,7 @@ vint8mf8_t test_vmul_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmul.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmul.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vmul_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, int8_t op2, size_t vl) {
@@ -2227,7 +2227,7 @@ vint8mf8_t test_vmul_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmul.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmul.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vmul_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, vint8mf4_t op2, size_t vl) {
@@ -2237,7 +2237,7 @@ vint8mf4_t test_vmul_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmul.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmul.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vmul_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, int8_t op2, size_t vl) {
@@ -2247,7 +2247,7 @@ vint8mf4_t test_vmul_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmul.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmul.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vmul_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, vint8mf2_t op2, size_t vl) {
@@ -2257,7 +2257,7 @@ vint8mf2_t test_vmul_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmul.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmul.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vmul_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, int8_t op2, size_t vl) {
@@ -2267,7 +2267,7 @@ vint8mf2_t test_vmul_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmul.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmul.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vmul_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, vint8m1_t op2, size_t vl) {
@@ -2277,7 +2277,7 @@ vint8m1_t test_vmul_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmul.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmul.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vmul_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, int8_t op2, size_t vl) {
@@ -2287,7 +2287,7 @@ vint8m1_t test_vmul_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmul.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmul.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vmul_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, vint8m2_t op2, size_t vl) {
@@ -2297,7 +2297,7 @@ vint8m2_t test_vmul_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmul.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmul.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vmul_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, int8_t op2, size_t vl) {
@@ -2307,7 +2307,7 @@ vint8m2_t test_vmul_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmul.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmul.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vmul_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, vint8m4_t op2, size_t vl) {
@@ -2317,7 +2317,7 @@ vint8m4_t test_vmul_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmul.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmul.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vmul_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, int8_t op2, size_t vl) {
@@ -2327,7 +2327,7 @@ vint8m4_t test_vmul_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmul.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmul.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vmul_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, vint8m8_t op2, size_t vl) {
@@ -2337,7 +2337,7 @@ vint8m8_t test_vmul_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmul.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmul.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vmul_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, int8_t op2, size_t vl) {
@@ -2347,7 +2347,7 @@ vint8m8_t test_vmul_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmul.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmul.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vmul_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, vint16mf4_t op2, size_t vl) {
@@ -2357,7 +2357,7 @@ vint16mf4_t test_vmul_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmul.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmul.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vmul_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, int16_t op2, size_t vl) {
@@ -2367,7 +2367,7 @@ vint16mf4_t test_vmul_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmul.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmul.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vmul_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, vint16mf2_t op2, size_t vl) {
@@ -2377,7 +2377,7 @@ vint16mf2_t test_vmul_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmul.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmul.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vmul_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, int16_t op2, size_t vl) {
@@ -2387,7 +2387,7 @@ vint16mf2_t test_vmul_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmul.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmul.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vmul_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, vint16m1_t op2, size_t vl) {
@@ -2397,7 +2397,7 @@ vint16m1_t test_vmul_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmul.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmul.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vmul_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, int16_t op2, size_t vl) {
@@ -2407,7 +2407,7 @@ vint16m1_t test_vmul_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmul.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmul.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vmul_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, vint16m2_t op2, size_t vl) {
@@ -2417,7 +2417,7 @@ vint16m2_t test_vmul_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmul.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmul.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vmul_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, int16_t op2, size_t vl) {
@@ -2427,7 +2427,7 @@ vint16m2_t test_vmul_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmul.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmul.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vmul_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, vint16m4_t op2, size_t vl) {
@@ -2437,7 +2437,7 @@ vint16m4_t test_vmul_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmul.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmul.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vmul_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, int16_t op2, size_t vl) {
@@ -2447,7 +2447,7 @@ vint16m4_t test_vmul_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmul.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmul.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vmul_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, vint16m8_t op2, size_t vl) {
@@ -2457,7 +2457,7 @@ vint16m8_t test_vmul_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmul.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmul.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vmul_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, int16_t op2, size_t vl) {
@@ -2467,7 +2467,7 @@ vint16m8_t test_vmul_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmul.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmul.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vmul_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, vint32mf2_t op2, size_t vl) {
@@ -2477,7 +2477,7 @@ vint32mf2_t test_vmul_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmul.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmul.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vmul_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, int32_t op2, size_t vl) {
@@ -2487,7 +2487,7 @@ vint32mf2_t test_vmul_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmul.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmul.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vmul_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, vint32m1_t op2, size_t vl) {
@@ -2497,7 +2497,7 @@ vint32m1_t test_vmul_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmul.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmul.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vmul_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, int32_t op2, size_t vl) {
@@ -2507,7 +2507,7 @@ vint32m1_t test_vmul_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmul.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmul.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vmul_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, vint32m2_t op2, size_t vl) {
@@ -2517,7 +2517,7 @@ vint32m2_t test_vmul_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmul.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmul.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vmul_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, int32_t op2, size_t vl) {
@@ -2527,7 +2527,7 @@ vint32m2_t test_vmul_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmul.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmul.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vmul_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, vint32m4_t op2, size_t vl) {
@@ -2537,7 +2537,7 @@ vint32m4_t test_vmul_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmul.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmul.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vmul_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, int32_t op2, size_t vl) {
@@ -2547,7 +2547,7 @@ vint32m4_t test_vmul_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmul.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmul.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vmul_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, vint32m8_t op2, size_t vl) {
@@ -2557,7 +2557,7 @@ vint32m8_t test_vmul_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmul.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmul.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vmul_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, int32_t op2, size_t vl) {
@@ -2567,7 +2567,7 @@ vint32m8_t test_vmul_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmul.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmul.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vmul_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, vint64m1_t op2, size_t vl) {
@@ -2577,7 +2577,7 @@ vint64m1_t test_vmul_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmul.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmul.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vmul_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, int64_t op2, size_t vl) {
@@ -2587,7 +2587,7 @@ vint64m1_t test_vmul_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmul.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmul.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vmul_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, vint64m2_t op2, size_t vl) {
@@ -2597,7 +2597,7 @@ vint64m2_t test_vmul_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmul.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmul.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vmul_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, int64_t op2, size_t vl) {
@@ -2607,7 +2607,7 @@ vint64m2_t test_vmul_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmul.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmul.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vmul_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, vint64m4_t op2, size_t vl) {
@@ -2617,7 +2617,7 @@ vint64m4_t test_vmul_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmul.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmul.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vmul_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, int64_t op2, size_t vl) {
@@ -2627,7 +2627,7 @@ vint64m4_t test_vmul_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmul.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmul.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vmul_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, vint64m8_t op2, size_t vl) {
@@ -2637,7 +2637,7 @@ vint64m8_t test_vmul_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmul.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmul.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vmul_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, int64_t op2, size_t vl) {
@@ -2647,7 +2647,7 @@ vint64m8_t test_vmul_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmul.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmul.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vmul_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -2657,7 +2657,7 @@ vuint8mf8_t test_vmul_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmul.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmul.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vmul_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, uint8_t op2, size_t vl) {
@@ -2667,7 +2667,7 @@ vuint8mf8_t test_vmul_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmul.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmul.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vmul_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, vuint8mf4_t op2, size_t vl) {
@@ -2677,7 +2677,7 @@ vuint8mf4_t test_vmul_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmul.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmul.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vmul_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, uint8_t op2, size_t vl) {
@@ -2687,7 +2687,7 @@ vuint8mf4_t test_vmul_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmul.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmul.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vmul_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, vuint8mf2_t op2, size_t vl) {
@@ -2697,7 +2697,7 @@ vuint8mf2_t test_vmul_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmul.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmul.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vmul_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, uint8_t op2, size_t vl) {
@@ -2707,7 +2707,7 @@ vuint8mf2_t test_vmul_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmul.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmul.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vmul_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, vuint8m1_t op2, size_t vl) {
@@ -2717,7 +2717,7 @@ vuint8m1_t test_vmul_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t o
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmul.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmul.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vmul_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, uint8_t op2, size_t vl) {
@@ -2727,7 +2727,7 @@ vuint8m1_t test_vmul_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t o
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmul.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmul.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vmul_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, vuint8m2_t op2, size_t vl) {
@@ -2737,7 +2737,7 @@ vuint8m2_t test_vmul_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t o
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmul.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmul.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vmul_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, uint8_t op2, size_t vl) {
@@ -2747,7 +2747,7 @@ vuint8m2_t test_vmul_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t o
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmul.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmul.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vmul_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, vuint8m4_t op2, size_t vl) {
@@ -2757,7 +2757,7 @@ vuint8m4_t test_vmul_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t o
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmul.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmul.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vmul_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, uint8_t op2, size_t vl) {
@@ -2767,7 +2767,7 @@ vuint8m4_t test_vmul_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t o
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmul.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmul.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vmul_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, vuint8m8_t op2, size_t vl) {
@@ -2777,7 +2777,7 @@ vuint8m8_t test_vmul_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t o
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmul.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmul.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vmul_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, uint8_t op2, size_t vl) {
@@ -2787,7 +2787,7 @@ vuint8m8_t test_vmul_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t o
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmul.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmul.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vmul_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, vuint16mf4_t op2, size_t vl) {
@@ -2797,7 +2797,7 @@ vuint16mf4_t test_vmul_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmul.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmul.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vmul_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, uint16_t op2, size_t vl) {
@@ -2807,7 +2807,7 @@ vuint16mf4_t test_vmul_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmul.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmul.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vmul_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, vuint16mf2_t op2, size_t vl) {
@@ -2817,7 +2817,7 @@ vuint16mf2_t test_vmul_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmul.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmul.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vmul_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, uint16_t op2, size_t vl) {
@@ -2827,7 +2827,7 @@ vuint16mf2_t test_vmul_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmul.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmul.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vmul_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, vuint16m1_t op2, size_t vl) {
@@ -2837,7 +2837,7 @@ vuint16m1_t test_vmul_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmul.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmul.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vmul_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, uint16_t op2, size_t vl) {
@@ -2847,7 +2847,7 @@ vuint16m1_t test_vmul_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmul.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmul.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vmul_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, vuint16m2_t op2, size_t vl) {
@@ -2857,7 +2857,7 @@ vuint16m2_t test_vmul_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmul.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmul.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vmul_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, uint16_t op2, size_t vl) {
@@ -2867,7 +2867,7 @@ vuint16m2_t test_vmul_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmul.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmul.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vmul_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, vuint16m4_t op2, size_t vl) {
@@ -2877,7 +2877,7 @@ vuint16m4_t test_vmul_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmul.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmul.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vmul_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, uint16_t op2, size_t vl) {
@@ -2887,7 +2887,7 @@ vuint16m4_t test_vmul_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmul.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmul.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vmul_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, vuint16m8_t op2, size_t vl) {
@@ -2897,7 +2897,7 @@ vuint16m8_t test_vmul_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmul.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmul.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vmul_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, uint16_t op2, size_t vl) {
@@ -2907,7 +2907,7 @@ vuint16m8_t test_vmul_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmul.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmul.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vmul_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, vuint32mf2_t op2, size_t vl) {
@@ -2917,7 +2917,7 @@ vuint32mf2_t test_vmul_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmul.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmul.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vmul_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, uint32_t op2, size_t vl) {
@@ -2927,7 +2927,7 @@ vuint32mf2_t test_vmul_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmul.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmul.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vmul_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, vuint32m1_t op2, size_t vl) {
@@ -2937,7 +2937,7 @@ vuint32m1_t test_vmul_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmul.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmul.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vmul_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, uint32_t op2, size_t vl) {
@@ -2947,7 +2947,7 @@ vuint32m1_t test_vmul_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmul.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmul.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vmul_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, vuint32m2_t op2, size_t vl) {
@@ -2957,7 +2957,7 @@ vuint32m2_t test_vmul_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmul.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmul.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vmul_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, uint32_t op2, size_t vl) {
@@ -2967,7 +2967,7 @@ vuint32m2_t test_vmul_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmul.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmul.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vmul_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, vuint32m4_t op2, size_t vl) {
@@ -2977,7 +2977,7 @@ vuint32m4_t test_vmul_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmul.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmul.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vmul_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, uint32_t op2, size_t vl) {
@@ -2987,7 +2987,7 @@ vuint32m4_t test_vmul_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmul.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmul.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vmul_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, vuint32m8_t op2, size_t vl) {
@@ -2997,7 +2997,7 @@ vuint32m8_t test_vmul_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmul.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmul.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vmul_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, uint32_t op2, size_t vl) {
@@ -3007,7 +3007,7 @@ vuint32m8_t test_vmul_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmul.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmul.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vmul_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, vuint64m1_t op2, size_t vl) {
@@ -3017,7 +3017,7 @@ vuint64m1_t test_vmul_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmul.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmul.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vmul_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, uint64_t op2, size_t vl) {
@@ -3027,7 +3027,7 @@ vuint64m1_t test_vmul_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmul.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmul.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vmul_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, vuint64m2_t op2, size_t vl) {
@@ -3037,7 +3037,7 @@ vuint64m2_t test_vmul_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmul.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmul.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vmul_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, uint64_t op2, size_t vl) {
@@ -3047,7 +3047,7 @@ vuint64m2_t test_vmul_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmul.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmul.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vmul_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, vuint64m4_t op2, size_t vl) {
@@ -3057,7 +3057,7 @@ vuint64m4_t test_vmul_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmul.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmul.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vmul_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, uint64_t op2, size_t vl) {
@@ -3067,7 +3067,7 @@ vuint64m4_t test_vmul_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vmul_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmul.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmul.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vmul_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, vuint64m8_t op2, size_t vl) {
@@ -3077,7 +3077,7 @@ vuint64m8_t test_vmul_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8
 //
 // CHECK-RV64-LABEL: @test_vmul_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmul.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmul.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vmul_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, uint64_t op2, size_t vl) {
@@ -3087,7 +3087,7 @@ vuint64m8_t test_vmul_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmulh.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmulh.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vmulh_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, vint8mf8_t op2, size_t vl) {
@@ -3097,7 +3097,7 @@ vint8mf8_t test_vmulh_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmulh.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmulh.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vmulh_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, int8_t op2, size_t vl) {
@@ -3107,7 +3107,7 @@ vint8mf8_t test_vmulh_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmulh.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmulh.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vmulh_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, vint8mf4_t op2, size_t vl) {
@@ -3117,7 +3117,7 @@ vint8mf4_t test_vmulh_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmulh.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmulh.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vmulh_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, int8_t op2, size_t vl) {
@@ -3127,7 +3127,7 @@ vint8mf4_t test_vmulh_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmulh.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmulh.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vmulh_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, vint8mf2_t op2, size_t vl) {
@@ -3137,7 +3137,7 @@ vint8mf2_t test_vmulh_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmulh.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmulh.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vmulh_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, int8_t op2, size_t vl) {
@@ -3147,7 +3147,7 @@ vint8mf2_t test_vmulh_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmulh.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmulh.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vmulh_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, vint8m1_t op2, size_t vl) {
@@ -3157,7 +3157,7 @@ vint8m1_t test_vmulh_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmulh.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmulh.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vmulh_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, int8_t op2, size_t vl) {
@@ -3167,7 +3167,7 @@ vint8m1_t test_vmulh_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmulh.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmulh.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vmulh_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, vint8m2_t op2, size_t vl) {
@@ -3177,7 +3177,7 @@ vint8m2_t test_vmulh_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmulh.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmulh.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vmulh_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, int8_t op2, size_t vl) {
@@ -3187,7 +3187,7 @@ vint8m2_t test_vmulh_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmulh.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmulh.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vmulh_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, vint8m4_t op2, size_t vl) {
@@ -3197,7 +3197,7 @@ vint8m4_t test_vmulh_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmulh.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmulh.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vmulh_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, int8_t op2, size_t vl) {
@@ -3207,7 +3207,7 @@ vint8m4_t test_vmulh_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmulh.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmulh.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vmulh_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, vint8m8_t op2, size_t vl) {
@@ -3217,7 +3217,7 @@ vint8m8_t test_vmulh_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmulh.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmulh.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vmulh_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, int8_t op2, size_t vl) {
@@ -3227,7 +3227,7 @@ vint8m8_t test_vmulh_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmulh.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmulh.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vmulh_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, vint16mf4_t op2, size_t vl) {
@@ -3237,7 +3237,7 @@ vint16mf4_t test_vmulh_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmulh.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmulh.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vmulh_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, int16_t op2, size_t vl) {
@@ -3247,7 +3247,7 @@ vint16mf4_t test_vmulh_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmulh.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmulh.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vmulh_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, vint16mf2_t op2, size_t vl) {
@@ -3257,7 +3257,7 @@ vint16mf2_t test_vmulh_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmulh.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmulh.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vmulh_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, int16_t op2, size_t vl) {
@@ -3267,7 +3267,7 @@ vint16mf2_t test_vmulh_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmulh.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmulh.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vmulh_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, vint16m1_t op2, size_t vl) {
@@ -3277,7 +3277,7 @@ vint16m1_t test_vmulh_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmulh.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmulh.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vmulh_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, int16_t op2, size_t vl) {
@@ -3287,7 +3287,7 @@ vint16m1_t test_vmulh_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmulh.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmulh.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vmulh_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, vint16m2_t op2, size_t vl) {
@@ -3297,7 +3297,7 @@ vint16m2_t test_vmulh_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmulh.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmulh.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vmulh_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, int16_t op2, size_t vl) {
@@ -3307,7 +3307,7 @@ vint16m2_t test_vmulh_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmulh.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmulh.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vmulh_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, vint16m4_t op2, size_t vl) {
@@ -3317,7 +3317,7 @@ vint16m4_t test_vmulh_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmulh.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmulh.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vmulh_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, int16_t op2, size_t vl) {
@@ -3327,7 +3327,7 @@ vint16m4_t test_vmulh_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmulh.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmulh.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vmulh_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, vint16m8_t op2, size_t vl) {
@@ -3337,7 +3337,7 @@ vint16m8_t test_vmulh_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmulh.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmulh.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vmulh_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, int16_t op2, size_t vl) {
@@ -3347,7 +3347,7 @@ vint16m8_t test_vmulh_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmulh.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmulh.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vmulh_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, vint32mf2_t op2, size_t vl) {
@@ -3357,7 +3357,7 @@ vint32mf2_t test_vmulh_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmulh.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmulh.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vmulh_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, int32_t op2, size_t vl) {
@@ -3367,7 +3367,7 @@ vint32mf2_t test_vmulh_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmulh.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmulh.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vmulh_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, vint32m1_t op2, size_t vl) {
@@ -3377,7 +3377,7 @@ vint32m1_t test_vmulh_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmulh.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmulh.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vmulh_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, int32_t op2, size_t vl) {
@@ -3387,7 +3387,7 @@ vint32m1_t test_vmulh_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmulh.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmulh.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vmulh_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, vint32m2_t op2, size_t vl) {
@@ -3397,7 +3397,7 @@ vint32m2_t test_vmulh_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmulh.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmulh.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vmulh_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, int32_t op2, size_t vl) {
@@ -3407,7 +3407,7 @@ vint32m2_t test_vmulh_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmulh.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmulh.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vmulh_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, vint32m4_t op2, size_t vl) {
@@ -3417,7 +3417,7 @@ vint32m4_t test_vmulh_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmulh.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmulh.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vmulh_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, int32_t op2, size_t vl) {
@@ -3427,7 +3427,7 @@ vint32m4_t test_vmulh_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmulh.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmulh.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vmulh_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, vint32m8_t op2, size_t vl) {
@@ -3437,7 +3437,7 @@ vint32m8_t test_vmulh_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmulh.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmulh.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vmulh_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, int32_t op2, size_t vl) {
@@ -3447,7 +3447,7 @@ vint32m8_t test_vmulh_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmulh.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmulh.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vmulh_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, vint64m1_t op2, size_t vl) {
@@ -3457,7 +3457,7 @@ vint64m1_t test_vmulh_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmulh.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmulh.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vmulh_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, int64_t op2, size_t vl) {
@@ -3467,7 +3467,7 @@ vint64m1_t test_vmulh_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmulh.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmulh.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vmulh_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, vint64m2_t op2, size_t vl) {
@@ -3477,7 +3477,7 @@ vint64m2_t test_vmulh_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmulh.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmulh.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vmulh_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, int64_t op2, size_t vl) {
@@ -3487,7 +3487,7 @@ vint64m2_t test_vmulh_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmulh.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmulh.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vmulh_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, vint64m4_t op2, size_t vl) {
@@ -3497,7 +3497,7 @@ vint64m4_t test_vmulh_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmulh.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmulh.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vmulh_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, int64_t op2, size_t vl) {
@@ -3507,7 +3507,7 @@ vint64m4_t test_vmulh_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_
 //
 // CHECK-RV64-LABEL: @test_vmulh_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmulh.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmulh.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vmulh_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, vint64m8_t op2, size_t vl) {
@@ -3517,7 +3517,7 @@ vint64m8_t test_vmulh_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vmulh_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmulh.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmulh.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vmulh_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, int64_t op2, size_t vl) {
@@ -3527,7 +3527,7 @@ vint64m8_t test_vmulh_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmulhu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmulhu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vmulhu_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -3537,7 +3537,7 @@ vuint8mf8_t test_vmulhu_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmulhu.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmulhu.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vmulhu_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, uint8_t op2, size_t vl) {
@@ -3547,7 +3547,7 @@ vuint8mf8_t test_vmulhu_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmulhu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmulhu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vmulhu_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, vuint8mf4_t op2, size_t vl) {
@@ -3557,7 +3557,7 @@ vuint8mf4_t test_vmulhu_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmulhu.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmulhu.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vmulhu_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, uint8_t op2, size_t vl) {
@@ -3567,7 +3567,7 @@ vuint8mf4_t test_vmulhu_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmulhu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmulhu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vmulhu_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, vuint8mf2_t op2, size_t vl) {
@@ -3577,7 +3577,7 @@ vuint8mf2_t test_vmulhu_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmulhu.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmulhu.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vmulhu_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, uint8_t op2, size_t vl) {
@@ -3587,7 +3587,7 @@ vuint8mf2_t test_vmulhu_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmulhu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmulhu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vmulhu_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, vuint8m1_t op2, size_t vl) {
@@ -3597,7 +3597,7 @@ vuint8m1_t test_vmulhu_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmulhu.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmulhu.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vmulhu_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, uint8_t op2, size_t vl) {
@@ -3607,7 +3607,7 @@ vuint8m1_t test_vmulhu_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmulhu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmulhu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vmulhu_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, vuint8m2_t op2, size_t vl) {
@@ -3617,7 +3617,7 @@ vuint8m2_t test_vmulhu_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmulhu.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmulhu.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vmulhu_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, uint8_t op2, size_t vl) {
@@ -3627,7 +3627,7 @@ vuint8m2_t test_vmulhu_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmulhu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmulhu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vmulhu_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, vuint8m4_t op2, size_t vl) {
@@ -3637,7 +3637,7 @@ vuint8m4_t test_vmulhu_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmulhu.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmulhu.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vmulhu_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, uint8_t op2, size_t vl) {
@@ -3647,7 +3647,7 @@ vuint8m4_t test_vmulhu_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmulhu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmulhu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vmulhu_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, vuint8m8_t op2, size_t vl) {
@@ -3657,7 +3657,7 @@ vuint8m8_t test_vmulhu_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmulhu.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmulhu.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vmulhu_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, uint8_t op2, size_t vl) {
@@ -3667,7 +3667,7 @@ vuint8m8_t test_vmulhu_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmulhu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmulhu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vmulhu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, vuint16mf4_t op2, size_t vl) {
@@ -3677,7 +3677,7 @@ vuint16mf4_t test_vmulhu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmulhu.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmulhu.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vmulhu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, uint16_t op2, size_t vl) {
@@ -3687,7 +3687,7 @@ vuint16mf4_t test_vmulhu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmulhu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmulhu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vmulhu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, vuint16mf2_t op2, size_t vl) {
@@ -3697,7 +3697,7 @@ vuint16mf2_t test_vmulhu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmulhu.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmulhu.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vmulhu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, uint16_t op2, size_t vl) {
@@ -3707,7 +3707,7 @@ vuint16mf2_t test_vmulhu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmulhu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmulhu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vmulhu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, vuint16m1_t op2, size_t vl) {
@@ -3717,7 +3717,7 @@ vuint16m1_t test_vmulhu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint1
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmulhu.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmulhu.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vmulhu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, uint16_t op2, size_t vl) {
@@ -3727,7 +3727,7 @@ vuint16m1_t test_vmulhu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint1
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmulhu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmulhu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vmulhu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, vuint16m2_t op2, size_t vl) {
@@ -3737,7 +3737,7 @@ vuint16m2_t test_vmulhu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmulhu.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmulhu.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vmulhu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, uint16_t op2, size_t vl) {
@@ -3747,7 +3747,7 @@ vuint16m2_t test_vmulhu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmulhu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmulhu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vmulhu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, vuint16m4_t op2, size_t vl) {
@@ -3757,7 +3757,7 @@ vuint16m4_t test_vmulhu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmulhu.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmulhu.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vmulhu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, uint16_t op2, size_t vl) {
@@ -3767,7 +3767,7 @@ vuint16m4_t test_vmulhu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmulhu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmulhu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vmulhu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, vuint16m8_t op2, size_t vl) {
@@ -3777,7 +3777,7 @@ vuint16m8_t test_vmulhu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmulhu.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmulhu.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vmulhu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, uint16_t op2, size_t vl) {
@@ -3787,7 +3787,7 @@ vuint16m8_t test_vmulhu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmulhu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmulhu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vmulhu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, vuint32mf2_t op2, size_t vl) {
@@ -3797,7 +3797,7 @@ vuint32mf2_t test_vmulhu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmulhu.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmulhu.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vmulhu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, uint32_t op2, size_t vl) {
@@ -3807,7 +3807,7 @@ vuint32mf2_t test_vmulhu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmulhu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmulhu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vmulhu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, vuint32m1_t op2, size_t vl) {
@@ -3817,7 +3817,7 @@ vuint32m1_t test_vmulhu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint3
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmulhu.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmulhu.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vmulhu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, uint32_t op2, size_t vl) {
@@ -3827,7 +3827,7 @@ vuint32m1_t test_vmulhu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint3
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmulhu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmulhu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vmulhu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, vuint32m2_t op2, size_t vl) {
@@ -3837,7 +3837,7 @@ vuint32m2_t test_vmulhu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint3
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmulhu.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmulhu.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vmulhu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, uint32_t op2, size_t vl) {
@@ -3847,7 +3847,7 @@ vuint32m2_t test_vmulhu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint3
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmulhu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmulhu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vmulhu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, vuint32m4_t op2, size_t vl) {
@@ -3857,7 +3857,7 @@ vuint32m4_t test_vmulhu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmulhu.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmulhu.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vmulhu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, uint32_t op2, size_t vl) {
@@ -3867,7 +3867,7 @@ vuint32m4_t test_vmulhu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmulhu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmulhu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vmulhu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, vuint32m8_t op2, size_t vl) {
@@ -3877,7 +3877,7 @@ vuint32m8_t test_vmulhu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmulhu.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmulhu.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vmulhu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, uint32_t op2, size_t vl) {
@@ -3887,7 +3887,7 @@ vuint32m8_t test_vmulhu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmulhu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmulhu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vmulhu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, vuint64m1_t op2, size_t vl) {
@@ -3897,7 +3897,7 @@ vuint64m1_t test_vmulhu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint6
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmulhu.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmulhu.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vmulhu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, uint64_t op2, size_t vl) {
@@ -3907,7 +3907,7 @@ vuint64m1_t test_vmulhu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint6
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmulhu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmulhu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vmulhu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, vuint64m2_t op2, size_t vl) {
@@ -3917,7 +3917,7 @@ vuint64m2_t test_vmulhu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint6
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmulhu.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmulhu.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vmulhu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, uint64_t op2, size_t vl) {
@@ -3927,7 +3927,7 @@ vuint64m2_t test_vmulhu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint6
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmulhu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmulhu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vmulhu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, vuint64m4_t op2, size_t vl) {
@@ -3937,7 +3937,7 @@ vuint64m4_t test_vmulhu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint6
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmulhu.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmulhu.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vmulhu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, uint64_t op2, size_t vl) {
@@ -3947,7 +3947,7 @@ vuint64m4_t test_vmulhu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint6
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmulhu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmulhu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vmulhu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, vuint64m8_t op2, size_t vl) {
@@ -3957,7 +3957,7 @@ vuint64m8_t test_vmulhu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vmulhu_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmulhu.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmulhu.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vmulhu_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, uint64_t op2, size_t vl) {
@@ -3967,7 +3967,7 @@ vuint64m8_t test_vmulhu_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmulhsu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmulhsu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vmulhsu_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -3977,7 +3977,7 @@ vint8mf8_t test_vmulhsu_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmulhsu.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vmulhsu.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vmulhsu_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, uint8_t op2, size_t vl) {
@@ -3987,7 +3987,7 @@ vint8mf8_t test_vmulhsu_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmulhsu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmulhsu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vmulhsu_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, vuint8mf4_t op2, size_t vl) {
@@ -3997,7 +3997,7 @@ vint8mf4_t test_vmulhsu_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmulhsu.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vmulhsu.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vmulhsu_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, uint8_t op2, size_t vl) {
@@ -4007,7 +4007,7 @@ vint8mf4_t test_vmulhsu_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmulhsu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmulhsu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vmulhsu_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, vuint8mf2_t op2, size_t vl) {
@@ -4017,7 +4017,7 @@ vint8mf2_t test_vmulhsu_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmulhsu.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vmulhsu.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vmulhsu_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, uint8_t op2, size_t vl) {
@@ -4027,7 +4027,7 @@ vint8mf2_t test_vmulhsu_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmulhsu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmulhsu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vmulhsu_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, vuint8m1_t op2, size_t vl) {
@@ -4037,7 +4037,7 @@ vint8m1_t test_vmulhsu_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t o
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmulhsu.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vmulhsu.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vmulhsu_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, uint8_t op2, size_t vl) {
@@ -4047,7 +4047,7 @@ vint8m1_t test_vmulhsu_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t o
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmulhsu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmulhsu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vmulhsu_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, vuint8m2_t op2, size_t vl) {
@@ -4057,7 +4057,7 @@ vint8m2_t test_vmulhsu_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t o
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmulhsu.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vmulhsu.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vmulhsu_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, uint8_t op2, size_t vl) {
@@ -4067,7 +4067,7 @@ vint8m2_t test_vmulhsu_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t o
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmulhsu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmulhsu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vmulhsu_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, vuint8m4_t op2, size_t vl) {
@@ -4077,7 +4077,7 @@ vint8m4_t test_vmulhsu_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t o
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmulhsu.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vmulhsu.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vmulhsu_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, uint8_t op2, size_t vl) {
@@ -4087,7 +4087,7 @@ vint8m4_t test_vmulhsu_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t o
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmulhsu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmulhsu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vmulhsu_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, vuint8m8_t op2, size_t vl) {
@@ -4097,7 +4097,7 @@ vint8m8_t test_vmulhsu_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t o
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmulhsu.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vmulhsu.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vmulhsu_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, uint8_t op2, size_t vl) {
@@ -4107,7 +4107,7 @@ vint8m8_t test_vmulhsu_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t o
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmulhsu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmulhsu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vmulhsu_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, vuint16mf4_t op2, size_t vl) {
@@ -4117,7 +4117,7 @@ vint16mf4_t test_vmulhsu_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmulhsu.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vmulhsu.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vmulhsu_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, uint16_t op2, size_t vl) {
@@ -4127,7 +4127,7 @@ vint16mf4_t test_vmulhsu_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmulhsu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmulhsu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vmulhsu_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, vuint16mf2_t op2, size_t vl) {
@@ -4137,7 +4137,7 @@ vint16mf2_t test_vmulhsu_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmulhsu.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vmulhsu.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vmulhsu_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, uint16_t op2, size_t vl) {
@@ -4147,7 +4147,7 @@ vint16mf2_t test_vmulhsu_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmulhsu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmulhsu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vmulhsu_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, vuint16m1_t op2, size_t vl) {
@@ -4157,7 +4157,7 @@ vint16m1_t test_vmulhsu_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmulhsu.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vmulhsu.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vmulhsu_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, uint16_t op2, size_t vl) {
@@ -4167,7 +4167,7 @@ vint16m1_t test_vmulhsu_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmulhsu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmulhsu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vmulhsu_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, vuint16m2_t op2, size_t vl) {
@@ -4177,7 +4177,7 @@ vint16m2_t test_vmulhsu_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmulhsu.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vmulhsu.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vmulhsu_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, uint16_t op2, size_t vl) {
@@ -4187,7 +4187,7 @@ vint16m2_t test_vmulhsu_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmulhsu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmulhsu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vmulhsu_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, vuint16m4_t op2, size_t vl) {
@@ -4197,7 +4197,7 @@ vint16m4_t test_vmulhsu_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmulhsu.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vmulhsu.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vmulhsu_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, uint16_t op2, size_t vl) {
@@ -4207,7 +4207,7 @@ vint16m4_t test_vmulhsu_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmulhsu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmulhsu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vmulhsu_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, vuint16m8_t op2, size_t vl) {
@@ -4217,7 +4217,7 @@ vint16m8_t test_vmulhsu_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmulhsu.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vmulhsu.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vmulhsu_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, uint16_t op2, size_t vl) {
@@ -4227,7 +4227,7 @@ vint16m8_t test_vmulhsu_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmulhsu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmulhsu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vmulhsu_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, vuint32mf2_t op2, size_t vl) {
@@ -4237,7 +4237,7 @@ vint32mf2_t test_vmulhsu_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmulhsu.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vmulhsu.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vmulhsu_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, uint32_t op2, size_t vl) {
@@ -4247,7 +4247,7 @@ vint32mf2_t test_vmulhsu_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmulhsu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmulhsu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vmulhsu_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, vuint32m1_t op2, size_t vl) {
@@ -4257,7 +4257,7 @@ vint32m1_t test_vmulhsu_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmulhsu.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vmulhsu.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vmulhsu_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, uint32_t op2, size_t vl) {
@@ -4267,7 +4267,7 @@ vint32m1_t test_vmulhsu_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmulhsu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmulhsu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vmulhsu_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, vuint32m2_t op2, size_t vl) {
@@ -4277,7 +4277,7 @@ vint32m2_t test_vmulhsu_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmulhsu.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vmulhsu.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vmulhsu_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, uint32_t op2, size_t vl) {
@@ -4287,7 +4287,7 @@ vint32m2_t test_vmulhsu_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmulhsu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmulhsu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vmulhsu_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, vuint32m4_t op2, size_t vl) {
@@ -4297,7 +4297,7 @@ vint32m4_t test_vmulhsu_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmulhsu.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vmulhsu.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vmulhsu_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, uint32_t op2, size_t vl) {
@@ -4307,7 +4307,7 @@ vint32m4_t test_vmulhsu_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmulhsu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmulhsu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vmulhsu_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, vuint32m8_t op2, size_t vl) {
@@ -4317,7 +4317,7 @@ vint32m8_t test_vmulhsu_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmulhsu.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vmulhsu.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vmulhsu_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, uint32_t op2, size_t vl) {
@@ -4327,7 +4327,7 @@ vint32m8_t test_vmulhsu_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmulhsu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmulhsu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vmulhsu_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, vuint64m1_t op2, size_t vl) {
@@ -4337,7 +4337,7 @@ vint64m1_t test_vmulhsu_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmulhsu.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vmulhsu.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vmulhsu_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, uint64_t op2, size_t vl) {
@@ -4347,7 +4347,7 @@ vint64m1_t test_vmulhsu_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmulhsu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmulhsu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vmulhsu_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, vuint64m2_t op2, size_t vl) {
@@ -4357,7 +4357,7 @@ vint64m2_t test_vmulhsu_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmulhsu.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vmulhsu.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vmulhsu_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, uint64_t op2, size_t vl) {
@@ -4367,7 +4367,7 @@ vint64m2_t test_vmulhsu_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmulhsu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmulhsu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vmulhsu_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, vuint64m4_t op2, size_t vl) {
@@ -4377,7 +4377,7 @@ vint64m4_t test_vmulhsu_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmulhsu.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vmulhsu.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vmulhsu_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, uint64_t op2, size_t vl) {
@@ -4387,7 +4387,7 @@ vint64m4_t test_vmulhsu_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmulhsu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmulhsu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vmulhsu_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, vuint64m8_t op2, size_t vl) {
@@ -4397,7 +4397,7 @@ vint64m8_t test_vmulhsu_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8
 //
 // CHECK-RV64-LABEL: @test_vmulhsu_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmulhsu.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vmulhsu.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vmulhsu_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, uint64_t op2, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnclip.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnclip.c
index fa08871..29c242a 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnclip.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnclip.c
@@ -622,7 +622,7 @@ vuint32m4_t test_vnclipu_wx_u32m4(vuint64m8_t op1, size_t shift, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vnclip_wv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnclip.mask.nxv1i8.nxv1i16.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i8> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnclip.mask.nxv1i8.nxv1i16.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i8> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vnclip_wv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
@@ -634,7 +634,7 @@ vint8mf8_t test_vnclip_wv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnclip.mask.nxv1i8.nxv1i16.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnclip.mask.nxv1i8.nxv1i16.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vnclip_wx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
@@ -645,7 +645,7 @@ vint8mf8_t test_vnclip_wx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnclip.mask.nxv2i8.nxv2i16.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i8> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnclip.mask.nxv2i8.nxv2i16.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i8> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vnclip_wv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
@@ -657,7 +657,7 @@ vint8mf4_t test_vnclip_wv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnclip.mask.nxv2i8.nxv2i16.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnclip.mask.nxv2i8.nxv2i16.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vnclip_wx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
@@ -668,7 +668,7 @@ vint8mf4_t test_vnclip_wx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnclip.mask.nxv4i8.nxv4i16.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i8> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnclip.mask.nxv4i8.nxv4i16.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i8> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vnclip_wv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
@@ -680,7 +680,7 @@ vint8mf2_t test_vnclip_wv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnclip.mask.nxv4i8.nxv4i16.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnclip.mask.nxv4i8.nxv4i16.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vnclip_wx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
@@ -691,7 +691,7 @@ vint8mf2_t test_vnclip_wx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnclip.mask.nxv8i8.nxv8i16.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i8> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnclip.mask.nxv8i8.nxv8i16.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i8> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vnclip_wv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
@@ -702,7 +702,7 @@ vint8m1_t test_vnclip_wv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnclip.mask.nxv8i8.nxv8i16.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnclip.mask.nxv8i8.nxv8i16.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vnclip_wx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
@@ -713,7 +713,7 @@ vint8m1_t test_vnclip_wx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnclip.mask.nxv16i8.nxv16i16.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i8> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnclip.mask.nxv16i8.nxv16i16.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i8> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vnclip_wv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
@@ -724,7 +724,7 @@ vint8m2_t test_vnclip_wv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnclip.mask.nxv16i8.nxv16i16.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnclip.mask.nxv16i8.nxv16i16.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vnclip_wx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
@@ -735,7 +735,7 @@ vint8m2_t test_vnclip_wx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnclip.mask.nxv32i8.nxv32i16.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i8> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnclip.mask.nxv32i8.nxv32i16.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i8> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vnclip_wv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
@@ -746,7 +746,7 @@ vint8m4_t test_vnclip_wv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnclip.mask.nxv32i8.nxv32i16.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnclip.mask.nxv32i8.nxv32i16.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vnclip_wx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
@@ -757,7 +757,7 @@ vint8m4_t test_vnclip_wx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnclip.mask.nxv1i16.nxv1i32.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i16> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnclip.mask.nxv1i16.nxv1i32.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i16> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vnclip_wv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -769,7 +769,7 @@ vint16mf4_t test_vnclip_wv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnclip.mask.nxv1i16.nxv1i32.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnclip.mask.nxv1i16.nxv1i32.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vnclip_wx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -780,7 +780,7 @@ vint16mf4_t test_vnclip_wx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnclip.mask.nxv2i16.nxv2i32.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i16> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnclip.mask.nxv2i16.nxv2i32.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i16> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vnclip_wv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -792,7 +792,7 @@ vint16mf2_t test_vnclip_wv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnclip.mask.nxv2i16.nxv2i32.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnclip.mask.nxv2i16.nxv2i32.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vnclip_wx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -803,7 +803,7 @@ vint16mf2_t test_vnclip_wx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnclip.mask.nxv4i16.nxv4i32.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i16> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnclip.mask.nxv4i16.nxv4i32.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i16> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vnclip_wv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -815,7 +815,7 @@ vint16m1_t test_vnclip_wv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnclip.mask.nxv4i16.nxv4i32.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnclip.mask.nxv4i16.nxv4i32.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vnclip_wx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -826,7 +826,7 @@ vint16m1_t test_vnclip_wx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnclip.mask.nxv8i16.nxv8i32.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i16> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnclip.mask.nxv8i16.nxv8i32.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i16> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vnclip_wv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -838,7 +838,7 @@ vint16m2_t test_vnclip_wv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnclip.mask.nxv8i16.nxv8i32.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnclip.mask.nxv8i16.nxv8i32.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vnclip_wx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -849,7 +849,7 @@ vint16m2_t test_vnclip_wx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnclip.mask.nxv16i16.nxv16i32.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i16> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnclip.mask.nxv16i16.nxv16i32.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i16> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vnclip_wv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -861,7 +861,7 @@ vint16m4_t test_vnclip_wv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnclip.mask.nxv16i16.nxv16i32.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnclip.mask.nxv16i16.nxv16i32.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vnclip_wx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -872,7 +872,7 @@ vint16m4_t test_vnclip_wx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnclip.mask.nxv1i32.nxv1i64.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i32> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnclip.mask.nxv1i32.nxv1i64.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i32> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vnclip_wv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -884,7 +884,7 @@ vint32mf2_t test_vnclip_wv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnclip.mask.nxv1i32.nxv1i64.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnclip.mask.nxv1i32.nxv1i64.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vnclip_wx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -895,7 +895,7 @@ vint32mf2_t test_vnclip_wx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnclip.mask.nxv2i32.nxv2i64.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i32> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnclip.mask.nxv2i32.nxv2i64.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i32> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vnclip_wv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -907,7 +907,7 @@ vint32m1_t test_vnclip_wv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnclip.mask.nxv2i32.nxv2i64.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnclip.mask.nxv2i32.nxv2i64.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vnclip_wx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -918,7 +918,7 @@ vint32m1_t test_vnclip_wx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnclip.mask.nxv4i32.nxv4i64.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i32> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnclip.mask.nxv4i32.nxv4i64.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i32> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vnclip_wv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -930,7 +930,7 @@ vint32m2_t test_vnclip_wv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnclip.mask.nxv4i32.nxv4i64.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnclip.mask.nxv4i32.nxv4i64.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vnclip_wx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -941,7 +941,7 @@ vint32m2_t test_vnclip_wx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnclip.mask.nxv8i32.nxv8i64.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i32> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnclip.mask.nxv8i32.nxv8i64.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i32> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vnclip_wv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -953,7 +953,7 @@ vint32m4_t test_vnclip_wv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclip_wx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnclip.mask.nxv8i32.nxv8i64.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnclip.mask.nxv8i32.nxv8i64.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vnclip_wx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -964,7 +964,7 @@ vint32m4_t test_vnclip_wx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnclipu.mask.nxv1i8.nxv1i16.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i8> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnclipu.mask.nxv1i8.nxv1i16.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i8> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vnclipu_wv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
@@ -976,7 +976,7 @@ vuint8mf8_t test_vnclipu_wv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnclipu.mask.nxv1i8.nxv1i16.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnclipu.mask.nxv1i8.nxv1i16.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vnclipu_wx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
@@ -987,7 +987,7 @@ vuint8mf8_t test_vnclipu_wx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnclipu.mask.nxv2i8.nxv2i16.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i8> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnclipu.mask.nxv2i8.nxv2i16.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i8> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vnclipu_wv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
@@ -999,7 +999,7 @@ vuint8mf4_t test_vnclipu_wv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnclipu.mask.nxv2i8.nxv2i16.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnclipu.mask.nxv2i8.nxv2i16.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vnclipu_wx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
@@ -1010,7 +1010,7 @@ vuint8mf4_t test_vnclipu_wx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnclipu.mask.nxv4i8.nxv4i16.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i8> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnclipu.mask.nxv4i8.nxv4i16.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i8> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vnclipu_wv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
@@ -1022,7 +1022,7 @@ vuint8mf2_t test_vnclipu_wv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnclipu.mask.nxv4i8.nxv4i16.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnclipu.mask.nxv4i8.nxv4i16.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vnclipu_wx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
@@ -1033,7 +1033,7 @@ vuint8mf2_t test_vnclipu_wx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnclipu.mask.nxv8i8.nxv8i16.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i8> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnclipu.mask.nxv8i8.nxv8i16.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i8> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vnclipu_wv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
@@ -1045,7 +1045,7 @@ vuint8m1_t test_vnclipu_wv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnclipu.mask.nxv8i8.nxv8i16.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnclipu.mask.nxv8i8.nxv8i16.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vnclipu_wx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
@@ -1056,7 +1056,7 @@ vuint8m1_t test_vnclipu_wx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnclipu.mask.nxv16i8.nxv16i16.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i8> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnclipu.mask.nxv16i8.nxv16i16.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i8> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vnclipu_wv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
@@ -1068,7 +1068,7 @@ vuint8m2_t test_vnclipu_wv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnclipu.mask.nxv16i8.nxv16i16.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnclipu.mask.nxv16i8.nxv16i16.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vnclipu_wx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
@@ -1079,7 +1079,7 @@ vuint8m2_t test_vnclipu_wx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnclipu.mask.nxv32i8.nxv32i16.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i8> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnclipu.mask.nxv32i8.nxv32i16.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i8> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vnclipu_wv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
@@ -1091,7 +1091,7 @@ vuint8m4_t test_vnclipu_wv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnclipu.mask.nxv32i8.nxv32i16.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnclipu.mask.nxv32i8.nxv32i16.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vnclipu_wx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
@@ -1102,7 +1102,7 @@ vuint8m4_t test_vnclipu_wx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnclipu.mask.nxv1i16.nxv1i32.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i16> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnclipu.mask.nxv1i16.nxv1i32.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i16> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vnclipu_wv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -1114,7 +1114,7 @@ vuint16mf4_t test_vnclipu_wv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnclipu.mask.nxv1i16.nxv1i32.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnclipu.mask.nxv1i16.nxv1i32.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vnclipu_wx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -1126,7 +1126,7 @@ vuint16mf4_t test_vnclipu_wx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnclipu.mask.nxv2i16.nxv2i32.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i16> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnclipu.mask.nxv2i16.nxv2i32.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i16> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vnclipu_wv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -1138,7 +1138,7 @@ vuint16mf2_t test_vnclipu_wv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnclipu.mask.nxv2i16.nxv2i32.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnclipu.mask.nxv2i16.nxv2i32.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vnclipu_wx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -1150,7 +1150,7 @@ vuint16mf2_t test_vnclipu_wx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnclipu.mask.nxv4i16.nxv4i32.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i16> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnclipu.mask.nxv4i16.nxv4i32.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i16> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vnclipu_wv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -1162,7 +1162,7 @@ vuint16m1_t test_vnclipu_wv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnclipu.mask.nxv4i16.nxv4i32.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnclipu.mask.nxv4i16.nxv4i32.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vnclipu_wx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -1173,7 +1173,7 @@ vuint16m1_t test_vnclipu_wx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnclipu.mask.nxv8i16.nxv8i32.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i16> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnclipu.mask.nxv8i16.nxv8i32.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i16> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vnclipu_wv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -1185,7 +1185,7 @@ vuint16m2_t test_vnclipu_wv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnclipu.mask.nxv8i16.nxv8i32.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnclipu.mask.nxv8i16.nxv8i32.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vnclipu_wx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -1196,7 +1196,7 @@ vuint16m2_t test_vnclipu_wx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnclipu.mask.nxv16i16.nxv16i32.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i16> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnclipu.mask.nxv16i16.nxv16i32.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i16> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vnclipu_wv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -1208,7 +1208,7 @@ vuint16m4_t test_vnclipu_wv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnclipu.mask.nxv16i16.nxv16i32.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnclipu.mask.nxv16i16.nxv16i32.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vnclipu_wx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -1219,7 +1219,7 @@ vuint16m4_t test_vnclipu_wx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnclipu.mask.nxv1i32.nxv1i64.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i32> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnclipu.mask.nxv1i32.nxv1i64.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i32> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vnclipu_wv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -1231,7 +1231,7 @@ vuint32mf2_t test_vnclipu_wv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnclipu.mask.nxv1i32.nxv1i64.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnclipu.mask.nxv1i32.nxv1i64.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vnclipu_wx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -1243,7 +1243,7 @@ vuint32mf2_t test_vnclipu_wx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnclipu.mask.nxv2i32.nxv2i64.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i32> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnclipu.mask.nxv2i32.nxv2i64.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i32> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vnclipu_wv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -1255,7 +1255,7 @@ vuint32m1_t test_vnclipu_wv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnclipu.mask.nxv2i32.nxv2i64.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnclipu.mask.nxv2i32.nxv2i64.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vnclipu_wx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -1266,7 +1266,7 @@ vuint32m1_t test_vnclipu_wx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnclipu.mask.nxv4i32.nxv4i64.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i32> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnclipu.mask.nxv4i32.nxv4i64.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i32> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vnclipu_wv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -1278,7 +1278,7 @@ vuint32m2_t test_vnclipu_wv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnclipu.mask.nxv4i32.nxv4i64.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnclipu.mask.nxv4i32.nxv4i64.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vnclipu_wx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -1289,7 +1289,7 @@ vuint32m2_t test_vnclipu_wx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnclipu.mask.nxv8i32.nxv8i64.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i32> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnclipu.mask.nxv8i32.nxv8i64.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i32> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vnclipu_wv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -1301,7 +1301,7 @@ vuint32m4_t test_vnclipu_wv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vnclipu_wx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnclipu.mask.nxv8i32.nxv8i64.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnclipu.mask.nxv8i32.nxv8i64.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vnclipu_wx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vncvt.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vncvt.c
index 7523092..e7a9510 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vncvt.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vncvt.c
@@ -307,7 +307,7 @@ vuint32m4_t test_vncvt_x_x_w_u32m4 (vuint64m8_t src, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnsrl.mask.nxv1i8.nxv1i16.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i64 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnsrl.mask.nxv1i8.nxv1i16.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i64 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vncvt_x_x_w_i8mf8_m (vbool64_t mask, vint8mf8_t maskedoff, vint16mf4_t src, size_t vl) {
@@ -317,7 +317,7 @@ vint8mf8_t test_vncvt_x_x_w_i8mf8_m (vbool64_t mask, vint8mf8_t maskedoff, vint1
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnsrl.mask.nxv2i8.nxv2i16.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i64 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnsrl.mask.nxv2i8.nxv2i16.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i64 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vncvt_x_x_w_i8mf4_m (vbool32_t mask, vint8mf4_t maskedoff, vint16mf2_t src, size_t vl) {
@@ -327,7 +327,7 @@ vint8mf4_t test_vncvt_x_x_w_i8mf4_m (vbool32_t mask, vint8mf4_t maskedoff, vint1
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnsrl.mask.nxv4i8.nxv4i16.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i64 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnsrl.mask.nxv4i8.nxv4i16.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i64 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vncvt_x_x_w_i8mf2_m (vbool16_t mask, vint8mf2_t maskedoff, vint16m1_t src, size_t vl) {
@@ -337,7 +337,7 @@ vint8mf2_t test_vncvt_x_x_w_i8mf2_m (vbool16_t mask, vint8mf2_t maskedoff, vint1
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnsrl.mask.nxv8i8.nxv8i16.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i64 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnsrl.mask.nxv8i8.nxv8i16.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i64 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vncvt_x_x_w_i8m1_m (vbool8_t mask, vint8m1_t maskedoff, vint16m2_t src, size_t vl) {
@@ -347,7 +347,7 @@ vint8m1_t test_vncvt_x_x_w_i8m1_m (vbool8_t mask, vint8m1_t maskedoff, vint16m2_
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnsrl.mask.nxv16i8.nxv16i16.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i64 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnsrl.mask.nxv16i8.nxv16i16.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i64 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vncvt_x_x_w_i8m2_m (vbool4_t mask, vint8m2_t maskedoff, vint16m4_t src, size_t vl) {
@@ -357,7 +357,7 @@ vint8m2_t test_vncvt_x_x_w_i8m2_m (vbool4_t mask, vint8m2_t maskedoff, vint16m4_
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnsrl.mask.nxv32i8.nxv32i16.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[SRC:%.*]], i64 0, <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnsrl.mask.nxv32i8.nxv32i16.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[SRC:%.*]], i64 0, <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vncvt_x_x_w_i8m4_m (vbool2_t mask, vint8m4_t maskedoff, vint16m8_t src, size_t vl) {
@@ -367,7 +367,7 @@ vint8m4_t test_vncvt_x_x_w_i8m4_m (vbool2_t mask, vint8m4_t maskedoff, vint16m8_
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnsrl.mask.nxv1i8.nxv1i16.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i64 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnsrl.mask.nxv1i8.nxv1i16.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i64 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vncvt_x_x_w_u8mf8_m (vbool64_t mask, vuint8mf8_t maskedoff, vuint16mf4_t src, size_t vl) {
@@ -377,7 +377,7 @@ vuint8mf8_t test_vncvt_x_x_w_u8mf8_m (vbool64_t mask, vuint8mf8_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnsrl.mask.nxv2i8.nxv2i16.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i64 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnsrl.mask.nxv2i8.nxv2i16.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i64 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vncvt_x_x_w_u8mf4_m (vbool32_t mask, vuint8mf4_t maskedoff, vuint16mf2_t src, size_t vl) {
@@ -387,7 +387,7 @@ vuint8mf4_t test_vncvt_x_x_w_u8mf4_m (vbool32_t mask, vuint8mf4_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnsrl.mask.nxv4i8.nxv4i16.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i64 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnsrl.mask.nxv4i8.nxv4i16.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i64 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vncvt_x_x_w_u8mf2_m (vbool16_t mask, vuint8mf2_t maskedoff, vuint16m1_t src, size_t vl) {
@@ -397,7 +397,7 @@ vuint8mf2_t test_vncvt_x_x_w_u8mf2_m (vbool16_t mask, vuint8mf2_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnsrl.mask.nxv8i8.nxv8i16.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i64 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnsrl.mask.nxv8i8.nxv8i16.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i64 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vncvt_x_x_w_u8m1_m (vbool8_t mask, vuint8m1_t maskedoff, vuint16m2_t src, size_t vl) {
@@ -407,7 +407,7 @@ vuint8m1_t test_vncvt_x_x_w_u8m1_m (vbool8_t mask, vuint8m1_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnsrl.mask.nxv16i8.nxv16i16.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i64 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnsrl.mask.nxv16i8.nxv16i16.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i64 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vncvt_x_x_w_u8m2_m (vbool4_t mask, vuint8m2_t maskedoff, vuint16m4_t src, size_t vl) {
@@ -417,7 +417,7 @@ vuint8m2_t test_vncvt_x_x_w_u8m2_m (vbool4_t mask, vuint8m2_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnsrl.mask.nxv32i8.nxv32i16.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[SRC:%.*]], i64 0, <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnsrl.mask.nxv32i8.nxv32i16.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[SRC:%.*]], i64 0, <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vncvt_x_x_w_u8m4_m (vbool2_t mask, vuint8m4_t maskedoff, vuint16m8_t src, size_t vl) {
@@ -427,7 +427,7 @@ vuint8m4_t test_vncvt_x_x_w_u8m4_m (vbool2_t mask, vuint8m4_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnsrl.mask.nxv1i16.nxv1i32.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i64 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnsrl.mask.nxv1i16.nxv1i32.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i64 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vncvt_x_x_w_i16mf4_m (vbool64_t mask, vint16mf4_t maskedoff, vint32mf2_t src, size_t vl) {
@@ -437,7 +437,7 @@ vint16mf4_t test_vncvt_x_x_w_i16mf4_m (vbool64_t mask, vint16mf4_t maskedoff, vi
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnsrl.mask.nxv2i16.nxv2i32.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i64 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnsrl.mask.nxv2i16.nxv2i32.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i64 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vncvt_x_x_w_i16mf2_m (vbool32_t mask, vint16mf2_t maskedoff, vint32m1_t src, size_t vl) {
@@ -447,7 +447,7 @@ vint16mf2_t test_vncvt_x_x_w_i16mf2_m (vbool32_t mask, vint16mf2_t maskedoff, vi
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnsrl.mask.nxv4i16.nxv4i32.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i64 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnsrl.mask.nxv4i16.nxv4i32.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i64 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vncvt_x_x_w_i16m1_m (vbool16_t mask, vint16m1_t maskedoff, vint32m2_t src, size_t vl) {
@@ -457,7 +457,7 @@ vint16m1_t test_vncvt_x_x_w_i16m1_m (vbool16_t mask, vint16m1_t maskedoff, vint3
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnsrl.mask.nxv8i16.nxv8i32.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i64 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnsrl.mask.nxv8i16.nxv8i32.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i64 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vncvt_x_x_w_i16m2_m (vbool8_t mask, vint16m2_t maskedoff, vint32m4_t src, size_t vl) {
@@ -467,7 +467,7 @@ vint16m2_t test_vncvt_x_x_w_i16m2_m (vbool8_t mask, vint16m2_t maskedoff, vint32
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnsrl.mask.nxv16i16.nxv16i32.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[SRC:%.*]], i64 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnsrl.mask.nxv16i16.nxv16i32.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[SRC:%.*]], i64 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vncvt_x_x_w_i16m4_m (vbool4_t mask, vint16m4_t maskedoff, vint32m8_t src, size_t vl) {
@@ -477,7 +477,7 @@ vint16m4_t test_vncvt_x_x_w_i16m4_m (vbool4_t mask, vint16m4_t maskedoff, vint32
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnsrl.mask.nxv1i16.nxv1i32.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i64 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnsrl.mask.nxv1i16.nxv1i32.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i64 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vncvt_x_x_w_u16mf4_m (vbool64_t mask, vuint16mf4_t maskedoff, vuint32mf2_t src, size_t vl) {
@@ -487,7 +487,7 @@ vuint16mf4_t test_vncvt_x_x_w_u16mf4_m (vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnsrl.mask.nxv2i16.nxv2i32.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i64 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnsrl.mask.nxv2i16.nxv2i32.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i64 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vncvt_x_x_w_u16mf2_m (vbool32_t mask, vuint16mf2_t maskedoff, vuint32m1_t src, size_t vl) {
@@ -497,7 +497,7 @@ vuint16mf2_t test_vncvt_x_x_w_u16mf2_m (vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnsrl.mask.nxv4i16.nxv4i32.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i64 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnsrl.mask.nxv4i16.nxv4i32.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i64 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vncvt_x_x_w_u16m1_m (vbool16_t mask, vuint16m1_t maskedoff, vuint32m2_t src, size_t vl) {
@@ -507,7 +507,7 @@ vuint16m1_t test_vncvt_x_x_w_u16m1_m (vbool16_t mask, vuint16m1_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnsrl.mask.nxv8i16.nxv8i32.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i64 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnsrl.mask.nxv8i16.nxv8i32.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i64 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vncvt_x_x_w_u16m2_m (vbool8_t mask, vuint16m2_t maskedoff, vuint32m4_t src, size_t vl) {
@@ -517,7 +517,7 @@ vuint16m2_t test_vncvt_x_x_w_u16m2_m (vbool8_t mask, vuint16m2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnsrl.mask.nxv16i16.nxv16i32.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[SRC:%.*]], i64 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnsrl.mask.nxv16i16.nxv16i32.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[SRC:%.*]], i64 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vncvt_x_x_w_u16m4_m (vbool4_t mask, vuint16m4_t maskedoff, vuint32m8_t src, size_t vl) {
@@ -527,7 +527,7 @@ vuint16m4_t test_vncvt_x_x_w_u16m4_m (vbool4_t mask, vuint16m4_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnsrl.mask.nxv1i32.nxv1i64.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[SRC:%.*]], i64 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnsrl.mask.nxv1i32.nxv1i64.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[SRC:%.*]], i64 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vncvt_x_x_w_i32mf2_m (vbool64_t mask, vint32mf2_t maskedoff, vint64m1_t src, size_t vl) {
@@ -537,7 +537,7 @@ vint32mf2_t test_vncvt_x_x_w_i32mf2_m (vbool64_t mask, vint32mf2_t maskedoff, vi
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnsrl.mask.nxv2i32.nxv2i64.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[SRC:%.*]], i64 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnsrl.mask.nxv2i32.nxv2i64.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[SRC:%.*]], i64 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vncvt_x_x_w_i32m1_m (vbool32_t mask, vint32m1_t maskedoff, vint64m2_t src, size_t vl) {
@@ -547,7 +547,7 @@ vint32m1_t test_vncvt_x_x_w_i32m1_m (vbool32_t mask, vint32m1_t maskedoff, vint6
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnsrl.mask.nxv4i32.nxv4i64.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[SRC:%.*]], i64 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnsrl.mask.nxv4i32.nxv4i64.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[SRC:%.*]], i64 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vncvt_x_x_w_i32m2_m (vbool16_t mask, vint32m2_t maskedoff, vint64m4_t src, size_t vl) {
@@ -557,7 +557,7 @@ vint32m2_t test_vncvt_x_x_w_i32m2_m (vbool16_t mask, vint32m2_t maskedoff, vint6
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnsrl.mask.nxv8i32.nxv8i64.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[SRC:%.*]], i64 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnsrl.mask.nxv8i32.nxv8i64.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[SRC:%.*]], i64 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vncvt_x_x_w_i32m4_m (vbool8_t mask, vint32m4_t maskedoff, vint64m8_t src, size_t vl) {
@@ -567,7 +567,7 @@ vint32m4_t test_vncvt_x_x_w_i32m4_m (vbool8_t mask, vint32m4_t maskedoff, vint64
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnsrl.mask.nxv1i32.nxv1i64.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[SRC:%.*]], i64 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnsrl.mask.nxv1i32.nxv1i64.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[SRC:%.*]], i64 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vncvt_x_x_w_u32mf2_m (vbool64_t mask, vuint32mf2_t maskedoff, vuint64m1_t src, size_t vl) {
@@ -577,7 +577,7 @@ vuint32mf2_t test_vncvt_x_x_w_u32mf2_m (vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnsrl.mask.nxv2i32.nxv2i64.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[SRC:%.*]], i64 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnsrl.mask.nxv2i32.nxv2i64.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[SRC:%.*]], i64 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vncvt_x_x_w_u32m1_m (vbool32_t mask, vuint32m1_t maskedoff, vuint64m2_t src, size_t vl) {
@@ -587,7 +587,7 @@ vuint32m1_t test_vncvt_x_x_w_u32m1_m (vbool32_t mask, vuint32m1_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnsrl.mask.nxv4i32.nxv4i64.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[SRC:%.*]], i64 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnsrl.mask.nxv4i32.nxv4i64.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[SRC:%.*]], i64 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vncvt_x_x_w_u32m2_m (vbool16_t mask, vuint32m2_t maskedoff, vuint64m4_t src, size_t vl) {
@@ -597,7 +597,7 @@ vuint32m2_t test_vncvt_x_x_w_u32m2_m (vbool16_t mask, vuint32m2_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vncvt_x_x_w_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnsrl.mask.nxv8i32.nxv8i64.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[SRC:%.*]], i64 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnsrl.mask.nxv8i32.nxv8i64.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[SRC:%.*]], i64 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vncvt_x_x_w_u32m4_m (vbool8_t mask, vuint32m4_t maskedoff, vuint64m8_t src, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vneg.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vneg.c
index 2e11915..ed5ab71 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vneg.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vneg.c
@@ -228,7 +228,7 @@ vint64m8_t test_vneg_v_i64m8 (vint64m8_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrsub.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrsub.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vneg_v_i8mf8_m (vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, size_t vl) {
@@ -238,7 +238,7 @@ vint8mf8_t test_vneg_v_i8mf8_m (vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrsub.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrsub.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vneg_v_i8mf4_m (vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, size_t vl) {
@@ -248,7 +248,7 @@ vint8mf4_t test_vneg_v_i8mf4_m (vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrsub.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrsub.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vneg_v_i8mf2_m (vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, size_t vl) {
@@ -258,7 +258,7 @@ vint8mf2_t test_vneg_v_i8mf2_m (vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrsub.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrsub.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vneg_v_i8m1_m (vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, size_t vl) {
@@ -268,7 +268,7 @@ vint8m1_t test_vneg_v_i8m1_m (vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrsub.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrsub.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vneg_v_i8m2_m (vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, size_t vl) {
@@ -278,7 +278,7 @@ vint8m2_t test_vneg_v_i8m2_m (vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrsub.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 0, <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrsub.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 0, <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vneg_v_i8m4_m (vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, size_t vl) {
@@ -288,7 +288,7 @@ vint8m4_t test_vneg_v_i8m4_m (vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vrsub.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 0, <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vrsub.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 0, <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vneg_v_i8m8_m (vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, size_t vl) {
@@ -298,7 +298,7 @@ vint8m8_t test_vneg_v_i8m8_m (vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrsub.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrsub.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vneg_v_i16mf4_m (vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, size_t vl) {
@@ -308,7 +308,7 @@ vint16mf4_t test_vneg_v_i16mf4_m (vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrsub.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrsub.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vneg_v_i16mf2_m (vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, size_t vl) {
@@ -318,7 +318,7 @@ vint16mf2_t test_vneg_v_i16mf2_m (vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrsub.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrsub.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vneg_v_i16m1_m (vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, size_t vl) {
@@ -328,7 +328,7 @@ vint16m1_t test_vneg_v_i16m1_m (vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrsub.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrsub.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vneg_v_i16m2_m (vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, size_t vl) {
@@ -338,7 +338,7 @@ vint16m2_t test_vneg_v_i16m2_m (vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrsub.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrsub.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vneg_v_i16m4_m (vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, size_t vl) {
@@ -348,7 +348,7 @@ vint16m4_t test_vneg_v_i16m4_m (vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrsub.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 0, <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrsub.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 0, <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vneg_v_i16m8_m (vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, size_t vl) {
@@ -358,7 +358,7 @@ vint16m8_t test_vneg_v_i16m8_m (vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrsub.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrsub.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vneg_v_i32mf2_m (vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, size_t vl) {
@@ -368,7 +368,7 @@ vint32mf2_t test_vneg_v_i32mf2_m (vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrsub.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrsub.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vneg_v_i32m1_m (vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, size_t vl) {
@@ -378,7 +378,7 @@ vint32m1_t test_vneg_v_i32m1_m (vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrsub.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrsub.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vneg_v_i32m2_m (vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, size_t vl) {
@@ -388,7 +388,7 @@ vint32m2_t test_vneg_v_i32m2_m (vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrsub.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrsub.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vneg_v_i32m4_m (vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, size_t vl) {
@@ -398,7 +398,7 @@ vint32m4_t test_vneg_v_i32m4_m (vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrsub.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrsub.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vneg_v_i32m8_m (vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, size_t vl) {
@@ -408,7 +408,7 @@ vint32m8_t test_vneg_v_i32m8_m (vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrsub.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrsub.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vneg_v_i64m1_m (vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, size_t vl) {
@@ -418,7 +418,7 @@ vint64m1_t test_vneg_v_i64m1_m (vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrsub.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrsub.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vneg_v_i64m2_m (vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, size_t vl) {
@@ -428,7 +428,7 @@ vint64m2_t test_vneg_v_i64m2_m (vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrsub.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrsub.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vneg_v_i64m4_m (vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, size_t vl) {
@@ -438,7 +438,7 @@ vint64m4_t test_vneg_v_i64m4_m (vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vneg_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrsub.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrsub.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vneg_v_i64m8_m (vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnmsac.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnmsac.c
index f8b12b2..1a4a68b 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnmsac.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnmsac.c
@@ -887,7 +887,7 @@ vuint64m8_t test_vnmsac_vx_u64m8(vuint64m8_t acc, uint64_t op1, vuint64m8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnmsac.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnmsac.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vnmsac_vv_i8mf8_m(vbool64_t mask, vint8mf8_t acc, vint8mf8_t op1, vint8mf8_t op2, size_t vl) {
@@ -897,7 +897,7 @@ vint8mf8_t test_vnmsac_vv_i8mf8_m(vbool64_t mask, vint8mf8_t acc, vint8mf8_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnmsac.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnmsac.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vnmsac_vx_i8mf8_m(vbool64_t mask, vint8mf8_t acc, int8_t op1, vint8mf8_t op2, size_t vl) {
@@ -907,7 +907,7 @@ vint8mf8_t test_vnmsac_vx_i8mf8_m(vbool64_t mask, vint8mf8_t acc, int8_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnmsac.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnmsac.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vnmsac_vv_i8mf4_m(vbool32_t mask, vint8mf4_t acc, vint8mf4_t op1, vint8mf4_t op2, size_t vl) {
@@ -917,7 +917,7 @@ vint8mf4_t test_vnmsac_vv_i8mf4_m(vbool32_t mask, vint8mf4_t acc, vint8mf4_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnmsac.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnmsac.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vnmsac_vx_i8mf4_m(vbool32_t mask, vint8mf4_t acc, int8_t op1, vint8mf4_t op2, size_t vl) {
@@ -927,7 +927,7 @@ vint8mf4_t test_vnmsac_vx_i8mf4_m(vbool32_t mask, vint8mf4_t acc, int8_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnmsac.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnmsac.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vnmsac_vv_i8mf2_m(vbool16_t mask, vint8mf2_t acc, vint8mf2_t op1, vint8mf2_t op2, size_t vl) {
@@ -937,7 +937,7 @@ vint8mf2_t test_vnmsac_vv_i8mf2_m(vbool16_t mask, vint8mf2_t acc, vint8mf2_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnmsac.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnmsac.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vnmsac_vx_i8mf2_m(vbool16_t mask, vint8mf2_t acc, int8_t op1, vint8mf2_t op2, size_t vl) {
@@ -947,7 +947,7 @@ vint8mf2_t test_vnmsac_vx_i8mf2_m(vbool16_t mask, vint8mf2_t acc, int8_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnmsac.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnmsac.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vnmsac_vv_i8m1_m(vbool8_t mask, vint8m1_t acc, vint8m1_t op1, vint8m1_t op2, size_t vl) {
@@ -957,7 +957,7 @@ vint8m1_t test_vnmsac_vv_i8m1_m(vbool8_t mask, vint8m1_t acc, vint8m1_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnmsac.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnmsac.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vnmsac_vx_i8m1_m(vbool8_t mask, vint8m1_t acc, int8_t op1, vint8m1_t op2, size_t vl) {
@@ -967,7 +967,7 @@ vint8m1_t test_vnmsac_vx_i8m1_m(vbool8_t mask, vint8m1_t acc, int8_t op1, vint8m
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnmsac.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnmsac.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vnmsac_vv_i8m2_m(vbool4_t mask, vint8m2_t acc, vint8m2_t op1, vint8m2_t op2, size_t vl) {
@@ -977,7 +977,7 @@ vint8m2_t test_vnmsac_vv_i8m2_m(vbool4_t mask, vint8m2_t acc, vint8m2_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnmsac.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnmsac.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vnmsac_vx_i8m2_m(vbool4_t mask, vint8m2_t acc, int8_t op1, vint8m2_t op2, size_t vl) {
@@ -987,7 +987,7 @@ vint8m2_t test_vnmsac_vx_i8m2_m(vbool4_t mask, vint8m2_t acc, int8_t op1, vint8m
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnmsac.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnmsac.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vnmsac_vv_i8m4_m(vbool2_t mask, vint8m4_t acc, vint8m4_t op1, vint8m4_t op2, size_t vl) {
@@ -997,7 +997,7 @@ vint8m4_t test_vnmsac_vv_i8m4_m(vbool2_t mask, vint8m4_t acc, vint8m4_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnmsac.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnmsac.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vnmsac_vx_i8m4_m(vbool2_t mask, vint8m4_t acc, int8_t op1, vint8m4_t op2, size_t vl) {
@@ -1007,7 +1007,7 @@ vint8m4_t test_vnmsac_vx_i8m4_m(vbool2_t mask, vint8m4_t acc, int8_t op1, vint8m
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vnmsac.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[ACC:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vnmsac.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[ACC:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vnmsac_vv_i8m8_m(vbool1_t mask, vint8m8_t acc, vint8m8_t op1, vint8m8_t op2, size_t vl) {
@@ -1017,7 +1017,7 @@ vint8m8_t test_vnmsac_vv_i8m8_m(vbool1_t mask, vint8m8_t acc, vint8m8_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vnmsac.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vnmsac.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vnmsac_vx_i8m8_m(vbool1_t mask, vint8m8_t acc, int8_t op1, vint8m8_t op2, size_t vl) {
@@ -1027,7 +1027,7 @@ vint8m8_t test_vnmsac_vx_i8m8_m(vbool1_t mask, vint8m8_t acc, int8_t op1, vint8m
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnmsac.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnmsac.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vnmsac_vv_i16mf4_m(vbool64_t mask, vint16mf4_t acc, vint16mf4_t op1, vint16mf4_t op2, size_t vl) {
@@ -1037,7 +1037,7 @@ vint16mf4_t test_vnmsac_vv_i16mf4_m(vbool64_t mask, vint16mf4_t acc, vint16mf4_t
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnmsac.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnmsac.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vnmsac_vx_i16mf4_m(vbool64_t mask, vint16mf4_t acc, int16_t op1, vint16mf4_t op2, size_t vl) {
@@ -1047,7 +1047,7 @@ vint16mf4_t test_vnmsac_vx_i16mf4_m(vbool64_t mask, vint16mf4_t acc, int16_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnmsac.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnmsac.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vnmsac_vv_i16mf2_m(vbool32_t mask, vint16mf2_t acc, vint16mf2_t op1, vint16mf2_t op2, size_t vl) {
@@ -1057,7 +1057,7 @@ vint16mf2_t test_vnmsac_vv_i16mf2_m(vbool32_t mask, vint16mf2_t acc, vint16mf2_t
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnmsac.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnmsac.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vnmsac_vx_i16mf2_m(vbool32_t mask, vint16mf2_t acc, int16_t op1, vint16mf2_t op2, size_t vl) {
@@ -1067,7 +1067,7 @@ vint16mf2_t test_vnmsac_vx_i16mf2_m(vbool32_t mask, vint16mf2_t acc, int16_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnmsac.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnmsac.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vnmsac_vv_i16m1_m(vbool16_t mask, vint16m1_t acc, vint16m1_t op1, vint16m1_t op2, size_t vl) {
@@ -1077,7 +1077,7 @@ vint16m1_t test_vnmsac_vv_i16m1_m(vbool16_t mask, vint16m1_t acc, vint16m1_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnmsac.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnmsac.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vnmsac_vx_i16m1_m(vbool16_t mask, vint16m1_t acc, int16_t op1, vint16m1_t op2, size_t vl) {
@@ -1087,7 +1087,7 @@ vint16m1_t test_vnmsac_vx_i16m1_m(vbool16_t mask, vint16m1_t acc, int16_t op1, v
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnmsac.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnmsac.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vnmsac_vv_i16m2_m(vbool8_t mask, vint16m2_t acc, vint16m2_t op1, vint16m2_t op2, size_t vl) {
@@ -1097,7 +1097,7 @@ vint16m2_t test_vnmsac_vv_i16m2_m(vbool8_t mask, vint16m2_t acc, vint16m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnmsac.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnmsac.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vnmsac_vx_i16m2_m(vbool8_t mask, vint16m2_t acc, int16_t op1, vint16m2_t op2, size_t vl) {
@@ -1107,7 +1107,7 @@ vint16m2_t test_vnmsac_vx_i16m2_m(vbool8_t mask, vint16m2_t acc, int16_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnmsac.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnmsac.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vnmsac_vv_i16m4_m(vbool4_t mask, vint16m4_t acc, vint16m4_t op1, vint16m4_t op2, size_t vl) {
@@ -1117,7 +1117,7 @@ vint16m4_t test_vnmsac_vv_i16m4_m(vbool4_t mask, vint16m4_t acc, vint16m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnmsac.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnmsac.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vnmsac_vx_i16m4_m(vbool4_t mask, vint16m4_t acc, int16_t op1, vint16m4_t op2, size_t vl) {
@@ -1127,7 +1127,7 @@ vint16m4_t test_vnmsac_vx_i16m4_m(vbool4_t mask, vint16m4_t acc, int16_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vnmsac.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vnmsac.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vnmsac_vv_i16m8_m(vbool2_t mask, vint16m8_t acc, vint16m8_t op1, vint16m8_t op2, size_t vl) {
@@ -1137,7 +1137,7 @@ vint16m8_t test_vnmsac_vv_i16m8_m(vbool2_t mask, vint16m8_t acc, vint16m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vnmsac.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vnmsac.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vnmsac_vx_i16m8_m(vbool2_t mask, vint16m8_t acc, int16_t op1, vint16m8_t op2, size_t vl) {
@@ -1147,7 +1147,7 @@ vint16m8_t test_vnmsac_vx_i16m8_m(vbool2_t mask, vint16m8_t acc, int16_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnmsac.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnmsac.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vnmsac_vv_i32mf2_m(vbool64_t mask, vint32mf2_t acc, vint32mf2_t op1, vint32mf2_t op2, size_t vl) {
@@ -1157,7 +1157,7 @@ vint32mf2_t test_vnmsac_vv_i32mf2_m(vbool64_t mask, vint32mf2_t acc, vint32mf2_t
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnmsac.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnmsac.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vnmsac_vx_i32mf2_m(vbool64_t mask, vint32mf2_t acc, int32_t op1, vint32mf2_t op2, size_t vl) {
@@ -1167,7 +1167,7 @@ vint32mf2_t test_vnmsac_vx_i32mf2_m(vbool64_t mask, vint32mf2_t acc, int32_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnmsac.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnmsac.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vnmsac_vv_i32m1_m(vbool32_t mask, vint32m1_t acc, vint32m1_t op1, vint32m1_t op2, size_t vl) {
@@ -1177,7 +1177,7 @@ vint32m1_t test_vnmsac_vv_i32m1_m(vbool32_t mask, vint32m1_t acc, vint32m1_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnmsac.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnmsac.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vnmsac_vx_i32m1_m(vbool32_t mask, vint32m1_t acc, int32_t op1, vint32m1_t op2, size_t vl) {
@@ -1187,7 +1187,7 @@ vint32m1_t test_vnmsac_vx_i32m1_m(vbool32_t mask, vint32m1_t acc, int32_t op1, v
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnmsac.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnmsac.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vnmsac_vv_i32m2_m(vbool16_t mask, vint32m2_t acc, vint32m2_t op1, vint32m2_t op2, size_t vl) {
@@ -1197,7 +1197,7 @@ vint32m2_t test_vnmsac_vv_i32m2_m(vbool16_t mask, vint32m2_t acc, vint32m2_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnmsac.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnmsac.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vnmsac_vx_i32m2_m(vbool16_t mask, vint32m2_t acc, int32_t op1, vint32m2_t op2, size_t vl) {
@@ -1207,7 +1207,7 @@ vint32m2_t test_vnmsac_vx_i32m2_m(vbool16_t mask, vint32m2_t acc, int32_t op1, v
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnmsac.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnmsac.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vnmsac_vv_i32m4_m(vbool8_t mask, vint32m4_t acc, vint32m4_t op1, vint32m4_t op2, size_t vl) {
@@ -1217,7 +1217,7 @@ vint32m4_t test_vnmsac_vv_i32m4_m(vbool8_t mask, vint32m4_t acc, vint32m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnmsac.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnmsac.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vnmsac_vx_i32m4_m(vbool8_t mask, vint32m4_t acc, int32_t op1, vint32m4_t op2, size_t vl) {
@@ -1227,7 +1227,7 @@ vint32m4_t test_vnmsac_vx_i32m4_m(vbool8_t mask, vint32m4_t acc, int32_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vnmsac.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vnmsac.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vnmsac_vv_i32m8_m(vbool4_t mask, vint32m8_t acc, vint32m8_t op1, vint32m8_t op2, size_t vl) {
@@ -1237,7 +1237,7 @@ vint32m8_t test_vnmsac_vv_i32m8_m(vbool4_t mask, vint32m8_t acc, vint32m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vnmsac.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vnmsac.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vnmsac_vx_i32m8_m(vbool4_t mask, vint32m8_t acc, int32_t op1, vint32m8_t op2, size_t vl) {
@@ -1247,7 +1247,7 @@ vint32m8_t test_vnmsac_vx_i32m8_m(vbool4_t mask, vint32m8_t acc, int32_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vnmsac.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vnmsac.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vnmsac_vv_i64m1_m(vbool64_t mask, vint64m1_t acc, vint64m1_t op1, vint64m1_t op2, size_t vl) {
@@ -1257,7 +1257,7 @@ vint64m1_t test_vnmsac_vv_i64m1_m(vbool64_t mask, vint64m1_t acc, vint64m1_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vnmsac.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vnmsac.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vnmsac_vx_i64m1_m(vbool64_t mask, vint64m1_t acc, int64_t op1, vint64m1_t op2, size_t vl) {
@@ -1267,7 +1267,7 @@ vint64m1_t test_vnmsac_vx_i64m1_m(vbool64_t mask, vint64m1_t acc, int64_t op1, v
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vnmsac.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vnmsac.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vnmsac_vv_i64m2_m(vbool32_t mask, vint64m2_t acc, vint64m2_t op1, vint64m2_t op2, size_t vl) {
@@ -1277,7 +1277,7 @@ vint64m2_t test_vnmsac_vv_i64m2_m(vbool32_t mask, vint64m2_t acc, vint64m2_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vnmsac.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vnmsac.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vnmsac_vx_i64m2_m(vbool32_t mask, vint64m2_t acc, int64_t op1, vint64m2_t op2, size_t vl) {
@@ -1287,7 +1287,7 @@ vint64m2_t test_vnmsac_vx_i64m2_m(vbool32_t mask, vint64m2_t acc, int64_t op1, v
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vnmsac.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vnmsac.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vnmsac_vv_i64m4_m(vbool16_t mask, vint64m4_t acc, vint64m4_t op1, vint64m4_t op2, size_t vl) {
@@ -1297,7 +1297,7 @@ vint64m4_t test_vnmsac_vv_i64m4_m(vbool16_t mask, vint64m4_t acc, vint64m4_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vnmsac.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vnmsac.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vnmsac_vx_i64m4_m(vbool16_t mask, vint64m4_t acc, int64_t op1, vint64m4_t op2, size_t vl) {
@@ -1307,7 +1307,7 @@ vint64m4_t test_vnmsac_vx_i64m4_m(vbool16_t mask, vint64m4_t acc, int64_t op1, v
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vnmsac.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vnmsac.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vnmsac_vv_i64m8_m(vbool8_t mask, vint64m8_t acc, vint64m8_t op1, vint64m8_t op2, size_t vl) {
@@ -1317,7 +1317,7 @@ vint64m8_t test_vnmsac_vv_i64m8_m(vbool8_t mask, vint64m8_t acc, vint64m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vnmsac.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vnmsac.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vnmsac_vx_i64m8_m(vbool8_t mask, vint64m8_t acc, int64_t op1, vint64m8_t op2, size_t vl) {
@@ -1327,7 +1327,7 @@ vint64m8_t test_vnmsac_vx_i64m8_m(vbool8_t mask, vint64m8_t acc, int64_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnmsac.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnmsac.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vnmsac_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t acc, vuint8mf8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -1337,7 +1337,7 @@ vuint8mf8_t test_vnmsac_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t acc, vuint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnmsac.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnmsac.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vnmsac_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t acc, uint8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -1347,7 +1347,7 @@ vuint8mf8_t test_vnmsac_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t acc, uint8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnmsac.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnmsac.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vnmsac_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t acc, vuint8mf4_t op1, vuint8mf4_t op2, size_t vl) {
@@ -1357,7 +1357,7 @@ vuint8mf4_t test_vnmsac_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t acc, vuint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnmsac.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnmsac.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vnmsac_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t acc, uint8_t op1, vuint8mf4_t op2, size_t vl) {
@@ -1367,7 +1367,7 @@ vuint8mf4_t test_vnmsac_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t acc, uint8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnmsac.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnmsac.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vnmsac_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t acc, vuint8mf2_t op1, vuint8mf2_t op2, size_t vl) {
@@ -1377,7 +1377,7 @@ vuint8mf2_t test_vnmsac_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t acc, vuint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnmsac.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnmsac.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vnmsac_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t acc, uint8_t op1, vuint8mf2_t op2, size_t vl) {
@@ -1387,7 +1387,7 @@ vuint8mf2_t test_vnmsac_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t acc, uint8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnmsac.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnmsac.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vnmsac_vv_u8m1_m(vbool8_t mask, vuint8m1_t acc, vuint8m1_t op1, vuint8m1_t op2, size_t vl) {
@@ -1397,7 +1397,7 @@ vuint8m1_t test_vnmsac_vv_u8m1_m(vbool8_t mask, vuint8m1_t acc, vuint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnmsac.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnmsac.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vnmsac_vx_u8m1_m(vbool8_t mask, vuint8m1_t acc, uint8_t op1, vuint8m1_t op2, size_t vl) {
@@ -1407,7 +1407,7 @@ vuint8m1_t test_vnmsac_vx_u8m1_m(vbool8_t mask, vuint8m1_t acc, uint8_t op1, vui
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnmsac.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnmsac.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vnmsac_vv_u8m2_m(vbool4_t mask, vuint8m2_t acc, vuint8m2_t op1, vuint8m2_t op2, size_t vl) {
@@ -1417,7 +1417,7 @@ vuint8m2_t test_vnmsac_vv_u8m2_m(vbool4_t mask, vuint8m2_t acc, vuint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnmsac.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnmsac.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vnmsac_vx_u8m2_m(vbool4_t mask, vuint8m2_t acc, uint8_t op1, vuint8m2_t op2, size_t vl) {
@@ -1427,7 +1427,7 @@ vuint8m2_t test_vnmsac_vx_u8m2_m(vbool4_t mask, vuint8m2_t acc, uint8_t op1, vui
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnmsac.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnmsac.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vnmsac_vv_u8m4_m(vbool2_t mask, vuint8m4_t acc, vuint8m4_t op1, vuint8m4_t op2, size_t vl) {
@@ -1437,7 +1437,7 @@ vuint8m4_t test_vnmsac_vv_u8m4_m(vbool2_t mask, vuint8m4_t acc, vuint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnmsac.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnmsac.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vnmsac_vx_u8m4_m(vbool2_t mask, vuint8m4_t acc, uint8_t op1, vuint8m4_t op2, size_t vl) {
@@ -1447,7 +1447,7 @@ vuint8m4_t test_vnmsac_vx_u8m4_m(vbool2_t mask, vuint8m4_t acc, uint8_t op1, vui
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vnmsac.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[ACC:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vnmsac.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[ACC:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vnmsac_vv_u8m8_m(vbool1_t mask, vuint8m8_t acc, vuint8m8_t op1, vuint8m8_t op2, size_t vl) {
@@ -1457,7 +1457,7 @@ vuint8m8_t test_vnmsac_vv_u8m8_m(vbool1_t mask, vuint8m8_t acc, vuint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vnmsac.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vnmsac.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vnmsac_vx_u8m8_m(vbool1_t mask, vuint8m8_t acc, uint8_t op1, vuint8m8_t op2, size_t vl) {
@@ -1467,7 +1467,7 @@ vuint8m8_t test_vnmsac_vx_u8m8_m(vbool1_t mask, vuint8m8_t acc, uint8_t op1, vui
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnmsac.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnmsac.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vnmsac_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t acc, vuint16mf4_t op1, vuint16mf4_t op2, size_t vl) {
@@ -1477,7 +1477,7 @@ vuint16mf4_t test_vnmsac_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t acc, vuint16mf
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnmsac.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnmsac.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vnmsac_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t acc, uint16_t op1, vuint16mf4_t op2, size_t vl) {
@@ -1487,7 +1487,7 @@ vuint16mf4_t test_vnmsac_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t acc, uint16_t
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnmsac.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnmsac.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vnmsac_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t acc, vuint16mf2_t op1, vuint16mf2_t op2, size_t vl) {
@@ -1497,7 +1497,7 @@ vuint16mf2_t test_vnmsac_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t acc, vuint16mf
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnmsac.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnmsac.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vnmsac_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t acc, uint16_t op1, vuint16mf2_t op2, size_t vl) {
@@ -1507,7 +1507,7 @@ vuint16mf2_t test_vnmsac_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t acc, uint16_t
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnmsac.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnmsac.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vnmsac_vv_u16m1_m(vbool16_t mask, vuint16m1_t acc, vuint16m1_t op1, vuint16m1_t op2, size_t vl) {
@@ -1517,7 +1517,7 @@ vuint16m1_t test_vnmsac_vv_u16m1_m(vbool16_t mask, vuint16m1_t acc, vuint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnmsac.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnmsac.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vnmsac_vx_u16m1_m(vbool16_t mask, vuint16m1_t acc, uint16_t op1, vuint16m1_t op2, size_t vl) {
@@ -1527,7 +1527,7 @@ vuint16m1_t test_vnmsac_vx_u16m1_m(vbool16_t mask, vuint16m1_t acc, uint16_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnmsac.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnmsac.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vnmsac_vv_u16m2_m(vbool8_t mask, vuint16m2_t acc, vuint16m2_t op1, vuint16m2_t op2, size_t vl) {
@@ -1537,7 +1537,7 @@ vuint16m2_t test_vnmsac_vv_u16m2_m(vbool8_t mask, vuint16m2_t acc, vuint16m2_t o
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnmsac.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnmsac.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vnmsac_vx_u16m2_m(vbool8_t mask, vuint16m2_t acc, uint16_t op1, vuint16m2_t op2, size_t vl) {
@@ -1547,7 +1547,7 @@ vuint16m2_t test_vnmsac_vx_u16m2_m(vbool8_t mask, vuint16m2_t acc, uint16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnmsac.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnmsac.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vnmsac_vv_u16m4_m(vbool4_t mask, vuint16m4_t acc, vuint16m4_t op1, vuint16m4_t op2, size_t vl) {
@@ -1557,7 +1557,7 @@ vuint16m4_t test_vnmsac_vv_u16m4_m(vbool4_t mask, vuint16m4_t acc, vuint16m4_t o
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnmsac.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnmsac.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vnmsac_vx_u16m4_m(vbool4_t mask, vuint16m4_t acc, uint16_t op1, vuint16m4_t op2, size_t vl) {
@@ -1567,7 +1567,7 @@ vuint16m4_t test_vnmsac_vx_u16m4_m(vbool4_t mask, vuint16m4_t acc, uint16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vnmsac.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vnmsac.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vnmsac_vv_u16m8_m(vbool2_t mask, vuint16m8_t acc, vuint16m8_t op1, vuint16m8_t op2, size_t vl) {
@@ -1577,7 +1577,7 @@ vuint16m8_t test_vnmsac_vv_u16m8_m(vbool2_t mask, vuint16m8_t acc, vuint16m8_t o
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vnmsac.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vnmsac.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vnmsac_vx_u16m8_m(vbool2_t mask, vuint16m8_t acc, uint16_t op1, vuint16m8_t op2, size_t vl) {
@@ -1587,7 +1587,7 @@ vuint16m8_t test_vnmsac_vx_u16m8_m(vbool2_t mask, vuint16m8_t acc, uint16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnmsac.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnmsac.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vnmsac_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t acc, vuint32mf2_t op1, vuint32mf2_t op2, size_t vl) {
@@ -1597,7 +1597,7 @@ vuint32mf2_t test_vnmsac_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t acc, vuint32mf
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnmsac.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnmsac.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vnmsac_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t acc, uint32_t op1, vuint32mf2_t op2, size_t vl) {
@@ -1607,7 +1607,7 @@ vuint32mf2_t test_vnmsac_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t acc, uint32_t
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnmsac.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnmsac.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vnmsac_vv_u32m1_m(vbool32_t mask, vuint32m1_t acc, vuint32m1_t op1, vuint32m1_t op2, size_t vl) {
@@ -1617,7 +1617,7 @@ vuint32m1_t test_vnmsac_vv_u32m1_m(vbool32_t mask, vuint32m1_t acc, vuint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnmsac.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnmsac.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vnmsac_vx_u32m1_m(vbool32_t mask, vuint32m1_t acc, uint32_t op1, vuint32m1_t op2, size_t vl) {
@@ -1627,7 +1627,7 @@ vuint32m1_t test_vnmsac_vx_u32m1_m(vbool32_t mask, vuint32m1_t acc, uint32_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnmsac.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnmsac.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vnmsac_vv_u32m2_m(vbool16_t mask, vuint32m2_t acc, vuint32m2_t op1, vuint32m2_t op2, size_t vl) {
@@ -1637,7 +1637,7 @@ vuint32m2_t test_vnmsac_vv_u32m2_m(vbool16_t mask, vuint32m2_t acc, vuint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnmsac.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnmsac.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vnmsac_vx_u32m2_m(vbool16_t mask, vuint32m2_t acc, uint32_t op1, vuint32m2_t op2, size_t vl) {
@@ -1647,7 +1647,7 @@ vuint32m2_t test_vnmsac_vx_u32m2_m(vbool16_t mask, vuint32m2_t acc, uint32_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnmsac.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnmsac.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vnmsac_vv_u32m4_m(vbool8_t mask, vuint32m4_t acc, vuint32m4_t op1, vuint32m4_t op2, size_t vl) {
@@ -1657,7 +1657,7 @@ vuint32m4_t test_vnmsac_vv_u32m4_m(vbool8_t mask, vuint32m4_t acc, vuint32m4_t o
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnmsac.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnmsac.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vnmsac_vx_u32m4_m(vbool8_t mask, vuint32m4_t acc, uint32_t op1, vuint32m4_t op2, size_t vl) {
@@ -1667,7 +1667,7 @@ vuint32m4_t test_vnmsac_vx_u32m4_m(vbool8_t mask, vuint32m4_t acc, uint32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vnmsac.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vnmsac.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vnmsac_vv_u32m8_m(vbool4_t mask, vuint32m8_t acc, vuint32m8_t op1, vuint32m8_t op2, size_t vl) {
@@ -1677,7 +1677,7 @@ vuint32m8_t test_vnmsac_vv_u32m8_m(vbool4_t mask, vuint32m8_t acc, vuint32m8_t o
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vnmsac.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vnmsac.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vnmsac_vx_u32m8_m(vbool4_t mask, vuint32m8_t acc, uint32_t op1, vuint32m8_t op2, size_t vl) {
@@ -1687,7 +1687,7 @@ vuint32m8_t test_vnmsac_vx_u32m8_m(vbool4_t mask, vuint32m8_t acc, uint32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vnmsac.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vnmsac.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vnmsac_vv_u64m1_m(vbool64_t mask, vuint64m1_t acc, vuint64m1_t op1, vuint64m1_t op2, size_t vl) {
@@ -1697,7 +1697,7 @@ vuint64m1_t test_vnmsac_vv_u64m1_m(vbool64_t mask, vuint64m1_t acc, vuint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vnmsac.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vnmsac.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vnmsac_vx_u64m1_m(vbool64_t mask, vuint64m1_t acc, uint64_t op1, vuint64m1_t op2, size_t vl) {
@@ -1707,7 +1707,7 @@ vuint64m1_t test_vnmsac_vx_u64m1_m(vbool64_t mask, vuint64m1_t acc, uint64_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vnmsac.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vnmsac.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vnmsac_vv_u64m2_m(vbool32_t mask, vuint64m2_t acc, vuint64m2_t op1, vuint64m2_t op2, size_t vl) {
@@ -1717,7 +1717,7 @@ vuint64m2_t test_vnmsac_vv_u64m2_m(vbool32_t mask, vuint64m2_t acc, vuint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vnmsac.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vnmsac.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vnmsac_vx_u64m2_m(vbool32_t mask, vuint64m2_t acc, uint64_t op1, vuint64m2_t op2, size_t vl) {
@@ -1727,7 +1727,7 @@ vuint64m2_t test_vnmsac_vx_u64m2_m(vbool32_t mask, vuint64m2_t acc, uint64_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vnmsac.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vnmsac.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vnmsac_vv_u64m4_m(vbool16_t mask, vuint64m4_t acc, vuint64m4_t op1, vuint64m4_t op2, size_t vl) {
@@ -1737,7 +1737,7 @@ vuint64m4_t test_vnmsac_vv_u64m4_m(vbool16_t mask, vuint64m4_t acc, vuint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vnmsac.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vnmsac.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vnmsac_vx_u64m4_m(vbool16_t mask, vuint64m4_t acc, uint64_t op1, vuint64m4_t op2, size_t vl) {
@@ -1747,7 +1747,7 @@ vuint64m4_t test_vnmsac_vx_u64m4_m(vbool16_t mask, vuint64m4_t acc, uint64_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vnmsac.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vnmsac.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vnmsac_vv_u64m8_m(vbool8_t mask, vuint64m8_t acc, vuint64m8_t op1, vuint64m8_t op2, size_t vl) {
@@ -1757,7 +1757,7 @@ vuint64m8_t test_vnmsac_vv_u64m8_m(vbool8_t mask, vuint64m8_t acc, vuint64m8_t o
 //
 // CHECK-RV64-LABEL: @test_vnmsac_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vnmsac.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vnmsac.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vnmsac_vx_u64m8_m(vbool8_t mask, vuint64m8_t acc, uint64_t op1, vuint64m8_t op2, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnmsub.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnmsub.c
index 2106b23..2e6c5e3 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnmsub.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnmsub.c
@@ -887,7 +887,7 @@ vuint64m8_t test_vnmsub_vx_u64m8(vuint64m8_t acc, uint64_t op1, vuint64m8_t op2,
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnmsub.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnmsub.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vnmsub_vv_i8mf8_m(vbool64_t mask, vint8mf8_t acc, vint8mf8_t op1, vint8mf8_t op2, size_t vl) {
@@ -897,7 +897,7 @@ vint8mf8_t test_vnmsub_vv_i8mf8_m(vbool64_t mask, vint8mf8_t acc, vint8mf8_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnmsub.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnmsub.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vnmsub_vx_i8mf8_m(vbool64_t mask, vint8mf8_t acc, int8_t op1, vint8mf8_t op2, size_t vl) {
@@ -907,7 +907,7 @@ vint8mf8_t test_vnmsub_vx_i8mf8_m(vbool64_t mask, vint8mf8_t acc, int8_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnmsub.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnmsub.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vnmsub_vv_i8mf4_m(vbool32_t mask, vint8mf4_t acc, vint8mf4_t op1, vint8mf4_t op2, size_t vl) {
@@ -917,7 +917,7 @@ vint8mf4_t test_vnmsub_vv_i8mf4_m(vbool32_t mask, vint8mf4_t acc, vint8mf4_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnmsub.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnmsub.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vnmsub_vx_i8mf4_m(vbool32_t mask, vint8mf4_t acc, int8_t op1, vint8mf4_t op2, size_t vl) {
@@ -927,7 +927,7 @@ vint8mf4_t test_vnmsub_vx_i8mf4_m(vbool32_t mask, vint8mf4_t acc, int8_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnmsub.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnmsub.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vnmsub_vv_i8mf2_m(vbool16_t mask, vint8mf2_t acc, vint8mf2_t op1, vint8mf2_t op2, size_t vl) {
@@ -937,7 +937,7 @@ vint8mf2_t test_vnmsub_vv_i8mf2_m(vbool16_t mask, vint8mf2_t acc, vint8mf2_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnmsub.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnmsub.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vnmsub_vx_i8mf2_m(vbool16_t mask, vint8mf2_t acc, int8_t op1, vint8mf2_t op2, size_t vl) {
@@ -947,7 +947,7 @@ vint8mf2_t test_vnmsub_vx_i8mf2_m(vbool16_t mask, vint8mf2_t acc, int8_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnmsub.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnmsub.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vnmsub_vv_i8m1_m(vbool8_t mask, vint8m1_t acc, vint8m1_t op1, vint8m1_t op2, size_t vl) {
@@ -957,7 +957,7 @@ vint8m1_t test_vnmsub_vv_i8m1_m(vbool8_t mask, vint8m1_t acc, vint8m1_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnmsub.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnmsub.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vnmsub_vx_i8m1_m(vbool8_t mask, vint8m1_t acc, int8_t op1, vint8m1_t op2, size_t vl) {
@@ -967,7 +967,7 @@ vint8m1_t test_vnmsub_vx_i8m1_m(vbool8_t mask, vint8m1_t acc, int8_t op1, vint8m
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnmsub.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnmsub.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vnmsub_vv_i8m2_m(vbool4_t mask, vint8m2_t acc, vint8m2_t op1, vint8m2_t op2, size_t vl) {
@@ -977,7 +977,7 @@ vint8m2_t test_vnmsub_vv_i8m2_m(vbool4_t mask, vint8m2_t acc, vint8m2_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnmsub.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnmsub.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vnmsub_vx_i8m2_m(vbool4_t mask, vint8m2_t acc, int8_t op1, vint8m2_t op2, size_t vl) {
@@ -987,7 +987,7 @@ vint8m2_t test_vnmsub_vx_i8m2_m(vbool4_t mask, vint8m2_t acc, int8_t op1, vint8m
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnmsub.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnmsub.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vnmsub_vv_i8m4_m(vbool2_t mask, vint8m4_t acc, vint8m4_t op1, vint8m4_t op2, size_t vl) {
@@ -997,7 +997,7 @@ vint8m4_t test_vnmsub_vv_i8m4_m(vbool2_t mask, vint8m4_t acc, vint8m4_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnmsub.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnmsub.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vnmsub_vx_i8m4_m(vbool2_t mask, vint8m4_t acc, int8_t op1, vint8m4_t op2, size_t vl) {
@@ -1007,7 +1007,7 @@ vint8m4_t test_vnmsub_vx_i8m4_m(vbool2_t mask, vint8m4_t acc, int8_t op1, vint8m
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vnmsub.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[ACC:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vnmsub.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[ACC:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vnmsub_vv_i8m8_m(vbool1_t mask, vint8m8_t acc, vint8m8_t op1, vint8m8_t op2, size_t vl) {
@@ -1017,7 +1017,7 @@ vint8m8_t test_vnmsub_vv_i8m8_m(vbool1_t mask, vint8m8_t acc, vint8m8_t op1, vin
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vnmsub.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vnmsub.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vnmsub_vx_i8m8_m(vbool1_t mask, vint8m8_t acc, int8_t op1, vint8m8_t op2, size_t vl) {
@@ -1027,7 +1027,7 @@ vint8m8_t test_vnmsub_vx_i8m8_m(vbool1_t mask, vint8m8_t acc, int8_t op1, vint8m
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnmsub.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnmsub.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vnmsub_vv_i16mf4_m(vbool64_t mask, vint16mf4_t acc, vint16mf4_t op1, vint16mf4_t op2, size_t vl) {
@@ -1037,7 +1037,7 @@ vint16mf4_t test_vnmsub_vv_i16mf4_m(vbool64_t mask, vint16mf4_t acc, vint16mf4_t
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnmsub.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnmsub.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vnmsub_vx_i16mf4_m(vbool64_t mask, vint16mf4_t acc, int16_t op1, vint16mf4_t op2, size_t vl) {
@@ -1047,7 +1047,7 @@ vint16mf4_t test_vnmsub_vx_i16mf4_m(vbool64_t mask, vint16mf4_t acc, int16_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnmsub.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnmsub.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vnmsub_vv_i16mf2_m(vbool32_t mask, vint16mf2_t acc, vint16mf2_t op1, vint16mf2_t op2, size_t vl) {
@@ -1057,7 +1057,7 @@ vint16mf2_t test_vnmsub_vv_i16mf2_m(vbool32_t mask, vint16mf2_t acc, vint16mf2_t
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnmsub.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnmsub.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vnmsub_vx_i16mf2_m(vbool32_t mask, vint16mf2_t acc, int16_t op1, vint16mf2_t op2, size_t vl) {
@@ -1067,7 +1067,7 @@ vint16mf2_t test_vnmsub_vx_i16mf2_m(vbool32_t mask, vint16mf2_t acc, int16_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnmsub.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnmsub.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vnmsub_vv_i16m1_m(vbool16_t mask, vint16m1_t acc, vint16m1_t op1, vint16m1_t op2, size_t vl) {
@@ -1077,7 +1077,7 @@ vint16m1_t test_vnmsub_vv_i16m1_m(vbool16_t mask, vint16m1_t acc, vint16m1_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnmsub.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnmsub.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vnmsub_vx_i16m1_m(vbool16_t mask, vint16m1_t acc, int16_t op1, vint16m1_t op2, size_t vl) {
@@ -1087,7 +1087,7 @@ vint16m1_t test_vnmsub_vx_i16m1_m(vbool16_t mask, vint16m1_t acc, int16_t op1, v
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnmsub.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnmsub.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vnmsub_vv_i16m2_m(vbool8_t mask, vint16m2_t acc, vint16m2_t op1, vint16m2_t op2, size_t vl) {
@@ -1097,7 +1097,7 @@ vint16m2_t test_vnmsub_vv_i16m2_m(vbool8_t mask, vint16m2_t acc, vint16m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnmsub.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnmsub.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vnmsub_vx_i16m2_m(vbool8_t mask, vint16m2_t acc, int16_t op1, vint16m2_t op2, size_t vl) {
@@ -1107,7 +1107,7 @@ vint16m2_t test_vnmsub_vx_i16m2_m(vbool8_t mask, vint16m2_t acc, int16_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnmsub.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnmsub.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vnmsub_vv_i16m4_m(vbool4_t mask, vint16m4_t acc, vint16m4_t op1, vint16m4_t op2, size_t vl) {
@@ -1117,7 +1117,7 @@ vint16m4_t test_vnmsub_vv_i16m4_m(vbool4_t mask, vint16m4_t acc, vint16m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnmsub.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnmsub.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vnmsub_vx_i16m4_m(vbool4_t mask, vint16m4_t acc, int16_t op1, vint16m4_t op2, size_t vl) {
@@ -1127,7 +1127,7 @@ vint16m4_t test_vnmsub_vx_i16m4_m(vbool4_t mask, vint16m4_t acc, int16_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vnmsub.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vnmsub.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vnmsub_vv_i16m8_m(vbool2_t mask, vint16m8_t acc, vint16m8_t op1, vint16m8_t op2, size_t vl) {
@@ -1137,7 +1137,7 @@ vint16m8_t test_vnmsub_vv_i16m8_m(vbool2_t mask, vint16m8_t acc, vint16m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vnmsub.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vnmsub.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vnmsub_vx_i16m8_m(vbool2_t mask, vint16m8_t acc, int16_t op1, vint16m8_t op2, size_t vl) {
@@ -1147,7 +1147,7 @@ vint16m8_t test_vnmsub_vx_i16m8_m(vbool2_t mask, vint16m8_t acc, int16_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnmsub.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnmsub.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vnmsub_vv_i32mf2_m(vbool64_t mask, vint32mf2_t acc, vint32mf2_t op1, vint32mf2_t op2, size_t vl) {
@@ -1157,7 +1157,7 @@ vint32mf2_t test_vnmsub_vv_i32mf2_m(vbool64_t mask, vint32mf2_t acc, vint32mf2_t
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnmsub.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnmsub.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vnmsub_vx_i32mf2_m(vbool64_t mask, vint32mf2_t acc, int32_t op1, vint32mf2_t op2, size_t vl) {
@@ -1167,7 +1167,7 @@ vint32mf2_t test_vnmsub_vx_i32mf2_m(vbool64_t mask, vint32mf2_t acc, int32_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnmsub.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnmsub.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vnmsub_vv_i32m1_m(vbool32_t mask, vint32m1_t acc, vint32m1_t op1, vint32m1_t op2, size_t vl) {
@@ -1177,7 +1177,7 @@ vint32m1_t test_vnmsub_vv_i32m1_m(vbool32_t mask, vint32m1_t acc, vint32m1_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnmsub.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnmsub.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vnmsub_vx_i32m1_m(vbool32_t mask, vint32m1_t acc, int32_t op1, vint32m1_t op2, size_t vl) {
@@ -1187,7 +1187,7 @@ vint32m1_t test_vnmsub_vx_i32m1_m(vbool32_t mask, vint32m1_t acc, int32_t op1, v
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnmsub.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnmsub.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vnmsub_vv_i32m2_m(vbool16_t mask, vint32m2_t acc, vint32m2_t op1, vint32m2_t op2, size_t vl) {
@@ -1197,7 +1197,7 @@ vint32m2_t test_vnmsub_vv_i32m2_m(vbool16_t mask, vint32m2_t acc, vint32m2_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnmsub.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnmsub.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vnmsub_vx_i32m2_m(vbool16_t mask, vint32m2_t acc, int32_t op1, vint32m2_t op2, size_t vl) {
@@ -1207,7 +1207,7 @@ vint32m2_t test_vnmsub_vx_i32m2_m(vbool16_t mask, vint32m2_t acc, int32_t op1, v
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnmsub.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnmsub.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vnmsub_vv_i32m4_m(vbool8_t mask, vint32m4_t acc, vint32m4_t op1, vint32m4_t op2, size_t vl) {
@@ -1217,7 +1217,7 @@ vint32m4_t test_vnmsub_vv_i32m4_m(vbool8_t mask, vint32m4_t acc, vint32m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnmsub.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnmsub.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vnmsub_vx_i32m4_m(vbool8_t mask, vint32m4_t acc, int32_t op1, vint32m4_t op2, size_t vl) {
@@ -1227,7 +1227,7 @@ vint32m4_t test_vnmsub_vx_i32m4_m(vbool8_t mask, vint32m4_t acc, int32_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vnmsub.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vnmsub.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vnmsub_vv_i32m8_m(vbool4_t mask, vint32m8_t acc, vint32m8_t op1, vint32m8_t op2, size_t vl) {
@@ -1237,7 +1237,7 @@ vint32m8_t test_vnmsub_vv_i32m8_m(vbool4_t mask, vint32m8_t acc, vint32m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vnmsub.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vnmsub.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vnmsub_vx_i32m8_m(vbool4_t mask, vint32m8_t acc, int32_t op1, vint32m8_t op2, size_t vl) {
@@ -1247,7 +1247,7 @@ vint32m8_t test_vnmsub_vx_i32m8_m(vbool4_t mask, vint32m8_t acc, int32_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vnmsub.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vnmsub.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vnmsub_vv_i64m1_m(vbool64_t mask, vint64m1_t acc, vint64m1_t op1, vint64m1_t op2, size_t vl) {
@@ -1257,7 +1257,7 @@ vint64m1_t test_vnmsub_vv_i64m1_m(vbool64_t mask, vint64m1_t acc, vint64m1_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vnmsub.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vnmsub.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vnmsub_vx_i64m1_m(vbool64_t mask, vint64m1_t acc, int64_t op1, vint64m1_t op2, size_t vl) {
@@ -1267,7 +1267,7 @@ vint64m1_t test_vnmsub_vx_i64m1_m(vbool64_t mask, vint64m1_t acc, int64_t op1, v
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vnmsub.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vnmsub.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vnmsub_vv_i64m2_m(vbool32_t mask, vint64m2_t acc, vint64m2_t op1, vint64m2_t op2, size_t vl) {
@@ -1277,7 +1277,7 @@ vint64m2_t test_vnmsub_vv_i64m2_m(vbool32_t mask, vint64m2_t acc, vint64m2_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vnmsub.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vnmsub.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vnmsub_vx_i64m2_m(vbool32_t mask, vint64m2_t acc, int64_t op1, vint64m2_t op2, size_t vl) {
@@ -1287,7 +1287,7 @@ vint64m2_t test_vnmsub_vx_i64m2_m(vbool32_t mask, vint64m2_t acc, int64_t op1, v
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vnmsub.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vnmsub.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vnmsub_vv_i64m4_m(vbool16_t mask, vint64m4_t acc, vint64m4_t op1, vint64m4_t op2, size_t vl) {
@@ -1297,7 +1297,7 @@ vint64m4_t test_vnmsub_vv_i64m4_m(vbool16_t mask, vint64m4_t acc, vint64m4_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vnmsub.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vnmsub.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vnmsub_vx_i64m4_m(vbool16_t mask, vint64m4_t acc, int64_t op1, vint64m4_t op2, size_t vl) {
@@ -1307,7 +1307,7 @@ vint64m4_t test_vnmsub_vx_i64m4_m(vbool16_t mask, vint64m4_t acc, int64_t op1, v
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vnmsub.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vnmsub.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vnmsub_vv_i64m8_m(vbool8_t mask, vint64m8_t acc, vint64m8_t op1, vint64m8_t op2, size_t vl) {
@@ -1317,7 +1317,7 @@ vint64m8_t test_vnmsub_vv_i64m8_m(vbool8_t mask, vint64m8_t acc, vint64m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vnmsub.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vnmsub.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vnmsub_vx_i64m8_m(vbool8_t mask, vint64m8_t acc, int64_t op1, vint64m8_t op2, size_t vl) {
@@ -1327,7 +1327,7 @@ vint64m8_t test_vnmsub_vx_i64m8_m(vbool8_t mask, vint64m8_t acc, int64_t op1, vi
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnmsub.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnmsub.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vnmsub_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t acc, vuint8mf8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -1337,7 +1337,7 @@ vuint8mf8_t test_vnmsub_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t acc, vuint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnmsub.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnmsub.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vnmsub_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t acc, uint8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -1347,7 +1347,7 @@ vuint8mf8_t test_vnmsub_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t acc, uint8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnmsub.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnmsub.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vnmsub_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t acc, vuint8mf4_t op1, vuint8mf4_t op2, size_t vl) {
@@ -1357,7 +1357,7 @@ vuint8mf4_t test_vnmsub_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t acc, vuint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnmsub.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnmsub.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vnmsub_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t acc, uint8_t op1, vuint8mf4_t op2, size_t vl) {
@@ -1367,7 +1367,7 @@ vuint8mf4_t test_vnmsub_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t acc, uint8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnmsub.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnmsub.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vnmsub_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t acc, vuint8mf2_t op1, vuint8mf2_t op2, size_t vl) {
@@ -1377,7 +1377,7 @@ vuint8mf2_t test_vnmsub_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t acc, vuint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnmsub.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnmsub.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vnmsub_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t acc, uint8_t op1, vuint8mf2_t op2, size_t vl) {
@@ -1387,7 +1387,7 @@ vuint8mf2_t test_vnmsub_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t acc, uint8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnmsub.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnmsub.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vnmsub_vv_u8m1_m(vbool8_t mask, vuint8m1_t acc, vuint8m1_t op1, vuint8m1_t op2, size_t vl) {
@@ -1397,7 +1397,7 @@ vuint8m1_t test_vnmsub_vv_u8m1_m(vbool8_t mask, vuint8m1_t acc, vuint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnmsub.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnmsub.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vnmsub_vx_u8m1_m(vbool8_t mask, vuint8m1_t acc, uint8_t op1, vuint8m1_t op2, size_t vl) {
@@ -1407,7 +1407,7 @@ vuint8m1_t test_vnmsub_vx_u8m1_m(vbool8_t mask, vuint8m1_t acc, uint8_t op1, vui
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnmsub.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnmsub.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vnmsub_vv_u8m2_m(vbool4_t mask, vuint8m2_t acc, vuint8m2_t op1, vuint8m2_t op2, size_t vl) {
@@ -1417,7 +1417,7 @@ vuint8m2_t test_vnmsub_vv_u8m2_m(vbool4_t mask, vuint8m2_t acc, vuint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnmsub.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnmsub.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vnmsub_vx_u8m2_m(vbool4_t mask, vuint8m2_t acc, uint8_t op1, vuint8m2_t op2, size_t vl) {
@@ -1427,7 +1427,7 @@ vuint8m2_t test_vnmsub_vx_u8m2_m(vbool4_t mask, vuint8m2_t acc, uint8_t op1, vui
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnmsub.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnmsub.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vnmsub_vv_u8m4_m(vbool2_t mask, vuint8m4_t acc, vuint8m4_t op1, vuint8m4_t op2, size_t vl) {
@@ -1437,7 +1437,7 @@ vuint8m4_t test_vnmsub_vv_u8m4_m(vbool2_t mask, vuint8m4_t acc, vuint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnmsub.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnmsub.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vnmsub_vx_u8m4_m(vbool2_t mask, vuint8m4_t acc, uint8_t op1, vuint8m4_t op2, size_t vl) {
@@ -1447,7 +1447,7 @@ vuint8m4_t test_vnmsub_vx_u8m4_m(vbool2_t mask, vuint8m4_t acc, uint8_t op1, vui
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vnmsub.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[ACC:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vnmsub.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[ACC:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vnmsub_vv_u8m8_m(vbool1_t mask, vuint8m8_t acc, vuint8m8_t op1, vuint8m8_t op2, size_t vl) {
@@ -1457,7 +1457,7 @@ vuint8m8_t test_vnmsub_vv_u8m8_m(vbool1_t mask, vuint8m8_t acc, vuint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vnmsub.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vnmsub.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vnmsub_vx_u8m8_m(vbool1_t mask, vuint8m8_t acc, uint8_t op1, vuint8m8_t op2, size_t vl) {
@@ -1467,7 +1467,7 @@ vuint8m8_t test_vnmsub_vx_u8m8_m(vbool1_t mask, vuint8m8_t acc, uint8_t op1, vui
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnmsub.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnmsub.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vnmsub_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t acc, vuint16mf4_t op1, vuint16mf4_t op2, size_t vl) {
@@ -1477,7 +1477,7 @@ vuint16mf4_t test_vnmsub_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t acc, vuint16mf
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnmsub.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnmsub.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vnmsub_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t acc, uint16_t op1, vuint16mf4_t op2, size_t vl) {
@@ -1487,7 +1487,7 @@ vuint16mf4_t test_vnmsub_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t acc, uint16_t
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnmsub.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnmsub.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vnmsub_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t acc, vuint16mf2_t op1, vuint16mf2_t op2, size_t vl) {
@@ -1497,7 +1497,7 @@ vuint16mf2_t test_vnmsub_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t acc, vuint16mf
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnmsub.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnmsub.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vnmsub_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t acc, uint16_t op1, vuint16mf2_t op2, size_t vl) {
@@ -1507,7 +1507,7 @@ vuint16mf2_t test_vnmsub_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t acc, uint16_t
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnmsub.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnmsub.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vnmsub_vv_u16m1_m(vbool16_t mask, vuint16m1_t acc, vuint16m1_t op1, vuint16m1_t op2, size_t vl) {
@@ -1517,7 +1517,7 @@ vuint16m1_t test_vnmsub_vv_u16m1_m(vbool16_t mask, vuint16m1_t acc, vuint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnmsub.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnmsub.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vnmsub_vx_u16m1_m(vbool16_t mask, vuint16m1_t acc, uint16_t op1, vuint16m1_t op2, size_t vl) {
@@ -1527,7 +1527,7 @@ vuint16m1_t test_vnmsub_vx_u16m1_m(vbool16_t mask, vuint16m1_t acc, uint16_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnmsub.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnmsub.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vnmsub_vv_u16m2_m(vbool8_t mask, vuint16m2_t acc, vuint16m2_t op1, vuint16m2_t op2, size_t vl) {
@@ -1537,7 +1537,7 @@ vuint16m2_t test_vnmsub_vv_u16m2_m(vbool8_t mask, vuint16m2_t acc, vuint16m2_t o
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnmsub.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnmsub.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vnmsub_vx_u16m2_m(vbool8_t mask, vuint16m2_t acc, uint16_t op1, vuint16m2_t op2, size_t vl) {
@@ -1547,7 +1547,7 @@ vuint16m2_t test_vnmsub_vx_u16m2_m(vbool8_t mask, vuint16m2_t acc, uint16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnmsub.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnmsub.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vnmsub_vv_u16m4_m(vbool4_t mask, vuint16m4_t acc, vuint16m4_t op1, vuint16m4_t op2, size_t vl) {
@@ -1557,7 +1557,7 @@ vuint16m4_t test_vnmsub_vv_u16m4_m(vbool4_t mask, vuint16m4_t acc, vuint16m4_t o
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnmsub.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnmsub.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vnmsub_vx_u16m4_m(vbool4_t mask, vuint16m4_t acc, uint16_t op1, vuint16m4_t op2, size_t vl) {
@@ -1567,7 +1567,7 @@ vuint16m4_t test_vnmsub_vx_u16m4_m(vbool4_t mask, vuint16m4_t acc, uint16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vnmsub.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vnmsub.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vnmsub_vv_u16m8_m(vbool2_t mask, vuint16m8_t acc, vuint16m8_t op1, vuint16m8_t op2, size_t vl) {
@@ -1577,7 +1577,7 @@ vuint16m8_t test_vnmsub_vv_u16m8_m(vbool2_t mask, vuint16m8_t acc, vuint16m8_t o
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vnmsub.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vnmsub.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vnmsub_vx_u16m8_m(vbool2_t mask, vuint16m8_t acc, uint16_t op1, vuint16m8_t op2, size_t vl) {
@@ -1587,7 +1587,7 @@ vuint16m8_t test_vnmsub_vx_u16m8_m(vbool2_t mask, vuint16m8_t acc, uint16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnmsub.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnmsub.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vnmsub_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t acc, vuint32mf2_t op1, vuint32mf2_t op2, size_t vl) {
@@ -1597,7 +1597,7 @@ vuint32mf2_t test_vnmsub_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t acc, vuint32mf
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnmsub.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnmsub.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vnmsub_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t acc, uint32_t op1, vuint32mf2_t op2, size_t vl) {
@@ -1607,7 +1607,7 @@ vuint32mf2_t test_vnmsub_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t acc, uint32_t
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnmsub.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnmsub.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vnmsub_vv_u32m1_m(vbool32_t mask, vuint32m1_t acc, vuint32m1_t op1, vuint32m1_t op2, size_t vl) {
@@ -1617,7 +1617,7 @@ vuint32m1_t test_vnmsub_vv_u32m1_m(vbool32_t mask, vuint32m1_t acc, vuint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnmsub.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnmsub.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vnmsub_vx_u32m1_m(vbool32_t mask, vuint32m1_t acc, uint32_t op1, vuint32m1_t op2, size_t vl) {
@@ -1627,7 +1627,7 @@ vuint32m1_t test_vnmsub_vx_u32m1_m(vbool32_t mask, vuint32m1_t acc, uint32_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnmsub.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnmsub.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vnmsub_vv_u32m2_m(vbool16_t mask, vuint32m2_t acc, vuint32m2_t op1, vuint32m2_t op2, size_t vl) {
@@ -1637,7 +1637,7 @@ vuint32m2_t test_vnmsub_vv_u32m2_m(vbool16_t mask, vuint32m2_t acc, vuint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnmsub.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnmsub.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vnmsub_vx_u32m2_m(vbool16_t mask, vuint32m2_t acc, uint32_t op1, vuint32m2_t op2, size_t vl) {
@@ -1647,7 +1647,7 @@ vuint32m2_t test_vnmsub_vx_u32m2_m(vbool16_t mask, vuint32m2_t acc, uint32_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnmsub.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnmsub.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vnmsub_vv_u32m4_m(vbool8_t mask, vuint32m4_t acc, vuint32m4_t op1, vuint32m4_t op2, size_t vl) {
@@ -1657,7 +1657,7 @@ vuint32m4_t test_vnmsub_vv_u32m4_m(vbool8_t mask, vuint32m4_t acc, vuint32m4_t o
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnmsub.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnmsub.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vnmsub_vx_u32m4_m(vbool8_t mask, vuint32m4_t acc, uint32_t op1, vuint32m4_t op2, size_t vl) {
@@ -1667,7 +1667,7 @@ vuint32m4_t test_vnmsub_vx_u32m4_m(vbool8_t mask, vuint32m4_t acc, uint32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vnmsub.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vnmsub.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vnmsub_vv_u32m8_m(vbool4_t mask, vuint32m8_t acc, vuint32m8_t op1, vuint32m8_t op2, size_t vl) {
@@ -1677,7 +1677,7 @@ vuint32m8_t test_vnmsub_vv_u32m8_m(vbool4_t mask, vuint32m8_t acc, vuint32m8_t o
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vnmsub.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vnmsub.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vnmsub_vx_u32m8_m(vbool4_t mask, vuint32m8_t acc, uint32_t op1, vuint32m8_t op2, size_t vl) {
@@ -1687,7 +1687,7 @@ vuint32m8_t test_vnmsub_vx_u32m8_m(vbool4_t mask, vuint32m8_t acc, uint32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vnmsub.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vnmsub.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vnmsub_vv_u64m1_m(vbool64_t mask, vuint64m1_t acc, vuint64m1_t op1, vuint64m1_t op2, size_t vl) {
@@ -1697,7 +1697,7 @@ vuint64m1_t test_vnmsub_vv_u64m1_m(vbool64_t mask, vuint64m1_t acc, vuint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vnmsub.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vnmsub.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vnmsub_vx_u64m1_m(vbool64_t mask, vuint64m1_t acc, uint64_t op1, vuint64m1_t op2, size_t vl) {
@@ -1707,7 +1707,7 @@ vuint64m1_t test_vnmsub_vx_u64m1_m(vbool64_t mask, vuint64m1_t acc, uint64_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vnmsub.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vnmsub.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vnmsub_vv_u64m2_m(vbool32_t mask, vuint64m2_t acc, vuint64m2_t op1, vuint64m2_t op2, size_t vl) {
@@ -1717,7 +1717,7 @@ vuint64m2_t test_vnmsub_vv_u64m2_m(vbool32_t mask, vuint64m2_t acc, vuint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vnmsub.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vnmsub.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vnmsub_vx_u64m2_m(vbool32_t mask, vuint64m2_t acc, uint64_t op1, vuint64m2_t op2, size_t vl) {
@@ -1727,7 +1727,7 @@ vuint64m2_t test_vnmsub_vx_u64m2_m(vbool32_t mask, vuint64m2_t acc, uint64_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vnmsub.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vnmsub.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vnmsub_vv_u64m4_m(vbool16_t mask, vuint64m4_t acc, vuint64m4_t op1, vuint64m4_t op2, size_t vl) {
@@ -1737,7 +1737,7 @@ vuint64m4_t test_vnmsub_vv_u64m4_m(vbool16_t mask, vuint64m4_t acc, vuint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vnmsub.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vnmsub.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vnmsub_vx_u64m4_m(vbool16_t mask, vuint64m4_t acc, uint64_t op1, vuint64m4_t op2, size_t vl) {
@@ -1747,7 +1747,7 @@ vuint64m4_t test_vnmsub_vx_u64m4_m(vbool16_t mask, vuint64m4_t acc, uint64_t op1
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vnmsub.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vnmsub.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vnmsub_vv_u64m8_m(vbool8_t mask, vuint64m8_t acc, vuint64m8_t op1, vuint64m8_t op2, size_t vl) {
@@ -1757,7 +1757,7 @@ vuint64m8_t test_vnmsub_vv_u64m8_m(vbool8_t mask, vuint64m8_t acc, vuint64m8_t o
 //
 // CHECK-RV64-LABEL: @test_vnmsub_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vnmsub.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vnmsub.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[ACC:%.*]], i64 [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vnmsub_vx_u64m8_m(vbool8_t mask, vuint64m8_t acc, uint64_t op1, vuint64m8_t op2, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnot.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnot.c
index 7a08334..391fc5c 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnot.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnot.c
@@ -448,7 +448,7 @@ vuint64m8_t test_vnot_v_u64m8 (vuint64m8_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vxor.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 -1, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vxor.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 -1, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vnot_v_i8mf8_m (vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, size_t vl) {
@@ -458,7 +458,7 @@ vint8mf8_t test_vnot_v_i8mf8_m (vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vxor.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 -1, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vxor.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 -1, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vnot_v_i8mf4_m (vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, size_t vl) {
@@ -468,7 +468,7 @@ vint8mf4_t test_vnot_v_i8mf4_m (vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vxor.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 -1, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vxor.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 -1, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vnot_v_i8mf2_m (vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, size_t vl) {
@@ -478,7 +478,7 @@ vint8mf2_t test_vnot_v_i8mf2_m (vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vxor.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 -1, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vxor.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 -1, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vnot_v_i8m1_m (vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, size_t vl) {
@@ -488,7 +488,7 @@ vint8m1_t test_vnot_v_i8m1_m (vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vxor.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 -1, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vxor.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 -1, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vnot_v_i8m2_m (vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, size_t vl) {
@@ -498,7 +498,7 @@ vint8m2_t test_vnot_v_i8m2_m (vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vxor.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 -1, <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vxor.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 -1, <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vnot_v_i8m4_m (vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, size_t vl) {
@@ -508,7 +508,7 @@ vint8m4_t test_vnot_v_i8m4_m (vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vxor.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 -1, <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vxor.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 -1, <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vnot_v_i8m8_m (vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, size_t vl) {
@@ -518,7 +518,7 @@ vint8m8_t test_vnot_v_i8m8_m (vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vxor.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 -1, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vxor.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 -1, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vnot_v_i16mf4_m (vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, size_t vl) {
@@ -528,7 +528,7 @@ vint16mf4_t test_vnot_v_i16mf4_m (vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vxor.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 -1, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vxor.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 -1, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vnot_v_i16mf2_m (vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, size_t vl) {
@@ -538,7 +538,7 @@ vint16mf2_t test_vnot_v_i16mf2_m (vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vxor.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 -1, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vxor.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 -1, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vnot_v_i16m1_m (vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, size_t vl) {
@@ -548,7 +548,7 @@ vint16m1_t test_vnot_v_i16m1_m (vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vxor.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 -1, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vxor.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 -1, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vnot_v_i16m2_m (vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, size_t vl) {
@@ -558,7 +558,7 @@ vint16m2_t test_vnot_v_i16m2_m (vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vxor.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 -1, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vxor.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 -1, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vnot_v_i16m4_m (vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, size_t vl) {
@@ -568,7 +568,7 @@ vint16m4_t test_vnot_v_i16m4_m (vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vxor.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 -1, <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vxor.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 -1, <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vnot_v_i16m8_m (vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, size_t vl) {
@@ -578,7 +578,7 @@ vint16m8_t test_vnot_v_i16m8_m (vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vxor.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 -1, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vxor.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 -1, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vnot_v_i32mf2_m (vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, size_t vl) {
@@ -588,7 +588,7 @@ vint32mf2_t test_vnot_v_i32mf2_m (vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vxor.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 -1, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vxor.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 -1, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vnot_v_i32m1_m (vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, size_t vl) {
@@ -598,7 +598,7 @@ vint32m1_t test_vnot_v_i32m1_m (vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vxor.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 -1, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vxor.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 -1, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vnot_v_i32m2_m (vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, size_t vl) {
@@ -608,7 +608,7 @@ vint32m2_t test_vnot_v_i32m2_m (vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vxor.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 -1, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vxor.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 -1, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vnot_v_i32m4_m (vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, size_t vl) {
@@ -618,7 +618,7 @@ vint32m4_t test_vnot_v_i32m4_m (vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vxor.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 -1, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vxor.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 -1, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vnot_v_i32m8_m (vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, size_t vl) {
@@ -628,7 +628,7 @@ vint32m8_t test_vnot_v_i32m8_m (vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vxor.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 -1, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vxor.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 -1, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vnot_v_i64m1_m (vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, size_t vl) {
@@ -638,7 +638,7 @@ vint64m1_t test_vnot_v_i64m1_m (vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vxor.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 -1, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vxor.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 -1, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vnot_v_i64m2_m (vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, size_t vl) {
@@ -648,7 +648,7 @@ vint64m2_t test_vnot_v_i64m2_m (vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vxor.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 -1, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vxor.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 -1, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vnot_v_i64m4_m (vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, size_t vl) {
@@ -658,7 +658,7 @@ vint64m4_t test_vnot_v_i64m4_m (vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vnot_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vxor.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 -1, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vxor.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 -1, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vnot_v_i64m8_m (vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, size_t vl) {
@@ -668,7 +668,7 @@ vint64m8_t test_vnot_v_i64m8_m (vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vxor.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 -1, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vxor.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 -1, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vnot_v_u8mf8_m (vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, size_t vl) {
@@ -678,7 +678,7 @@ vuint8mf8_t test_vnot_v_u8mf8_m (vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vxor.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 -1, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vxor.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 -1, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vnot_v_u8mf4_m (vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, size_t vl) {
@@ -688,7 +688,7 @@ vuint8mf4_t test_vnot_v_u8mf4_m (vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vxor.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 -1, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vxor.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 -1, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vnot_v_u8mf2_m (vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, size_t vl) {
@@ -698,7 +698,7 @@ vuint8mf2_t test_vnot_v_u8mf2_m (vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vxor.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 -1, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vxor.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 -1, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vnot_v_u8m1_m (vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, size_t vl) {
@@ -708,7 +708,7 @@ vuint8m1_t test_vnot_v_u8m1_m (vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t o
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vxor.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 -1, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vxor.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 -1, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vnot_v_u8m2_m (vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, size_t vl) {
@@ -718,7 +718,7 @@ vuint8m2_t test_vnot_v_u8m2_m (vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t o
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vxor.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 -1, <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vxor.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 -1, <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vnot_v_u8m4_m (vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, size_t vl) {
@@ -728,7 +728,7 @@ vuint8m4_t test_vnot_v_u8m4_m (vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t o
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vxor.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 -1, <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vxor.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 -1, <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vnot_v_u8m8_m (vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, size_t vl) {
@@ -738,7 +738,7 @@ vuint8m8_t test_vnot_v_u8m8_m (vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t o
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vxor.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 -1, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vxor.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 -1, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vnot_v_u16mf4_m (vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, size_t vl) {
@@ -748,7 +748,7 @@ vuint16mf4_t test_vnot_v_u16mf4_m (vbool64_t mask, vuint16mf4_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vxor.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 -1, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vxor.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 -1, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vnot_v_u16mf2_m (vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, size_t vl) {
@@ -758,7 +758,7 @@ vuint16mf2_t test_vnot_v_u16mf2_m (vbool32_t mask, vuint16mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vxor.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 -1, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vxor.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 -1, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vnot_v_u16m1_m (vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, size_t vl) {
@@ -768,7 +768,7 @@ vuint16m1_t test_vnot_v_u16m1_m (vbool16_t mask, vuint16m1_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vxor.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 -1, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vxor.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 -1, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vnot_v_u16m2_m (vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, size_t vl) {
@@ -778,7 +778,7 @@ vuint16m2_t test_vnot_v_u16m2_m (vbool8_t mask, vuint16m2_t maskedoff, vuint16m2
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vxor.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 -1, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vxor.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 -1, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vnot_v_u16m4_m (vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, size_t vl) {
@@ -788,7 +788,7 @@ vuint16m4_t test_vnot_v_u16m4_m (vbool4_t mask, vuint16m4_t maskedoff, vuint16m4
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vxor.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 -1, <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vxor.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 -1, <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vnot_v_u16m8_m (vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, size_t vl) {
@@ -798,7 +798,7 @@ vuint16m8_t test_vnot_v_u16m8_m (vbool2_t mask, vuint16m8_t maskedoff, vuint16m8
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vxor.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 -1, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vxor.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 -1, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vnot_v_u32mf2_m (vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, size_t vl) {
@@ -808,7 +808,7 @@ vuint32mf2_t test_vnot_v_u32mf2_m (vbool64_t mask, vuint32mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vxor.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 -1, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vxor.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 -1, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vnot_v_u32m1_m (vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, size_t vl) {
@@ -818,7 +818,7 @@ vuint32m1_t test_vnot_v_u32m1_m (vbool32_t mask, vuint32m1_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vxor.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 -1, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vxor.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 -1, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vnot_v_u32m2_m (vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, size_t vl) {
@@ -828,7 +828,7 @@ vuint32m2_t test_vnot_v_u32m2_m (vbool16_t mask, vuint32m2_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vxor.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 -1, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vxor.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 -1, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vnot_v_u32m4_m (vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, size_t vl) {
@@ -838,7 +838,7 @@ vuint32m4_t test_vnot_v_u32m4_m (vbool8_t mask, vuint32m4_t maskedoff, vuint32m4
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vxor.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 -1, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vxor.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 -1, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vnot_v_u32m8_m (vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, size_t vl) {
@@ -848,7 +848,7 @@ vuint32m8_t test_vnot_v_u32m8_m (vbool4_t mask, vuint32m8_t maskedoff, vuint32m8
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vxor.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 -1, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vxor.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 -1, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vnot_v_u64m1_m (vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, size_t vl) {
@@ -858,7 +858,7 @@ vuint64m1_t test_vnot_v_u64m1_m (vbool64_t mask, vuint64m1_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vxor.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 -1, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vxor.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 -1, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vnot_v_u64m2_m (vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, size_t vl) {
@@ -868,7 +868,7 @@ vuint64m2_t test_vnot_v_u64m2_m (vbool32_t mask, vuint64m2_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vxor.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 -1, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vxor.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 -1, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vnot_v_u64m4_m (vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, size_t vl) {
@@ -878,7 +878,7 @@ vuint64m4_t test_vnot_v_u64m4_m (vbool16_t mask, vuint64m4_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vnot_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vxor.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 -1, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vxor.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 -1, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vnot_v_u64m8_m (vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnsra.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnsra.c
index a90d69e..e76b076 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnsra.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnsra.c
@@ -307,7 +307,7 @@ vint32m4_t test_vnsra_wx_i32m4(vint64m8_t op1, size_t shift, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vnsra_wv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnsra.mask.nxv1i8.nxv1i16.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i8> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnsra.mask.nxv1i8.nxv1i16.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i8> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vnsra_wv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint16mf4_t op1, vuint8mf8_t shift, size_t vl) {
@@ -317,7 +317,7 @@ vint8mf8_t test_vnsra_wv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint16mf4
 //
 // CHECK-RV64-LABEL: @test_vnsra_wx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnsra.mask.nxv1i8.nxv1i16.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnsra.mask.nxv1i8.nxv1i16.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vnsra_wx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint16mf4_t op1, size_t shift, size_t vl) {
@@ -327,7 +327,7 @@ vint8mf8_t test_vnsra_wx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint16mf4
 //
 // CHECK-RV64-LABEL: @test_vnsra_wv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnsra.mask.nxv2i8.nxv2i16.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i8> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnsra.mask.nxv2i8.nxv2i16.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i8> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vnsra_wv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint16mf2_t op1, vuint8mf4_t shift, size_t vl) {
@@ -337,7 +337,7 @@ vint8mf4_t test_vnsra_wv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint16mf2
 //
 // CHECK-RV64-LABEL: @test_vnsra_wx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnsra.mask.nxv2i8.nxv2i16.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnsra.mask.nxv2i8.nxv2i16.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vnsra_wx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint16mf2_t op1, size_t shift, size_t vl) {
@@ -347,7 +347,7 @@ vint8mf4_t test_vnsra_wx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint16mf2
 //
 // CHECK-RV64-LABEL: @test_vnsra_wv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnsra.mask.nxv4i8.nxv4i16.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i8> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnsra.mask.nxv4i8.nxv4i16.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i8> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vnsra_wv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint16m1_t op1, vuint8mf2_t shift, size_t vl) {
@@ -357,7 +357,7 @@ vint8mf2_t test_vnsra_wv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint16m1_
 //
 // CHECK-RV64-LABEL: @test_vnsra_wx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnsra.mask.nxv4i8.nxv4i16.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnsra.mask.nxv4i8.nxv4i16.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vnsra_wx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint16m1_t op1, size_t shift, size_t vl) {
@@ -367,7 +367,7 @@ vint8mf2_t test_vnsra_wx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint16m1_
 //
 // CHECK-RV64-LABEL: @test_vnsra_wv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnsra.mask.nxv8i8.nxv8i16.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i8> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnsra.mask.nxv8i8.nxv8i16.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i8> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vnsra_wv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint16m2_t op1, vuint8m1_t shift, size_t vl) {
@@ -377,7 +377,7 @@ vint8m1_t test_vnsra_wv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint16m2_t op
 //
 // CHECK-RV64-LABEL: @test_vnsra_wx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnsra.mask.nxv8i8.nxv8i16.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnsra.mask.nxv8i8.nxv8i16.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vnsra_wx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint16m2_t op1, size_t shift, size_t vl) {
@@ -387,7 +387,7 @@ vint8m1_t test_vnsra_wx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint16m2_t op
 //
 // CHECK-RV64-LABEL: @test_vnsra_wv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnsra.mask.nxv16i8.nxv16i16.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i8> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnsra.mask.nxv16i8.nxv16i16.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i8> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vnsra_wv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint16m4_t op1, vuint8m2_t shift, size_t vl) {
@@ -397,7 +397,7 @@ vint8m2_t test_vnsra_wv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint16m4_t op
 //
 // CHECK-RV64-LABEL: @test_vnsra_wx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnsra.mask.nxv16i8.nxv16i16.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnsra.mask.nxv16i8.nxv16i16.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vnsra_wx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint16m4_t op1, size_t shift, size_t vl) {
@@ -407,7 +407,7 @@ vint8m2_t test_vnsra_wx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint16m4_t op
 //
 // CHECK-RV64-LABEL: @test_vnsra_wv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnsra.mask.nxv32i8.nxv32i16.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i8> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnsra.mask.nxv32i8.nxv32i16.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i8> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vnsra_wv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint16m8_t op1, vuint8m4_t shift, size_t vl) {
@@ -417,7 +417,7 @@ vint8m4_t test_vnsra_wv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint16m8_t op
 //
 // CHECK-RV64-LABEL: @test_vnsra_wx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnsra.mask.nxv32i8.nxv32i16.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnsra.mask.nxv32i8.nxv32i16.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vnsra_wx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint16m8_t op1, size_t shift, size_t vl) {
@@ -427,7 +427,7 @@ vint8m4_t test_vnsra_wx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint16m8_t op
 //
 // CHECK-RV64-LABEL: @test_vnsra_wv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnsra.mask.nxv1i16.nxv1i32.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i16> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnsra.mask.nxv1i16.nxv1i32.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i16> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vnsra_wv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint32mf2_t op1, vuint16mf4_t shift, size_t vl) {
@@ -437,7 +437,7 @@ vint16mf4_t test_vnsra_wv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint32
 //
 // CHECK-RV64-LABEL: @test_vnsra_wx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnsra.mask.nxv1i16.nxv1i32.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnsra.mask.nxv1i16.nxv1i32.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vnsra_wx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint32mf2_t op1, size_t shift, size_t vl) {
@@ -447,7 +447,7 @@ vint16mf4_t test_vnsra_wx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint32
 //
 // CHECK-RV64-LABEL: @test_vnsra_wv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnsra.mask.nxv2i16.nxv2i32.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i16> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnsra.mask.nxv2i16.nxv2i32.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i16> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vnsra_wv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint32m1_t op1, vuint16mf2_t shift, size_t vl) {
@@ -457,7 +457,7 @@ vint16mf2_t test_vnsra_wv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint32
 //
 // CHECK-RV64-LABEL: @test_vnsra_wx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnsra.mask.nxv2i16.nxv2i32.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnsra.mask.nxv2i16.nxv2i32.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vnsra_wx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint32m1_t op1, size_t shift, size_t vl) {
@@ -467,7 +467,7 @@ vint16mf2_t test_vnsra_wx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint32
 //
 // CHECK-RV64-LABEL: @test_vnsra_wv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnsra.mask.nxv4i16.nxv4i32.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i16> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnsra.mask.nxv4i16.nxv4i32.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i16> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vnsra_wv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint32m2_t op1, vuint16m1_t shift, size_t vl) {
@@ -477,7 +477,7 @@ vint16m1_t test_vnsra_wv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint32m2_
 //
 // CHECK-RV64-LABEL: @test_vnsra_wx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnsra.mask.nxv4i16.nxv4i32.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnsra.mask.nxv4i16.nxv4i32.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vnsra_wx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint32m2_t op1, size_t shift, size_t vl) {
@@ -487,7 +487,7 @@ vint16m1_t test_vnsra_wx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint32m2_
 //
 // CHECK-RV64-LABEL: @test_vnsra_wv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnsra.mask.nxv8i16.nxv8i32.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i16> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnsra.mask.nxv8i16.nxv8i32.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i16> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vnsra_wv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint32m4_t op1, vuint16m2_t shift, size_t vl) {
@@ -497,7 +497,7 @@ vint16m2_t test_vnsra_wv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vnsra_wx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnsra.mask.nxv8i16.nxv8i32.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnsra.mask.nxv8i16.nxv8i32.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vnsra_wx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint32m4_t op1, size_t shift, size_t vl) {
@@ -507,7 +507,7 @@ vint16m2_t test_vnsra_wx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vnsra_wv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnsra.mask.nxv16i16.nxv16i32.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i16> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnsra.mask.nxv16i16.nxv16i32.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i16> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vnsra_wv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint32m8_t op1, vuint16m4_t shift, size_t vl) {
@@ -517,7 +517,7 @@ vint16m4_t test_vnsra_wv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vnsra_wx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnsra.mask.nxv16i16.nxv16i32.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnsra.mask.nxv16i16.nxv16i32.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vnsra_wx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint32m8_t op1, size_t shift, size_t vl) {
@@ -527,7 +527,7 @@ vint16m4_t test_vnsra_wx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vnsra_wv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnsra.mask.nxv1i32.nxv1i64.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i32> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnsra.mask.nxv1i32.nxv1i64.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i32> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vnsra_wv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint64m1_t op1, vuint32mf2_t shift, size_t vl) {
@@ -537,7 +537,7 @@ vint32mf2_t test_vnsra_wv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint64
 //
 // CHECK-RV64-LABEL: @test_vnsra_wx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnsra.mask.nxv1i32.nxv1i64.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnsra.mask.nxv1i32.nxv1i64.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vnsra_wx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint64m1_t op1, size_t shift, size_t vl) {
@@ -547,7 +547,7 @@ vint32mf2_t test_vnsra_wx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint64
 //
 // CHECK-RV64-LABEL: @test_vnsra_wv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnsra.mask.nxv2i32.nxv2i64.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i32> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnsra.mask.nxv2i32.nxv2i64.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i32> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vnsra_wv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint64m2_t op1, vuint32m1_t shift, size_t vl) {
@@ -557,7 +557,7 @@ vint32m1_t test_vnsra_wv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint64m2_
 //
 // CHECK-RV64-LABEL: @test_vnsra_wx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnsra.mask.nxv2i32.nxv2i64.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnsra.mask.nxv2i32.nxv2i64.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vnsra_wx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint64m2_t op1, size_t shift, size_t vl) {
@@ -567,7 +567,7 @@ vint32m1_t test_vnsra_wx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint64m2_
 //
 // CHECK-RV64-LABEL: @test_vnsra_wv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnsra.mask.nxv4i32.nxv4i64.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i32> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnsra.mask.nxv4i32.nxv4i64.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i32> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vnsra_wv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint64m4_t op1, vuint32m2_t shift, size_t vl) {
@@ -577,7 +577,7 @@ vint32m2_t test_vnsra_wv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint64m4_
 //
 // CHECK-RV64-LABEL: @test_vnsra_wx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnsra.mask.nxv4i32.nxv4i64.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnsra.mask.nxv4i32.nxv4i64.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vnsra_wx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint64m4_t op1, size_t shift, size_t vl) {
@@ -587,7 +587,7 @@ vint32m2_t test_vnsra_wx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint64m4_
 //
 // CHECK-RV64-LABEL: @test_vnsra_wv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnsra.mask.nxv8i32.nxv8i64.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i32> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnsra.mask.nxv8i32.nxv8i64.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i32> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vnsra_wv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint64m8_t op1, vuint32m4_t shift, size_t vl) {
@@ -597,7 +597,7 @@ vint32m4_t test_vnsra_wv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vnsra_wx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnsra.mask.nxv8i32.nxv8i64.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnsra.mask.nxv8i32.nxv8i64.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vnsra_wx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint64m8_t op1, size_t shift, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnsrl.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnsrl.c
index 12e0482..e478bec 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnsrl.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vnsrl.c
@@ -307,7 +307,7 @@ vuint32m4_t test_vnsrl_wx_u32m4(vuint64m8_t op1, size_t shift, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnsrl.mask.nxv1i8.nxv1i16.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i8> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnsrl.mask.nxv1i8.nxv1i16.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i8> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vnsrl_wv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint16mf4_t op1, vuint8mf8_t shift, size_t vl) {
@@ -317,7 +317,7 @@ vuint8mf8_t test_vnsrl_wv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnsrl.mask.nxv1i8.nxv1i16.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vnsrl.mask.nxv1i8.nxv1i16.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vnsrl_wx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint16mf4_t op1, size_t shift, size_t vl) {
@@ -327,7 +327,7 @@ vuint8mf8_t test_vnsrl_wx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnsrl.mask.nxv2i8.nxv2i16.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i8> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnsrl.mask.nxv2i8.nxv2i16.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i8> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vnsrl_wv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint16mf2_t op1, vuint8mf4_t shift, size_t vl) {
@@ -337,7 +337,7 @@ vuint8mf4_t test_vnsrl_wv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnsrl.mask.nxv2i8.nxv2i16.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vnsrl.mask.nxv2i8.nxv2i16.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vnsrl_wx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint16mf2_t op1, size_t shift, size_t vl) {
@@ -347,7 +347,7 @@ vuint8mf4_t test_vnsrl_wx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnsrl.mask.nxv4i8.nxv4i16.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i8> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnsrl.mask.nxv4i8.nxv4i16.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i8> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vnsrl_wv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint16m1_t op1, vuint8mf2_t shift, size_t vl) {
@@ -357,7 +357,7 @@ vuint8mf2_t test_vnsrl_wv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnsrl.mask.nxv4i8.nxv4i16.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vnsrl.mask.nxv4i8.nxv4i16.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vnsrl_wx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint16m1_t op1, size_t shift, size_t vl) {
@@ -367,7 +367,7 @@ vuint8mf2_t test_vnsrl_wx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnsrl.mask.nxv8i8.nxv8i16.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i8> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnsrl.mask.nxv8i8.nxv8i16.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i8> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vnsrl_wv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint16m2_t op1, vuint8m1_t shift, size_t vl) {
@@ -377,7 +377,7 @@ vuint8m1_t test_vnsrl_wv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnsrl.mask.nxv8i8.nxv8i16.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vnsrl.mask.nxv8i8.nxv8i16.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vnsrl_wx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint16m2_t op1, size_t shift, size_t vl) {
@@ -387,7 +387,7 @@ vuint8m1_t test_vnsrl_wx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnsrl.mask.nxv16i8.nxv16i16.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i8> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnsrl.mask.nxv16i8.nxv16i16.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i8> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vnsrl_wv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint16m4_t op1, vuint8m2_t shift, size_t vl) {
@@ -397,7 +397,7 @@ vuint8m2_t test_vnsrl_wv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnsrl.mask.nxv16i8.nxv16i16.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vnsrl.mask.nxv16i8.nxv16i16.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vnsrl_wx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint16m4_t op1, size_t shift, size_t vl) {
@@ -407,7 +407,7 @@ vuint8m2_t test_vnsrl_wx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnsrl.mask.nxv32i8.nxv32i16.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i8> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnsrl.mask.nxv32i8.nxv32i16.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i8> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vnsrl_wv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint16m8_t op1, vuint8m4_t shift, size_t vl) {
@@ -417,7 +417,7 @@ vuint8m4_t test_vnsrl_wv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnsrl.mask.nxv32i8.nxv32i16.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vnsrl.mask.nxv32i8.nxv32i16.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vnsrl_wx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint16m8_t op1, size_t shift, size_t vl) {
@@ -427,7 +427,7 @@ vuint8m4_t test_vnsrl_wx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnsrl.mask.nxv1i16.nxv1i32.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i16> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnsrl.mask.nxv1i16.nxv1i32.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i16> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vnsrl_wv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint32mf2_t op1, vuint16mf4_t shift, size_t vl) {
@@ -437,7 +437,7 @@ vuint16mf4_t test_vnsrl_wv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnsrl.mask.nxv1i16.nxv1i32.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vnsrl.mask.nxv1i16.nxv1i32.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vnsrl_wx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint32mf2_t op1, size_t shift, size_t vl) {
@@ -447,7 +447,7 @@ vuint16mf4_t test_vnsrl_wx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnsrl.mask.nxv2i16.nxv2i32.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i16> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnsrl.mask.nxv2i16.nxv2i32.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i16> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vnsrl_wv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint32m1_t op1, vuint16mf2_t shift, size_t vl) {
@@ -457,7 +457,7 @@ vuint16mf2_t test_vnsrl_wv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnsrl.mask.nxv2i16.nxv2i32.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vnsrl.mask.nxv2i16.nxv2i32.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vnsrl_wx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint32m1_t op1, size_t shift, size_t vl) {
@@ -467,7 +467,7 @@ vuint16mf2_t test_vnsrl_wx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnsrl.mask.nxv4i16.nxv4i32.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i16> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnsrl.mask.nxv4i16.nxv4i32.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i16> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vnsrl_wv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint32m2_t op1, vuint16m1_t shift, size_t vl) {
@@ -477,7 +477,7 @@ vuint16m1_t test_vnsrl_wv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnsrl.mask.nxv4i16.nxv4i32.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vnsrl.mask.nxv4i16.nxv4i32.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vnsrl_wx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint32m2_t op1, size_t shift, size_t vl) {
@@ -487,7 +487,7 @@ vuint16m1_t test_vnsrl_wx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnsrl.mask.nxv8i16.nxv8i32.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i16> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnsrl.mask.nxv8i16.nxv8i32.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i16> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vnsrl_wv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint32m4_t op1, vuint16m2_t shift, size_t vl) {
@@ -497,7 +497,7 @@ vuint16m2_t test_vnsrl_wv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnsrl.mask.nxv8i16.nxv8i32.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vnsrl.mask.nxv8i16.nxv8i32.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vnsrl_wx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint32m4_t op1, size_t shift, size_t vl) {
@@ -507,7 +507,7 @@ vuint16m2_t test_vnsrl_wx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnsrl.mask.nxv16i16.nxv16i32.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i16> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnsrl.mask.nxv16i16.nxv16i32.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i16> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vnsrl_wv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint32m8_t op1, vuint16m4_t shift, size_t vl) {
@@ -517,7 +517,7 @@ vuint16m4_t test_vnsrl_wv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnsrl.mask.nxv16i16.nxv16i32.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vnsrl.mask.nxv16i16.nxv16i32.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vnsrl_wx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint32m8_t op1, size_t shift, size_t vl) {
@@ -527,7 +527,7 @@ vuint16m4_t test_vnsrl_wx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnsrl.mask.nxv1i32.nxv1i64.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i32> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnsrl.mask.nxv1i32.nxv1i64.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i32> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vnsrl_wv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint64m1_t op1, vuint32mf2_t shift, size_t vl) {
@@ -537,7 +537,7 @@ vuint32mf2_t test_vnsrl_wv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnsrl.mask.nxv1i32.nxv1i64.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vnsrl.mask.nxv1i32.nxv1i64.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vnsrl_wx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint64m1_t op1, size_t shift, size_t vl) {
@@ -547,7 +547,7 @@ vuint32mf2_t test_vnsrl_wx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnsrl.mask.nxv2i32.nxv2i64.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i32> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnsrl.mask.nxv2i32.nxv2i64.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i32> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vnsrl_wv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint64m2_t op1, vuint32m1_t shift, size_t vl) {
@@ -557,7 +557,7 @@ vuint32m1_t test_vnsrl_wv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnsrl.mask.nxv2i32.nxv2i64.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vnsrl.mask.nxv2i32.nxv2i64.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vnsrl_wx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint64m2_t op1, size_t shift, size_t vl) {
@@ -567,7 +567,7 @@ vuint32m1_t test_vnsrl_wx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnsrl.mask.nxv4i32.nxv4i64.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i32> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnsrl.mask.nxv4i32.nxv4i64.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i32> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vnsrl_wv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint64m4_t op1, vuint32m2_t shift, size_t vl) {
@@ -577,7 +577,7 @@ vuint32m2_t test_vnsrl_wv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnsrl.mask.nxv4i32.nxv4i64.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vnsrl.mask.nxv4i32.nxv4i64.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vnsrl_wx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint64m4_t op1, size_t shift, size_t vl) {
@@ -587,7 +587,7 @@ vuint32m2_t test_vnsrl_wx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnsrl.mask.nxv8i32.nxv8i64.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i32> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnsrl.mask.nxv8i32.nxv8i64.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i32> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vnsrl_wv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint64m8_t op1, vuint32m4_t shift, size_t vl) {
@@ -597,7 +597,7 @@ vuint32m4_t test_vnsrl_wv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vnsrl_wx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnsrl.mask.nxv8i32.nxv8i64.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vnsrl.mask.nxv8i32.nxv8i64.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vnsrl_wx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint64m8_t op1, size_t shift, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vor.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vor.c
index b7dc8c8..6a84e87 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vor.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vor.c
@@ -887,7 +887,7 @@ vuint64m8_t test_vor_vx_u64m8(vuint64m8_t op1, uint64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vor.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vor.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vor_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, vint8mf8_t op2, size_t vl) {
@@ -897,7 +897,7 @@ vint8mf8_t test_vor_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vor.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vor.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vor_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, int8_t op2, size_t vl) {
@@ -907,7 +907,7 @@ vint8mf8_t test_vor_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vor.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vor.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vor_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, vint8mf4_t op2, size_t vl) {
@@ -917,7 +917,7 @@ vint8mf4_t test_vor_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vor.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vor.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vor_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, int8_t op2, size_t vl) {
@@ -927,7 +927,7 @@ vint8mf4_t test_vor_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vor.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vor.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vor_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, vint8mf2_t op2, size_t vl) {
@@ -937,7 +937,7 @@ vint8mf2_t test_vor_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vor.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vor.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vor_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, int8_t op2, size_t vl) {
@@ -947,7 +947,7 @@ vint8mf2_t test_vor_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vor.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vor.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vor_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, vint8m1_t op2, size_t vl) {
@@ -957,7 +957,7 @@ vint8m1_t test_vor_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vor.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vor.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vor_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, int8_t op2, size_t vl) {
@@ -967,7 +967,7 @@ vint8m1_t test_vor_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vor.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vor.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vor_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, vint8m2_t op2, size_t vl) {
@@ -977,7 +977,7 @@ vint8m2_t test_vor_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vor.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vor.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vor_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, int8_t op2, size_t vl) {
@@ -987,7 +987,7 @@ vint8m2_t test_vor_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vor.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vor.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vor_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, vint8m4_t op2, size_t vl) {
@@ -997,7 +997,7 @@ vint8m4_t test_vor_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vor.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vor.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vor_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, int8_t op2, size_t vl) {
@@ -1007,7 +1007,7 @@ vint8m4_t test_vor_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vor.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vor.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vor_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, vint8m8_t op2, size_t vl) {
@@ -1017,7 +1017,7 @@ vint8m8_t test_vor_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vor.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vor.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vor_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, int8_t op2, size_t vl) {
@@ -1027,7 +1027,7 @@ vint8m8_t test_vor_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vor.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vor.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vor_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, vint16mf4_t op2, size_t vl) {
@@ -1037,7 +1037,7 @@ vint16mf4_t test_vor_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vor.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vor.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vor_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, int16_t op2, size_t vl) {
@@ -1047,7 +1047,7 @@ vint16mf4_t test_vor_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vor.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vor.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vor_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, vint16mf2_t op2, size_t vl) {
@@ -1057,7 +1057,7 @@ vint16mf2_t test_vor_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vor.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vor.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vor_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, int16_t op2, size_t vl) {
@@ -1067,7 +1067,7 @@ vint16mf2_t test_vor_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vor.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vor.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vor_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, vint16m1_t op2, size_t vl) {
@@ -1077,7 +1077,7 @@ vint16m1_t test_vor_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vor.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vor.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vor_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, int16_t op2, size_t vl) {
@@ -1087,7 +1087,7 @@ vint16m1_t test_vor_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vor.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vor.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vor_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, vint16m2_t op2, size_t vl) {
@@ -1097,7 +1097,7 @@ vint16m2_t test_vor_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t o
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vor.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vor.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vor_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, int16_t op2, size_t vl) {
@@ -1107,7 +1107,7 @@ vint16m2_t test_vor_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t o
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vor.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vor.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vor_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, vint16m4_t op2, size_t vl) {
@@ -1117,7 +1117,7 @@ vint16m4_t test_vor_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t o
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vor.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vor.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vor_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, int16_t op2, size_t vl) {
@@ -1127,7 +1127,7 @@ vint16m4_t test_vor_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t o
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vor.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vor.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vor_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, vint16m8_t op2, size_t vl) {
@@ -1137,7 +1137,7 @@ vint16m8_t test_vor_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t o
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vor.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vor.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vor_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, int16_t op2, size_t vl) {
@@ -1147,7 +1147,7 @@ vint16m8_t test_vor_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t o
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vor.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vor.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vor_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, vint32mf2_t op2, size_t vl) {
@@ -1157,7 +1157,7 @@ vint32mf2_t test_vor_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vor.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vor.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vor_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, int32_t op2, size_t vl) {
@@ -1167,7 +1167,7 @@ vint32mf2_t test_vor_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vor_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, vint32m1_t op2, size_t vl) {
@@ -1177,7 +1177,7 @@ vint32m1_t test_vor_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vor.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vor.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vor_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, int32_t op2, size_t vl) {
@@ -1187,7 +1187,7 @@ vint32m1_t test_vor_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vor.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vor.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vor_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, vint32m2_t op2, size_t vl) {
@@ -1197,7 +1197,7 @@ vint32m2_t test_vor_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vor.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vor.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vor_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, int32_t op2, size_t vl) {
@@ -1207,7 +1207,7 @@ vint32m2_t test_vor_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vor.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vor.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vor_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, vint32m4_t op2, size_t vl) {
@@ -1217,7 +1217,7 @@ vint32m4_t test_vor_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t o
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vor.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vor.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vor_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, int32_t op2, size_t vl) {
@@ -1227,7 +1227,7 @@ vint32m4_t test_vor_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t o
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vor.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vor.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vor_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, vint32m8_t op2, size_t vl) {
@@ -1237,7 +1237,7 @@ vint32m8_t test_vor_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t o
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vor.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vor.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vor_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, int32_t op2, size_t vl) {
@@ -1247,7 +1247,7 @@ vint32m8_t test_vor_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t o
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vor_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, vint64m1_t op2, size_t vl) {
@@ -1257,7 +1257,7 @@ vint64m1_t test_vor_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vor.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vor.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vor_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, int64_t op2, size_t vl) {
@@ -1267,7 +1267,7 @@ vint64m1_t test_vor_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vor.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vor.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vor_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, vint64m2_t op2, size_t vl) {
@@ -1277,7 +1277,7 @@ vint64m2_t test_vor_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vor.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vor.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vor_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, int64_t op2, size_t vl) {
@@ -1287,7 +1287,7 @@ vint64m2_t test_vor_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vor.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vor.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vor_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, vint64m4_t op2, size_t vl) {
@@ -1297,7 +1297,7 @@ vint64m4_t test_vor_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vor.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vor.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vor_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, int64_t op2, size_t vl) {
@@ -1307,7 +1307,7 @@ vint64m4_t test_vor_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vor_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vor.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vor.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vor_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, vint64m8_t op2, size_t vl) {
@@ -1317,7 +1317,7 @@ vint64m8_t test_vor_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t o
 //
 // CHECK-RV64-LABEL: @test_vor_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vor.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vor.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vor_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, int64_t op2, size_t vl) {
@@ -1327,7 +1327,7 @@ vint64m8_t test_vor_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t o
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vor.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vor.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vor_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -1337,7 +1337,7 @@ vuint8mf8_t test_vor_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vor.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vor.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vor_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, uint8_t op2, size_t vl) {
@@ -1347,7 +1347,7 @@ vuint8mf8_t test_vor_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vor.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vor.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vor_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, vuint8mf4_t op2, size_t vl) {
@@ -1357,7 +1357,7 @@ vuint8mf4_t test_vor_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vor.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vor.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vor_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, uint8_t op2, size_t vl) {
@@ -1367,7 +1367,7 @@ vuint8mf4_t test_vor_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vor.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vor.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vor_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, vuint8mf2_t op2, size_t vl) {
@@ -1377,7 +1377,7 @@ vuint8mf2_t test_vor_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vor.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vor.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vor_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, uint8_t op2, size_t vl) {
@@ -1387,7 +1387,7 @@ vuint8mf2_t test_vor_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vor.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vor.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vor_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, vuint8m1_t op2, size_t vl) {
@@ -1397,7 +1397,7 @@ vuint8m1_t test_vor_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vor.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vor.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vor_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, uint8_t op2, size_t vl) {
@@ -1407,7 +1407,7 @@ vuint8m1_t test_vor_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vor.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vor.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vor_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, vuint8m2_t op2, size_t vl) {
@@ -1417,7 +1417,7 @@ vuint8m2_t test_vor_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vor.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vor.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vor_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, uint8_t op2, size_t vl) {
@@ -1427,7 +1427,7 @@ vuint8m2_t test_vor_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vor.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vor.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vor_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, vuint8m4_t op2, size_t vl) {
@@ -1437,7 +1437,7 @@ vuint8m4_t test_vor_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vor.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vor.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vor_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, uint8_t op2, size_t vl) {
@@ -1447,7 +1447,7 @@ vuint8m4_t test_vor_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vor.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vor.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vor_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, vuint8m8_t op2, size_t vl) {
@@ -1457,7 +1457,7 @@ vuint8m8_t test_vor_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vor.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vor.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vor_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, uint8_t op2, size_t vl) {
@@ -1467,7 +1467,7 @@ vuint8m8_t test_vor_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vor.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vor.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vor_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, vuint16mf4_t op2, size_t vl) {
@@ -1477,7 +1477,7 @@ vuint16mf4_t test_vor_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint1
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vor.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vor.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vor_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, uint16_t op2, size_t vl) {
@@ -1487,7 +1487,7 @@ vuint16mf4_t test_vor_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint1
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vor.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vor.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vor_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, vuint16mf2_t op2, size_t vl) {
@@ -1497,7 +1497,7 @@ vuint16mf2_t test_vor_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint1
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vor.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vor.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vor_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, uint16_t op2, size_t vl) {
@@ -1507,7 +1507,7 @@ vuint16mf2_t test_vor_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint1
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vor.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vor.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vor_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, vuint16m1_t op2, size_t vl) {
@@ -1517,7 +1517,7 @@ vuint16m1_t test_vor_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vor.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vor.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vor_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, uint16_t op2, size_t vl) {
@@ -1527,7 +1527,7 @@ vuint16m1_t test_vor_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vor.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vor.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vor_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, vuint16m2_t op2, size_t vl) {
@@ -1537,7 +1537,7 @@ vuint16m2_t test_vor_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vor.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vor.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vor_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, uint16_t op2, size_t vl) {
@@ -1547,7 +1547,7 @@ vuint16m2_t test_vor_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vor.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vor.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vor_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, vuint16m4_t op2, size_t vl) {
@@ -1557,7 +1557,7 @@ vuint16m4_t test_vor_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vor.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vor.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vor_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, uint16_t op2, size_t vl) {
@@ -1567,7 +1567,7 @@ vuint16m4_t test_vor_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vor.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vor.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vor_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, vuint16m8_t op2, size_t vl) {
@@ -1577,7 +1577,7 @@ vuint16m8_t test_vor_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vor.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vor.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vor_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, uint16_t op2, size_t vl) {
@@ -1587,7 +1587,7 @@ vuint16m8_t test_vor_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vor.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vor.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vor_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, vuint32mf2_t op2, size_t vl) {
@@ -1597,7 +1597,7 @@ vuint32mf2_t test_vor_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint3
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vor.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vor.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vor_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, uint32_t op2, size_t vl) {
@@ -1607,7 +1607,7 @@ vuint32mf2_t test_vor_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint3
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vor_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, vuint32m1_t op2, size_t vl) {
@@ -1617,7 +1617,7 @@ vuint32m1_t test_vor_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vor.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vor.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vor_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, uint32_t op2, size_t vl) {
@@ -1627,7 +1627,7 @@ vuint32m1_t test_vor_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vor.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vor.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vor_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, vuint32m2_t op2, size_t vl) {
@@ -1637,7 +1637,7 @@ vuint32m2_t test_vor_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vor.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vor.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vor_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, uint32_t op2, size_t vl) {
@@ -1647,7 +1647,7 @@ vuint32m2_t test_vor_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vor.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vor.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vor_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, vuint32m4_t op2, size_t vl) {
@@ -1657,7 +1657,7 @@ vuint32m4_t test_vor_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vor.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vor.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vor_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, uint32_t op2, size_t vl) {
@@ -1667,7 +1667,7 @@ vuint32m4_t test_vor_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vor.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vor.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vor_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, vuint32m8_t op2, size_t vl) {
@@ -1677,7 +1677,7 @@ vuint32m8_t test_vor_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vor.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vor.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vor_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, uint32_t op2, size_t vl) {
@@ -1687,7 +1687,7 @@ vuint32m8_t test_vor_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vor_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, vuint64m1_t op2, size_t vl) {
@@ -1697,7 +1697,7 @@ vuint64m1_t test_vor_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vor.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vor.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vor_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, uint64_t op2, size_t vl) {
@@ -1707,7 +1707,7 @@ vuint64m1_t test_vor_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vor.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vor.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vor_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, vuint64m2_t op2, size_t vl) {
@@ -1717,7 +1717,7 @@ vuint64m2_t test_vor_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vor.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vor.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vor_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, uint64_t op2, size_t vl) {
@@ -1727,7 +1727,7 @@ vuint64m2_t test_vor_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vor.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vor.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vor_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, vuint64m4_t op2, size_t vl) {
@@ -1737,7 +1737,7 @@ vuint64m4_t test_vor_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vor.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vor.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vor_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, uint64_t op2, size_t vl) {
@@ -1747,7 +1747,7 @@ vuint64m4_t test_vor_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4
 //
 // CHECK-RV64-LABEL: @test_vor_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vor.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vor.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vor_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, vuint64m8_t op2, size_t vl) {
@@ -1757,7 +1757,7 @@ vuint64m8_t test_vor_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_
 //
 // CHECK-RV64-LABEL: @test_vor_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vor.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vor.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vor_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, uint64_t op2, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vpopc.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vpopc.c
index f64f4f2..4a7d9cb 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vpopc.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vpopc.c
@@ -77,7 +77,7 @@ unsigned long test_vpopc_m_b64(vbool64_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vpopc_m_b1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vpopc.mask.nxv64i1.i64(<vscale x 64 x i1> [[OP1:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vpopc.mask.nxv64i1.i64(<vscale x 64 x i1> [[OP1:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret i64 [[TMP0]]
 //
 unsigned long test_vpopc_m_b1_m(vbool1_t mask, vbool1_t op1, size_t vl) {
@@ -87,7 +87,7 @@ unsigned long test_vpopc_m_b1_m(vbool1_t mask, vbool1_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vpopc_m_b2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vpopc.mask.nxv32i1.i64(<vscale x 32 x i1> [[OP1:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vpopc.mask.nxv32i1.i64(<vscale x 32 x i1> [[OP1:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret i64 [[TMP0]]
 //
 unsigned long test_vpopc_m_b2_m(vbool2_t mask, vbool2_t op1, size_t vl) {
@@ -97,7 +97,7 @@ unsigned long test_vpopc_m_b2_m(vbool2_t mask, vbool2_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vpopc_m_b4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vpopc.mask.nxv16i1.i64(<vscale x 16 x i1> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vpopc.mask.nxv16i1.i64(<vscale x 16 x i1> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret i64 [[TMP0]]
 //
 unsigned long test_vpopc_m_b4_m(vbool4_t mask, vbool4_t op1, size_t vl) {
@@ -107,7 +107,7 @@ unsigned long test_vpopc_m_b4_m(vbool4_t mask, vbool4_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vpopc_m_b8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vpopc.mask.nxv8i1.i64(<vscale x 8 x i1> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vpopc.mask.nxv8i1.i64(<vscale x 8 x i1> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret i64 [[TMP0]]
 //
 unsigned long test_vpopc_m_b8_m(vbool8_t mask, vbool8_t op1, size_t vl) {
@@ -117,7 +117,7 @@ unsigned long test_vpopc_m_b8_m(vbool8_t mask, vbool8_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vpopc_m_b16_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vpopc.mask.nxv4i1.i64(<vscale x 4 x i1> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vpopc.mask.nxv4i1.i64(<vscale x 4 x i1> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret i64 [[TMP0]]
 //
 unsigned long test_vpopc_m_b16_m(vbool16_t mask, vbool16_t op1, size_t vl) {
@@ -127,7 +127,7 @@ unsigned long test_vpopc_m_b16_m(vbool16_t mask, vbool16_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vpopc_m_b32_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vpopc.mask.nxv2i1.i64(<vscale x 2 x i1> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vpopc.mask.nxv2i1.i64(<vscale x 2 x i1> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret i64 [[TMP0]]
 //
 unsigned long test_vpopc_m_b32_m(vbool32_t mask, vbool32_t op1, size_t vl) {
@@ -137,7 +137,7 @@ unsigned long test_vpopc_m_b32_m(vbool32_t mask, vbool32_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vpopc_m_b64_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vpopc.mask.nxv1i1.i64(<vscale x 1 x i1> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call i64 @llvm.riscv.vpopc.mask.nxv1i1.i64(<vscale x 1 x i1> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret i64 [[TMP0]]
 //
 unsigned long test_vpopc_m_b64_m(vbool64_t mask, vbool64_t op1, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredand.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredand.c
index cc070db6..6aa7874 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredand.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredand.c
@@ -491,7 +491,7 @@ vuint64m1_t test_vredand_vs_u64m8_u64m1(vuint64m1_t dst, vuint64m8_t vector,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i8mf8_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredand_vs_i8mf8_i8m1_m(vbool64_t mask, vint8m1_t dst,
@@ -503,7 +503,7 @@ vint8m1_t test_vredand_vs_i8mf8_i8m1_m(vbool64_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i8mf4_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredand_vs_i8mf4_i8m1_m(vbool32_t mask, vint8m1_t dst,
@@ -515,7 +515,7 @@ vint8m1_t test_vredand_vs_i8mf4_i8m1_m(vbool32_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i8mf2_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredand_vs_i8mf2_i8m1_m(vbool16_t mask, vint8m1_t dst,
@@ -527,7 +527,7 @@ vint8m1_t test_vredand_vs_i8mf2_i8m1_m(vbool16_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i8m1_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredand_vs_i8m1_i8m1_m(vbool8_t mask, vint8m1_t dst,
@@ -539,7 +539,7 @@ vint8m1_t test_vredand_vs_i8m1_i8m1_m(vbool8_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i8m2_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredand_vs_i8m2_i8m1_m(vbool4_t mask, vint8m1_t dst,
@@ -551,7 +551,7 @@ vint8m1_t test_vredand_vs_i8m2_i8m1_m(vbool4_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i8m4_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredand_vs_i8m4_i8m1_m(vbool2_t mask, vint8m1_t dst,
@@ -563,7 +563,7 @@ vint8m1_t test_vredand_vs_i8m4_i8m1_m(vbool2_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i8m8_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredand_vs_i8m8_i8m1_m(vbool1_t mask, vint8m1_t dst,
@@ -575,7 +575,7 @@ vint8m1_t test_vredand_vs_i8m8_i8m1_m(vbool1_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i16mf4_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredand_vs_i16mf4_i16m1_m(vbool64_t mask, vint16m1_t dst,
@@ -587,7 +587,7 @@ vint16m1_t test_vredand_vs_i16mf4_i16m1_m(vbool64_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i16mf2_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredand_vs_i16mf2_i16m1_m(vbool32_t mask, vint16m1_t dst,
@@ -599,7 +599,7 @@ vint16m1_t test_vredand_vs_i16mf2_i16m1_m(vbool32_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i16m1_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredand_vs_i16m1_i16m1_m(vbool16_t mask, vint16m1_t dst,
@@ -611,7 +611,7 @@ vint16m1_t test_vredand_vs_i16m1_i16m1_m(vbool16_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i16m2_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredand_vs_i16m2_i16m1_m(vbool8_t mask, vint16m1_t dst,
@@ -623,7 +623,7 @@ vint16m1_t test_vredand_vs_i16m2_i16m1_m(vbool8_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i16m4_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredand_vs_i16m4_i16m1_m(vbool4_t mask, vint16m1_t dst,
@@ -635,7 +635,7 @@ vint16m1_t test_vredand_vs_i16m4_i16m1_m(vbool4_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i16m8_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredand_vs_i16m8_i16m1_m(vbool2_t mask, vint16m1_t dst,
@@ -647,7 +647,7 @@ vint16m1_t test_vredand_vs_i16m8_i16m1_m(vbool2_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i32mf2_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredand.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredand.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredand_vs_i32mf2_i32m1_m(vbool64_t mask, vint32m1_t dst,
@@ -659,7 +659,7 @@ vint32m1_t test_vredand_vs_i32mf2_i32m1_m(vbool64_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i32m1_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredand.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredand.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredand_vs_i32m1_i32m1_m(vbool32_t mask, vint32m1_t dst,
@@ -671,7 +671,7 @@ vint32m1_t test_vredand_vs_i32m1_i32m1_m(vbool32_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i32m2_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredand.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredand.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredand_vs_i32m2_i32m1_m(vbool16_t mask, vint32m1_t dst,
@@ -683,7 +683,7 @@ vint32m1_t test_vredand_vs_i32m2_i32m1_m(vbool16_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i32m4_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredand.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredand.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredand_vs_i32m4_i32m1_m(vbool8_t mask, vint32m1_t dst,
@@ -695,7 +695,7 @@ vint32m1_t test_vredand_vs_i32m4_i32m1_m(vbool8_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i32m8_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredand.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredand.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredand_vs_i32m8_i32m1_m(vbool4_t mask, vint32m1_t dst,
@@ -707,7 +707,7 @@ vint32m1_t test_vredand_vs_i32m8_i32m1_m(vbool4_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i64m1_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredand.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredand.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredand_vs_i64m1_i64m1_m(vbool64_t mask, vint64m1_t dst,
@@ -719,7 +719,7 @@ vint64m1_t test_vredand_vs_i64m1_i64m1_m(vbool64_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i64m2_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredand.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredand.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredand_vs_i64m2_i64m1_m(vbool32_t mask, vint64m1_t dst,
@@ -731,7 +731,7 @@ vint64m1_t test_vredand_vs_i64m2_i64m1_m(vbool32_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i64m4_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredand.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredand.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredand_vs_i64m4_i64m1_m(vbool16_t mask, vint64m1_t dst,
@@ -743,7 +743,7 @@ vint64m1_t test_vredand_vs_i64m4_i64m1_m(vbool16_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_i64m8_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredand.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredand.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredand_vs_i64m8_i64m1_m(vbool8_t mask, vint64m1_t dst,
@@ -755,7 +755,7 @@ vint64m1_t test_vredand_vs_i64m8_i64m1_m(vbool8_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u8mf8_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredand_vs_u8mf8_u8m1_m(vbool64_t mask, vuint8m1_t dst,
@@ -767,7 +767,7 @@ vuint8m1_t test_vredand_vs_u8mf8_u8m1_m(vbool64_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u8mf4_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredand_vs_u8mf4_u8m1_m(vbool32_t mask, vuint8m1_t dst,
@@ -779,7 +779,7 @@ vuint8m1_t test_vredand_vs_u8mf4_u8m1_m(vbool32_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u8mf2_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredand_vs_u8mf2_u8m1_m(vbool16_t mask, vuint8m1_t dst,
@@ -791,7 +791,7 @@ vuint8m1_t test_vredand_vs_u8mf2_u8m1_m(vbool16_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u8m1_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredand_vs_u8m1_u8m1_m(vbool8_t mask, vuint8m1_t dst,
@@ -803,7 +803,7 @@ vuint8m1_t test_vredand_vs_u8m1_u8m1_m(vbool8_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u8m2_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredand_vs_u8m2_u8m1_m(vbool4_t mask, vuint8m1_t dst,
@@ -815,7 +815,7 @@ vuint8m1_t test_vredand_vs_u8m2_u8m1_m(vbool4_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u8m4_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredand_vs_u8m4_u8m1_m(vbool2_t mask, vuint8m1_t dst,
@@ -827,7 +827,7 @@ vuint8m1_t test_vredand_vs_u8m4_u8m1_m(vbool2_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u8m8_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredand.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredand_vs_u8m8_u8m1_m(vbool1_t mask, vuint8m1_t dst,
@@ -839,7 +839,7 @@ vuint8m1_t test_vredand_vs_u8m8_u8m1_m(vbool1_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u16mf4_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredand_vs_u16mf4_u16m1_m(vbool64_t mask, vuint16m1_t dst,
@@ -851,7 +851,7 @@ vuint16m1_t test_vredand_vs_u16mf4_u16m1_m(vbool64_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u16mf2_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredand_vs_u16mf2_u16m1_m(vbool32_t mask, vuint16m1_t dst,
@@ -863,7 +863,7 @@ vuint16m1_t test_vredand_vs_u16mf2_u16m1_m(vbool32_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u16m1_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredand_vs_u16m1_u16m1_m(vbool16_t mask, vuint16m1_t dst,
@@ -875,7 +875,7 @@ vuint16m1_t test_vredand_vs_u16m1_u16m1_m(vbool16_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u16m2_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredand_vs_u16m2_u16m1_m(vbool8_t mask, vuint16m1_t dst,
@@ -887,7 +887,7 @@ vuint16m1_t test_vredand_vs_u16m2_u16m1_m(vbool8_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u16m4_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredand_vs_u16m4_u16m1_m(vbool4_t mask, vuint16m1_t dst,
@@ -899,7 +899,7 @@ vuint16m1_t test_vredand_vs_u16m4_u16m1_m(vbool4_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u16m8_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredand.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredand_vs_u16m8_u16m1_m(vbool2_t mask, vuint16m1_t dst,
@@ -911,7 +911,7 @@ vuint16m1_t test_vredand_vs_u16m8_u16m1_m(vbool2_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u32mf2_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredand.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredand.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredand_vs_u32mf2_u32m1_m(vbool64_t mask, vuint32m1_t dst,
@@ -923,7 +923,7 @@ vuint32m1_t test_vredand_vs_u32mf2_u32m1_m(vbool64_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u32m1_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredand.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredand.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredand_vs_u32m1_u32m1_m(vbool32_t mask, vuint32m1_t dst,
@@ -935,7 +935,7 @@ vuint32m1_t test_vredand_vs_u32m1_u32m1_m(vbool32_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u32m2_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredand.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredand.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredand_vs_u32m2_u32m1_m(vbool16_t mask, vuint32m1_t dst,
@@ -947,7 +947,7 @@ vuint32m1_t test_vredand_vs_u32m2_u32m1_m(vbool16_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u32m4_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredand.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredand.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredand_vs_u32m4_u32m1_m(vbool8_t mask, vuint32m1_t dst,
@@ -959,7 +959,7 @@ vuint32m1_t test_vredand_vs_u32m4_u32m1_m(vbool8_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u32m8_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredand.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredand.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredand_vs_u32m8_u32m1_m(vbool4_t mask, vuint32m1_t dst,
@@ -971,7 +971,7 @@ vuint32m1_t test_vredand_vs_u32m8_u32m1_m(vbool4_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u64m1_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredand.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredand.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredand_vs_u64m1_u64m1_m(vbool64_t mask, vuint64m1_t dst,
@@ -983,7 +983,7 @@ vuint64m1_t test_vredand_vs_u64m1_u64m1_m(vbool64_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u64m2_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredand.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredand.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredand_vs_u64m2_u64m1_m(vbool32_t mask, vuint64m1_t dst,
@@ -995,7 +995,7 @@ vuint64m1_t test_vredand_vs_u64m2_u64m1_m(vbool32_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u64m4_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredand.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredand.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredand_vs_u64m4_u64m1_m(vbool16_t mask, vuint64m1_t dst,
@@ -1007,7 +1007,7 @@ vuint64m1_t test_vredand_vs_u64m4_u64m1_m(vbool16_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredand_vs_u64m8_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredand.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredand.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredand_vs_u64m8_u64m1_m(vbool8_t mask, vuint64m1_t dst,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredmax.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredmax.c
index aff604a..ed1ae50 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredmax.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredmax.c
@@ -491,7 +491,7 @@ vuint64m1_t test_vredmaxu_vs_u64m8_u64m1(vuint64m1_t dst, vuint64m8_t vector,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i8mf8_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmax.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmax.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredmax_vs_i8mf8_i8m1_m(vbool64_t mask, vint8m1_t dst,
@@ -503,7 +503,7 @@ vint8m1_t test_vredmax_vs_i8mf8_i8m1_m(vbool64_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i8mf4_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmax.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmax.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredmax_vs_i8mf4_i8m1_m(vbool32_t mask, vint8m1_t dst,
@@ -515,7 +515,7 @@ vint8m1_t test_vredmax_vs_i8mf4_i8m1_m(vbool32_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i8mf2_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmax.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmax.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredmax_vs_i8mf2_i8m1_m(vbool16_t mask, vint8m1_t dst,
@@ -527,7 +527,7 @@ vint8m1_t test_vredmax_vs_i8mf2_i8m1_m(vbool16_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i8m1_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmax.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmax.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredmax_vs_i8m1_i8m1_m(vbool8_t mask, vint8m1_t dst,
@@ -539,7 +539,7 @@ vint8m1_t test_vredmax_vs_i8m1_i8m1_m(vbool8_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i8m2_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmax.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmax.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredmax_vs_i8m2_i8m1_m(vbool4_t mask, vint8m1_t dst,
@@ -551,7 +551,7 @@ vint8m1_t test_vredmax_vs_i8m2_i8m1_m(vbool4_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i8m4_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmax.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmax.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredmax_vs_i8m4_i8m1_m(vbool2_t mask, vint8m1_t dst,
@@ -563,7 +563,7 @@ vint8m1_t test_vredmax_vs_i8m4_i8m1_m(vbool2_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i8m8_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmax.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmax.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredmax_vs_i8m8_i8m1_m(vbool1_t mask, vint8m1_t dst,
@@ -575,7 +575,7 @@ vint8m1_t test_vredmax_vs_i8m8_i8m1_m(vbool1_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i16mf4_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmax.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmax.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredmax_vs_i16mf4_i16m1_m(vbool64_t mask, vint16m1_t dst,
@@ -587,7 +587,7 @@ vint16m1_t test_vredmax_vs_i16mf4_i16m1_m(vbool64_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i16mf2_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmax.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmax.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredmax_vs_i16mf2_i16m1_m(vbool32_t mask, vint16m1_t dst,
@@ -599,7 +599,7 @@ vint16m1_t test_vredmax_vs_i16mf2_i16m1_m(vbool32_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i16m1_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmax.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmax.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredmax_vs_i16m1_i16m1_m(vbool16_t mask, vint16m1_t dst,
@@ -611,7 +611,7 @@ vint16m1_t test_vredmax_vs_i16m1_i16m1_m(vbool16_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i16m2_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmax.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmax.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredmax_vs_i16m2_i16m1_m(vbool8_t mask, vint16m1_t dst,
@@ -623,7 +623,7 @@ vint16m1_t test_vredmax_vs_i16m2_i16m1_m(vbool8_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i16m4_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmax.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmax.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredmax_vs_i16m4_i16m1_m(vbool4_t mask, vint16m1_t dst,
@@ -635,7 +635,7 @@ vint16m1_t test_vredmax_vs_i16m4_i16m1_m(vbool4_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i16m8_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmax.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmax.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredmax_vs_i16m8_i16m1_m(vbool2_t mask, vint16m1_t dst,
@@ -647,7 +647,7 @@ vint16m1_t test_vredmax_vs_i16m8_i16m1_m(vbool2_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i32mf2_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmax.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmax.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredmax_vs_i32mf2_i32m1_m(vbool64_t mask, vint32m1_t dst,
@@ -659,7 +659,7 @@ vint32m1_t test_vredmax_vs_i32mf2_i32m1_m(vbool64_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i32m1_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmax.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmax.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredmax_vs_i32m1_i32m1_m(vbool32_t mask, vint32m1_t dst,
@@ -671,7 +671,7 @@ vint32m1_t test_vredmax_vs_i32m1_i32m1_m(vbool32_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i32m2_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmax.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmax.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredmax_vs_i32m2_i32m1_m(vbool16_t mask, vint32m1_t dst,
@@ -683,7 +683,7 @@ vint32m1_t test_vredmax_vs_i32m2_i32m1_m(vbool16_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i32m4_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmax.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmax.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredmax_vs_i32m4_i32m1_m(vbool8_t mask, vint32m1_t dst,
@@ -695,7 +695,7 @@ vint32m1_t test_vredmax_vs_i32m4_i32m1_m(vbool8_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i32m8_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmax.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmax.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredmax_vs_i32m8_i32m1_m(vbool4_t mask, vint32m1_t dst,
@@ -707,7 +707,7 @@ vint32m1_t test_vredmax_vs_i32m8_i32m1_m(vbool4_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i64m1_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmax.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmax.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredmax_vs_i64m1_i64m1_m(vbool64_t mask, vint64m1_t dst,
@@ -719,7 +719,7 @@ vint64m1_t test_vredmax_vs_i64m1_i64m1_m(vbool64_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i64m2_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmax.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmax.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredmax_vs_i64m2_i64m1_m(vbool32_t mask, vint64m1_t dst,
@@ -731,7 +731,7 @@ vint64m1_t test_vredmax_vs_i64m2_i64m1_m(vbool32_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i64m4_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmax.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmax.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredmax_vs_i64m4_i64m1_m(vbool16_t mask, vint64m1_t dst,
@@ -743,7 +743,7 @@ vint64m1_t test_vredmax_vs_i64m4_i64m1_m(vbool16_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmax_vs_i64m8_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmax.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmax.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredmax_vs_i64m8_i64m1_m(vbool8_t mask, vint64m1_t dst,
@@ -755,7 +755,7 @@ vint64m1_t test_vredmax_vs_i64m8_i64m1_m(vbool8_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u8mf8_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmaxu.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmaxu.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredmaxu_vs_u8mf8_u8m1_m(vbool64_t mask, vuint8m1_t dst,
@@ -767,7 +767,7 @@ vuint8m1_t test_vredmaxu_vs_u8mf8_u8m1_m(vbool64_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u8mf4_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmaxu.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmaxu.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredmaxu_vs_u8mf4_u8m1_m(vbool32_t mask, vuint8m1_t dst,
@@ -779,7 +779,7 @@ vuint8m1_t test_vredmaxu_vs_u8mf4_u8m1_m(vbool32_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u8mf2_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmaxu.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmaxu.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredmaxu_vs_u8mf2_u8m1_m(vbool16_t mask, vuint8m1_t dst,
@@ -791,7 +791,7 @@ vuint8m1_t test_vredmaxu_vs_u8mf2_u8m1_m(vbool16_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u8m1_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmaxu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmaxu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredmaxu_vs_u8m1_u8m1_m(vbool8_t mask, vuint8m1_t dst,
@@ -803,7 +803,7 @@ vuint8m1_t test_vredmaxu_vs_u8m1_u8m1_m(vbool8_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u8m2_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmaxu.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmaxu.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredmaxu_vs_u8m2_u8m1_m(vbool4_t mask, vuint8m1_t dst,
@@ -815,7 +815,7 @@ vuint8m1_t test_vredmaxu_vs_u8m2_u8m1_m(vbool4_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u8m4_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmaxu.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmaxu.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredmaxu_vs_u8m4_u8m1_m(vbool2_t mask, vuint8m1_t dst,
@@ -827,7 +827,7 @@ vuint8m1_t test_vredmaxu_vs_u8m4_u8m1_m(vbool2_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u8m8_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmaxu.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmaxu.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredmaxu_vs_u8m8_u8m1_m(vbool1_t mask, vuint8m1_t dst,
@@ -839,7 +839,7 @@ vuint8m1_t test_vredmaxu_vs_u8m8_u8m1_m(vbool1_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u16mf4_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmaxu.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmaxu.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredmaxu_vs_u16mf4_u16m1_m(vbool64_t mask, vuint16m1_t dst,
@@ -851,7 +851,7 @@ vuint16m1_t test_vredmaxu_vs_u16mf4_u16m1_m(vbool64_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u16mf2_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmaxu.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmaxu.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredmaxu_vs_u16mf2_u16m1_m(vbool32_t mask, vuint16m1_t dst,
@@ -863,7 +863,7 @@ vuint16m1_t test_vredmaxu_vs_u16mf2_u16m1_m(vbool32_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u16m1_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmaxu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmaxu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredmaxu_vs_u16m1_u16m1_m(vbool16_t mask, vuint16m1_t dst,
@@ -875,7 +875,7 @@ vuint16m1_t test_vredmaxu_vs_u16m1_u16m1_m(vbool16_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u16m2_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmaxu.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmaxu.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredmaxu_vs_u16m2_u16m1_m(vbool8_t mask, vuint16m1_t dst,
@@ -887,7 +887,7 @@ vuint16m1_t test_vredmaxu_vs_u16m2_u16m1_m(vbool8_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u16m4_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmaxu.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmaxu.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredmaxu_vs_u16m4_u16m1_m(vbool4_t mask, vuint16m1_t dst,
@@ -899,7 +899,7 @@ vuint16m1_t test_vredmaxu_vs_u16m4_u16m1_m(vbool4_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u16m8_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmaxu.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmaxu.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredmaxu_vs_u16m8_u16m1_m(vbool2_t mask, vuint16m1_t dst,
@@ -911,7 +911,7 @@ vuint16m1_t test_vredmaxu_vs_u16m8_u16m1_m(vbool2_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u32mf2_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmaxu.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmaxu.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredmaxu_vs_u32mf2_u32m1_m(vbool64_t mask, vuint32m1_t dst,
@@ -923,7 +923,7 @@ vuint32m1_t test_vredmaxu_vs_u32mf2_u32m1_m(vbool64_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u32m1_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmaxu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmaxu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredmaxu_vs_u32m1_u32m1_m(vbool32_t mask, vuint32m1_t dst,
@@ -935,7 +935,7 @@ vuint32m1_t test_vredmaxu_vs_u32m1_u32m1_m(vbool32_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u32m2_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmaxu.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmaxu.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredmaxu_vs_u32m2_u32m1_m(vbool16_t mask, vuint32m1_t dst,
@@ -947,7 +947,7 @@ vuint32m1_t test_vredmaxu_vs_u32m2_u32m1_m(vbool16_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u32m4_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmaxu.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmaxu.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredmaxu_vs_u32m4_u32m1_m(vbool8_t mask, vuint32m1_t dst,
@@ -959,7 +959,7 @@ vuint32m1_t test_vredmaxu_vs_u32m4_u32m1_m(vbool8_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u32m8_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmaxu.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmaxu.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredmaxu_vs_u32m8_u32m1_m(vbool4_t mask, vuint32m1_t dst,
@@ -971,7 +971,7 @@ vuint32m1_t test_vredmaxu_vs_u32m8_u32m1_m(vbool4_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u64m1_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmaxu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmaxu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredmaxu_vs_u64m1_u64m1_m(vbool64_t mask, vuint64m1_t dst,
@@ -983,7 +983,7 @@ vuint64m1_t test_vredmaxu_vs_u64m1_u64m1_m(vbool64_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u64m2_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmaxu.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmaxu.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredmaxu_vs_u64m2_u64m1_m(vbool32_t mask, vuint64m1_t dst,
@@ -995,7 +995,7 @@ vuint64m1_t test_vredmaxu_vs_u64m2_u64m1_m(vbool32_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u64m4_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmaxu.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmaxu.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredmaxu_vs_u64m4_u64m1_m(vbool16_t mask, vuint64m1_t dst,
@@ -1007,7 +1007,7 @@ vuint64m1_t test_vredmaxu_vs_u64m4_u64m1_m(vbool16_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmaxu_vs_u64m8_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmaxu.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmaxu.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredmaxu_vs_u64m8_u64m1_m(vbool8_t mask, vuint64m1_t dst,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredmin.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredmin.c
index 19d8456..bdf2205 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredmin.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredmin.c
@@ -491,7 +491,7 @@ vuint64m1_t test_vredminu_vs_u64m8_u64m1(vuint64m1_t dst, vuint64m8_t vector,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i8mf8_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmin.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmin.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredmin_vs_i8mf8_i8m1_m(vbool64_t mask, vint8m1_t dst,
@@ -503,7 +503,7 @@ vint8m1_t test_vredmin_vs_i8mf8_i8m1_m(vbool64_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i8mf4_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmin.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmin.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredmin_vs_i8mf4_i8m1_m(vbool32_t mask, vint8m1_t dst,
@@ -515,7 +515,7 @@ vint8m1_t test_vredmin_vs_i8mf4_i8m1_m(vbool32_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i8mf2_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmin.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmin.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredmin_vs_i8mf2_i8m1_m(vbool16_t mask, vint8m1_t dst,
@@ -527,7 +527,7 @@ vint8m1_t test_vredmin_vs_i8mf2_i8m1_m(vbool16_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i8m1_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmin.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmin.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredmin_vs_i8m1_i8m1_m(vbool8_t mask, vint8m1_t dst,
@@ -539,7 +539,7 @@ vint8m1_t test_vredmin_vs_i8m1_i8m1_m(vbool8_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i8m2_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmin.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmin.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredmin_vs_i8m2_i8m1_m(vbool4_t mask, vint8m1_t dst,
@@ -551,7 +551,7 @@ vint8m1_t test_vredmin_vs_i8m2_i8m1_m(vbool4_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i8m4_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmin.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmin.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredmin_vs_i8m4_i8m1_m(vbool2_t mask, vint8m1_t dst,
@@ -563,7 +563,7 @@ vint8m1_t test_vredmin_vs_i8m4_i8m1_m(vbool2_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i8m8_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmin.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredmin.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredmin_vs_i8m8_i8m1_m(vbool1_t mask, vint8m1_t dst,
@@ -575,7 +575,7 @@ vint8m1_t test_vredmin_vs_i8m8_i8m1_m(vbool1_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i16mf4_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmin.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmin.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredmin_vs_i16mf4_i16m1_m(vbool64_t mask, vint16m1_t dst,
@@ -587,7 +587,7 @@ vint16m1_t test_vredmin_vs_i16mf4_i16m1_m(vbool64_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i16mf2_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmin.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmin.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredmin_vs_i16mf2_i16m1_m(vbool32_t mask, vint16m1_t dst,
@@ -599,7 +599,7 @@ vint16m1_t test_vredmin_vs_i16mf2_i16m1_m(vbool32_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i16m1_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmin.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmin.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredmin_vs_i16m1_i16m1_m(vbool16_t mask, vint16m1_t dst,
@@ -611,7 +611,7 @@ vint16m1_t test_vredmin_vs_i16m1_i16m1_m(vbool16_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i16m2_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmin.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmin.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredmin_vs_i16m2_i16m1_m(vbool8_t mask, vint16m1_t dst,
@@ -623,7 +623,7 @@ vint16m1_t test_vredmin_vs_i16m2_i16m1_m(vbool8_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i16m4_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmin.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmin.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredmin_vs_i16m4_i16m1_m(vbool4_t mask, vint16m1_t dst,
@@ -635,7 +635,7 @@ vint16m1_t test_vredmin_vs_i16m4_i16m1_m(vbool4_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i16m8_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmin.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredmin.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredmin_vs_i16m8_i16m1_m(vbool2_t mask, vint16m1_t dst,
@@ -647,7 +647,7 @@ vint16m1_t test_vredmin_vs_i16m8_i16m1_m(vbool2_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i32mf2_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmin.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmin.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredmin_vs_i32mf2_i32m1_m(vbool64_t mask, vint32m1_t dst,
@@ -659,7 +659,7 @@ vint32m1_t test_vredmin_vs_i32mf2_i32m1_m(vbool64_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i32m1_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmin.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmin.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredmin_vs_i32m1_i32m1_m(vbool32_t mask, vint32m1_t dst,
@@ -671,7 +671,7 @@ vint32m1_t test_vredmin_vs_i32m1_i32m1_m(vbool32_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i32m2_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmin.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmin.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredmin_vs_i32m2_i32m1_m(vbool16_t mask, vint32m1_t dst,
@@ -683,7 +683,7 @@ vint32m1_t test_vredmin_vs_i32m2_i32m1_m(vbool16_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i32m4_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmin.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmin.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredmin_vs_i32m4_i32m1_m(vbool8_t mask, vint32m1_t dst,
@@ -695,7 +695,7 @@ vint32m1_t test_vredmin_vs_i32m4_i32m1_m(vbool8_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i32m8_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmin.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredmin.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredmin_vs_i32m8_i32m1_m(vbool4_t mask, vint32m1_t dst,
@@ -707,7 +707,7 @@ vint32m1_t test_vredmin_vs_i32m8_i32m1_m(vbool4_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i64m1_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmin.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmin.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredmin_vs_i64m1_i64m1_m(vbool64_t mask, vint64m1_t dst,
@@ -719,7 +719,7 @@ vint64m1_t test_vredmin_vs_i64m1_i64m1_m(vbool64_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i64m2_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmin.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmin.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredmin_vs_i64m2_i64m1_m(vbool32_t mask, vint64m1_t dst,
@@ -731,7 +731,7 @@ vint64m1_t test_vredmin_vs_i64m2_i64m1_m(vbool32_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i64m4_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmin.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmin.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredmin_vs_i64m4_i64m1_m(vbool16_t mask, vint64m1_t dst,
@@ -743,7 +743,7 @@ vint64m1_t test_vredmin_vs_i64m4_i64m1_m(vbool16_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredmin_vs_i64m8_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmin.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredmin.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredmin_vs_i64m8_i64m1_m(vbool8_t mask, vint64m1_t dst,
@@ -755,7 +755,7 @@ vint64m1_t test_vredmin_vs_i64m8_i64m1_m(vbool8_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u8mf8_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredminu.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredminu.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredminu_vs_u8mf8_u8m1_m(vbool64_t mask, vuint8m1_t dst,
@@ -767,7 +767,7 @@ vuint8m1_t test_vredminu_vs_u8mf8_u8m1_m(vbool64_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u8mf4_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredminu.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredminu.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredminu_vs_u8mf4_u8m1_m(vbool32_t mask, vuint8m1_t dst,
@@ -779,7 +779,7 @@ vuint8m1_t test_vredminu_vs_u8mf4_u8m1_m(vbool32_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u8mf2_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredminu.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredminu.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredminu_vs_u8mf2_u8m1_m(vbool16_t mask, vuint8m1_t dst,
@@ -791,7 +791,7 @@ vuint8m1_t test_vredminu_vs_u8mf2_u8m1_m(vbool16_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u8m1_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredminu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredminu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredminu_vs_u8m1_u8m1_m(vbool8_t mask, vuint8m1_t dst,
@@ -803,7 +803,7 @@ vuint8m1_t test_vredminu_vs_u8m1_u8m1_m(vbool8_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u8m2_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredminu.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredminu.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredminu_vs_u8m2_u8m1_m(vbool4_t mask, vuint8m1_t dst,
@@ -815,7 +815,7 @@ vuint8m1_t test_vredminu_vs_u8m2_u8m1_m(vbool4_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u8m4_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredminu.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredminu.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredminu_vs_u8m4_u8m1_m(vbool2_t mask, vuint8m1_t dst,
@@ -827,7 +827,7 @@ vuint8m1_t test_vredminu_vs_u8m4_u8m1_m(vbool2_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u8m8_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredminu.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredminu.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredminu_vs_u8m8_u8m1_m(vbool1_t mask, vuint8m1_t dst,
@@ -839,7 +839,7 @@ vuint8m1_t test_vredminu_vs_u8m8_u8m1_m(vbool1_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u16mf4_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredminu.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredminu.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredminu_vs_u16mf4_u16m1_m(vbool64_t mask, vuint16m1_t dst,
@@ -851,7 +851,7 @@ vuint16m1_t test_vredminu_vs_u16mf4_u16m1_m(vbool64_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u16mf2_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredminu.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredminu.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredminu_vs_u16mf2_u16m1_m(vbool32_t mask, vuint16m1_t dst,
@@ -863,7 +863,7 @@ vuint16m1_t test_vredminu_vs_u16mf2_u16m1_m(vbool32_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u16m1_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredminu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredminu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredminu_vs_u16m1_u16m1_m(vbool16_t mask, vuint16m1_t dst,
@@ -875,7 +875,7 @@ vuint16m1_t test_vredminu_vs_u16m1_u16m1_m(vbool16_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u16m2_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredminu.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredminu.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredminu_vs_u16m2_u16m1_m(vbool8_t mask, vuint16m1_t dst,
@@ -887,7 +887,7 @@ vuint16m1_t test_vredminu_vs_u16m2_u16m1_m(vbool8_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u16m4_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredminu.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredminu.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredminu_vs_u16m4_u16m1_m(vbool4_t mask, vuint16m1_t dst,
@@ -899,7 +899,7 @@ vuint16m1_t test_vredminu_vs_u16m4_u16m1_m(vbool4_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u16m8_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredminu.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredminu.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredminu_vs_u16m8_u16m1_m(vbool2_t mask, vuint16m1_t dst,
@@ -911,7 +911,7 @@ vuint16m1_t test_vredminu_vs_u16m8_u16m1_m(vbool2_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u32mf2_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredminu.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredminu.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredminu_vs_u32mf2_u32m1_m(vbool64_t mask, vuint32m1_t dst,
@@ -923,7 +923,7 @@ vuint32m1_t test_vredminu_vs_u32mf2_u32m1_m(vbool64_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u32m1_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredminu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredminu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredminu_vs_u32m1_u32m1_m(vbool32_t mask, vuint32m1_t dst,
@@ -935,7 +935,7 @@ vuint32m1_t test_vredminu_vs_u32m1_u32m1_m(vbool32_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u32m2_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredminu.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredminu.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredminu_vs_u32m2_u32m1_m(vbool16_t mask, vuint32m1_t dst,
@@ -947,7 +947,7 @@ vuint32m1_t test_vredminu_vs_u32m2_u32m1_m(vbool16_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u32m4_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredminu.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredminu.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredminu_vs_u32m4_u32m1_m(vbool8_t mask, vuint32m1_t dst,
@@ -959,7 +959,7 @@ vuint32m1_t test_vredminu_vs_u32m4_u32m1_m(vbool8_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u32m8_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredminu.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredminu.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredminu_vs_u32m8_u32m1_m(vbool4_t mask, vuint32m1_t dst,
@@ -971,7 +971,7 @@ vuint32m1_t test_vredminu_vs_u32m8_u32m1_m(vbool4_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u64m1_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredminu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredminu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredminu_vs_u64m1_u64m1_m(vbool64_t mask, vuint64m1_t dst,
@@ -983,7 +983,7 @@ vuint64m1_t test_vredminu_vs_u64m1_u64m1_m(vbool64_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u64m2_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredminu.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredminu.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredminu_vs_u64m2_u64m1_m(vbool32_t mask, vuint64m1_t dst,
@@ -995,7 +995,7 @@ vuint64m1_t test_vredminu_vs_u64m2_u64m1_m(vbool32_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u64m4_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredminu.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredminu.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredminu_vs_u64m4_u64m1_m(vbool16_t mask, vuint64m1_t dst,
@@ -1007,7 +1007,7 @@ vuint64m1_t test_vredminu_vs_u64m4_u64m1_m(vbool16_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredminu_vs_u64m8_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredminu.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredminu.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredminu_vs_u64m8_u64m1_m(vbool8_t mask, vuint64m1_t dst,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredor.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredor.c
index 4b6c8c1..60c303b 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredor.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredor.c
@@ -491,7 +491,7 @@ vuint64m1_t test_vredor_vs_u64m8_u64m1(vuint64m1_t dst, vuint64m8_t vector,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i8mf8_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredor_vs_i8mf8_i8m1_m(vbool64_t mask, vint8m1_t dst,
@@ -503,7 +503,7 @@ vint8m1_t test_vredor_vs_i8mf8_i8m1_m(vbool64_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i8mf4_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredor_vs_i8mf4_i8m1_m(vbool32_t mask, vint8m1_t dst,
@@ -515,7 +515,7 @@ vint8m1_t test_vredor_vs_i8mf4_i8m1_m(vbool32_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i8mf2_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredor_vs_i8mf2_i8m1_m(vbool16_t mask, vint8m1_t dst,
@@ -527,7 +527,7 @@ vint8m1_t test_vredor_vs_i8mf2_i8m1_m(vbool16_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i8m1_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredor_vs_i8m1_i8m1_m(vbool8_t mask, vint8m1_t dst,
@@ -539,7 +539,7 @@ vint8m1_t test_vredor_vs_i8m1_i8m1_m(vbool8_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i8m2_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredor_vs_i8m2_i8m1_m(vbool4_t mask, vint8m1_t dst,
@@ -551,7 +551,7 @@ vint8m1_t test_vredor_vs_i8m2_i8m1_m(vbool4_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i8m4_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredor_vs_i8m4_i8m1_m(vbool2_t mask, vint8m1_t dst,
@@ -563,7 +563,7 @@ vint8m1_t test_vredor_vs_i8m4_i8m1_m(vbool2_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i8m8_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredor_vs_i8m8_i8m1_m(vbool1_t mask, vint8m1_t dst,
@@ -575,7 +575,7 @@ vint8m1_t test_vredor_vs_i8m8_i8m1_m(vbool1_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i16mf4_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredor_vs_i16mf4_i16m1_m(vbool64_t mask, vint16m1_t dst,
@@ -587,7 +587,7 @@ vint16m1_t test_vredor_vs_i16mf4_i16m1_m(vbool64_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i16mf2_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredor_vs_i16mf2_i16m1_m(vbool32_t mask, vint16m1_t dst,
@@ -599,7 +599,7 @@ vint16m1_t test_vredor_vs_i16mf2_i16m1_m(vbool32_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i16m1_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredor_vs_i16m1_i16m1_m(vbool16_t mask, vint16m1_t dst,
@@ -611,7 +611,7 @@ vint16m1_t test_vredor_vs_i16m1_i16m1_m(vbool16_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i16m2_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredor_vs_i16m2_i16m1_m(vbool8_t mask, vint16m1_t dst,
@@ -623,7 +623,7 @@ vint16m1_t test_vredor_vs_i16m2_i16m1_m(vbool8_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i16m4_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredor_vs_i16m4_i16m1_m(vbool4_t mask, vint16m1_t dst,
@@ -635,7 +635,7 @@ vint16m1_t test_vredor_vs_i16m4_i16m1_m(vbool4_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i16m8_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredor_vs_i16m8_i16m1_m(vbool2_t mask, vint16m1_t dst,
@@ -647,7 +647,7 @@ vint16m1_t test_vredor_vs_i16m8_i16m1_m(vbool2_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i32mf2_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredor.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredor.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredor_vs_i32mf2_i32m1_m(vbool64_t mask, vint32m1_t dst,
@@ -659,7 +659,7 @@ vint32m1_t test_vredor_vs_i32mf2_i32m1_m(vbool64_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i32m1_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredor_vs_i32m1_i32m1_m(vbool32_t mask, vint32m1_t dst,
@@ -671,7 +671,7 @@ vint32m1_t test_vredor_vs_i32m1_i32m1_m(vbool32_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i32m2_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredor.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredor.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredor_vs_i32m2_i32m1_m(vbool16_t mask, vint32m1_t dst,
@@ -683,7 +683,7 @@ vint32m1_t test_vredor_vs_i32m2_i32m1_m(vbool16_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i32m4_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredor.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredor.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredor_vs_i32m4_i32m1_m(vbool8_t mask, vint32m1_t dst,
@@ -695,7 +695,7 @@ vint32m1_t test_vredor_vs_i32m4_i32m1_m(vbool8_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i32m8_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredor.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredor.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredor_vs_i32m8_i32m1_m(vbool4_t mask, vint32m1_t dst,
@@ -707,7 +707,7 @@ vint32m1_t test_vredor_vs_i32m8_i32m1_m(vbool4_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i64m1_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredor_vs_i64m1_i64m1_m(vbool64_t mask, vint64m1_t dst,
@@ -719,7 +719,7 @@ vint64m1_t test_vredor_vs_i64m1_i64m1_m(vbool64_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i64m2_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredor.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredor.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredor_vs_i64m2_i64m1_m(vbool32_t mask, vint64m1_t dst,
@@ -731,7 +731,7 @@ vint64m1_t test_vredor_vs_i64m2_i64m1_m(vbool32_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i64m4_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredor.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredor.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredor_vs_i64m4_i64m1_m(vbool16_t mask, vint64m1_t dst,
@@ -743,7 +743,7 @@ vint64m1_t test_vredor_vs_i64m4_i64m1_m(vbool16_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_i64m8_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredor.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredor.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredor_vs_i64m8_i64m1_m(vbool8_t mask, vint64m1_t dst,
@@ -755,7 +755,7 @@ vint64m1_t test_vredor_vs_i64m8_i64m1_m(vbool8_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u8mf8_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredor_vs_u8mf8_u8m1_m(vbool64_t mask, vuint8m1_t dst,
@@ -767,7 +767,7 @@ vuint8m1_t test_vredor_vs_u8mf8_u8m1_m(vbool64_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u8mf4_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredor_vs_u8mf4_u8m1_m(vbool32_t mask, vuint8m1_t dst,
@@ -779,7 +779,7 @@ vuint8m1_t test_vredor_vs_u8mf4_u8m1_m(vbool32_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u8mf2_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredor_vs_u8mf2_u8m1_m(vbool16_t mask, vuint8m1_t dst,
@@ -791,7 +791,7 @@ vuint8m1_t test_vredor_vs_u8mf2_u8m1_m(vbool16_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u8m1_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredor_vs_u8m1_u8m1_m(vbool8_t mask, vuint8m1_t dst,
@@ -803,7 +803,7 @@ vuint8m1_t test_vredor_vs_u8m1_u8m1_m(vbool8_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u8m2_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredor_vs_u8m2_u8m1_m(vbool4_t mask, vuint8m1_t dst,
@@ -815,7 +815,7 @@ vuint8m1_t test_vredor_vs_u8m2_u8m1_m(vbool4_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u8m4_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredor_vs_u8m4_u8m1_m(vbool2_t mask, vuint8m1_t dst,
@@ -827,7 +827,7 @@ vuint8m1_t test_vredor_vs_u8m4_u8m1_m(vbool2_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u8m8_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredor.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredor_vs_u8m8_u8m1_m(vbool1_t mask, vuint8m1_t dst,
@@ -839,7 +839,7 @@ vuint8m1_t test_vredor_vs_u8m8_u8m1_m(vbool1_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u16mf4_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredor_vs_u16mf4_u16m1_m(vbool64_t mask, vuint16m1_t dst,
@@ -851,7 +851,7 @@ vuint16m1_t test_vredor_vs_u16mf4_u16m1_m(vbool64_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u16mf2_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredor_vs_u16mf2_u16m1_m(vbool32_t mask, vuint16m1_t dst,
@@ -863,7 +863,7 @@ vuint16m1_t test_vredor_vs_u16mf2_u16m1_m(vbool32_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u16m1_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredor_vs_u16m1_u16m1_m(vbool16_t mask, vuint16m1_t dst,
@@ -875,7 +875,7 @@ vuint16m1_t test_vredor_vs_u16m1_u16m1_m(vbool16_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u16m2_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredor_vs_u16m2_u16m1_m(vbool8_t mask, vuint16m1_t dst,
@@ -887,7 +887,7 @@ vuint16m1_t test_vredor_vs_u16m2_u16m1_m(vbool8_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u16m4_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredor_vs_u16m4_u16m1_m(vbool4_t mask, vuint16m1_t dst,
@@ -899,7 +899,7 @@ vuint16m1_t test_vredor_vs_u16m4_u16m1_m(vbool4_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u16m8_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredor.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredor_vs_u16m8_u16m1_m(vbool2_t mask, vuint16m1_t dst,
@@ -911,7 +911,7 @@ vuint16m1_t test_vredor_vs_u16m8_u16m1_m(vbool2_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u32mf2_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredor.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredor.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredor_vs_u32mf2_u32m1_m(vbool64_t mask, vuint32m1_t dst,
@@ -923,7 +923,7 @@ vuint32m1_t test_vredor_vs_u32mf2_u32m1_m(vbool64_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u32m1_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredor_vs_u32m1_u32m1_m(vbool32_t mask, vuint32m1_t dst,
@@ -935,7 +935,7 @@ vuint32m1_t test_vredor_vs_u32m1_u32m1_m(vbool32_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u32m2_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredor.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredor.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredor_vs_u32m2_u32m1_m(vbool16_t mask, vuint32m1_t dst,
@@ -947,7 +947,7 @@ vuint32m1_t test_vredor_vs_u32m2_u32m1_m(vbool16_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u32m4_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredor.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredor.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredor_vs_u32m4_u32m1_m(vbool8_t mask, vuint32m1_t dst,
@@ -959,7 +959,7 @@ vuint32m1_t test_vredor_vs_u32m4_u32m1_m(vbool8_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u32m8_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredor.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredor.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredor_vs_u32m8_u32m1_m(vbool4_t mask, vuint32m1_t dst,
@@ -971,7 +971,7 @@ vuint32m1_t test_vredor_vs_u32m8_u32m1_m(vbool4_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u64m1_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredor_vs_u64m1_u64m1_m(vbool64_t mask, vuint64m1_t dst,
@@ -983,7 +983,7 @@ vuint64m1_t test_vredor_vs_u64m1_u64m1_m(vbool64_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u64m2_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredor.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredor.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredor_vs_u64m2_u64m1_m(vbool32_t mask, vuint64m1_t dst,
@@ -995,7 +995,7 @@ vuint64m1_t test_vredor_vs_u64m2_u64m1_m(vbool32_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u64m4_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredor.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredor.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredor_vs_u64m4_u64m1_m(vbool16_t mask, vuint64m1_t dst,
@@ -1007,7 +1007,7 @@ vuint64m1_t test_vredor_vs_u64m4_u64m1_m(vbool16_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredor_vs_u64m8_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredor.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredor.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredor_vs_u64m8_u64m1_m(vbool8_t mask, vuint64m1_t dst,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredsum.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredsum.c
index 72e3a09..a770eb9 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredsum.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredsum.c
@@ -491,7 +491,7 @@ vuint64m1_t test_vredsum_vs_u64m8_u64m1(vuint64m1_t dst, vuint64m8_t vector,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i8mf8_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredsum_vs_i8mf8_i8m1_m(vbool64_t mask, vint8m1_t dst,
@@ -503,7 +503,7 @@ vint8m1_t test_vredsum_vs_i8mf8_i8m1_m(vbool64_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i8mf4_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredsum_vs_i8mf4_i8m1_m(vbool32_t mask, vint8m1_t dst,
@@ -515,7 +515,7 @@ vint8m1_t test_vredsum_vs_i8mf4_i8m1_m(vbool32_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i8mf2_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredsum_vs_i8mf2_i8m1_m(vbool16_t mask, vint8m1_t dst,
@@ -527,7 +527,7 @@ vint8m1_t test_vredsum_vs_i8mf2_i8m1_m(vbool16_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i8m1_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredsum_vs_i8m1_i8m1_m(vbool8_t mask, vint8m1_t dst,
@@ -539,7 +539,7 @@ vint8m1_t test_vredsum_vs_i8m1_i8m1_m(vbool8_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i8m2_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredsum_vs_i8m2_i8m1_m(vbool4_t mask, vint8m1_t dst,
@@ -551,7 +551,7 @@ vint8m1_t test_vredsum_vs_i8m2_i8m1_m(vbool4_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i8m4_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredsum_vs_i8m4_i8m1_m(vbool2_t mask, vint8m1_t dst,
@@ -563,7 +563,7 @@ vint8m1_t test_vredsum_vs_i8m4_i8m1_m(vbool2_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i8m8_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredsum_vs_i8m8_i8m1_m(vbool1_t mask, vint8m1_t dst,
@@ -575,7 +575,7 @@ vint8m1_t test_vredsum_vs_i8m8_i8m1_m(vbool1_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i16mf4_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredsum_vs_i16mf4_i16m1_m(vbool64_t mask, vint16m1_t dst,
@@ -587,7 +587,7 @@ vint16m1_t test_vredsum_vs_i16mf4_i16m1_m(vbool64_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i16mf2_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredsum_vs_i16mf2_i16m1_m(vbool32_t mask, vint16m1_t dst,
@@ -599,7 +599,7 @@ vint16m1_t test_vredsum_vs_i16mf2_i16m1_m(vbool32_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i16m1_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredsum_vs_i16m1_i16m1_m(vbool16_t mask, vint16m1_t dst,
@@ -611,7 +611,7 @@ vint16m1_t test_vredsum_vs_i16m1_i16m1_m(vbool16_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i16m2_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredsum_vs_i16m2_i16m1_m(vbool8_t mask, vint16m1_t dst,
@@ -623,7 +623,7 @@ vint16m1_t test_vredsum_vs_i16m2_i16m1_m(vbool8_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i16m4_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredsum_vs_i16m4_i16m1_m(vbool4_t mask, vint16m1_t dst,
@@ -635,7 +635,7 @@ vint16m1_t test_vredsum_vs_i16m4_i16m1_m(vbool4_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i16m8_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredsum_vs_i16m8_i16m1_m(vbool2_t mask, vint16m1_t dst,
@@ -647,7 +647,7 @@ vint16m1_t test_vredsum_vs_i16m8_i16m1_m(vbool2_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i32mf2_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredsum.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredsum.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredsum_vs_i32mf2_i32m1_m(vbool64_t mask, vint32m1_t dst,
@@ -659,7 +659,7 @@ vint32m1_t test_vredsum_vs_i32mf2_i32m1_m(vbool64_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i32m1_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredsum.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredsum.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredsum_vs_i32m1_i32m1_m(vbool32_t mask, vint32m1_t dst,
@@ -671,7 +671,7 @@ vint32m1_t test_vredsum_vs_i32m1_i32m1_m(vbool32_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i32m2_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredsum.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredsum.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredsum_vs_i32m2_i32m1_m(vbool16_t mask, vint32m1_t dst,
@@ -683,7 +683,7 @@ vint32m1_t test_vredsum_vs_i32m2_i32m1_m(vbool16_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i32m4_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredsum.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredsum.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredsum_vs_i32m4_i32m1_m(vbool8_t mask, vint32m1_t dst,
@@ -695,7 +695,7 @@ vint32m1_t test_vredsum_vs_i32m4_i32m1_m(vbool8_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i32m8_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredsum.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredsum.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredsum_vs_i32m8_i32m1_m(vbool4_t mask, vint32m1_t dst,
@@ -707,7 +707,7 @@ vint32m1_t test_vredsum_vs_i32m8_i32m1_m(vbool4_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i64m1_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredsum.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredsum.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredsum_vs_i64m1_i64m1_m(vbool64_t mask, vint64m1_t dst,
@@ -719,7 +719,7 @@ vint64m1_t test_vredsum_vs_i64m1_i64m1_m(vbool64_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i64m2_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredsum.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredsum.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredsum_vs_i64m2_i64m1_m(vbool32_t mask, vint64m1_t dst,
@@ -731,7 +731,7 @@ vint64m1_t test_vredsum_vs_i64m2_i64m1_m(vbool32_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i64m4_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredsum.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredsum.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredsum_vs_i64m4_i64m1_m(vbool16_t mask, vint64m1_t dst,
@@ -743,7 +743,7 @@ vint64m1_t test_vredsum_vs_i64m4_i64m1_m(vbool16_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_i64m8_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredsum.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredsum.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredsum_vs_i64m8_i64m1_m(vbool8_t mask, vint64m1_t dst,
@@ -755,7 +755,7 @@ vint64m1_t test_vredsum_vs_i64m8_i64m1_m(vbool8_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u8mf8_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredsum_vs_u8mf8_u8m1_m(vbool64_t mask, vuint8m1_t dst,
@@ -767,7 +767,7 @@ vuint8m1_t test_vredsum_vs_u8mf8_u8m1_m(vbool64_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u8mf4_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredsum_vs_u8mf4_u8m1_m(vbool32_t mask, vuint8m1_t dst,
@@ -779,7 +779,7 @@ vuint8m1_t test_vredsum_vs_u8mf4_u8m1_m(vbool32_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u8mf2_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredsum_vs_u8mf2_u8m1_m(vbool16_t mask, vuint8m1_t dst,
@@ -791,7 +791,7 @@ vuint8m1_t test_vredsum_vs_u8mf2_u8m1_m(vbool16_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u8m1_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredsum_vs_u8m1_u8m1_m(vbool8_t mask, vuint8m1_t dst,
@@ -803,7 +803,7 @@ vuint8m1_t test_vredsum_vs_u8m1_u8m1_m(vbool8_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u8m2_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredsum_vs_u8m2_u8m1_m(vbool4_t mask, vuint8m1_t dst,
@@ -815,7 +815,7 @@ vuint8m1_t test_vredsum_vs_u8m2_u8m1_m(vbool4_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u8m4_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredsum_vs_u8m4_u8m1_m(vbool2_t mask, vuint8m1_t dst,
@@ -827,7 +827,7 @@ vuint8m1_t test_vredsum_vs_u8m4_u8m1_m(vbool2_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u8m8_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredsum.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredsum_vs_u8m8_u8m1_m(vbool1_t mask, vuint8m1_t dst,
@@ -839,7 +839,7 @@ vuint8m1_t test_vredsum_vs_u8m8_u8m1_m(vbool1_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u16mf4_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredsum_vs_u16mf4_u16m1_m(vbool64_t mask, vuint16m1_t dst,
@@ -851,7 +851,7 @@ vuint16m1_t test_vredsum_vs_u16mf4_u16m1_m(vbool64_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u16mf2_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredsum_vs_u16mf2_u16m1_m(vbool32_t mask, vuint16m1_t dst,
@@ -863,7 +863,7 @@ vuint16m1_t test_vredsum_vs_u16mf2_u16m1_m(vbool32_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u16m1_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredsum_vs_u16m1_u16m1_m(vbool16_t mask, vuint16m1_t dst,
@@ -875,7 +875,7 @@ vuint16m1_t test_vredsum_vs_u16m1_u16m1_m(vbool16_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u16m2_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredsum_vs_u16m2_u16m1_m(vbool8_t mask, vuint16m1_t dst,
@@ -887,7 +887,7 @@ vuint16m1_t test_vredsum_vs_u16m2_u16m1_m(vbool8_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u16m4_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredsum_vs_u16m4_u16m1_m(vbool4_t mask, vuint16m1_t dst,
@@ -899,7 +899,7 @@ vuint16m1_t test_vredsum_vs_u16m4_u16m1_m(vbool4_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u16m8_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredsum.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredsum_vs_u16m8_u16m1_m(vbool2_t mask, vuint16m1_t dst,
@@ -911,7 +911,7 @@ vuint16m1_t test_vredsum_vs_u16m8_u16m1_m(vbool2_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u32mf2_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredsum.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredsum.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredsum_vs_u32mf2_u32m1_m(vbool64_t mask, vuint32m1_t dst,
@@ -923,7 +923,7 @@ vuint32m1_t test_vredsum_vs_u32mf2_u32m1_m(vbool64_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u32m1_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredsum.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredsum.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredsum_vs_u32m1_u32m1_m(vbool32_t mask, vuint32m1_t dst,
@@ -935,7 +935,7 @@ vuint32m1_t test_vredsum_vs_u32m1_u32m1_m(vbool32_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u32m2_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredsum.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredsum.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredsum_vs_u32m2_u32m1_m(vbool16_t mask, vuint32m1_t dst,
@@ -947,7 +947,7 @@ vuint32m1_t test_vredsum_vs_u32m2_u32m1_m(vbool16_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u32m4_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredsum.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredsum.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredsum_vs_u32m4_u32m1_m(vbool8_t mask, vuint32m1_t dst,
@@ -959,7 +959,7 @@ vuint32m1_t test_vredsum_vs_u32m4_u32m1_m(vbool8_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u32m8_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredsum.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredsum.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredsum_vs_u32m8_u32m1_m(vbool4_t mask, vuint32m1_t dst,
@@ -971,7 +971,7 @@ vuint32m1_t test_vredsum_vs_u32m8_u32m1_m(vbool4_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u64m1_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredsum.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredsum.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredsum_vs_u64m1_u64m1_m(vbool64_t mask, vuint64m1_t dst,
@@ -983,7 +983,7 @@ vuint64m1_t test_vredsum_vs_u64m1_u64m1_m(vbool64_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u64m2_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredsum.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredsum.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredsum_vs_u64m2_u64m1_m(vbool32_t mask, vuint64m1_t dst,
@@ -995,7 +995,7 @@ vuint64m1_t test_vredsum_vs_u64m2_u64m1_m(vbool32_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u64m4_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredsum.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredsum.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredsum_vs_u64m4_u64m1_m(vbool16_t mask, vuint64m1_t dst,
@@ -1007,7 +1007,7 @@ vuint64m1_t test_vredsum_vs_u64m4_u64m1_m(vbool16_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredsum_vs_u64m8_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredsum.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredsum.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredsum_vs_u64m8_u64m1_m(vbool8_t mask, vuint64m1_t dst,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredxor.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredxor.c
index 1773411..9ad3e71 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredxor.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vredxor.c
@@ -491,7 +491,7 @@ vuint64m1_t test_vredxor_vs_u64m8_u64m1(vuint64m1_t dst, vuint64m8_t vector,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i8mf8_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredxor_vs_i8mf8_i8m1_m(vbool64_t mask, vint8m1_t dst,
@@ -503,7 +503,7 @@ vint8m1_t test_vredxor_vs_i8mf8_i8m1_m(vbool64_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i8mf4_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredxor_vs_i8mf4_i8m1_m(vbool32_t mask, vint8m1_t dst,
@@ -515,7 +515,7 @@ vint8m1_t test_vredxor_vs_i8mf4_i8m1_m(vbool32_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i8mf2_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredxor_vs_i8mf2_i8m1_m(vbool16_t mask, vint8m1_t dst,
@@ -527,7 +527,7 @@ vint8m1_t test_vredxor_vs_i8mf2_i8m1_m(vbool16_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i8m1_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredxor_vs_i8m1_i8m1_m(vbool8_t mask, vint8m1_t dst,
@@ -539,7 +539,7 @@ vint8m1_t test_vredxor_vs_i8m1_i8m1_m(vbool8_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i8m2_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredxor_vs_i8m2_i8m1_m(vbool4_t mask, vint8m1_t dst,
@@ -551,7 +551,7 @@ vint8m1_t test_vredxor_vs_i8m2_i8m1_m(vbool4_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i8m4_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredxor_vs_i8m4_i8m1_m(vbool2_t mask, vint8m1_t dst,
@@ -563,7 +563,7 @@ vint8m1_t test_vredxor_vs_i8m4_i8m1_m(vbool2_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i8m8_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vredxor_vs_i8m8_i8m1_m(vbool1_t mask, vint8m1_t dst,
@@ -575,7 +575,7 @@ vint8m1_t test_vredxor_vs_i8m8_i8m1_m(vbool1_t mask, vint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i16mf4_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredxor_vs_i16mf4_i16m1_m(vbool64_t mask, vint16m1_t dst,
@@ -587,7 +587,7 @@ vint16m1_t test_vredxor_vs_i16mf4_i16m1_m(vbool64_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i16mf2_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredxor_vs_i16mf2_i16m1_m(vbool32_t mask, vint16m1_t dst,
@@ -599,7 +599,7 @@ vint16m1_t test_vredxor_vs_i16mf2_i16m1_m(vbool32_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i16m1_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredxor_vs_i16m1_i16m1_m(vbool16_t mask, vint16m1_t dst,
@@ -611,7 +611,7 @@ vint16m1_t test_vredxor_vs_i16m1_i16m1_m(vbool16_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i16m2_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredxor_vs_i16m2_i16m1_m(vbool8_t mask, vint16m1_t dst,
@@ -623,7 +623,7 @@ vint16m1_t test_vredxor_vs_i16m2_i16m1_m(vbool8_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i16m4_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredxor_vs_i16m4_i16m1_m(vbool4_t mask, vint16m1_t dst,
@@ -635,7 +635,7 @@ vint16m1_t test_vredxor_vs_i16m4_i16m1_m(vbool4_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i16m8_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vredxor_vs_i16m8_i16m1_m(vbool2_t mask, vint16m1_t dst,
@@ -647,7 +647,7 @@ vint16m1_t test_vredxor_vs_i16m8_i16m1_m(vbool2_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i32mf2_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredxor.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredxor.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredxor_vs_i32mf2_i32m1_m(vbool64_t mask, vint32m1_t dst,
@@ -659,7 +659,7 @@ vint32m1_t test_vredxor_vs_i32mf2_i32m1_m(vbool64_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i32m1_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredxor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredxor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredxor_vs_i32m1_i32m1_m(vbool32_t mask, vint32m1_t dst,
@@ -671,7 +671,7 @@ vint32m1_t test_vredxor_vs_i32m1_i32m1_m(vbool32_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i32m2_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredxor.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredxor.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredxor_vs_i32m2_i32m1_m(vbool16_t mask, vint32m1_t dst,
@@ -683,7 +683,7 @@ vint32m1_t test_vredxor_vs_i32m2_i32m1_m(vbool16_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i32m4_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredxor.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredxor.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredxor_vs_i32m4_i32m1_m(vbool8_t mask, vint32m1_t dst,
@@ -695,7 +695,7 @@ vint32m1_t test_vredxor_vs_i32m4_i32m1_m(vbool8_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i32m8_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredxor.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredxor.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vredxor_vs_i32m8_i32m1_m(vbool4_t mask, vint32m1_t dst,
@@ -707,7 +707,7 @@ vint32m1_t test_vredxor_vs_i32m8_i32m1_m(vbool4_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i64m1_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredxor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredxor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredxor_vs_i64m1_i64m1_m(vbool64_t mask, vint64m1_t dst,
@@ -719,7 +719,7 @@ vint64m1_t test_vredxor_vs_i64m1_i64m1_m(vbool64_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i64m2_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredxor.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredxor.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredxor_vs_i64m2_i64m1_m(vbool32_t mask, vint64m1_t dst,
@@ -731,7 +731,7 @@ vint64m1_t test_vredxor_vs_i64m2_i64m1_m(vbool32_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i64m4_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredxor.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredxor.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredxor_vs_i64m4_i64m1_m(vbool16_t mask, vint64m1_t dst,
@@ -743,7 +743,7 @@ vint64m1_t test_vredxor_vs_i64m4_i64m1_m(vbool16_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_i64m8_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredxor.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredxor.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vredxor_vs_i64m8_i64m1_m(vbool8_t mask, vint64m1_t dst,
@@ -755,7 +755,7 @@ vint64m1_t test_vredxor_vs_i64m8_i64m1_m(vbool8_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u8mf8_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv1i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredxor_vs_u8mf8_u8m1_m(vbool64_t mask, vuint8m1_t dst,
@@ -767,7 +767,7 @@ vuint8m1_t test_vredxor_vs_u8mf8_u8m1_m(vbool64_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u8mf4_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv2i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredxor_vs_u8mf4_u8m1_m(vbool32_t mask, vuint8m1_t dst,
@@ -779,7 +779,7 @@ vuint8m1_t test_vredxor_vs_u8mf4_u8m1_m(vbool32_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u8mf2_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv4i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredxor_vs_u8mf2_u8m1_m(vbool16_t mask, vuint8m1_t dst,
@@ -791,7 +791,7 @@ vuint8m1_t test_vredxor_vs_u8mf2_u8m1_m(vbool16_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u8m1_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredxor_vs_u8m1_u8m1_m(vbool8_t mask, vuint8m1_t dst,
@@ -803,7 +803,7 @@ vuint8m1_t test_vredxor_vs_u8m1_u8m1_m(vbool8_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u8m2_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv16i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredxor_vs_u8m2_u8m1_m(vbool4_t mask, vuint8m1_t dst,
@@ -815,7 +815,7 @@ vuint8m1_t test_vredxor_vs_u8m2_u8m1_m(vbool4_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u8m4_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv32i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredxor_vs_u8m4_u8m1_m(vbool2_t mask, vuint8m1_t dst,
@@ -827,7 +827,7 @@ vuint8m1_t test_vredxor_vs_u8m4_u8m1_m(vbool2_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u8m8_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vredxor.mask.nxv8i8.nxv64i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 8 x i8> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vredxor_vs_u8m8_u8m1_m(vbool1_t mask, vuint8m1_t dst,
@@ -839,7 +839,7 @@ vuint8m1_t test_vredxor_vs_u8m8_u8m1_m(vbool1_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u16mf4_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv1i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredxor_vs_u16mf4_u16m1_m(vbool64_t mask, vuint16m1_t dst,
@@ -851,7 +851,7 @@ vuint16m1_t test_vredxor_vs_u16mf4_u16m1_m(vbool64_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u16mf2_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv2i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredxor_vs_u16mf2_u16m1_m(vbool32_t mask, vuint16m1_t dst,
@@ -863,7 +863,7 @@ vuint16m1_t test_vredxor_vs_u16mf2_u16m1_m(vbool32_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u16m1_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredxor_vs_u16m1_u16m1_m(vbool16_t mask, vuint16m1_t dst,
@@ -875,7 +875,7 @@ vuint16m1_t test_vredxor_vs_u16m1_u16m1_m(vbool16_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u16m2_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv8i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredxor_vs_u16m2_u16m1_m(vbool8_t mask, vuint16m1_t dst,
@@ -887,7 +887,7 @@ vuint16m1_t test_vredxor_vs_u16m2_u16m1_m(vbool8_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u16m4_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv16i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredxor_vs_u16m4_u16m1_m(vbool4_t mask, vuint16m1_t dst,
@@ -899,7 +899,7 @@ vuint16m1_t test_vredxor_vs_u16m4_u16m1_m(vbool4_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u16m8_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vredxor.mask.nxv4i16.nxv32i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vredxor_vs_u16m8_u16m1_m(vbool2_t mask, vuint16m1_t dst,
@@ -911,7 +911,7 @@ vuint16m1_t test_vredxor_vs_u16m8_u16m1_m(vbool2_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u32mf2_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredxor.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredxor.mask.nxv2i32.nxv1i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredxor_vs_u32mf2_u32m1_m(vbool64_t mask, vuint32m1_t dst,
@@ -923,7 +923,7 @@ vuint32m1_t test_vredxor_vs_u32mf2_u32m1_m(vbool64_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u32m1_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredxor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredxor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredxor_vs_u32m1_u32m1_m(vbool32_t mask, vuint32m1_t dst,
@@ -935,7 +935,7 @@ vuint32m1_t test_vredxor_vs_u32m1_u32m1_m(vbool32_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u32m2_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredxor.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredxor.mask.nxv2i32.nxv4i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredxor_vs_u32m2_u32m1_m(vbool16_t mask, vuint32m1_t dst,
@@ -947,7 +947,7 @@ vuint32m1_t test_vredxor_vs_u32m2_u32m1_m(vbool16_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u32m4_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredxor.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredxor.mask.nxv2i32.nxv8i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredxor_vs_u32m4_u32m1_m(vbool8_t mask, vuint32m1_t dst,
@@ -959,7 +959,7 @@ vuint32m1_t test_vredxor_vs_u32m4_u32m1_m(vbool8_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u32m8_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredxor.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vredxor.mask.nxv2i32.nxv16i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vredxor_vs_u32m8_u32m1_m(vbool4_t mask, vuint32m1_t dst,
@@ -971,7 +971,7 @@ vuint32m1_t test_vredxor_vs_u32m8_u32m1_m(vbool4_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u64m1_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredxor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredxor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredxor_vs_u64m1_u64m1_m(vbool64_t mask, vuint64m1_t dst,
@@ -983,7 +983,7 @@ vuint64m1_t test_vredxor_vs_u64m1_u64m1_m(vbool64_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u64m2_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredxor.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredxor.mask.nxv1i64.nxv2i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredxor_vs_u64m2_u64m1_m(vbool32_t mask, vuint64m1_t dst,
@@ -995,7 +995,7 @@ vuint64m1_t test_vredxor_vs_u64m2_u64m1_m(vbool32_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u64m4_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredxor.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredxor.mask.nxv1i64.nxv4i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredxor_vs_u64m4_u64m1_m(vbool16_t mask, vuint64m1_t dst,
@@ -1007,7 +1007,7 @@ vuint64m1_t test_vredxor_vs_u64m4_u64m1_m(vbool16_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vredxor_vs_u64m8_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredxor.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vredxor.mask.nxv1i64.nxv8i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i64> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vredxor_vs_u64m8_u64m1_m(vbool8_t mask, vuint64m1_t dst,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vrem.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vrem.c
index 2571b6b..d791364 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vrem.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vrem.c
@@ -887,7 +887,7 @@ vuint64m8_t test_vremu_vx_u64m8(vuint64m8_t op1, uint64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrem.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrem.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vrem_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, vint8mf8_t op2, size_t vl) {
@@ -897,7 +897,7 @@ vint8mf8_t test_vrem_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrem.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrem.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vrem_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, int8_t op2, size_t vl) {
@@ -907,7 +907,7 @@ vint8mf8_t test_vrem_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrem.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrem.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vrem_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, vint8mf4_t op2, size_t vl) {
@@ -917,7 +917,7 @@ vint8mf4_t test_vrem_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrem.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrem.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vrem_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, int8_t op2, size_t vl) {
@@ -927,7 +927,7 @@ vint8mf4_t test_vrem_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrem.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrem.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vrem_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, vint8mf2_t op2, size_t vl) {
@@ -937,7 +937,7 @@ vint8mf2_t test_vrem_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrem.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrem.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vrem_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, int8_t op2, size_t vl) {
@@ -947,7 +947,7 @@ vint8mf2_t test_vrem_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrem.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrem.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vrem_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, vint8m1_t op2, size_t vl) {
@@ -957,7 +957,7 @@ vint8m1_t test_vrem_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrem.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrem.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vrem_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, int8_t op2, size_t vl) {
@@ -967,7 +967,7 @@ vint8m1_t test_vrem_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrem.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrem.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vrem_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, vint8m2_t op2, size_t vl) {
@@ -977,7 +977,7 @@ vint8m2_t test_vrem_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrem.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrem.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vrem_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, int8_t op2, size_t vl) {
@@ -987,7 +987,7 @@ vint8m2_t test_vrem_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrem.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrem.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vrem_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, vint8m4_t op2, size_t vl) {
@@ -997,7 +997,7 @@ vint8m4_t test_vrem_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrem.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrem.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vrem_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, int8_t op2, size_t vl) {
@@ -1007,7 +1007,7 @@ vint8m4_t test_vrem_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vrem.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vrem.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vrem_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, vint8m8_t op2, size_t vl) {
@@ -1017,7 +1017,7 @@ vint8m8_t test_vrem_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vrem.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vrem.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vrem_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, int8_t op2, size_t vl) {
@@ -1027,7 +1027,7 @@ vint8m8_t test_vrem_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrem.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrem.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vrem_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, vint16mf4_t op2, size_t vl) {
@@ -1037,7 +1037,7 @@ vint16mf4_t test_vrem_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrem.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrem.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vrem_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, int16_t op2, size_t vl) {
@@ -1047,7 +1047,7 @@ vint16mf4_t test_vrem_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrem.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrem.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vrem_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, vint16mf2_t op2, size_t vl) {
@@ -1057,7 +1057,7 @@ vint16mf2_t test_vrem_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrem.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrem.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vrem_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, int16_t op2, size_t vl) {
@@ -1067,7 +1067,7 @@ vint16mf2_t test_vrem_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrem.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrem.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vrem_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, vint16m1_t op2, size_t vl) {
@@ -1077,7 +1077,7 @@ vint16m1_t test_vrem_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrem.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrem.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vrem_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, int16_t op2, size_t vl) {
@@ -1087,7 +1087,7 @@ vint16m1_t test_vrem_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrem.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrem.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vrem_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, vint16m2_t op2, size_t vl) {
@@ -1097,7 +1097,7 @@ vint16m2_t test_vrem_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrem.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrem.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vrem_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, int16_t op2, size_t vl) {
@@ -1107,7 +1107,7 @@ vint16m2_t test_vrem_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrem.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrem.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vrem_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, vint16m4_t op2, size_t vl) {
@@ -1117,7 +1117,7 @@ vint16m4_t test_vrem_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrem.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrem.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vrem_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, int16_t op2, size_t vl) {
@@ -1127,7 +1127,7 @@ vint16m4_t test_vrem_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrem.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrem.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vrem_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, vint16m8_t op2, size_t vl) {
@@ -1137,7 +1137,7 @@ vint16m8_t test_vrem_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrem.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrem.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vrem_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, int16_t op2, size_t vl) {
@@ -1147,7 +1147,7 @@ vint16m8_t test_vrem_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrem.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrem.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vrem_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, vint32mf2_t op2, size_t vl) {
@@ -1157,7 +1157,7 @@ vint32mf2_t test_vrem_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrem.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrem.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vrem_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, int32_t op2, size_t vl) {
@@ -1167,7 +1167,7 @@ vint32mf2_t test_vrem_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrem.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrem.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vrem_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, vint32m1_t op2, size_t vl) {
@@ -1177,7 +1177,7 @@ vint32m1_t test_vrem_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrem.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrem.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vrem_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, int32_t op2, size_t vl) {
@@ -1187,7 +1187,7 @@ vint32m1_t test_vrem_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrem.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrem.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vrem_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, vint32m2_t op2, size_t vl) {
@@ -1197,7 +1197,7 @@ vint32m2_t test_vrem_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrem.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrem.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vrem_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, int32_t op2, size_t vl) {
@@ -1207,7 +1207,7 @@ vint32m2_t test_vrem_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrem.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrem.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vrem_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, vint32m4_t op2, size_t vl) {
@@ -1217,7 +1217,7 @@ vint32m4_t test_vrem_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrem.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrem.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vrem_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, int32_t op2, size_t vl) {
@@ -1227,7 +1227,7 @@ vint32m4_t test_vrem_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrem.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrem.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vrem_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, vint32m8_t op2, size_t vl) {
@@ -1237,7 +1237,7 @@ vint32m8_t test_vrem_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrem.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrem.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vrem_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, int32_t op2, size_t vl) {
@@ -1247,7 +1247,7 @@ vint32m8_t test_vrem_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrem.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrem.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vrem_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, vint64m1_t op2, size_t vl) {
@@ -1257,7 +1257,7 @@ vint64m1_t test_vrem_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrem.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrem.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vrem_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, int64_t op2, size_t vl) {
@@ -1267,7 +1267,7 @@ vint64m1_t test_vrem_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrem.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrem.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vrem_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, vint64m2_t op2, size_t vl) {
@@ -1277,7 +1277,7 @@ vint64m2_t test_vrem_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrem.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrem.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vrem_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, int64_t op2, size_t vl) {
@@ -1287,7 +1287,7 @@ vint64m2_t test_vrem_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrem.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrem.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vrem_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, vint64m4_t op2, size_t vl) {
@@ -1297,7 +1297,7 @@ vint64m4_t test_vrem_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrem.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrem.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vrem_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, int64_t op2, size_t vl) {
@@ -1307,7 +1307,7 @@ vint64m4_t test_vrem_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrem.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrem.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vrem_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, vint64m8_t op2, size_t vl) {
@@ -1317,7 +1317,7 @@ vint64m8_t test_vrem_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vrem_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrem.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrem.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vrem_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, int64_t op2, size_t vl) {
@@ -1327,7 +1327,7 @@ vint64m8_t test_vrem_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vremu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vremu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vremu_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -1337,7 +1337,7 @@ vuint8mf8_t test_vremu_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vremu.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vremu.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vremu_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, uint8_t op2, size_t vl) {
@@ -1347,7 +1347,7 @@ vuint8mf8_t test_vremu_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vremu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vremu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vremu_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, vuint8mf4_t op2, size_t vl) {
@@ -1357,7 +1357,7 @@ vuint8mf4_t test_vremu_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vremu.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vremu.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vremu_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, uint8_t op2, size_t vl) {
@@ -1367,7 +1367,7 @@ vuint8mf4_t test_vremu_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vremu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vremu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vremu_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, vuint8mf2_t op2, size_t vl) {
@@ -1377,7 +1377,7 @@ vuint8mf2_t test_vremu_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vremu.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vremu.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vremu_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, uint8_t op2, size_t vl) {
@@ -1387,7 +1387,7 @@ vuint8mf2_t test_vremu_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vremu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vremu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vremu_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, vuint8m1_t op2, size_t vl) {
@@ -1397,7 +1397,7 @@ vuint8m1_t test_vremu_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vremu.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vremu.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vremu_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, uint8_t op2, size_t vl) {
@@ -1407,7 +1407,7 @@ vuint8m1_t test_vremu_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vremu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vremu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vremu_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, vuint8m2_t op2, size_t vl) {
@@ -1417,7 +1417,7 @@ vuint8m2_t test_vremu_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vremu.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vremu.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vremu_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, uint8_t op2, size_t vl) {
@@ -1427,7 +1427,7 @@ vuint8m2_t test_vremu_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vremu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vremu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vremu_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, vuint8m4_t op2, size_t vl) {
@@ -1437,7 +1437,7 @@ vuint8m4_t test_vremu_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vremu.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vremu.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vremu_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, uint8_t op2, size_t vl) {
@@ -1447,7 +1447,7 @@ vuint8m4_t test_vremu_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vremu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vremu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vremu_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, vuint8m8_t op2, size_t vl) {
@@ -1457,7 +1457,7 @@ vuint8m8_t test_vremu_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vremu.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vremu.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vremu_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, uint8_t op2, size_t vl) {
@@ -1467,7 +1467,7 @@ vuint8m8_t test_vremu_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vremu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vremu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vremu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, vuint16mf4_t op2, size_t vl) {
@@ -1477,7 +1477,7 @@ vuint16mf4_t test_vremu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vremu.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vremu.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vremu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, uint16_t op2, size_t vl) {
@@ -1487,7 +1487,7 @@ vuint16mf4_t test_vremu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vremu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vremu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vremu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, vuint16mf2_t op2, size_t vl) {
@@ -1497,7 +1497,7 @@ vuint16mf2_t test_vremu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vremu.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vremu.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vremu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, uint16_t op2, size_t vl) {
@@ -1507,7 +1507,7 @@ vuint16mf2_t test_vremu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vremu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vremu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vremu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, vuint16m1_t op2, size_t vl) {
@@ -1517,7 +1517,7 @@ vuint16m1_t test_vremu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vremu.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vremu.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vremu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, uint16_t op2, size_t vl) {
@@ -1527,7 +1527,7 @@ vuint16m1_t test_vremu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vremu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vremu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vremu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, vuint16m2_t op2, size_t vl) {
@@ -1537,7 +1537,7 @@ vuint16m2_t test_vremu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vremu.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vremu.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vremu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, uint16_t op2, size_t vl) {
@@ -1547,7 +1547,7 @@ vuint16m2_t test_vremu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vremu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vremu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vremu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, vuint16m4_t op2, size_t vl) {
@@ -1557,7 +1557,7 @@ vuint16m4_t test_vremu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vremu.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vremu.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vremu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, uint16_t op2, size_t vl) {
@@ -1567,7 +1567,7 @@ vuint16m4_t test_vremu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vremu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vremu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vremu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, vuint16m8_t op2, size_t vl) {
@@ -1577,7 +1577,7 @@ vuint16m8_t test_vremu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vremu.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vremu.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vremu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, uint16_t op2, size_t vl) {
@@ -1587,7 +1587,7 @@ vuint16m8_t test_vremu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vremu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vremu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vremu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, vuint32mf2_t op2, size_t vl) {
@@ -1597,7 +1597,7 @@ vuint32mf2_t test_vremu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vremu.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vremu.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vremu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, uint32_t op2, size_t vl) {
@@ -1607,7 +1607,7 @@ vuint32mf2_t test_vremu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vremu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vremu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vremu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, vuint32m1_t op2, size_t vl) {
@@ -1617,7 +1617,7 @@ vuint32m1_t test_vremu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vremu.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vremu.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vremu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, uint32_t op2, size_t vl) {
@@ -1627,7 +1627,7 @@ vuint32m1_t test_vremu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vremu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vremu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vremu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, vuint32m2_t op2, size_t vl) {
@@ -1637,7 +1637,7 @@ vuint32m2_t test_vremu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vremu.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vremu.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vremu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, uint32_t op2, size_t vl) {
@@ -1647,7 +1647,7 @@ vuint32m2_t test_vremu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vremu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vremu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vremu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, vuint32m4_t op2, size_t vl) {
@@ -1657,7 +1657,7 @@ vuint32m4_t test_vremu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vremu.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vremu.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vremu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, uint32_t op2, size_t vl) {
@@ -1667,7 +1667,7 @@ vuint32m4_t test_vremu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vremu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vremu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vremu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, vuint32m8_t op2, size_t vl) {
@@ -1677,7 +1677,7 @@ vuint32m8_t test_vremu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vremu.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vremu.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vremu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, uint32_t op2, size_t vl) {
@@ -1687,7 +1687,7 @@ vuint32m8_t test_vremu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vremu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vremu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vremu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, vuint64m1_t op2, size_t vl) {
@@ -1697,7 +1697,7 @@ vuint64m1_t test_vremu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vremu.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vremu.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vremu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, uint64_t op2, size_t vl) {
@@ -1707,7 +1707,7 @@ vuint64m1_t test_vremu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vremu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vremu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vremu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, vuint64m2_t op2, size_t vl) {
@@ -1717,7 +1717,7 @@ vuint64m2_t test_vremu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vremu.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vremu.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vremu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, uint64_t op2, size_t vl) {
@@ -1727,7 +1727,7 @@ vuint64m2_t test_vremu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vremu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vremu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vremu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, vuint64m4_t op2, size_t vl) {
@@ -1737,7 +1737,7 @@ vuint64m4_t test_vremu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vremu.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vremu.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vremu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, uint64_t op2, size_t vl) {
@@ -1747,7 +1747,7 @@ vuint64m4_t test_vremu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vremu_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vremu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vremu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vremu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, vuint64m8_t op2, size_t vl) {
@@ -1757,7 +1757,7 @@ vuint64m8_t test_vremu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vremu_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vremu.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vremu.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vremu_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, uint64_t op2, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vrgather.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vrgather.c
index d536156..04bfd70 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vrgather.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vrgather.c
@@ -1675,7 +1675,7 @@ vfloat64m8_t test_vrgatherei16_vv_f64m8(vfloat64m8_t op1, vuint16m2_t op2,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrgather.vv.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrgather.vv.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vrgather_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
@@ -1687,7 +1687,7 @@ vint8mf8_t test_vrgather_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrgather.vx.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrgather.vx.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vrgather_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
@@ -1698,7 +1698,7 @@ vint8mf8_t test_vrgather_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrgather.vv.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrgather.vv.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vrgather_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
@@ -1710,7 +1710,7 @@ vint8mf4_t test_vrgather_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrgather.vx.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrgather.vx.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vrgather_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
@@ -1721,7 +1721,7 @@ vint8mf4_t test_vrgather_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrgather.vv.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrgather.vv.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vrgather_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
@@ -1733,7 +1733,7 @@ vint8mf2_t test_vrgather_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrgather.vx.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrgather.vx.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vrgather_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
@@ -1744,7 +1744,7 @@ vint8mf2_t test_vrgather_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrgather.vv.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrgather.vv.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vrgather_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
@@ -1755,7 +1755,7 @@ vint8m1_t test_vrgather_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrgather.vx.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrgather.vx.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vrgather_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
@@ -1766,7 +1766,7 @@ vint8m1_t test_vrgather_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrgather.vv.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrgather.vv.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vrgather_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
@@ -1777,7 +1777,7 @@ vint8m2_t test_vrgather_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrgather.vx.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrgather.vx.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vrgather_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
@@ -1788,7 +1788,7 @@ vint8m2_t test_vrgather_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrgather.vv.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[INDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrgather.vv.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[INDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vrgather_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
@@ -1799,7 +1799,7 @@ vint8m4_t test_vrgather_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrgather.vx.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrgather.vx.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vrgather_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
@@ -1810,7 +1810,7 @@ vint8m4_t test_vrgather_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vrgather.vv.mask.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[INDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vrgather.vv.mask.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[INDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vrgather_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
@@ -1821,7 +1821,7 @@ vint8m8_t test_vrgather_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vrgather.vx.mask.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vrgather.vx.mask.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vrgather_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
@@ -1832,7 +1832,7 @@ vint8m8_t test_vrgather_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrgather.vv.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrgather.vv.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vrgather_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -1844,7 +1844,7 @@ vint16mf4_t test_vrgather_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrgather.vx.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrgather.vx.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vrgather_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -1856,7 +1856,7 @@ vint16mf4_t test_vrgather_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrgather.vv.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrgather.vv.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vrgather_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -1868,7 +1868,7 @@ vint16mf2_t test_vrgather_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrgather.vx.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrgather.vx.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vrgather_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -1880,7 +1880,7 @@ vint16mf2_t test_vrgather_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrgather.vv.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrgather.vv.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vrgather_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -1892,7 +1892,7 @@ vint16m1_t test_vrgather_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrgather.vx.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrgather.vx.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vrgather_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -1903,7 +1903,7 @@ vint16m1_t test_vrgather_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrgather.vv.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrgather.vv.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vrgather_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -1915,7 +1915,7 @@ vint16m2_t test_vrgather_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrgather.vx.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrgather.vx.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vrgather_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -1926,7 +1926,7 @@ vint16m2_t test_vrgather_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrgather.vv.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrgather.vv.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vrgather_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -1938,7 +1938,7 @@ vint16m4_t test_vrgather_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrgather.vx.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrgather.vx.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vrgather_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -1949,7 +1949,7 @@ vint16m4_t test_vrgather_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrgather.vv.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[INDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrgather.vv.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[INDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vrgather_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -1961,7 +1961,7 @@ vint16m8_t test_vrgather_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrgather.vx.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrgather.vx.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vrgather_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -1972,7 +1972,7 @@ vint16m8_t test_vrgather_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrgather.vv.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrgather.vv.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vrgather_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -1984,7 +1984,7 @@ vint32mf2_t test_vrgather_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrgather.vx.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrgather.vx.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vrgather_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -1996,7 +1996,7 @@ vint32mf2_t test_vrgather_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrgather.vv.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrgather.vv.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vrgather_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -2008,7 +2008,7 @@ vint32m1_t test_vrgather_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrgather.vx.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrgather.vx.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vrgather_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -2019,7 +2019,7 @@ vint32m1_t test_vrgather_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrgather.vv.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrgather.vv.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vrgather_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -2031,7 +2031,7 @@ vint32m2_t test_vrgather_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrgather.vx.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrgather.vx.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vrgather_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -2042,7 +2042,7 @@ vint32m2_t test_vrgather_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrgather.vv.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrgather.vv.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vrgather_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -2054,7 +2054,7 @@ vint32m4_t test_vrgather_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrgather.vx.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrgather.vx.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vrgather_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -2065,7 +2065,7 @@ vint32m4_t test_vrgather_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrgather.vv.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrgather.vv.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vrgather_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -2077,7 +2077,7 @@ vint32m8_t test_vrgather_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrgather.vx.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrgather.vx.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vrgather_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -2088,7 +2088,7 @@ vint32m8_t test_vrgather_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrgather.vv.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrgather.vv.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vrgather_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -2100,7 +2100,7 @@ vint64m1_t test_vrgather_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrgather.vx.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrgather.vx.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vrgather_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -2111,7 +2111,7 @@ vint64m1_t test_vrgather_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrgather.vv.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrgather.vv.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vrgather_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -2123,7 +2123,7 @@ vint64m2_t test_vrgather_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrgather.vx.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrgather.vx.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vrgather_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -2134,7 +2134,7 @@ vint64m2_t test_vrgather_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrgather.vv.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrgather.vv.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vrgather_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -2146,7 +2146,7 @@ vint64m4_t test_vrgather_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrgather.vx.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrgather.vx.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vrgather_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -2157,7 +2157,7 @@ vint64m4_t test_vrgather_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrgather.vv.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrgather.vv.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vrgather_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -2169,7 +2169,7 @@ vint64m8_t test_vrgather_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrgather.vx.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrgather.vx.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vrgather_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -2180,7 +2180,7 @@ vint64m8_t test_vrgather_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrgather.vv.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrgather.vv.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vrgather_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
@@ -2192,7 +2192,7 @@ vuint8mf8_t test_vrgather_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrgather.vx.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrgather.vx.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vrgather_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
@@ -2203,7 +2203,7 @@ vuint8mf8_t test_vrgather_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrgather.vv.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrgather.vv.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vrgather_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
@@ -2215,7 +2215,7 @@ vuint8mf4_t test_vrgather_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrgather.vx.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrgather.vx.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vrgather_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
@@ -2226,7 +2226,7 @@ vuint8mf4_t test_vrgather_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrgather.vv.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrgather.vv.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vrgather_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
@@ -2238,7 +2238,7 @@ vuint8mf2_t test_vrgather_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrgather.vx.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrgather.vx.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vrgather_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
@@ -2249,7 +2249,7 @@ vuint8mf2_t test_vrgather_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrgather.vv.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrgather.vv.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vrgather_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
@@ -2261,7 +2261,7 @@ vuint8m1_t test_vrgather_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrgather.vx.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrgather.vx.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vrgather_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
@@ -2272,7 +2272,7 @@ vuint8m1_t test_vrgather_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrgather.vv.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrgather.vv.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vrgather_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
@@ -2284,7 +2284,7 @@ vuint8m2_t test_vrgather_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrgather.vx.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrgather.vx.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vrgather_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
@@ -2295,7 +2295,7 @@ vuint8m2_t test_vrgather_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrgather.vv.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[INDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrgather.vv.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[INDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vrgather_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
@@ -2307,7 +2307,7 @@ vuint8m4_t test_vrgather_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrgather.vx.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrgather.vx.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vrgather_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
@@ -2318,7 +2318,7 @@ vuint8m4_t test_vrgather_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vrgather.vv.mask.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[INDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vrgather.vv.mask.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[INDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vrgather_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
@@ -2330,7 +2330,7 @@ vuint8m8_t test_vrgather_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vrgather.vx.mask.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vrgather.vx.mask.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vrgather_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
@@ -2341,7 +2341,7 @@ vuint8m8_t test_vrgather_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrgather.vv.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrgather.vv.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vrgather_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -2353,7 +2353,7 @@ vuint16mf4_t test_vrgather_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrgather.vx.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrgather.vx.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vrgather_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -2365,7 +2365,7 @@ vuint16mf4_t test_vrgather_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrgather.vv.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrgather.vv.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vrgather_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -2377,7 +2377,7 @@ vuint16mf2_t test_vrgather_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrgather.vx.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrgather.vx.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vrgather_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -2389,7 +2389,7 @@ vuint16mf2_t test_vrgather_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrgather.vv.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrgather.vv.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vrgather_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -2401,7 +2401,7 @@ vuint16m1_t test_vrgather_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrgather.vx.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrgather.vx.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vrgather_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -2412,7 +2412,7 @@ vuint16m1_t test_vrgather_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrgather.vv.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrgather.vv.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vrgather_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -2424,7 +2424,7 @@ vuint16m2_t test_vrgather_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrgather.vx.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrgather.vx.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vrgather_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -2435,7 +2435,7 @@ vuint16m2_t test_vrgather_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrgather.vv.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrgather.vv.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vrgather_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -2447,7 +2447,7 @@ vuint16m4_t test_vrgather_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrgather.vx.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrgather.vx.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vrgather_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -2458,7 +2458,7 @@ vuint16m4_t test_vrgather_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrgather.vv.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[INDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrgather.vv.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[INDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vrgather_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -2470,7 +2470,7 @@ vuint16m8_t test_vrgather_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrgather.vx.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrgather.vx.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vrgather_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -2481,7 +2481,7 @@ vuint16m8_t test_vrgather_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrgather.vv.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrgather.vv.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vrgather_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -2493,7 +2493,7 @@ vuint32mf2_t test_vrgather_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrgather.vx.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrgather.vx.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vrgather_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -2505,7 +2505,7 @@ vuint32mf2_t test_vrgather_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrgather.vv.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrgather.vv.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vrgather_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -2517,7 +2517,7 @@ vuint32m1_t test_vrgather_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrgather.vx.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrgather.vx.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vrgather_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -2528,7 +2528,7 @@ vuint32m1_t test_vrgather_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrgather.vv.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrgather.vv.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vrgather_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -2540,7 +2540,7 @@ vuint32m2_t test_vrgather_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrgather.vx.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrgather.vx.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vrgather_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -2551,7 +2551,7 @@ vuint32m2_t test_vrgather_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrgather.vv.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrgather.vv.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vrgather_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -2563,7 +2563,7 @@ vuint32m4_t test_vrgather_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrgather.vx.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrgather.vx.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vrgather_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -2574,7 +2574,7 @@ vuint32m4_t test_vrgather_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrgather.vv.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrgather.vv.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vrgather_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -2586,7 +2586,7 @@ vuint32m8_t test_vrgather_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrgather.vx.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrgather.vx.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vrgather_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -2597,7 +2597,7 @@ vuint32m8_t test_vrgather_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrgather.vv.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrgather.vv.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vrgather_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -2609,7 +2609,7 @@ vuint64m1_t test_vrgather_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrgather.vx.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrgather.vx.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vrgather_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -2620,7 +2620,7 @@ vuint64m1_t test_vrgather_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrgather.vv.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrgather.vv.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vrgather_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -2632,7 +2632,7 @@ vuint64m2_t test_vrgather_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrgather.vx.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrgather.vx.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vrgather_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -2643,7 +2643,7 @@ vuint64m2_t test_vrgather_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrgather.vv.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrgather.vv.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vrgather_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -2655,7 +2655,7 @@ vuint64m4_t test_vrgather_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrgather.vx.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrgather.vx.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vrgather_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -2666,7 +2666,7 @@ vuint64m4_t test_vrgather_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrgather.vv.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrgather.vv.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vrgather_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
@@ -2678,7 +2678,7 @@ vuint64m8_t test_vrgather_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrgather.vx.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrgather.vx.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vrgather_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
@@ -2689,7 +2689,7 @@ vuint64m8_t test_vrgather_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vrgather.vv.mask.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x i32> [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vrgather.vv.mask.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x i32> [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vrgather_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -2701,7 +2701,7 @@ vfloat32mf2_t test_vrgather_vv_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vrgather.vx.mask.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vrgather.vx.mask.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vrgather_vx_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
@@ -2713,7 +2713,7 @@ vfloat32mf2_t test_vrgather_vx_f32mf2_m(vbool64_t mask, vfloat32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vrgather.vv.mask.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x i32> [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vrgather.vv.mask.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x i32> [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vrgather_vv_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -2725,7 +2725,7 @@ vfloat32m1_t test_vrgather_vv_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vrgather.vx.mask.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vrgather.vx.mask.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vrgather_vx_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
@@ -2737,7 +2737,7 @@ vfloat32m1_t test_vrgather_vx_f32m1_m(vbool32_t mask, vfloat32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vrgather.vv.mask.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x i32> [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vrgather.vv.mask.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x i32> [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vrgather_vv_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -2749,7 +2749,7 @@ vfloat32m2_t test_vrgather_vv_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vrgather.vx.mask.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vrgather.vx.mask.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vrgather_vx_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
@@ -2761,7 +2761,7 @@ vfloat32m2_t test_vrgather_vx_f32m2_m(vbool16_t mask, vfloat32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vrgather.vv.mask.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x i32> [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vrgather.vv.mask.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x i32> [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vrgather_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -2773,7 +2773,7 @@ vfloat32m4_t test_vrgather_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vrgather.vx.mask.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vrgather.vx.mask.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vrgather_vx_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -2785,7 +2785,7 @@ vfloat32m4_t test_vrgather_vx_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vrgather.vv.mask.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x i32> [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vrgather.vv.mask.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x i32> [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vrgather_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -2797,7 +2797,7 @@ vfloat32m8_t test_vrgather_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vrgather.vx.mask.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vrgather.vx.mask.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vrgather_vx_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -2809,7 +2809,7 @@ vfloat32m8_t test_vrgather_vx_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vrgather.vv.mask.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x i64> [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vrgather.vv.mask.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x i64> [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vrgather_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -2821,7 +2821,7 @@ vfloat64m1_t test_vrgather_vv_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vrgather.vx.mask.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vrgather.vx.mask.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vrgather_vx_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
@@ -2833,7 +2833,7 @@ vfloat64m1_t test_vrgather_vx_f64m1_m(vbool64_t mask, vfloat64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vrgather.vv.mask.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x i64> [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vrgather.vv.mask.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x i64> [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vrgather_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -2845,7 +2845,7 @@ vfloat64m2_t test_vrgather_vv_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vrgather.vx.mask.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vrgather.vx.mask.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vrgather_vx_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
@@ -2857,7 +2857,7 @@ vfloat64m2_t test_vrgather_vx_f64m2_m(vbool32_t mask, vfloat64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vrgather.vv.mask.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x i64> [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vrgather.vv.mask.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x i64> [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vrgather_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -2869,7 +2869,7 @@ vfloat64m4_t test_vrgather_vv_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vrgather.vx.mask.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vrgather.vx.mask.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vrgather_vx_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
@@ -2881,7 +2881,7 @@ vfloat64m4_t test_vrgather_vx_f64m4_m(vbool16_t mask, vfloat64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vrgather.vv.mask.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x i64> [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vrgather.vv.mask.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x i64> [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vrgather_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -2893,7 +2893,7 @@ vfloat64m8_t test_vrgather_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgather_vx_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vrgather.vx.mask.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vrgather.vx.mask.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], i64 [[INDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vrgather_vx_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
@@ -2905,7 +2905,7 @@ vfloat64m8_t test_vrgather_vx_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vrgatherei16_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
@@ -2917,7 +2917,7 @@ vint8mf8_t test_vrgatherei16_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vrgatherei16_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
@@ -2929,7 +2929,7 @@ vint8mf4_t test_vrgatherei16_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vrgatherei16_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
@@ -2941,7 +2941,7 @@ vint8mf2_t test_vrgatherei16_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vrgatherei16_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
@@ -2953,7 +2953,7 @@ vint8m1_t test_vrgatherei16_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vrgatherei16_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
@@ -2965,7 +2965,7 @@ vint8m2_t test_vrgatherei16_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vrgatherei16_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
@@ -2977,7 +2977,7 @@ vint8m4_t test_vrgatherei16_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vrgatherei16_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -2989,7 +2989,7 @@ vint16mf4_t test_vrgatherei16_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vrgatherei16_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -3001,7 +3001,7 @@ vint16mf2_t test_vrgatherei16_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vrgatherei16_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -3013,7 +3013,7 @@ vint16m1_t test_vrgatherei16_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vrgatherei16_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -3025,7 +3025,7 @@ vint16m2_t test_vrgatherei16_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vrgatherei16_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -3037,7 +3037,7 @@ vint16m4_t test_vrgatherei16_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vrgatherei16_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -3049,7 +3049,7 @@ vint16m8_t test_vrgatherei16_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrgatherei16.vv.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrgatherei16.vv.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vrgatherei16_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -3061,7 +3061,7 @@ vint32mf2_t test_vrgatherei16_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrgatherei16.vv.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrgatherei16.vv.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vrgatherei16_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -3073,7 +3073,7 @@ vint32m1_t test_vrgatherei16_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrgatherei16.vv.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrgatherei16.vv.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vrgatherei16_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -3085,7 +3085,7 @@ vint32m2_t test_vrgatherei16_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrgatherei16.vv.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrgatherei16.vv.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vrgatherei16_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -3097,7 +3097,7 @@ vint32m4_t test_vrgatherei16_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrgatherei16.vv.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrgatherei16.vv.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vrgatherei16_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -3109,7 +3109,7 @@ vint32m8_t test_vrgatherei16_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrgatherei16.vv.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrgatherei16.vv.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vrgatherei16_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -3121,7 +3121,7 @@ vint64m1_t test_vrgatherei16_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrgatherei16.vv.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrgatherei16.vv.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vrgatherei16_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -3133,7 +3133,7 @@ vint64m2_t test_vrgatherei16_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrgatherei16.vv.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrgatherei16.vv.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vrgatherei16_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -3145,7 +3145,7 @@ vint64m4_t test_vrgatherei16_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrgatherei16.vv.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrgatherei16.vv.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vrgatherei16_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -3157,7 +3157,7 @@ vint64m8_t test_vrgatherei16_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vrgatherei16_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
@@ -3169,7 +3169,7 @@ vuint8mf8_t test_vrgatherei16_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vrgatherei16_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
@@ -3181,7 +3181,7 @@ vuint8mf4_t test_vrgatherei16_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vrgatherei16_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
@@ -3193,7 +3193,7 @@ vuint8mf2_t test_vrgatherei16_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vrgatherei16_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
@@ -3205,7 +3205,7 @@ vuint8m1_t test_vrgatherei16_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vrgatherei16_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
@@ -3217,7 +3217,7 @@ vuint8m2_t test_vrgatherei16_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrgatherei16.vv.mask.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vrgatherei16_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
@@ -3229,7 +3229,7 @@ vuint8m4_t test_vrgatherei16_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vrgatherei16_vv_u16mf4_m(vbool64_t mask,
@@ -3242,7 +3242,7 @@ vuint16mf4_t test_vrgatherei16_vv_u16mf4_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vrgatherei16_vv_u16mf2_m(vbool32_t mask,
@@ -3255,7 +3255,7 @@ vuint16mf2_t test_vrgatherei16_vv_u16mf2_m(vbool32_t mask,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vrgatherei16_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -3267,7 +3267,7 @@ vuint16m1_t test_vrgatherei16_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vrgatherei16_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -3279,7 +3279,7 @@ vuint16m2_t test_vrgatherei16_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vrgatherei16_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -3291,7 +3291,7 @@ vuint16m4_t test_vrgatherei16_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrgatherei16.vv.mask.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vrgatherei16_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -3303,7 +3303,7 @@ vuint16m8_t test_vrgatherei16_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrgatherei16.vv.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrgatherei16.vv.mask.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vrgatherei16_vv_u32mf2_m(vbool64_t mask,
@@ -3316,7 +3316,7 @@ vuint32mf2_t test_vrgatherei16_vv_u32mf2_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrgatherei16.vv.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrgatherei16.vv.mask.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vrgatherei16_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -3328,7 +3328,7 @@ vuint32m1_t test_vrgatherei16_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrgatherei16.vv.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrgatherei16.vv.mask.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vrgatherei16_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -3340,7 +3340,7 @@ vuint32m2_t test_vrgatherei16_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrgatherei16.vv.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrgatherei16.vv.mask.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vrgatherei16_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -3352,7 +3352,7 @@ vuint32m4_t test_vrgatherei16_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrgatherei16.vv.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrgatherei16.vv.mask.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vrgatherei16_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -3364,7 +3364,7 @@ vuint32m8_t test_vrgatherei16_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrgatherei16.vv.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrgatherei16.vv.mask.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vrgatherei16_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -3376,7 +3376,7 @@ vuint64m1_t test_vrgatherei16_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrgatherei16.vv.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrgatherei16.vv.mask.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vrgatherei16_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -3388,7 +3388,7 @@ vuint64m2_t test_vrgatherei16_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrgatherei16.vv.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrgatherei16.vv.mask.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vrgatherei16_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -3400,7 +3400,7 @@ vuint64m4_t test_vrgatherei16_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrgatherei16.vv.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrgatherei16.vv.mask.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vrgatherei16_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
@@ -3412,7 +3412,7 @@ vuint64m8_t test_vrgatherei16_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vrgatherei16.vv.mask.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vrgatherei16.vv.mask.nxv1f32.i64(<vscale x 1 x float> [[MASKEDOFF:%.*]], <vscale x 1 x float> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vrgatherei16_vv_f32mf2_m(vbool64_t mask,
@@ -3425,7 +3425,7 @@ vfloat32mf2_t test_vrgatherei16_vv_f32mf2_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vrgatherei16.vv.mask.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vrgatherei16.vv.mask.nxv2f32.i64(<vscale x 2 x float> [[MASKEDOFF:%.*]], <vscale x 2 x float> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vrgatherei16_vv_f32m1_m(vbool32_t mask,
@@ -3438,7 +3438,7 @@ vfloat32m1_t test_vrgatherei16_vv_f32m1_m(vbool32_t mask,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vrgatherei16.vv.mask.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vrgatherei16.vv.mask.nxv4f32.i64(<vscale x 4 x float> [[MASKEDOFF:%.*]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vrgatherei16_vv_f32m2_m(vbool16_t mask,
@@ -3451,7 +3451,7 @@ vfloat32m2_t test_vrgatherei16_vv_f32m2_m(vbool16_t mask,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vrgatherei16.vv.mask.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vrgatherei16.vv.mask.nxv8f32.i64(<vscale x 8 x float> [[MASKEDOFF:%.*]], <vscale x 8 x float> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vrgatherei16_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
@@ -3463,7 +3463,7 @@ vfloat32m4_t test_vrgatherei16_vv_f32m4_m(vbool8_t mask, vfloat32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vrgatherei16.vv.mask.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vrgatherei16.vv.mask.nxv16f32.i64(<vscale x 16 x float> [[MASKEDOFF:%.*]], <vscale x 16 x float> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vrgatherei16_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
@@ -3475,7 +3475,7 @@ vfloat32m8_t test_vrgatherei16_vv_f32m8_m(vbool4_t mask, vfloat32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vrgatherei16.vv.mask.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vrgatherei16.vv.mask.nxv1f64.i64(<vscale x 1 x double> [[MASKEDOFF:%.*]], <vscale x 1 x double> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vrgatherei16_vv_f64m1_m(vbool64_t mask,
@@ -3488,7 +3488,7 @@ vfloat64m1_t test_vrgatherei16_vv_f64m1_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vrgatherei16.vv.mask.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vrgatherei16.vv.mask.nxv2f64.i64(<vscale x 2 x double> [[MASKEDOFF:%.*]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vrgatherei16_vv_f64m2_m(vbool32_t mask,
@@ -3501,7 +3501,7 @@ vfloat64m2_t test_vrgatherei16_vv_f64m2_m(vbool32_t mask,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vrgatherei16.vv.mask.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vrgatherei16.vv.mask.nxv4f64.i64(<vscale x 4 x double> [[MASKEDOFF:%.*]], <vscale x 4 x double> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vrgatherei16_vv_f64m4_m(vbool16_t mask,
@@ -3514,7 +3514,7 @@ vfloat64m4_t test_vrgatherei16_vv_f64m4_m(vbool16_t mask,
 //
 // CHECK-RV64-LABEL: @test_vrgatherei16_vv_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vrgatherei16.vv.mask.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vrgatherei16.vv.mask.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vrgatherei16_vv_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vrsub.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vrsub.c
index 2f5b379..8ecd01f 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vrsub.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vrsub.c
@@ -447,7 +447,7 @@ vuint64m8_t test_vrsub_vx_u64m8(vuint64m8_t op1, uint64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrsub.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrsub.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vrsub_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, int8_t op2, size_t vl) {
@@ -457,7 +457,7 @@ vint8mf8_t test_vrsub_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrsub.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrsub.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vrsub_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, int8_t op2, size_t vl) {
@@ -467,7 +467,7 @@ vint8mf4_t test_vrsub_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrsub.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrsub.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vrsub_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, int8_t op2, size_t vl) {
@@ -477,7 +477,7 @@ vint8mf2_t test_vrsub_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrsub.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrsub.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vrsub_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, int8_t op2, size_t vl) {
@@ -487,7 +487,7 @@ vint8m1_t test_vrsub_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrsub.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrsub.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vrsub_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, int8_t op2, size_t vl) {
@@ -497,7 +497,7 @@ vint8m2_t test_vrsub_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrsub.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrsub.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vrsub_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, int8_t op2, size_t vl) {
@@ -507,7 +507,7 @@ vint8m4_t test_vrsub_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vrsub.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vrsub.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vrsub_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, int8_t op2, size_t vl) {
@@ -517,7 +517,7 @@ vint8m8_t test_vrsub_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrsub.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrsub.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vrsub_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, int16_t op2, size_t vl) {
@@ -527,7 +527,7 @@ vint16mf4_t test_vrsub_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrsub.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrsub.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vrsub_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, int16_t op2, size_t vl) {
@@ -537,7 +537,7 @@ vint16mf2_t test_vrsub_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrsub.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrsub.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vrsub_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, int16_t op2, size_t vl) {
@@ -547,7 +547,7 @@ vint16m1_t test_vrsub_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrsub.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrsub.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vrsub_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, int16_t op2, size_t vl) {
@@ -557,7 +557,7 @@ vint16m2_t test_vrsub_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrsub.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrsub.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vrsub_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, int16_t op2, size_t vl) {
@@ -567,7 +567,7 @@ vint16m4_t test_vrsub_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrsub.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrsub.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vrsub_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, int16_t op2, size_t vl) {
@@ -577,7 +577,7 @@ vint16m8_t test_vrsub_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrsub.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrsub.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vrsub_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, int32_t op2, size_t vl) {
@@ -587,7 +587,7 @@ vint32mf2_t test_vrsub_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrsub.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrsub.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vrsub_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, int32_t op2, size_t vl) {
@@ -597,7 +597,7 @@ vint32m1_t test_vrsub_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrsub.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrsub.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vrsub_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, int32_t op2, size_t vl) {
@@ -607,7 +607,7 @@ vint32m2_t test_vrsub_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrsub.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrsub.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vrsub_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, int32_t op2, size_t vl) {
@@ -617,7 +617,7 @@ vint32m4_t test_vrsub_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrsub.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrsub.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vrsub_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, int32_t op2, size_t vl) {
@@ -627,7 +627,7 @@ vint32m8_t test_vrsub_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrsub.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrsub.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vrsub_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, int64_t op2, size_t vl) {
@@ -637,7 +637,7 @@ vint64m1_t test_vrsub_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrsub.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrsub.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vrsub_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, int64_t op2, size_t vl) {
@@ -647,7 +647,7 @@ vint64m2_t test_vrsub_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrsub.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrsub.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vrsub_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, int64_t op2, size_t vl) {
@@ -657,7 +657,7 @@ vint64m4_t test_vrsub_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrsub.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrsub.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vrsub_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, int64_t op2, size_t vl) {
@@ -667,7 +667,7 @@ vint64m8_t test_vrsub_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrsub.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vrsub.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vrsub_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, uint8_t op2, size_t vl) {
@@ -677,7 +677,7 @@ vuint8mf8_t test_vrsub_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrsub.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vrsub.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vrsub_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, uint8_t op2, size_t vl) {
@@ -687,7 +687,7 @@ vuint8mf4_t test_vrsub_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrsub.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vrsub.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vrsub_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, uint8_t op2, size_t vl) {
@@ -697,7 +697,7 @@ vuint8mf2_t test_vrsub_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrsub.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vrsub.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vrsub_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, uint8_t op2, size_t vl) {
@@ -707,7 +707,7 @@ vuint8m1_t test_vrsub_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrsub.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vrsub.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vrsub_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, uint8_t op2, size_t vl) {
@@ -717,7 +717,7 @@ vuint8m2_t test_vrsub_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrsub.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vrsub.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vrsub_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, uint8_t op2, size_t vl) {
@@ -727,7 +727,7 @@ vuint8m4_t test_vrsub_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vrsub.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vrsub.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vrsub_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, uint8_t op2, size_t vl) {
@@ -737,7 +737,7 @@ vuint8m8_t test_vrsub_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrsub.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vrsub.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vrsub_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, uint16_t op2, size_t vl) {
@@ -747,7 +747,7 @@ vuint16mf4_t test_vrsub_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrsub.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vrsub.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vrsub_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, uint16_t op2, size_t vl) {
@@ -757,7 +757,7 @@ vuint16mf2_t test_vrsub_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrsub.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vrsub.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vrsub_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, uint16_t op2, size_t vl) {
@@ -767,7 +767,7 @@ vuint16m1_t test_vrsub_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrsub.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vrsub.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vrsub_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, uint16_t op2, size_t vl) {
@@ -777,7 +777,7 @@ vuint16m2_t test_vrsub_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrsub.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vrsub.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vrsub_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, uint16_t op2, size_t vl) {
@@ -787,7 +787,7 @@ vuint16m4_t test_vrsub_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrsub.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vrsub.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vrsub_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, uint16_t op2, size_t vl) {
@@ -797,7 +797,7 @@ vuint16m8_t test_vrsub_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrsub.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vrsub.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vrsub_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, uint32_t op2, size_t vl) {
@@ -807,7 +807,7 @@ vuint32mf2_t test_vrsub_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuin
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrsub.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vrsub.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vrsub_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, uint32_t op2, size_t vl) {
@@ -817,7 +817,7 @@ vuint32m1_t test_vrsub_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrsub.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vrsub.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vrsub_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, uint32_t op2, size_t vl) {
@@ -827,7 +827,7 @@ vuint32m2_t test_vrsub_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrsub.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vrsub.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vrsub_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, uint32_t op2, size_t vl) {
@@ -837,7 +837,7 @@ vuint32m4_t test_vrsub_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrsub.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vrsub.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vrsub_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, uint32_t op2, size_t vl) {
@@ -847,7 +847,7 @@ vuint32m8_t test_vrsub_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrsub.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vrsub.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vrsub_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, uint64_t op2, size_t vl) {
@@ -857,7 +857,7 @@ vuint64m1_t test_vrsub_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrsub.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vrsub.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vrsub_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, uint64_t op2, size_t vl) {
@@ -867,7 +867,7 @@ vuint64m2_t test_vrsub_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrsub.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vrsub.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vrsub_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, uint64_t op2, size_t vl) {
@@ -877,7 +877,7 @@ vuint64m4_t test_vrsub_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64
 //
 // CHECK-RV64-LABEL: @test_vrsub_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrsub.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vrsub.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vrsub_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, uint64_t op2, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsadd.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsadd.c
index 197f7fc..2763370 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsadd.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsadd.c
@@ -890,7 +890,7 @@ vuint64m8_t test_vsaddu_vx_u64m8(vuint64m8_t op1, uint64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsadd.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsadd.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vsadd_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
@@ -901,7 +901,7 @@ vint8mf8_t test_vsadd_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsadd.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsadd.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vsadd_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
@@ -912,7 +912,7 @@ vint8mf8_t test_vsadd_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsadd.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsadd.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vsadd_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
@@ -923,7 +923,7 @@ vint8mf4_t test_vsadd_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsadd.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsadd.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vsadd_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
@@ -934,7 +934,7 @@ vint8mf4_t test_vsadd_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsadd.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsadd.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vsadd_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
@@ -945,7 +945,7 @@ vint8mf2_t test_vsadd_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsadd.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsadd.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vsadd_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
@@ -956,7 +956,7 @@ vint8mf2_t test_vsadd_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsadd.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsadd.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vsadd_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
@@ -967,7 +967,7 @@ vint8m1_t test_vsadd_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsadd.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsadd.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vsadd_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
@@ -978,7 +978,7 @@ vint8m1_t test_vsadd_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsadd.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsadd.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vsadd_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
@@ -989,7 +989,7 @@ vint8m2_t test_vsadd_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsadd.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsadd.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vsadd_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
@@ -1000,7 +1000,7 @@ vint8m2_t test_vsadd_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsadd.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsadd.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vsadd_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
@@ -1011,7 +1011,7 @@ vint8m4_t test_vsadd_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsadd.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsadd.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vsadd_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
@@ -1022,7 +1022,7 @@ vint8m4_t test_vsadd_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsadd.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsadd.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vsadd_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
@@ -1033,7 +1033,7 @@ vint8m8_t test_vsadd_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsadd.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsadd.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vsadd_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
@@ -1044,7 +1044,7 @@ vint8m8_t test_vsadd_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsadd.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsadd.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vsadd_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -1056,7 +1056,7 @@ vint16mf4_t test_vsadd_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsadd.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsadd.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vsadd_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -1067,7 +1067,7 @@ vint16mf4_t test_vsadd_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsadd.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsadd.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vsadd_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -1079,7 +1079,7 @@ vint16mf2_t test_vsadd_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsadd.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsadd.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vsadd_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -1090,7 +1090,7 @@ vint16mf2_t test_vsadd_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsadd.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsadd.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vsadd_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -1101,7 +1101,7 @@ vint16m1_t test_vsadd_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsadd.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsadd.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vsadd_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -1112,7 +1112,7 @@ vint16m1_t test_vsadd_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsadd.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsadd.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vsadd_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -1123,7 +1123,7 @@ vint16m2_t test_vsadd_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsadd.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsadd.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vsadd_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -1134,7 +1134,7 @@ vint16m2_t test_vsadd_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsadd.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsadd.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vsadd_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -1145,7 +1145,7 @@ vint16m4_t test_vsadd_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsadd.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsadd.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vsadd_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -1156,7 +1156,7 @@ vint16m4_t test_vsadd_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsadd.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsadd.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vsadd_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -1167,7 +1167,7 @@ vint16m8_t test_vsadd_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsadd.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsadd.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vsadd_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -1178,7 +1178,7 @@ vint16m8_t test_vsadd_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsadd.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsadd.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vsadd_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -1190,7 +1190,7 @@ vint32mf2_t test_vsadd_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsadd.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsadd.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vsadd_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -1201,7 +1201,7 @@ vint32mf2_t test_vsadd_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsadd.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsadd.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vsadd_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -1212,7 +1212,7 @@ vint32m1_t test_vsadd_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsadd.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsadd.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vsadd_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -1223,7 +1223,7 @@ vint32m1_t test_vsadd_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsadd.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsadd.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vsadd_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -1234,7 +1234,7 @@ vint32m2_t test_vsadd_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsadd.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsadd.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vsadd_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -1245,7 +1245,7 @@ vint32m2_t test_vsadd_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsadd.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsadd.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vsadd_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -1256,7 +1256,7 @@ vint32m4_t test_vsadd_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsadd.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsadd.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vsadd_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -1267,7 +1267,7 @@ vint32m4_t test_vsadd_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsadd.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsadd.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vsadd_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -1278,7 +1278,7 @@ vint32m8_t test_vsadd_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsadd.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsadd.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vsadd_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -1289,7 +1289,7 @@ vint32m8_t test_vsadd_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsadd.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsadd.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vsadd_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -1300,7 +1300,7 @@ vint64m1_t test_vsadd_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsadd.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsadd.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vsadd_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -1311,7 +1311,7 @@ vint64m1_t test_vsadd_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsadd.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsadd.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vsadd_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -1322,7 +1322,7 @@ vint64m2_t test_vsadd_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsadd.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsadd.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vsadd_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -1333,7 +1333,7 @@ vint64m2_t test_vsadd_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsadd.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsadd.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vsadd_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -1344,7 +1344,7 @@ vint64m4_t test_vsadd_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsadd.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsadd.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vsadd_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -1355,7 +1355,7 @@ vint64m4_t test_vsadd_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsadd.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsadd.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vsadd_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -1366,7 +1366,7 @@ vint64m8_t test_vsadd_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsadd_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsadd.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsadd.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vsadd_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -1377,7 +1377,7 @@ vint64m8_t test_vsadd_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsaddu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsaddu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vsaddu_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
@@ -1389,7 +1389,7 @@ vuint8mf8_t test_vsaddu_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsaddu.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsaddu.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vsaddu_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
@@ -1400,7 +1400,7 @@ vuint8mf8_t test_vsaddu_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsaddu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsaddu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vsaddu_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
@@ -1412,7 +1412,7 @@ vuint8mf4_t test_vsaddu_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsaddu.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsaddu.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vsaddu_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
@@ -1423,7 +1423,7 @@ vuint8mf4_t test_vsaddu_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsaddu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsaddu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vsaddu_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
@@ -1435,7 +1435,7 @@ vuint8mf2_t test_vsaddu_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsaddu.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsaddu.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vsaddu_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
@@ -1446,7 +1446,7 @@ vuint8mf2_t test_vsaddu_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsaddu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsaddu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vsaddu_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
@@ -1457,7 +1457,7 @@ vuint8m1_t test_vsaddu_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsaddu.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsaddu.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vsaddu_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
@@ -1468,7 +1468,7 @@ vuint8m1_t test_vsaddu_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsaddu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsaddu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vsaddu_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
@@ -1479,7 +1479,7 @@ vuint8m2_t test_vsaddu_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsaddu.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsaddu.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vsaddu_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
@@ -1490,7 +1490,7 @@ vuint8m2_t test_vsaddu_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsaddu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsaddu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vsaddu_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
@@ -1501,7 +1501,7 @@ vuint8m4_t test_vsaddu_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsaddu.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsaddu.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vsaddu_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
@@ -1512,7 +1512,7 @@ vuint8m4_t test_vsaddu_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsaddu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsaddu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vsaddu_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
@@ -1523,7 +1523,7 @@ vuint8m8_t test_vsaddu_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsaddu.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsaddu.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vsaddu_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
@@ -1534,7 +1534,7 @@ vuint8m8_t test_vsaddu_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsaddu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsaddu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vsaddu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -1546,7 +1546,7 @@ vuint16mf4_t test_vsaddu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsaddu.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsaddu.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vsaddu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -1558,7 +1558,7 @@ vuint16mf4_t test_vsaddu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsaddu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsaddu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vsaddu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -1570,7 +1570,7 @@ vuint16mf2_t test_vsaddu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsaddu.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsaddu.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vsaddu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -1582,7 +1582,7 @@ vuint16mf2_t test_vsaddu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsaddu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsaddu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vsaddu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -1594,7 +1594,7 @@ vuint16m1_t test_vsaddu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsaddu.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsaddu.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vsaddu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -1605,7 +1605,7 @@ vuint16m1_t test_vsaddu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsaddu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsaddu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vsaddu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -1617,7 +1617,7 @@ vuint16m2_t test_vsaddu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsaddu.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsaddu.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vsaddu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -1628,7 +1628,7 @@ vuint16m2_t test_vsaddu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsaddu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsaddu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vsaddu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -1640,7 +1640,7 @@ vuint16m4_t test_vsaddu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsaddu.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsaddu.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vsaddu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -1651,7 +1651,7 @@ vuint16m4_t test_vsaddu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsaddu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsaddu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vsaddu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -1663,7 +1663,7 @@ vuint16m8_t test_vsaddu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsaddu.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsaddu.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vsaddu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -1674,7 +1674,7 @@ vuint16m8_t test_vsaddu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsaddu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsaddu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vsaddu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -1686,7 +1686,7 @@ vuint32mf2_t test_vsaddu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsaddu.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsaddu.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vsaddu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -1698,7 +1698,7 @@ vuint32mf2_t test_vsaddu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsaddu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsaddu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vsaddu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -1710,7 +1710,7 @@ vuint32m1_t test_vsaddu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsaddu.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsaddu.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vsaddu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -1721,7 +1721,7 @@ vuint32m1_t test_vsaddu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsaddu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsaddu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vsaddu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -1733,7 +1733,7 @@ vuint32m2_t test_vsaddu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsaddu.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsaddu.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vsaddu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -1744,7 +1744,7 @@ vuint32m2_t test_vsaddu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsaddu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsaddu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vsaddu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -1756,7 +1756,7 @@ vuint32m4_t test_vsaddu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsaddu.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsaddu.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vsaddu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -1767,7 +1767,7 @@ vuint32m4_t test_vsaddu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsaddu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsaddu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vsaddu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -1779,7 +1779,7 @@ vuint32m8_t test_vsaddu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsaddu.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsaddu.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vsaddu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -1790,7 +1790,7 @@ vuint32m8_t test_vsaddu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsaddu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsaddu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vsaddu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -1802,7 +1802,7 @@ vuint64m1_t test_vsaddu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsaddu.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsaddu.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vsaddu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -1813,7 +1813,7 @@ vuint64m1_t test_vsaddu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsaddu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsaddu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vsaddu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -1825,7 +1825,7 @@ vuint64m2_t test_vsaddu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsaddu.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsaddu.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vsaddu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -1836,7 +1836,7 @@ vuint64m2_t test_vsaddu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsaddu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsaddu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vsaddu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -1848,7 +1848,7 @@ vuint64m4_t test_vsaddu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsaddu.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsaddu.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vsaddu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -1859,7 +1859,7 @@ vuint64m4_t test_vsaddu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsaddu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsaddu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vsaddu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
@@ -1871,7 +1871,7 @@ vuint64m8_t test_vsaddu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsaddu_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsaddu.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsaddu.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vsaddu_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vse.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vse.c
index 84fd860..4e7bb76 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vse.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vse.c
@@ -9,7 +9,7 @@
 // CHECK-RV64-LABEL: @test_vse8_v_i8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11:[0-9]+]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_i8mf8(int8_t *base, vint8mf8_t value, size_t vl) {
@@ -20,7 +20,7 @@ void test_vse8_v_i8mf8(int8_t *base, vint8mf8_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse8_v_i8mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_i8mf4(int8_t *base, vint8mf4_t value, size_t vl) {
@@ -31,7 +31,7 @@ void test_vse8_v_i8mf4(int8_t *base, vint8mf4_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse8_v_i8mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_i8mf2(int8_t *base, vint8mf2_t value, size_t vl) {
@@ -42,7 +42,7 @@ void test_vse8_v_i8mf2(int8_t *base, vint8mf2_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse8_v_i8m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_i8m1(int8_t *base, vint8m1_t value, size_t vl) {
@@ -53,7 +53,7 @@ void test_vse8_v_i8m1(int8_t *base, vint8m1_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse8_v_i8m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_i8m2(int8_t *base, vint8m2_t value, size_t vl) {
@@ -64,7 +64,7 @@ void test_vse8_v_i8m2(int8_t *base, vint8m2_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse8_v_i8m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_i8m4(int8_t *base, vint8m4_t value, size_t vl) {
@@ -75,7 +75,7 @@ void test_vse8_v_i8m4(int8_t *base, vint8m4_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse8_v_i8m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_i8m8(int8_t *base, vint8m8_t value, size_t vl) {
@@ -86,7 +86,7 @@ void test_vse8_v_i8m8(int8_t *base, vint8m8_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse16_v_i16mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_i16mf4(int16_t *base, vint16mf4_t value, size_t vl) {
@@ -97,7 +97,7 @@ void test_vse16_v_i16mf4(int16_t *base, vint16mf4_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse16_v_i16mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_i16mf2(int16_t *base, vint16mf2_t value, size_t vl) {
@@ -108,7 +108,7 @@ void test_vse16_v_i16mf2(int16_t *base, vint16mf2_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse16_v_i16m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_i16m1(int16_t *base, vint16m1_t value, size_t vl) {
@@ -119,7 +119,7 @@ void test_vse16_v_i16m1(int16_t *base, vint16m1_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse16_v_i16m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_i16m2(int16_t *base, vint16m2_t value, size_t vl) {
@@ -130,7 +130,7 @@ void test_vse16_v_i16m2(int16_t *base, vint16m2_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse16_v_i16m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_i16m4(int16_t *base, vint16m4_t value, size_t vl) {
@@ -141,7 +141,7 @@ void test_vse16_v_i16m4(int16_t *base, vint16m4_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse16_v_i16m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_i16m8(int16_t *base, vint16m8_t value, size_t vl) {
@@ -152,7 +152,7 @@ void test_vse16_v_i16m8(int16_t *base, vint16m8_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse32_v_i32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_i32mf2(int32_t *base, vint32mf2_t value, size_t vl) {
@@ -163,7 +163,7 @@ void test_vse32_v_i32mf2(int32_t *base, vint32mf2_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse32_v_i32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_i32m1(int32_t *base, vint32m1_t value, size_t vl) {
@@ -174,7 +174,7 @@ void test_vse32_v_i32m1(int32_t *base, vint32m1_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse32_v_i32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_i32m2(int32_t *base, vint32m2_t value, size_t vl) {
@@ -185,7 +185,7 @@ void test_vse32_v_i32m2(int32_t *base, vint32m2_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse32_v_i32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_i32m4(int32_t *base, vint32m4_t value, size_t vl) {
@@ -196,7 +196,7 @@ void test_vse32_v_i32m4(int32_t *base, vint32m4_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse32_v_i32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_i32m8(int32_t *base, vint32m8_t value, size_t vl) {
@@ -207,7 +207,7 @@ void test_vse32_v_i32m8(int32_t *base, vint32m8_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse64_v_i64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_i64m1(int64_t *base, vint64m1_t value, size_t vl) {
@@ -218,7 +218,7 @@ void test_vse64_v_i64m1(int64_t *base, vint64m1_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse64_v_i64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_i64m2(int64_t *base, vint64m2_t value, size_t vl) {
@@ -229,7 +229,7 @@ void test_vse64_v_i64m2(int64_t *base, vint64m2_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse64_v_i64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_i64m4(int64_t *base, vint64m4_t value, size_t vl) {
@@ -240,7 +240,7 @@ void test_vse64_v_i64m4(int64_t *base, vint64m4_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse64_v_i64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_i64m8(int64_t *base, vint64m8_t value, size_t vl) {
@@ -251,7 +251,7 @@ void test_vse64_v_i64m8(int64_t *base, vint64m8_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse8_v_u8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_u8mf8(uint8_t *base, vuint8mf8_t value, size_t vl) {
@@ -262,7 +262,7 @@ void test_vse8_v_u8mf8(uint8_t *base, vuint8mf8_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse8_v_u8mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_u8mf4(uint8_t *base, vuint8mf4_t value, size_t vl) {
@@ -273,7 +273,7 @@ void test_vse8_v_u8mf4(uint8_t *base, vuint8mf4_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse8_v_u8mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_u8mf2(uint8_t *base, vuint8mf2_t value, size_t vl) {
@@ -284,7 +284,7 @@ void test_vse8_v_u8mf2(uint8_t *base, vuint8mf2_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse8_v_u8m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_u8m1(uint8_t *base, vuint8m1_t value, size_t vl) {
@@ -295,7 +295,7 @@ void test_vse8_v_u8m1(uint8_t *base, vuint8m1_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse8_v_u8m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_u8m2(uint8_t *base, vuint8m2_t value, size_t vl) {
@@ -306,7 +306,7 @@ void test_vse8_v_u8m2(uint8_t *base, vuint8m2_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse8_v_u8m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_u8m4(uint8_t *base, vuint8m4_t value, size_t vl) {
@@ -317,7 +317,7 @@ void test_vse8_v_u8m4(uint8_t *base, vuint8m4_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse8_v_u8m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_u8m8(uint8_t *base, vuint8m8_t value, size_t vl) {
@@ -328,7 +328,7 @@ void test_vse8_v_u8m8(uint8_t *base, vuint8m8_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse16_v_u16mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_u16mf4(uint16_t *base, vuint16mf4_t value, size_t vl) {
@@ -339,7 +339,7 @@ void test_vse16_v_u16mf4(uint16_t *base, vuint16mf4_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse16_v_u16mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_u16mf2(uint16_t *base, vuint16mf2_t value, size_t vl) {
@@ -350,7 +350,7 @@ void test_vse16_v_u16mf2(uint16_t *base, vuint16mf2_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse16_v_u16m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_u16m1(uint16_t *base, vuint16m1_t value, size_t vl) {
@@ -361,7 +361,7 @@ void test_vse16_v_u16m1(uint16_t *base, vuint16m1_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse16_v_u16m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_u16m2(uint16_t *base, vuint16m2_t value, size_t vl) {
@@ -372,7 +372,7 @@ void test_vse16_v_u16m2(uint16_t *base, vuint16m2_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse16_v_u16m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_u16m4(uint16_t *base, vuint16m4_t value, size_t vl) {
@@ -383,7 +383,7 @@ void test_vse16_v_u16m4(uint16_t *base, vuint16m4_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse16_v_u16m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_u16m8(uint16_t *base, vuint16m8_t value, size_t vl) {
@@ -394,7 +394,7 @@ void test_vse16_v_u16m8(uint16_t *base, vuint16m8_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse32_v_u32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_u32mf2(uint32_t *base, vuint32mf2_t value, size_t vl) {
@@ -405,7 +405,7 @@ void test_vse32_v_u32mf2(uint32_t *base, vuint32mf2_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse32_v_u32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_u32m1(uint32_t *base, vuint32m1_t value, size_t vl) {
@@ -416,7 +416,7 @@ void test_vse32_v_u32m1(uint32_t *base, vuint32m1_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse32_v_u32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_u32m2(uint32_t *base, vuint32m2_t value, size_t vl) {
@@ -427,7 +427,7 @@ void test_vse32_v_u32m2(uint32_t *base, vuint32m2_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse32_v_u32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_u32m4(uint32_t *base, vuint32m4_t value, size_t vl) {
@@ -438,7 +438,7 @@ void test_vse32_v_u32m4(uint32_t *base, vuint32m4_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse32_v_u32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_u32m8(uint32_t *base, vuint32m8_t value, size_t vl) {
@@ -449,7 +449,7 @@ void test_vse32_v_u32m8(uint32_t *base, vuint32m8_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse64_v_u64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_u64m1(uint64_t *base, vuint64m1_t value, size_t vl) {
@@ -460,7 +460,7 @@ void test_vse64_v_u64m1(uint64_t *base, vuint64m1_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse64_v_u64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_u64m2(uint64_t *base, vuint64m2_t value, size_t vl) {
@@ -471,7 +471,7 @@ void test_vse64_v_u64m2(uint64_t *base, vuint64m2_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse64_v_u64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_u64m4(uint64_t *base, vuint64m4_t value, size_t vl) {
@@ -482,7 +482,7 @@ void test_vse64_v_u64m4(uint64_t *base, vuint64m4_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse64_v_u64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_u64m8(uint64_t *base, vuint64m8_t value, size_t vl) {
@@ -493,7 +493,7 @@ void test_vse64_v_u64m8(uint64_t *base, vuint64m8_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse32_v_f32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1f32.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1f32.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_f32mf2(float *base, vfloat32mf2_t value, size_t vl) {
@@ -504,7 +504,7 @@ void test_vse32_v_f32mf2(float *base, vfloat32mf2_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse32_v_f32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv2f32.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv2f32.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_f32m1(float *base, vfloat32m1_t value, size_t vl) {
@@ -515,7 +515,7 @@ void test_vse32_v_f32m1(float *base, vfloat32m1_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse32_v_f32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv4f32.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv4f32.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_f32m2(float *base, vfloat32m2_t value, size_t vl) {
@@ -526,7 +526,7 @@ void test_vse32_v_f32m2(float *base, vfloat32m2_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse32_v_f32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv8f32.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv8f32.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_f32m4(float *base, vfloat32m4_t value, size_t vl) {
@@ -537,7 +537,7 @@ void test_vse32_v_f32m4(float *base, vfloat32m4_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse32_v_f32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv16f32.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv16f32.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_f32m8(float *base, vfloat32m8_t value, size_t vl) {
@@ -548,7 +548,7 @@ void test_vse32_v_f32m8(float *base, vfloat32m8_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse64_v_f64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1f64.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1f64.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_f64m1(double *base, vfloat64m1_t value, size_t vl) {
@@ -559,7 +559,7 @@ void test_vse64_v_f64m1(double *base, vfloat64m1_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse64_v_f64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv2f64.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv2f64.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_f64m2(double *base, vfloat64m2_t value, size_t vl) {
@@ -570,7 +570,7 @@ void test_vse64_v_f64m2(double *base, vfloat64m2_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse64_v_f64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv4f64.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv4f64.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_f64m4(double *base, vfloat64m4_t value, size_t vl) {
@@ -581,7 +581,7 @@ void test_vse64_v_f64m4(double *base, vfloat64m4_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse64_v_f64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv8f64.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv8f64.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_f64m8(double *base, vfloat64m8_t value, size_t vl) {
@@ -592,7 +592,7 @@ void test_vse64_v_f64m8(double *base, vfloat64m8_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse8_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_i8mf8_m(vbool64_t mask, int8_t *base, vint8mf8_t value, size_t vl) {
@@ -603,7 +603,7 @@ void test_vse8_v_i8mf8_m(vbool64_t mask, int8_t *base, vint8mf8_t value, size_t
 // CHECK-RV64-LABEL: @test_vse8_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_i8mf4_m(vbool32_t mask, int8_t *base, vint8mf4_t value, size_t vl) {
@@ -614,7 +614,7 @@ void test_vse8_v_i8mf4_m(vbool32_t mask, int8_t *base, vint8mf4_t value, size_t
 // CHECK-RV64-LABEL: @test_vse8_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_i8mf2_m(vbool16_t mask, int8_t *base, vint8mf2_t value, size_t vl) {
@@ -625,7 +625,7 @@ void test_vse8_v_i8mf2_m(vbool16_t mask, int8_t *base, vint8mf2_t value, size_t
 // CHECK-RV64-LABEL: @test_vse8_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_i8m1_m(vbool8_t mask, int8_t *base, vint8m1_t value, size_t vl) {
@@ -636,7 +636,7 @@ void test_vse8_v_i8m1_m(vbool8_t mask, int8_t *base, vint8m1_t value, size_t vl)
 // CHECK-RV64-LABEL: @test_vse8_v_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_i8m2_m(vbool4_t mask, int8_t *base, vint8m2_t value, size_t vl) {
@@ -647,7 +647,7 @@ void test_vse8_v_i8m2_m(vbool4_t mask, int8_t *base, vint8m2_t value, size_t vl)
 // CHECK-RV64-LABEL: @test_vse8_v_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_i8m4_m(vbool2_t mask, int8_t *base, vint8m4_t value, size_t vl) {
@@ -658,7 +658,7 @@ void test_vse8_v_i8m4_m(vbool2_t mask, int8_t *base, vint8m4_t value, size_t vl)
 // CHECK-RV64-LABEL: @test_vse8_v_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_i8m8_m(vbool1_t mask, int8_t *base, vint8m8_t value, size_t vl) {
@@ -669,7 +669,7 @@ void test_vse8_v_i8m8_m(vbool1_t mask, int8_t *base, vint8m8_t value, size_t vl)
 // CHECK-RV64-LABEL: @test_vse16_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_i16mf4_m(vbool64_t mask, int16_t *base, vint16mf4_t value, size_t vl) {
@@ -680,7 +680,7 @@ void test_vse16_v_i16mf4_m(vbool64_t mask, int16_t *base, vint16mf4_t value, siz
 // CHECK-RV64-LABEL: @test_vse16_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_i16mf2_m(vbool32_t mask, int16_t *base, vint16mf2_t value, size_t vl) {
@@ -691,7 +691,7 @@ void test_vse16_v_i16mf2_m(vbool32_t mask, int16_t *base, vint16mf2_t value, siz
 // CHECK-RV64-LABEL: @test_vse16_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_i16m1_m(vbool16_t mask, int16_t *base, vint16m1_t value, size_t vl) {
@@ -702,7 +702,7 @@ void test_vse16_v_i16m1_m(vbool16_t mask, int16_t *base, vint16m1_t value, size_
 // CHECK-RV64-LABEL: @test_vse16_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_i16m2_m(vbool8_t mask, int16_t *base, vint16m2_t value, size_t vl) {
@@ -713,7 +713,7 @@ void test_vse16_v_i16m2_m(vbool8_t mask, int16_t *base, vint16m2_t value, size_t
 // CHECK-RV64-LABEL: @test_vse16_v_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_i16m4_m(vbool4_t mask, int16_t *base, vint16m4_t value, size_t vl) {
@@ -724,7 +724,7 @@ void test_vse16_v_i16m4_m(vbool4_t mask, int16_t *base, vint16m4_t value, size_t
 // CHECK-RV64-LABEL: @test_vse16_v_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_i16m8_m(vbool2_t mask, int16_t *base, vint16m8_t value, size_t vl) {
@@ -735,7 +735,7 @@ void test_vse16_v_i16m8_m(vbool2_t mask, int16_t *base, vint16m8_t value, size_t
 // CHECK-RV64-LABEL: @test_vse32_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_i32mf2_m(vbool64_t mask, int32_t *base, vint32mf2_t value, size_t vl) {
@@ -746,7 +746,7 @@ void test_vse32_v_i32mf2_m(vbool64_t mask, int32_t *base, vint32mf2_t value, siz
 // CHECK-RV64-LABEL: @test_vse32_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_i32m1_m(vbool32_t mask, int32_t *base, vint32m1_t value, size_t vl) {
@@ -757,7 +757,7 @@ void test_vse32_v_i32m1_m(vbool32_t mask, int32_t *base, vint32m1_t value, size_
 // CHECK-RV64-LABEL: @test_vse32_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_i32m2_m(vbool16_t mask, int32_t *base, vint32m2_t value, size_t vl) {
@@ -768,7 +768,7 @@ void test_vse32_v_i32m2_m(vbool16_t mask, int32_t *base, vint32m2_t value, size_
 // CHECK-RV64-LABEL: @test_vse32_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_i32m4_m(vbool8_t mask, int32_t *base, vint32m4_t value, size_t vl) {
@@ -779,7 +779,7 @@ void test_vse32_v_i32m4_m(vbool8_t mask, int32_t *base, vint32m4_t value, size_t
 // CHECK-RV64-LABEL: @test_vse32_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_i32m8_m(vbool4_t mask, int32_t *base, vint32m8_t value, size_t vl) {
@@ -790,7 +790,7 @@ void test_vse32_v_i32m8_m(vbool4_t mask, int32_t *base, vint32m8_t value, size_t
 // CHECK-RV64-LABEL: @test_vse64_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_i64m1_m(vbool64_t mask, int64_t *base, vint64m1_t value, size_t vl) {
@@ -801,7 +801,7 @@ void test_vse64_v_i64m1_m(vbool64_t mask, int64_t *base, vint64m1_t value, size_
 // CHECK-RV64-LABEL: @test_vse64_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_i64m2_m(vbool32_t mask, int64_t *base, vint64m2_t value, size_t vl) {
@@ -812,7 +812,7 @@ void test_vse64_v_i64m2_m(vbool32_t mask, int64_t *base, vint64m2_t value, size_
 // CHECK-RV64-LABEL: @test_vse64_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_i64m4_m(vbool16_t mask, int64_t *base, vint64m4_t value, size_t vl) {
@@ -823,7 +823,7 @@ void test_vse64_v_i64m4_m(vbool16_t mask, int64_t *base, vint64m4_t value, size_
 // CHECK-RV64-LABEL: @test_vse64_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_i64m8_m(vbool8_t mask, int64_t *base, vint64m8_t value, size_t vl) {
@@ -834,7 +834,7 @@ void test_vse64_v_i64m8_m(vbool8_t mask, int64_t *base, vint64m8_t value, size_t
 // CHECK-RV64-LABEL: @test_vse8_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_u8mf8_m(vbool64_t mask, uint8_t *base, vuint8mf8_t value, size_t vl) {
@@ -845,7 +845,7 @@ void test_vse8_v_u8mf8_m(vbool64_t mask, uint8_t *base, vuint8mf8_t value, size_
 // CHECK-RV64-LABEL: @test_vse8_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_u8mf4_m(vbool32_t mask, uint8_t *base, vuint8mf4_t value, size_t vl) {
@@ -856,7 +856,7 @@ void test_vse8_v_u8mf4_m(vbool32_t mask, uint8_t *base, vuint8mf4_t value, size_
 // CHECK-RV64-LABEL: @test_vse8_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_u8mf2_m(vbool16_t mask, uint8_t *base, vuint8mf2_t value, size_t vl) {
@@ -867,7 +867,7 @@ void test_vse8_v_u8mf2_m(vbool16_t mask, uint8_t *base, vuint8mf2_t value, size_
 // CHECK-RV64-LABEL: @test_vse8_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_u8m1_m(vbool8_t mask, uint8_t *base, vuint8m1_t value, size_t vl) {
@@ -878,7 +878,7 @@ void test_vse8_v_u8m1_m(vbool8_t mask, uint8_t *base, vuint8m1_t value, size_t v
 // CHECK-RV64-LABEL: @test_vse8_v_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_u8m2_m(vbool4_t mask, uint8_t *base, vuint8m2_t value, size_t vl) {
@@ -889,7 +889,7 @@ void test_vse8_v_u8m2_m(vbool4_t mask, uint8_t *base, vuint8m2_t value, size_t v
 // CHECK-RV64-LABEL: @test_vse8_v_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_u8m4_m(vbool2_t mask, uint8_t *base, vuint8m4_t value, size_t vl) {
@@ -900,7 +900,7 @@ void test_vse8_v_u8m4_m(vbool2_t mask, uint8_t *base, vuint8m4_t value, size_t v
 // CHECK-RV64-LABEL: @test_vse8_v_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse8_v_u8m8_m(vbool1_t mask, uint8_t *base, vuint8m8_t value, size_t vl) {
@@ -911,7 +911,7 @@ void test_vse8_v_u8m8_m(vbool1_t mask, uint8_t *base, vuint8m8_t value, size_t v
 // CHECK-RV64-LABEL: @test_vse16_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_u16mf4_m(vbool64_t mask, uint16_t *base, vuint16mf4_t value, size_t vl) {
@@ -922,7 +922,7 @@ void test_vse16_v_u16mf4_m(vbool64_t mask, uint16_t *base, vuint16mf4_t value, s
 // CHECK-RV64-LABEL: @test_vse16_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_u16mf2_m(vbool32_t mask, uint16_t *base, vuint16mf2_t value, size_t vl) {
@@ -933,7 +933,7 @@ void test_vse16_v_u16mf2_m(vbool32_t mask, uint16_t *base, vuint16mf2_t value, s
 // CHECK-RV64-LABEL: @test_vse16_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_u16m1_m(vbool16_t mask, uint16_t *base, vuint16m1_t value, size_t vl) {
@@ -944,7 +944,7 @@ void test_vse16_v_u16m1_m(vbool16_t mask, uint16_t *base, vuint16m1_t value, siz
 // CHECK-RV64-LABEL: @test_vse16_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_u16m2_m(vbool8_t mask, uint16_t *base, vuint16m2_t value, size_t vl) {
@@ -955,7 +955,7 @@ void test_vse16_v_u16m2_m(vbool8_t mask, uint16_t *base, vuint16m2_t value, size
 // CHECK-RV64-LABEL: @test_vse16_v_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_u16m4_m(vbool4_t mask, uint16_t *base, vuint16m4_t value, size_t vl) {
@@ -966,7 +966,7 @@ void test_vse16_v_u16m4_m(vbool4_t mask, uint16_t *base, vuint16m4_t value, size
 // CHECK-RV64-LABEL: @test_vse16_v_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse16_v_u16m8_m(vbool2_t mask, uint16_t *base, vuint16m8_t value, size_t vl) {
@@ -977,7 +977,7 @@ void test_vse16_v_u16m8_m(vbool2_t mask, uint16_t *base, vuint16m8_t value, size
 // CHECK-RV64-LABEL: @test_vse32_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_u32mf2_m(vbool64_t mask, uint32_t *base, vuint32mf2_t value, size_t vl) {
@@ -988,7 +988,7 @@ void test_vse32_v_u32mf2_m(vbool64_t mask, uint32_t *base, vuint32mf2_t value, s
 // CHECK-RV64-LABEL: @test_vse32_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_u32m1_m(vbool32_t mask, uint32_t *base, vuint32m1_t value, size_t vl) {
@@ -999,7 +999,7 @@ void test_vse32_v_u32m1_m(vbool32_t mask, uint32_t *base, vuint32m1_t value, siz
 // CHECK-RV64-LABEL: @test_vse32_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_u32m2_m(vbool16_t mask, uint32_t *base, vuint32m2_t value, size_t vl) {
@@ -1010,7 +1010,7 @@ void test_vse32_v_u32m2_m(vbool16_t mask, uint32_t *base, vuint32m2_t value, siz
 // CHECK-RV64-LABEL: @test_vse32_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_u32m4_m(vbool8_t mask, uint32_t *base, vuint32m4_t value, size_t vl) {
@@ -1021,7 +1021,7 @@ void test_vse32_v_u32m4_m(vbool8_t mask, uint32_t *base, vuint32m4_t value, size
 // CHECK-RV64-LABEL: @test_vse32_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_u32m8_m(vbool4_t mask, uint32_t *base, vuint32m8_t value, size_t vl) {
@@ -1032,7 +1032,7 @@ void test_vse32_v_u32m8_m(vbool4_t mask, uint32_t *base, vuint32m8_t value, size
 // CHECK-RV64-LABEL: @test_vse64_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_u64m1_m(vbool64_t mask, uint64_t *base, vuint64m1_t value, size_t vl) {
@@ -1043,7 +1043,7 @@ void test_vse64_v_u64m1_m(vbool64_t mask, uint64_t *base, vuint64m1_t value, siz
 // CHECK-RV64-LABEL: @test_vse64_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_u64m2_m(vbool32_t mask, uint64_t *base, vuint64m2_t value, size_t vl) {
@@ -1054,7 +1054,7 @@ void test_vse64_v_u64m2_m(vbool32_t mask, uint64_t *base, vuint64m2_t value, siz
 // CHECK-RV64-LABEL: @test_vse64_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_u64m4_m(vbool16_t mask, uint64_t *base, vuint64m4_t value, size_t vl) {
@@ -1065,7 +1065,7 @@ void test_vse64_v_u64m4_m(vbool16_t mask, uint64_t *base, vuint64m4_t value, siz
 // CHECK-RV64-LABEL: @test_vse64_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_u64m8_m(vbool8_t mask, uint64_t *base, vuint64m8_t value, size_t vl) {
@@ -1076,7 +1076,7 @@ void test_vse64_v_u64m8_m(vbool8_t mask, uint64_t *base, vuint64m8_t value, size
 // CHECK-RV64-LABEL: @test_vse32_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv1f32.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv1f32.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_f32mf2_m(vbool64_t mask, float *base, vfloat32mf2_t value, size_t vl) {
@@ -1087,7 +1087,7 @@ void test_vse32_v_f32mf2_m(vbool64_t mask, float *base, vfloat32mf2_t value, siz
 // CHECK-RV64-LABEL: @test_vse32_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv2f32.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv2f32.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_f32m1_m(vbool32_t mask, float *base, vfloat32m1_t value, size_t vl) {
@@ -1098,7 +1098,7 @@ void test_vse32_v_f32m1_m(vbool32_t mask, float *base, vfloat32m1_t value, size_
 // CHECK-RV64-LABEL: @test_vse32_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv4f32.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv4f32.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_f32m2_m(vbool16_t mask, float *base, vfloat32m2_t value, size_t vl) {
@@ -1109,7 +1109,7 @@ void test_vse32_v_f32m2_m(vbool16_t mask, float *base, vfloat32m2_t value, size_
 // CHECK-RV64-LABEL: @test_vse32_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv8f32.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv8f32.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_f32m4_m(vbool8_t mask, float *base, vfloat32m4_t value, size_t vl) {
@@ -1120,7 +1120,7 @@ void test_vse32_v_f32m4_m(vbool8_t mask, float *base, vfloat32m4_t value, size_t
 // CHECK-RV64-LABEL: @test_vse32_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv16f32.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv16f32.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse32_v_f32m8_m(vbool4_t mask, float *base, vfloat32m8_t value, size_t vl) {
@@ -1131,7 +1131,7 @@ void test_vse32_v_f32m8_m(vbool4_t mask, float *base, vfloat32m8_t value, size_t
 // CHECK-RV64-LABEL: @test_vse64_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv1f64.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv1f64.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_f64m1_m(vbool64_t mask, double *base, vfloat64m1_t value, size_t vl) {
@@ -1142,7 +1142,7 @@ void test_vse64_v_f64m1_m(vbool64_t mask, double *base, vfloat64m1_t value, size
 // CHECK-RV64-LABEL: @test_vse64_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv2f64.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv2f64.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_f64m2_m(vbool32_t mask, double *base, vfloat64m2_t value, size_t vl) {
@@ -1153,7 +1153,7 @@ void test_vse64_v_f64m2_m(vbool32_t mask, double *base, vfloat64m2_t value, size
 // CHECK-RV64-LABEL: @test_vse64_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv4f64.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv4f64.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_f64m4_m(vbool16_t mask, double *base, vfloat64m4_t value, size_t vl) {
@@ -1164,7 +1164,7 @@ void test_vse64_v_f64m4_m(vbool16_t mask, double *base, vfloat64m4_t value, size
 // CHECK-RV64-LABEL: @test_vse64_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv8f64.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv8f64.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse64_v_f64m8_m(vbool8_t mask, double *base, vfloat64m8_t value, size_t vl) {
@@ -1175,7 +1175,7 @@ void test_vse64_v_f64m8_m(vbool8_t mask, double *base, vfloat64m8_t value, size_
 // CHECK-RV64-LABEL: @test_vse1_v_b1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i1>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse1.nxv64i1.i64(<vscale x 64 x i1> [[VALUE:%.*]], <vscale x 64 x i1>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse1.nxv64i1.i64(<vscale x 64 x i1> [[VALUE:%.*]], <vscale x 64 x i1>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse1_v_b1(uint8_t *base, vbool1_t value, size_t vl) {
@@ -1186,7 +1186,7 @@ void test_vse1_v_b1(uint8_t *base, vbool1_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse1_v_b2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i1>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse1.nxv32i1.i64(<vscale x 32 x i1> [[VALUE:%.*]], <vscale x 32 x i1>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse1.nxv32i1.i64(<vscale x 32 x i1> [[VALUE:%.*]], <vscale x 32 x i1>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse1_v_b2(uint8_t *base, vbool2_t value, size_t vl) {
@@ -1197,7 +1197,7 @@ void test_vse1_v_b2(uint8_t *base, vbool2_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse1_v_b4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i1>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse1.nxv16i1.i64(<vscale x 16 x i1> [[VALUE:%.*]], <vscale x 16 x i1>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse1.nxv16i1.i64(<vscale x 16 x i1> [[VALUE:%.*]], <vscale x 16 x i1>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse1_v_b4(uint8_t *base, vbool4_t value, size_t vl) {
@@ -1208,7 +1208,7 @@ void test_vse1_v_b4(uint8_t *base, vbool4_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse1_v_b8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i1>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse1.nxv8i1.i64(<vscale x 8 x i1> [[VALUE:%.*]], <vscale x 8 x i1>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse1.nxv8i1.i64(<vscale x 8 x i1> [[VALUE:%.*]], <vscale x 8 x i1>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse1_v_b8(uint8_t *base, vbool8_t value, size_t vl) {
@@ -1219,7 +1219,7 @@ void test_vse1_v_b8(uint8_t *base, vbool8_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse1_v_b16(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i1>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse1.nxv4i1.i64(<vscale x 4 x i1> [[VALUE:%.*]], <vscale x 4 x i1>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse1.nxv4i1.i64(<vscale x 4 x i1> [[VALUE:%.*]], <vscale x 4 x i1>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse1_v_b16(uint8_t *base, vbool16_t value, size_t vl) {
@@ -1230,7 +1230,7 @@ void test_vse1_v_b16(uint8_t *base, vbool16_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse1_v_b32(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i1>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse1.nxv2i1.i64(<vscale x 2 x i1> [[VALUE:%.*]], <vscale x 2 x i1>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse1.nxv2i1.i64(<vscale x 2 x i1> [[VALUE:%.*]], <vscale x 2 x i1>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse1_v_b32(uint8_t *base, vbool32_t value, size_t vl) {
@@ -1241,7 +1241,7 @@ void test_vse1_v_b32(uint8_t *base, vbool32_t value, size_t vl) {
 // CHECK-RV64-LABEL: @test_vse1_v_b64(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i1>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vse1.nxv1i1.i64(<vscale x 1 x i1> [[VALUE:%.*]], <vscale x 1 x i1>* [[TMP0]], i64 [[VL:%.*]]) #[[ATTR11]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse1.nxv1i1.i64(<vscale x 1 x i1> [[VALUE:%.*]], <vscale x 1 x i1>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vse1_v_b64(uint8_t *base, vbool64_t value, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsext.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsext.c
index b5ca5da..64f8ab7 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsext.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsext.c
@@ -287,7 +287,7 @@ vint64m8_t test_vsext_vf2_i64m8(vint32m4_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vsext_vf2_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsext.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsext.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vsext_vf2_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -298,7 +298,7 @@ vint16mf4_t test_vsext_vf2_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf2_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsext.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsext.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vsext_vf2_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -309,7 +309,7 @@ vint16mf2_t test_vsext_vf2_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf2_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsext.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsext.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vsext_vf2_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -320,7 +320,7 @@ vint16m1_t test_vsext_vf2_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf2_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsext.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsext.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vsext_vf2_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -331,7 +331,7 @@ vint16m2_t test_vsext_vf2_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf2_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsext.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsext.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vsext_vf2_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -342,7 +342,7 @@ vint16m4_t test_vsext_vf2_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf2_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsext.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsext.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vsext_vf2_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -353,7 +353,7 @@ vint16m8_t test_vsext_vf2_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf4_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsext.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsext.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vsext_vf4_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -364,7 +364,7 @@ vint32mf2_t test_vsext_vf4_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf4_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsext.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsext.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vsext_vf4_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -375,7 +375,7 @@ vint32m1_t test_vsext_vf4_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf4_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsext.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsext.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vsext_vf4_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -386,7 +386,7 @@ vint32m2_t test_vsext_vf4_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf4_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsext.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsext.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vsext_vf4_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -397,7 +397,7 @@ vint32m4_t test_vsext_vf4_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf4_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsext.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsext.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vsext_vf4_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -408,7 +408,7 @@ vint32m8_t test_vsext_vf4_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf8_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsext.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsext.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vsext_vf8_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -419,7 +419,7 @@ vint64m1_t test_vsext_vf8_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf8_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsext.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsext.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vsext_vf8_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -430,7 +430,7 @@ vint64m2_t test_vsext_vf8_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf8_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsext.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsext.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vsext_vf8_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -441,7 +441,7 @@ vint64m4_t test_vsext_vf8_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf8_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsext.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsext.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vsext_vf8_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -452,7 +452,7 @@ vint64m8_t test_vsext_vf8_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf2_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsext.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsext.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vsext_vf2_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -463,7 +463,7 @@ vint32mf2_t test_vsext_vf2_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf2_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsext.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsext.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vsext_vf2_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -474,7 +474,7 @@ vint32m1_t test_vsext_vf2_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf2_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsext.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsext.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vsext_vf2_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -485,7 +485,7 @@ vint32m2_t test_vsext_vf2_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf2_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsext.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsext.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vsext_vf2_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -496,7 +496,7 @@ vint32m4_t test_vsext_vf2_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf2_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsext.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsext.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vsext_vf2_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -507,7 +507,7 @@ vint32m8_t test_vsext_vf2_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf4_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsext.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsext.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vsext_vf4_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -518,7 +518,7 @@ vint64m1_t test_vsext_vf4_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf4_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsext.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsext.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vsext_vf4_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -529,7 +529,7 @@ vint64m2_t test_vsext_vf4_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf4_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsext.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsext.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vsext_vf4_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -540,7 +540,7 @@ vint64m4_t test_vsext_vf4_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf4_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsext.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsext.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vsext_vf4_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -551,7 +551,7 @@ vint64m8_t test_vsext_vf4_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf2_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsext.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsext.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vsext_vf2_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -562,7 +562,7 @@ vint64m1_t test_vsext_vf2_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf2_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsext.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsext.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vsext_vf2_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -573,7 +573,7 @@ vint64m2_t test_vsext_vf2_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf2_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsext.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsext.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vsext_vf2_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -584,7 +584,7 @@ vint64m4_t test_vsext_vf2_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsext_vf2_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsext.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsext.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vsext_vf2_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vslide1down.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vslide1down.c
index d424e2f..1d9d3c5 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vslide1down.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vslide1down.c
@@ -468,7 +468,7 @@ vuint64m8_t test_vslide1down_vx_u64m8(vuint64m8_t src, uint64_t value,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vslide1down.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vslide1down.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vslide1down_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
@@ -480,7 +480,7 @@ vint8mf8_t test_vslide1down_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vslide1down.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vslide1down.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vslide1down_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
@@ -492,7 +492,7 @@ vint8mf4_t test_vslide1down_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vslide1down.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vslide1down.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vslide1down_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
@@ -504,7 +504,7 @@ vint8mf2_t test_vslide1down_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vslide1down.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vslide1down.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vslide1down_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
@@ -515,7 +515,7 @@ vint8m1_t test_vslide1down_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vslide1down.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vslide1down.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vslide1down_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
@@ -526,7 +526,7 @@ vint8m2_t test_vslide1down_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vslide1down.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vslide1down.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vslide1down_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
@@ -537,7 +537,7 @@ vint8m4_t test_vslide1down_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vslide1down.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vslide1down.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vslide1down_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
@@ -548,7 +548,7 @@ vint8m8_t test_vslide1down_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vslide1down.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vslide1down.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vslide1down_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -560,7 +560,7 @@ vint16mf4_t test_vslide1down_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vslide1down.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vslide1down.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vslide1down_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -572,7 +572,7 @@ vint16mf2_t test_vslide1down_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vslide1down.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vslide1down.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vslide1down_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -584,7 +584,7 @@ vint16m1_t test_vslide1down_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vslide1down.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vslide1down.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vslide1down_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -596,7 +596,7 @@ vint16m2_t test_vslide1down_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vslide1down.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vslide1down.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vslide1down_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -608,7 +608,7 @@ vint16m4_t test_vslide1down_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vslide1down.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vslide1down.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vslide1down_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -620,7 +620,7 @@ vint16m8_t test_vslide1down_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vslide1down.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vslide1down.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vslide1down_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -632,7 +632,7 @@ vint32mf2_t test_vslide1down_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vslide1down.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vslide1down.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vslide1down_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -644,7 +644,7 @@ vint32m1_t test_vslide1down_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vslide1down.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vslide1down.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vslide1down_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -656,7 +656,7 @@ vint32m2_t test_vslide1down_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vslide1down.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vslide1down.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vslide1down_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -668,7 +668,7 @@ vint32m4_t test_vslide1down_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vslide1down.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vslide1down.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vslide1down_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -680,7 +680,7 @@ vint32m8_t test_vslide1down_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vslide1down.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vslide1down.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vslide1down_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -692,7 +692,7 @@ vint64m1_t test_vslide1down_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vslide1down.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vslide1down.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vslide1down_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -704,7 +704,7 @@ vint64m2_t test_vslide1down_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vslide1down.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vslide1down.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vslide1down_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -716,7 +716,7 @@ vint64m4_t test_vslide1down_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vslide1down.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vslide1down.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vslide1down_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -728,7 +728,7 @@ vint64m8_t test_vslide1down_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vslide1down.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vslide1down.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vslide1down_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
@@ -740,7 +740,7 @@ vuint8mf8_t test_vslide1down_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vslide1down.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vslide1down.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vslide1down_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
@@ -752,7 +752,7 @@ vuint8mf4_t test_vslide1down_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vslide1down.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vslide1down.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vslide1down_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
@@ -764,7 +764,7 @@ vuint8mf2_t test_vslide1down_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vslide1down.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vslide1down.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vslide1down_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
@@ -776,7 +776,7 @@ vuint8m1_t test_vslide1down_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vslide1down.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vslide1down.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vslide1down_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
@@ -788,7 +788,7 @@ vuint8m2_t test_vslide1down_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vslide1down.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vslide1down.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vslide1down_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
@@ -800,7 +800,7 @@ vuint8m4_t test_vslide1down_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vslide1down.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vslide1down.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vslide1down_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
@@ -812,7 +812,7 @@ vuint8m8_t test_vslide1down_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vslide1down.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vslide1down.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vslide1down_vx_u16mf4_m(vbool64_t mask,
@@ -825,7 +825,7 @@ vuint16mf4_t test_vslide1down_vx_u16mf4_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vslide1down.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vslide1down.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vslide1down_vx_u16mf2_m(vbool32_t mask,
@@ -838,7 +838,7 @@ vuint16mf2_t test_vslide1down_vx_u16mf2_m(vbool32_t mask,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vslide1down.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vslide1down.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vslide1down_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -850,7 +850,7 @@ vuint16m1_t test_vslide1down_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vslide1down.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vslide1down.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vslide1down_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -862,7 +862,7 @@ vuint16m2_t test_vslide1down_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vslide1down.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vslide1down.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vslide1down_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -874,7 +874,7 @@ vuint16m4_t test_vslide1down_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vslide1down.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vslide1down.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vslide1down_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -886,7 +886,7 @@ vuint16m8_t test_vslide1down_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vslide1down.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vslide1down.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vslide1down_vx_u32mf2_m(vbool64_t mask,
@@ -899,7 +899,7 @@ vuint32mf2_t test_vslide1down_vx_u32mf2_m(vbool64_t mask,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vslide1down.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vslide1down.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vslide1down_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -911,7 +911,7 @@ vuint32m1_t test_vslide1down_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vslide1down.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vslide1down.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vslide1down_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -923,7 +923,7 @@ vuint32m2_t test_vslide1down_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vslide1down.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vslide1down.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vslide1down_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -935,7 +935,7 @@ vuint32m4_t test_vslide1down_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vslide1down.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vslide1down.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vslide1down_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -947,7 +947,7 @@ vuint32m8_t test_vslide1down_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vslide1down.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vslide1down.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vslide1down_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -959,7 +959,7 @@ vuint64m1_t test_vslide1down_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vslide1down.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vslide1down.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vslide1down_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -971,7 +971,7 @@ vuint64m2_t test_vslide1down_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vslide1down.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vslide1down.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vslide1down_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -983,7 +983,7 @@ vuint64m4_t test_vslide1down_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1down_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vslide1down.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vslide1down.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vslide1down_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vslide1up.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vslide1up.c
index 478b061..051d976 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vslide1up.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vslide1up.c
@@ -465,7 +465,7 @@ vuint64m8_t test_vslide1up_vx_u64m8(vuint64m8_t src, uint64_t value,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vslide1up.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vslide1up.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vslide1up_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
@@ -476,7 +476,7 @@ vint8mf8_t test_vslide1up_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vslide1up.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vslide1up.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vslide1up_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
@@ -487,7 +487,7 @@ vint8mf4_t test_vslide1up_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vslide1up.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vslide1up.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vslide1up_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
@@ -498,7 +498,7 @@ vint8mf2_t test_vslide1up_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vslide1up.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vslide1up.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vslide1up_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
@@ -509,7 +509,7 @@ vint8m1_t test_vslide1up_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vslide1up.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vslide1up.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vslide1up_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
@@ -520,7 +520,7 @@ vint8m2_t test_vslide1up_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vslide1up.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vslide1up.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vslide1up_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
@@ -531,7 +531,7 @@ vint8m4_t test_vslide1up_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vslide1up.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vslide1up.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vslide1up_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
@@ -542,7 +542,7 @@ vint8m8_t test_vslide1up_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vslide1up.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vslide1up.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vslide1up_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -554,7 +554,7 @@ vint16mf4_t test_vslide1up_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vslide1up.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vslide1up.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vslide1up_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -566,7 +566,7 @@ vint16mf2_t test_vslide1up_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vslide1up.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vslide1up.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vslide1up_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -577,7 +577,7 @@ vint16m1_t test_vslide1up_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vslide1up.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vslide1up.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vslide1up_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -588,7 +588,7 @@ vint16m2_t test_vslide1up_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vslide1up.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vslide1up.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vslide1up_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -599,7 +599,7 @@ vint16m4_t test_vslide1up_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vslide1up.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vslide1up.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vslide1up_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -610,7 +610,7 @@ vint16m8_t test_vslide1up_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vslide1up.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vslide1up.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vslide1up_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -622,7 +622,7 @@ vint32mf2_t test_vslide1up_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vslide1up.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vslide1up.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vslide1up_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -633,7 +633,7 @@ vint32m1_t test_vslide1up_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vslide1up.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vslide1up.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vslide1up_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -644,7 +644,7 @@ vint32m2_t test_vslide1up_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vslide1up.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vslide1up.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vslide1up_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -655,7 +655,7 @@ vint32m4_t test_vslide1up_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vslide1up.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vslide1up.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vslide1up_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -666,7 +666,7 @@ vint32m8_t test_vslide1up_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vslide1up.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vslide1up.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vslide1up_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -677,7 +677,7 @@ vint64m1_t test_vslide1up_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vslide1up.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vslide1up.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vslide1up_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -688,7 +688,7 @@ vint64m2_t test_vslide1up_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vslide1up.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vslide1up.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vslide1up_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -699,7 +699,7 @@ vint64m4_t test_vslide1up_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vslide1up.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vslide1up.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vslide1up_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -710,7 +710,7 @@ vint64m8_t test_vslide1up_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vslide1up.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vslide1up.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vslide1up_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
@@ -722,7 +722,7 @@ vuint8mf8_t test_vslide1up_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vslide1up.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vslide1up.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vslide1up_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
@@ -734,7 +734,7 @@ vuint8mf4_t test_vslide1up_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vslide1up.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vslide1up.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vslide1up_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
@@ -746,7 +746,7 @@ vuint8mf2_t test_vslide1up_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vslide1up.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vslide1up.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vslide1up_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
@@ -757,7 +757,7 @@ vuint8m1_t test_vslide1up_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vslide1up.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vslide1up.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vslide1up_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
@@ -768,7 +768,7 @@ vuint8m2_t test_vslide1up_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vslide1up.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vslide1up.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vslide1up_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
@@ -779,7 +779,7 @@ vuint8m4_t test_vslide1up_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vslide1up.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vslide1up.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[SRC:%.*]], i8 [[VALUE:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vslide1up_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
@@ -790,7 +790,7 @@ vuint8m8_t test_vslide1up_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vslide1up.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vslide1up.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vslide1up_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -802,7 +802,7 @@ vuint16mf4_t test_vslide1up_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vslide1up.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vslide1up.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vslide1up_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -814,7 +814,7 @@ vuint16mf2_t test_vslide1up_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vslide1up.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vslide1up.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vslide1up_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -826,7 +826,7 @@ vuint16m1_t test_vslide1up_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vslide1up.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vslide1up.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vslide1up_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -838,7 +838,7 @@ vuint16m2_t test_vslide1up_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vslide1up.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vslide1up.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vslide1up_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -850,7 +850,7 @@ vuint16m4_t test_vslide1up_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vslide1up.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vslide1up.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[SRC:%.*]], i16 [[VALUE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vslide1up_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -862,7 +862,7 @@ vuint16m8_t test_vslide1up_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vslide1up.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vslide1up.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vslide1up_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -874,7 +874,7 @@ vuint32mf2_t test_vslide1up_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vslide1up.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vslide1up.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vslide1up_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -886,7 +886,7 @@ vuint32m1_t test_vslide1up_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vslide1up.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vslide1up.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vslide1up_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -898,7 +898,7 @@ vuint32m2_t test_vslide1up_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vslide1up.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vslide1up.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vslide1up_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -910,7 +910,7 @@ vuint32m4_t test_vslide1up_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vslide1up.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vslide1up.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[SRC:%.*]], i32 [[VALUE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vslide1up_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -922,7 +922,7 @@ vuint32m8_t test_vslide1up_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vslide1up.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vslide1up.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vslide1up_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -934,7 +934,7 @@ vuint64m1_t test_vslide1up_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vslide1up.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vslide1up.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vslide1up_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -946,7 +946,7 @@ vuint64m2_t test_vslide1up_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vslide1up.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vslide1up.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vslide1up_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -958,7 +958,7 @@ vuint64m4_t test_vslide1up_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vslide1up_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vslide1up.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vslide1up.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[SRC:%.*]], i64 [[VALUE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vslide1up_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vslidedown.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vslidedown.c
index a3793bf..b5e0603 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vslidedown.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vslidedown.c
@@ -591,7 +591,7 @@ vfloat64m8_t test_vslidedown_vx_f64m8(vfloat64m8_t dst, vfloat64m8_t src,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vslidedown.mask.nxv1i8.i64(<vscale x 1 x i8> [[DST:%.*]], <vscale x 1 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vslidedown.mask.nxv1i8.i64(<vscale x 1 x i8> [[DST:%.*]], <vscale x 1 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vslidedown_vx_i8mf8_m(vbool64_t mask, vint8mf8_t dst,
@@ -603,7 +603,7 @@ vint8mf8_t test_vslidedown_vx_i8mf8_m(vbool64_t mask, vint8mf8_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vslidedown.mask.nxv2i8.i64(<vscale x 2 x i8> [[DST:%.*]], <vscale x 2 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vslidedown.mask.nxv2i8.i64(<vscale x 2 x i8> [[DST:%.*]], <vscale x 2 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vslidedown_vx_i8mf4_m(vbool32_t mask, vint8mf4_t dst,
@@ -615,7 +615,7 @@ vint8mf4_t test_vslidedown_vx_i8mf4_m(vbool32_t mask, vint8mf4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vslidedown.mask.nxv4i8.i64(<vscale x 4 x i8> [[DST:%.*]], <vscale x 4 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vslidedown.mask.nxv4i8.i64(<vscale x 4 x i8> [[DST:%.*]], <vscale x 4 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vslidedown_vx_i8mf2_m(vbool16_t mask, vint8mf2_t dst,
@@ -627,7 +627,7 @@ vint8mf2_t test_vslidedown_vx_i8mf2_m(vbool16_t mask, vint8mf2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vslidedown.mask.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vslidedown.mask.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vslidedown_vx_i8m1_m(vbool8_t mask, vint8m1_t dst, vint8m1_t src,
@@ -638,7 +638,7 @@ vint8m1_t test_vslidedown_vx_i8m1_m(vbool8_t mask, vint8m1_t dst, vint8m1_t src,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vslidedown.mask.nxv16i8.i64(<vscale x 16 x i8> [[DST:%.*]], <vscale x 16 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vslidedown.mask.nxv16i8.i64(<vscale x 16 x i8> [[DST:%.*]], <vscale x 16 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vslidedown_vx_i8m2_m(vbool4_t mask, vint8m2_t dst, vint8m2_t src,
@@ -649,7 +649,7 @@ vint8m2_t test_vslidedown_vx_i8m2_m(vbool4_t mask, vint8m2_t dst, vint8m2_t src,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vslidedown.mask.nxv32i8.i64(<vscale x 32 x i8> [[DST:%.*]], <vscale x 32 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vslidedown.mask.nxv32i8.i64(<vscale x 32 x i8> [[DST:%.*]], <vscale x 32 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vslidedown_vx_i8m4_m(vbool2_t mask, vint8m4_t dst, vint8m4_t src,
@@ -660,7 +660,7 @@ vint8m4_t test_vslidedown_vx_i8m4_m(vbool2_t mask, vint8m4_t dst, vint8m4_t src,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vslidedown.mask.nxv64i8.i64(<vscale x 64 x i8> [[DST:%.*]], <vscale x 64 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vslidedown.mask.nxv64i8.i64(<vscale x 64 x i8> [[DST:%.*]], <vscale x 64 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vslidedown_vx_i8m8_m(vbool1_t mask, vint8m8_t dst, vint8m8_t src,
@@ -671,7 +671,7 @@ vint8m8_t test_vslidedown_vx_i8m8_m(vbool1_t mask, vint8m8_t dst, vint8m8_t src,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vslidedown.mask.nxv1i16.i64(<vscale x 1 x i16> [[DST:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vslidedown.mask.nxv1i16.i64(<vscale x 1 x i16> [[DST:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vslidedown_vx_i16mf4_m(vbool64_t mask, vint16mf4_t dst,
@@ -683,7 +683,7 @@ vint16mf4_t test_vslidedown_vx_i16mf4_m(vbool64_t mask, vint16mf4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vslidedown.mask.nxv2i16.i64(<vscale x 2 x i16> [[DST:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vslidedown.mask.nxv2i16.i64(<vscale x 2 x i16> [[DST:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vslidedown_vx_i16mf2_m(vbool32_t mask, vint16mf2_t dst,
@@ -695,7 +695,7 @@ vint16mf2_t test_vslidedown_vx_i16mf2_m(vbool32_t mask, vint16mf2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vslidedown.mask.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vslidedown.mask.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vslidedown_vx_i16m1_m(vbool16_t mask, vint16m1_t dst,
@@ -707,7 +707,7 @@ vint16m1_t test_vslidedown_vx_i16m1_m(vbool16_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vslidedown.mask.nxv8i16.i64(<vscale x 8 x i16> [[DST:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vslidedown.mask.nxv8i16.i64(<vscale x 8 x i16> [[DST:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vslidedown_vx_i16m2_m(vbool8_t mask, vint16m2_t dst,
@@ -719,7 +719,7 @@ vint16m2_t test_vslidedown_vx_i16m2_m(vbool8_t mask, vint16m2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vslidedown.mask.nxv16i16.i64(<vscale x 16 x i16> [[DST:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vslidedown.mask.nxv16i16.i64(<vscale x 16 x i16> [[DST:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vslidedown_vx_i16m4_m(vbool4_t mask, vint16m4_t dst,
@@ -731,7 +731,7 @@ vint16m4_t test_vslidedown_vx_i16m4_m(vbool4_t mask, vint16m4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vslidedown.mask.nxv32i16.i64(<vscale x 32 x i16> [[DST:%.*]], <vscale x 32 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vslidedown.mask.nxv32i16.i64(<vscale x 32 x i16> [[DST:%.*]], <vscale x 32 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vslidedown_vx_i16m8_m(vbool2_t mask, vint16m8_t dst,
@@ -743,7 +743,7 @@ vint16m8_t test_vslidedown_vx_i16m8_m(vbool2_t mask, vint16m8_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vslidedown.mask.nxv1i32.i64(<vscale x 1 x i32> [[DST:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vslidedown.mask.nxv1i32.i64(<vscale x 1 x i32> [[DST:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vslidedown_vx_i32mf2_m(vbool64_t mask, vint32mf2_t dst,
@@ -755,7 +755,7 @@ vint32mf2_t test_vslidedown_vx_i32mf2_m(vbool64_t mask, vint32mf2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vslidedown.mask.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vslidedown.mask.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vslidedown_vx_i32m1_m(vbool32_t mask, vint32m1_t dst,
@@ -767,7 +767,7 @@ vint32m1_t test_vslidedown_vx_i32m1_m(vbool32_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vslidedown.mask.nxv4i32.i64(<vscale x 4 x i32> [[DST:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vslidedown.mask.nxv4i32.i64(<vscale x 4 x i32> [[DST:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vslidedown_vx_i32m2_m(vbool16_t mask, vint32m2_t dst,
@@ -779,7 +779,7 @@ vint32m2_t test_vslidedown_vx_i32m2_m(vbool16_t mask, vint32m2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vslidedown.mask.nxv8i32.i64(<vscale x 8 x i32> [[DST:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vslidedown.mask.nxv8i32.i64(<vscale x 8 x i32> [[DST:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vslidedown_vx_i32m4_m(vbool8_t mask, vint32m4_t dst,
@@ -791,7 +791,7 @@ vint32m4_t test_vslidedown_vx_i32m4_m(vbool8_t mask, vint32m4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vslidedown.mask.nxv16i32.i64(<vscale x 16 x i32> [[DST:%.*]], <vscale x 16 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vslidedown.mask.nxv16i32.i64(<vscale x 16 x i32> [[DST:%.*]], <vscale x 16 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vslidedown_vx_i32m8_m(vbool4_t mask, vint32m8_t dst,
@@ -803,7 +803,7 @@ vint32m8_t test_vslidedown_vx_i32m8_m(vbool4_t mask, vint32m8_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vslidedown.mask.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vslidedown.mask.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vslidedown_vx_i64m1_m(vbool64_t mask, vint64m1_t dst,
@@ -815,7 +815,7 @@ vint64m1_t test_vslidedown_vx_i64m1_m(vbool64_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vslidedown.mask.nxv2i64.i64(<vscale x 2 x i64> [[DST:%.*]], <vscale x 2 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vslidedown.mask.nxv2i64.i64(<vscale x 2 x i64> [[DST:%.*]], <vscale x 2 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vslidedown_vx_i64m2_m(vbool32_t mask, vint64m2_t dst,
@@ -827,7 +827,7 @@ vint64m2_t test_vslidedown_vx_i64m2_m(vbool32_t mask, vint64m2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vslidedown.mask.nxv4i64.i64(<vscale x 4 x i64> [[DST:%.*]], <vscale x 4 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vslidedown.mask.nxv4i64.i64(<vscale x 4 x i64> [[DST:%.*]], <vscale x 4 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vslidedown_vx_i64m4_m(vbool16_t mask, vint64m4_t dst,
@@ -839,7 +839,7 @@ vint64m4_t test_vslidedown_vx_i64m4_m(vbool16_t mask, vint64m4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vslidedown.mask.nxv8i64.i64(<vscale x 8 x i64> [[DST:%.*]], <vscale x 8 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vslidedown.mask.nxv8i64.i64(<vscale x 8 x i64> [[DST:%.*]], <vscale x 8 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vslidedown_vx_i64m8_m(vbool8_t mask, vint64m8_t dst,
@@ -851,7 +851,7 @@ vint64m8_t test_vslidedown_vx_i64m8_m(vbool8_t mask, vint64m8_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vslidedown.mask.nxv1i8.i64(<vscale x 1 x i8> [[DST:%.*]], <vscale x 1 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vslidedown.mask.nxv1i8.i64(<vscale x 1 x i8> [[DST:%.*]], <vscale x 1 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vslidedown_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t dst,
@@ -863,7 +863,7 @@ vuint8mf8_t test_vslidedown_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vslidedown.mask.nxv2i8.i64(<vscale x 2 x i8> [[DST:%.*]], <vscale x 2 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vslidedown.mask.nxv2i8.i64(<vscale x 2 x i8> [[DST:%.*]], <vscale x 2 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vslidedown_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t dst,
@@ -875,7 +875,7 @@ vuint8mf4_t test_vslidedown_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vslidedown.mask.nxv4i8.i64(<vscale x 4 x i8> [[DST:%.*]], <vscale x 4 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vslidedown.mask.nxv4i8.i64(<vscale x 4 x i8> [[DST:%.*]], <vscale x 4 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vslidedown_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t dst,
@@ -887,7 +887,7 @@ vuint8mf2_t test_vslidedown_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vslidedown.mask.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vslidedown.mask.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vslidedown_vx_u8m1_m(vbool8_t mask, vuint8m1_t dst,
@@ -898,7 +898,7 @@ vuint8m1_t test_vslidedown_vx_u8m1_m(vbool8_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vslidedown.mask.nxv16i8.i64(<vscale x 16 x i8> [[DST:%.*]], <vscale x 16 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vslidedown.mask.nxv16i8.i64(<vscale x 16 x i8> [[DST:%.*]], <vscale x 16 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vslidedown_vx_u8m2_m(vbool4_t mask, vuint8m2_t dst,
@@ -909,7 +909,7 @@ vuint8m2_t test_vslidedown_vx_u8m2_m(vbool4_t mask, vuint8m2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vslidedown.mask.nxv32i8.i64(<vscale x 32 x i8> [[DST:%.*]], <vscale x 32 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vslidedown.mask.nxv32i8.i64(<vscale x 32 x i8> [[DST:%.*]], <vscale x 32 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vslidedown_vx_u8m4_m(vbool2_t mask, vuint8m4_t dst,
@@ -920,7 +920,7 @@ vuint8m4_t test_vslidedown_vx_u8m4_m(vbool2_t mask, vuint8m4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vslidedown.mask.nxv64i8.i64(<vscale x 64 x i8> [[DST:%.*]], <vscale x 64 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vslidedown.mask.nxv64i8.i64(<vscale x 64 x i8> [[DST:%.*]], <vscale x 64 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vslidedown_vx_u8m8_m(vbool1_t mask, vuint8m8_t dst,
@@ -931,7 +931,7 @@ vuint8m8_t test_vslidedown_vx_u8m8_m(vbool1_t mask, vuint8m8_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vslidedown.mask.nxv1i16.i64(<vscale x 1 x i16> [[DST:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vslidedown.mask.nxv1i16.i64(<vscale x 1 x i16> [[DST:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vslidedown_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t dst,
@@ -943,7 +943,7 @@ vuint16mf4_t test_vslidedown_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vslidedown.mask.nxv2i16.i64(<vscale x 2 x i16> [[DST:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vslidedown.mask.nxv2i16.i64(<vscale x 2 x i16> [[DST:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vslidedown_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t dst,
@@ -955,7 +955,7 @@ vuint16mf2_t test_vslidedown_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vslidedown.mask.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vslidedown.mask.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vslidedown_vx_u16m1_m(vbool16_t mask, vuint16m1_t dst,
@@ -967,7 +967,7 @@ vuint16m1_t test_vslidedown_vx_u16m1_m(vbool16_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vslidedown.mask.nxv8i16.i64(<vscale x 8 x i16> [[DST:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vslidedown.mask.nxv8i16.i64(<vscale x 8 x i16> [[DST:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vslidedown_vx_u16m2_m(vbool8_t mask, vuint16m2_t dst,
@@ -979,7 +979,7 @@ vuint16m2_t test_vslidedown_vx_u16m2_m(vbool8_t mask, vuint16m2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vslidedown.mask.nxv16i16.i64(<vscale x 16 x i16> [[DST:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vslidedown.mask.nxv16i16.i64(<vscale x 16 x i16> [[DST:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vslidedown_vx_u16m4_m(vbool4_t mask, vuint16m4_t dst,
@@ -991,7 +991,7 @@ vuint16m4_t test_vslidedown_vx_u16m4_m(vbool4_t mask, vuint16m4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vslidedown.mask.nxv32i16.i64(<vscale x 32 x i16> [[DST:%.*]], <vscale x 32 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vslidedown.mask.nxv32i16.i64(<vscale x 32 x i16> [[DST:%.*]], <vscale x 32 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vslidedown_vx_u16m8_m(vbool2_t mask, vuint16m8_t dst,
@@ -1003,7 +1003,7 @@ vuint16m8_t test_vslidedown_vx_u16m8_m(vbool2_t mask, vuint16m8_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vslidedown.mask.nxv1i32.i64(<vscale x 1 x i32> [[DST:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vslidedown.mask.nxv1i32.i64(<vscale x 1 x i32> [[DST:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vslidedown_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t dst,
@@ -1015,7 +1015,7 @@ vuint32mf2_t test_vslidedown_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vslidedown.mask.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vslidedown.mask.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vslidedown_vx_u32m1_m(vbool32_t mask, vuint32m1_t dst,
@@ -1027,7 +1027,7 @@ vuint32m1_t test_vslidedown_vx_u32m1_m(vbool32_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vslidedown.mask.nxv4i32.i64(<vscale x 4 x i32> [[DST:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vslidedown.mask.nxv4i32.i64(<vscale x 4 x i32> [[DST:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vslidedown_vx_u32m2_m(vbool16_t mask, vuint32m2_t dst,
@@ -1039,7 +1039,7 @@ vuint32m2_t test_vslidedown_vx_u32m2_m(vbool16_t mask, vuint32m2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vslidedown.mask.nxv8i32.i64(<vscale x 8 x i32> [[DST:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vslidedown.mask.nxv8i32.i64(<vscale x 8 x i32> [[DST:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vslidedown_vx_u32m4_m(vbool8_t mask, vuint32m4_t dst,
@@ -1051,7 +1051,7 @@ vuint32m4_t test_vslidedown_vx_u32m4_m(vbool8_t mask, vuint32m4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vslidedown.mask.nxv16i32.i64(<vscale x 16 x i32> [[DST:%.*]], <vscale x 16 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vslidedown.mask.nxv16i32.i64(<vscale x 16 x i32> [[DST:%.*]], <vscale x 16 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vslidedown_vx_u32m8_m(vbool4_t mask, vuint32m8_t dst,
@@ -1063,7 +1063,7 @@ vuint32m8_t test_vslidedown_vx_u32m8_m(vbool4_t mask, vuint32m8_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vslidedown.mask.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vslidedown.mask.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vslidedown_vx_u64m1_m(vbool64_t mask, vuint64m1_t dst,
@@ -1075,7 +1075,7 @@ vuint64m1_t test_vslidedown_vx_u64m1_m(vbool64_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vslidedown.mask.nxv2i64.i64(<vscale x 2 x i64> [[DST:%.*]], <vscale x 2 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vslidedown.mask.nxv2i64.i64(<vscale x 2 x i64> [[DST:%.*]], <vscale x 2 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vslidedown_vx_u64m2_m(vbool32_t mask, vuint64m2_t dst,
@@ -1087,7 +1087,7 @@ vuint64m2_t test_vslidedown_vx_u64m2_m(vbool32_t mask, vuint64m2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vslidedown.mask.nxv4i64.i64(<vscale x 4 x i64> [[DST:%.*]], <vscale x 4 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vslidedown.mask.nxv4i64.i64(<vscale x 4 x i64> [[DST:%.*]], <vscale x 4 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vslidedown_vx_u64m4_m(vbool16_t mask, vuint64m4_t dst,
@@ -1099,7 +1099,7 @@ vuint64m4_t test_vslidedown_vx_u64m4_m(vbool16_t mask, vuint64m4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vslidedown.mask.nxv8i64.i64(<vscale x 8 x i64> [[DST:%.*]], <vscale x 8 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vslidedown.mask.nxv8i64.i64(<vscale x 8 x i64> [[DST:%.*]], <vscale x 8 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vslidedown_vx_u64m8_m(vbool8_t mask, vuint64m8_t dst,
@@ -1111,7 +1111,7 @@ vuint64m8_t test_vslidedown_vx_u64m8_m(vbool8_t mask, vuint64m8_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vslidedown.mask.nxv1f32.i64(<vscale x 1 x float> [[DST:%.*]], <vscale x 1 x float> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vslidedown.mask.nxv1f32.i64(<vscale x 1 x float> [[DST:%.*]], <vscale x 1 x float> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vslidedown_vx_f32mf2_m(vbool64_t mask, vfloat32mf2_t dst,
@@ -1123,7 +1123,7 @@ vfloat32mf2_t test_vslidedown_vx_f32mf2_m(vbool64_t mask, vfloat32mf2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vslidedown.mask.nxv2f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 2 x float> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vslidedown.mask.nxv2f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 2 x float> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vslidedown_vx_f32m1_m(vbool32_t mask, vfloat32m1_t dst,
@@ -1135,7 +1135,7 @@ vfloat32m1_t test_vslidedown_vx_f32m1_m(vbool32_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vslidedown.mask.nxv4f32.i64(<vscale x 4 x float> [[DST:%.*]], <vscale x 4 x float> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vslidedown.mask.nxv4f32.i64(<vscale x 4 x float> [[DST:%.*]], <vscale x 4 x float> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vslidedown_vx_f32m2_m(vbool16_t mask, vfloat32m2_t dst,
@@ -1147,7 +1147,7 @@ vfloat32m2_t test_vslidedown_vx_f32m2_m(vbool16_t mask, vfloat32m2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vslidedown.mask.nxv8f32.i64(<vscale x 8 x float> [[DST:%.*]], <vscale x 8 x float> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vslidedown.mask.nxv8f32.i64(<vscale x 8 x float> [[DST:%.*]], <vscale x 8 x float> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vslidedown_vx_f32m4_m(vbool8_t mask, vfloat32m4_t dst,
@@ -1159,7 +1159,7 @@ vfloat32m4_t test_vslidedown_vx_f32m4_m(vbool8_t mask, vfloat32m4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vslidedown.mask.nxv16f32.i64(<vscale x 16 x float> [[DST:%.*]], <vscale x 16 x float> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vslidedown.mask.nxv16f32.i64(<vscale x 16 x float> [[DST:%.*]], <vscale x 16 x float> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vslidedown_vx_f32m8_m(vbool4_t mask, vfloat32m8_t dst,
@@ -1171,7 +1171,7 @@ vfloat32m8_t test_vslidedown_vx_f32m8_m(vbool4_t mask, vfloat32m8_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vslidedown.mask.nxv1f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 1 x double> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vslidedown.mask.nxv1f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 1 x double> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vslidedown_vx_f64m1_m(vbool64_t mask, vfloat64m1_t dst,
@@ -1183,7 +1183,7 @@ vfloat64m1_t test_vslidedown_vx_f64m1_m(vbool64_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vslidedown.mask.nxv2f64.i64(<vscale x 2 x double> [[DST:%.*]], <vscale x 2 x double> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vslidedown.mask.nxv2f64.i64(<vscale x 2 x double> [[DST:%.*]], <vscale x 2 x double> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vslidedown_vx_f64m2_m(vbool32_t mask, vfloat64m2_t dst,
@@ -1195,7 +1195,7 @@ vfloat64m2_t test_vslidedown_vx_f64m2_m(vbool32_t mask, vfloat64m2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vslidedown.mask.nxv4f64.i64(<vscale x 4 x double> [[DST:%.*]], <vscale x 4 x double> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vslidedown.mask.nxv4f64.i64(<vscale x 4 x double> [[DST:%.*]], <vscale x 4 x double> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vslidedown_vx_f64m4_m(vbool16_t mask, vfloat64m4_t dst,
@@ -1207,7 +1207,7 @@ vfloat64m4_t test_vslidedown_vx_f64m4_m(vbool16_t mask, vfloat64m4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslidedown_vx_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vslidedown.mask.nxv8f64.i64(<vscale x 8 x double> [[DST:%.*]], <vscale x 8 x double> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vslidedown.mask.nxv8f64.i64(<vscale x 8 x double> [[DST:%.*]], <vscale x 8 x double> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vslidedown_vx_f64m8_m(vbool8_t mask, vfloat64m8_t dst,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vslideup.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vslideup.c
index 49d954d..0dcc739 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vslideup.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vslideup.c
@@ -591,7 +591,7 @@ vfloat64m8_t test_vslideup_vx_f64m8(vfloat64m8_t dst, vfloat64m8_t src,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vslideup.mask.nxv1i8.i64(<vscale x 1 x i8> [[DST:%.*]], <vscale x 1 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vslideup.mask.nxv1i8.i64(<vscale x 1 x i8> [[DST:%.*]], <vscale x 1 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vslideup_vx_i8mf8_m(vbool64_t mask, vint8mf8_t dst,
@@ -602,7 +602,7 @@ vint8mf8_t test_vslideup_vx_i8mf8_m(vbool64_t mask, vint8mf8_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vslideup.mask.nxv2i8.i64(<vscale x 2 x i8> [[DST:%.*]], <vscale x 2 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vslideup.mask.nxv2i8.i64(<vscale x 2 x i8> [[DST:%.*]], <vscale x 2 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vslideup_vx_i8mf4_m(vbool32_t mask, vint8mf4_t dst,
@@ -613,7 +613,7 @@ vint8mf4_t test_vslideup_vx_i8mf4_m(vbool32_t mask, vint8mf4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vslideup.mask.nxv4i8.i64(<vscale x 4 x i8> [[DST:%.*]], <vscale x 4 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vslideup.mask.nxv4i8.i64(<vscale x 4 x i8> [[DST:%.*]], <vscale x 4 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vslideup_vx_i8mf2_m(vbool16_t mask, vint8mf2_t dst,
@@ -624,7 +624,7 @@ vint8mf2_t test_vslideup_vx_i8mf2_m(vbool16_t mask, vint8mf2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vslideup.mask.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vslideup.mask.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vslideup_vx_i8m1_m(vbool8_t mask, vint8m1_t dst, vint8m1_t src,
@@ -635,7 +635,7 @@ vint8m1_t test_vslideup_vx_i8m1_m(vbool8_t mask, vint8m1_t dst, vint8m1_t src,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vslideup.mask.nxv16i8.i64(<vscale x 16 x i8> [[DST:%.*]], <vscale x 16 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vslideup.mask.nxv16i8.i64(<vscale x 16 x i8> [[DST:%.*]], <vscale x 16 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vslideup_vx_i8m2_m(vbool4_t mask, vint8m2_t dst, vint8m2_t src,
@@ -646,7 +646,7 @@ vint8m2_t test_vslideup_vx_i8m2_m(vbool4_t mask, vint8m2_t dst, vint8m2_t src,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vslideup.mask.nxv32i8.i64(<vscale x 32 x i8> [[DST:%.*]], <vscale x 32 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vslideup.mask.nxv32i8.i64(<vscale x 32 x i8> [[DST:%.*]], <vscale x 32 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vslideup_vx_i8m4_m(vbool2_t mask, vint8m4_t dst, vint8m4_t src,
@@ -657,7 +657,7 @@ vint8m4_t test_vslideup_vx_i8m4_m(vbool2_t mask, vint8m4_t dst, vint8m4_t src,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vslideup.mask.nxv64i8.i64(<vscale x 64 x i8> [[DST:%.*]], <vscale x 64 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vslideup.mask.nxv64i8.i64(<vscale x 64 x i8> [[DST:%.*]], <vscale x 64 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vslideup_vx_i8m8_m(vbool1_t mask, vint8m8_t dst, vint8m8_t src,
@@ -668,7 +668,7 @@ vint8m8_t test_vslideup_vx_i8m8_m(vbool1_t mask, vint8m8_t dst, vint8m8_t src,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vslideup.mask.nxv1i16.i64(<vscale x 1 x i16> [[DST:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vslideup.mask.nxv1i16.i64(<vscale x 1 x i16> [[DST:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vslideup_vx_i16mf4_m(vbool64_t mask, vint16mf4_t dst,
@@ -680,7 +680,7 @@ vint16mf4_t test_vslideup_vx_i16mf4_m(vbool64_t mask, vint16mf4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vslideup.mask.nxv2i16.i64(<vscale x 2 x i16> [[DST:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vslideup.mask.nxv2i16.i64(<vscale x 2 x i16> [[DST:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vslideup_vx_i16mf2_m(vbool32_t mask, vint16mf2_t dst,
@@ -692,7 +692,7 @@ vint16mf2_t test_vslideup_vx_i16mf2_m(vbool32_t mask, vint16mf2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vslideup.mask.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vslideup.mask.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vslideup_vx_i16m1_m(vbool16_t mask, vint16m1_t dst,
@@ -703,7 +703,7 @@ vint16m1_t test_vslideup_vx_i16m1_m(vbool16_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vslideup.mask.nxv8i16.i64(<vscale x 8 x i16> [[DST:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vslideup.mask.nxv8i16.i64(<vscale x 8 x i16> [[DST:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vslideup_vx_i16m2_m(vbool8_t mask, vint16m2_t dst,
@@ -714,7 +714,7 @@ vint16m2_t test_vslideup_vx_i16m2_m(vbool8_t mask, vint16m2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vslideup.mask.nxv16i16.i64(<vscale x 16 x i16> [[DST:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vslideup.mask.nxv16i16.i64(<vscale x 16 x i16> [[DST:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vslideup_vx_i16m4_m(vbool4_t mask, vint16m4_t dst,
@@ -725,7 +725,7 @@ vint16m4_t test_vslideup_vx_i16m4_m(vbool4_t mask, vint16m4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vslideup.mask.nxv32i16.i64(<vscale x 32 x i16> [[DST:%.*]], <vscale x 32 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vslideup.mask.nxv32i16.i64(<vscale x 32 x i16> [[DST:%.*]], <vscale x 32 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vslideup_vx_i16m8_m(vbool2_t mask, vint16m8_t dst,
@@ -736,7 +736,7 @@ vint16m8_t test_vslideup_vx_i16m8_m(vbool2_t mask, vint16m8_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vslideup.mask.nxv1i32.i64(<vscale x 1 x i32> [[DST:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vslideup.mask.nxv1i32.i64(<vscale x 1 x i32> [[DST:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vslideup_vx_i32mf2_m(vbool64_t mask, vint32mf2_t dst,
@@ -748,7 +748,7 @@ vint32mf2_t test_vslideup_vx_i32mf2_m(vbool64_t mask, vint32mf2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vslideup.mask.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vslideup.mask.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vslideup_vx_i32m1_m(vbool32_t mask, vint32m1_t dst,
@@ -759,7 +759,7 @@ vint32m1_t test_vslideup_vx_i32m1_m(vbool32_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vslideup.mask.nxv4i32.i64(<vscale x 4 x i32> [[DST:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vslideup.mask.nxv4i32.i64(<vscale x 4 x i32> [[DST:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vslideup_vx_i32m2_m(vbool16_t mask, vint32m2_t dst,
@@ -770,7 +770,7 @@ vint32m2_t test_vslideup_vx_i32m2_m(vbool16_t mask, vint32m2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vslideup.mask.nxv8i32.i64(<vscale x 8 x i32> [[DST:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vslideup.mask.nxv8i32.i64(<vscale x 8 x i32> [[DST:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vslideup_vx_i32m4_m(vbool8_t mask, vint32m4_t dst,
@@ -781,7 +781,7 @@ vint32m4_t test_vslideup_vx_i32m4_m(vbool8_t mask, vint32m4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vslideup.mask.nxv16i32.i64(<vscale x 16 x i32> [[DST:%.*]], <vscale x 16 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vslideup.mask.nxv16i32.i64(<vscale x 16 x i32> [[DST:%.*]], <vscale x 16 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vslideup_vx_i32m8_m(vbool4_t mask, vint32m8_t dst,
@@ -792,7 +792,7 @@ vint32m8_t test_vslideup_vx_i32m8_m(vbool4_t mask, vint32m8_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vslideup.mask.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vslideup.mask.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vslideup_vx_i64m1_m(vbool64_t mask, vint64m1_t dst,
@@ -803,7 +803,7 @@ vint64m1_t test_vslideup_vx_i64m1_m(vbool64_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vslideup.mask.nxv2i64.i64(<vscale x 2 x i64> [[DST:%.*]], <vscale x 2 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vslideup.mask.nxv2i64.i64(<vscale x 2 x i64> [[DST:%.*]], <vscale x 2 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vslideup_vx_i64m2_m(vbool32_t mask, vint64m2_t dst,
@@ -814,7 +814,7 @@ vint64m2_t test_vslideup_vx_i64m2_m(vbool32_t mask, vint64m2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vslideup.mask.nxv4i64.i64(<vscale x 4 x i64> [[DST:%.*]], <vscale x 4 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vslideup.mask.nxv4i64.i64(<vscale x 4 x i64> [[DST:%.*]], <vscale x 4 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vslideup_vx_i64m4_m(vbool16_t mask, vint64m4_t dst,
@@ -825,7 +825,7 @@ vint64m4_t test_vslideup_vx_i64m4_m(vbool16_t mask, vint64m4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vslideup.mask.nxv8i64.i64(<vscale x 8 x i64> [[DST:%.*]], <vscale x 8 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vslideup.mask.nxv8i64.i64(<vscale x 8 x i64> [[DST:%.*]], <vscale x 8 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vslideup_vx_i64m8_m(vbool8_t mask, vint64m8_t dst,
@@ -836,7 +836,7 @@ vint64m8_t test_vslideup_vx_i64m8_m(vbool8_t mask, vint64m8_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vslideup.mask.nxv1i8.i64(<vscale x 1 x i8> [[DST:%.*]], <vscale x 1 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vslideup.mask.nxv1i8.i64(<vscale x 1 x i8> [[DST:%.*]], <vscale x 1 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vslideup_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t dst,
@@ -848,7 +848,7 @@ vuint8mf8_t test_vslideup_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vslideup.mask.nxv2i8.i64(<vscale x 2 x i8> [[DST:%.*]], <vscale x 2 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vslideup.mask.nxv2i8.i64(<vscale x 2 x i8> [[DST:%.*]], <vscale x 2 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vslideup_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t dst,
@@ -860,7 +860,7 @@ vuint8mf4_t test_vslideup_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vslideup.mask.nxv4i8.i64(<vscale x 4 x i8> [[DST:%.*]], <vscale x 4 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vslideup.mask.nxv4i8.i64(<vscale x 4 x i8> [[DST:%.*]], <vscale x 4 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vslideup_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t dst,
@@ -872,7 +872,7 @@ vuint8mf2_t test_vslideup_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vslideup.mask.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vslideup.mask.nxv8i8.i64(<vscale x 8 x i8> [[DST:%.*]], <vscale x 8 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vslideup_vx_u8m1_m(vbool8_t mask, vuint8m1_t dst,
@@ -883,7 +883,7 @@ vuint8m1_t test_vslideup_vx_u8m1_m(vbool8_t mask, vuint8m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vslideup.mask.nxv16i8.i64(<vscale x 16 x i8> [[DST:%.*]], <vscale x 16 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vslideup.mask.nxv16i8.i64(<vscale x 16 x i8> [[DST:%.*]], <vscale x 16 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vslideup_vx_u8m2_m(vbool4_t mask, vuint8m2_t dst,
@@ -894,7 +894,7 @@ vuint8m2_t test_vslideup_vx_u8m2_m(vbool4_t mask, vuint8m2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vslideup.mask.nxv32i8.i64(<vscale x 32 x i8> [[DST:%.*]], <vscale x 32 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vslideup.mask.nxv32i8.i64(<vscale x 32 x i8> [[DST:%.*]], <vscale x 32 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vslideup_vx_u8m4_m(vbool2_t mask, vuint8m4_t dst,
@@ -905,7 +905,7 @@ vuint8m4_t test_vslideup_vx_u8m4_m(vbool2_t mask, vuint8m4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vslideup.mask.nxv64i8.i64(<vscale x 64 x i8> [[DST:%.*]], <vscale x 64 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vslideup.mask.nxv64i8.i64(<vscale x 64 x i8> [[DST:%.*]], <vscale x 64 x i8> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vslideup_vx_u8m8_m(vbool1_t mask, vuint8m8_t dst,
@@ -916,7 +916,7 @@ vuint8m8_t test_vslideup_vx_u8m8_m(vbool1_t mask, vuint8m8_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vslideup.mask.nxv1i16.i64(<vscale x 1 x i16> [[DST:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vslideup.mask.nxv1i16.i64(<vscale x 1 x i16> [[DST:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vslideup_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t dst,
@@ -928,7 +928,7 @@ vuint16mf4_t test_vslideup_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vslideup.mask.nxv2i16.i64(<vscale x 2 x i16> [[DST:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vslideup.mask.nxv2i16.i64(<vscale x 2 x i16> [[DST:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vslideup_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t dst,
@@ -940,7 +940,7 @@ vuint16mf2_t test_vslideup_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vslideup.mask.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vslideup.mask.nxv4i16.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vslideup_vx_u16m1_m(vbool16_t mask, vuint16m1_t dst,
@@ -952,7 +952,7 @@ vuint16m1_t test_vslideup_vx_u16m1_m(vbool16_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vslideup.mask.nxv8i16.i64(<vscale x 8 x i16> [[DST:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vslideup.mask.nxv8i16.i64(<vscale x 8 x i16> [[DST:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vslideup_vx_u16m2_m(vbool8_t mask, vuint16m2_t dst,
@@ -964,7 +964,7 @@ vuint16m2_t test_vslideup_vx_u16m2_m(vbool8_t mask, vuint16m2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vslideup.mask.nxv16i16.i64(<vscale x 16 x i16> [[DST:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vslideup.mask.nxv16i16.i64(<vscale x 16 x i16> [[DST:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vslideup_vx_u16m4_m(vbool4_t mask, vuint16m4_t dst,
@@ -976,7 +976,7 @@ vuint16m4_t test_vslideup_vx_u16m4_m(vbool4_t mask, vuint16m4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vslideup.mask.nxv32i16.i64(<vscale x 32 x i16> [[DST:%.*]], <vscale x 32 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vslideup.mask.nxv32i16.i64(<vscale x 32 x i16> [[DST:%.*]], <vscale x 32 x i16> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vslideup_vx_u16m8_m(vbool2_t mask, vuint16m8_t dst,
@@ -988,7 +988,7 @@ vuint16m8_t test_vslideup_vx_u16m8_m(vbool2_t mask, vuint16m8_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vslideup.mask.nxv1i32.i64(<vscale x 1 x i32> [[DST:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vslideup.mask.nxv1i32.i64(<vscale x 1 x i32> [[DST:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vslideup_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t dst,
@@ -1000,7 +1000,7 @@ vuint32mf2_t test_vslideup_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vslideup.mask.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vslideup.mask.nxv2i32.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vslideup_vx_u32m1_m(vbool32_t mask, vuint32m1_t dst,
@@ -1012,7 +1012,7 @@ vuint32m1_t test_vslideup_vx_u32m1_m(vbool32_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vslideup.mask.nxv4i32.i64(<vscale x 4 x i32> [[DST:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vslideup.mask.nxv4i32.i64(<vscale x 4 x i32> [[DST:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vslideup_vx_u32m2_m(vbool16_t mask, vuint32m2_t dst,
@@ -1024,7 +1024,7 @@ vuint32m2_t test_vslideup_vx_u32m2_m(vbool16_t mask, vuint32m2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vslideup.mask.nxv8i32.i64(<vscale x 8 x i32> [[DST:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vslideup.mask.nxv8i32.i64(<vscale x 8 x i32> [[DST:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vslideup_vx_u32m4_m(vbool8_t mask, vuint32m4_t dst,
@@ -1036,7 +1036,7 @@ vuint32m4_t test_vslideup_vx_u32m4_m(vbool8_t mask, vuint32m4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vslideup.mask.nxv16i32.i64(<vscale x 16 x i32> [[DST:%.*]], <vscale x 16 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vslideup.mask.nxv16i32.i64(<vscale x 16 x i32> [[DST:%.*]], <vscale x 16 x i32> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vslideup_vx_u32m8_m(vbool4_t mask, vuint32m8_t dst,
@@ -1048,7 +1048,7 @@ vuint32m8_t test_vslideup_vx_u32m8_m(vbool4_t mask, vuint32m8_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vslideup.mask.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vslideup.mask.nxv1i64.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vslideup_vx_u64m1_m(vbool64_t mask, vuint64m1_t dst,
@@ -1060,7 +1060,7 @@ vuint64m1_t test_vslideup_vx_u64m1_m(vbool64_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vslideup.mask.nxv2i64.i64(<vscale x 2 x i64> [[DST:%.*]], <vscale x 2 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vslideup.mask.nxv2i64.i64(<vscale x 2 x i64> [[DST:%.*]], <vscale x 2 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vslideup_vx_u64m2_m(vbool32_t mask, vuint64m2_t dst,
@@ -1072,7 +1072,7 @@ vuint64m2_t test_vslideup_vx_u64m2_m(vbool32_t mask, vuint64m2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vslideup.mask.nxv4i64.i64(<vscale x 4 x i64> [[DST:%.*]], <vscale x 4 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vslideup.mask.nxv4i64.i64(<vscale x 4 x i64> [[DST:%.*]], <vscale x 4 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vslideup_vx_u64m4_m(vbool16_t mask, vuint64m4_t dst,
@@ -1084,7 +1084,7 @@ vuint64m4_t test_vslideup_vx_u64m4_m(vbool16_t mask, vuint64m4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vslideup.mask.nxv8i64.i64(<vscale x 8 x i64> [[DST:%.*]], <vscale x 8 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vslideup.mask.nxv8i64.i64(<vscale x 8 x i64> [[DST:%.*]], <vscale x 8 x i64> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vslideup_vx_u64m8_m(vbool8_t mask, vuint64m8_t dst,
@@ -1096,7 +1096,7 @@ vuint64m8_t test_vslideup_vx_u64m8_m(vbool8_t mask, vuint64m8_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vslideup.mask.nxv1f32.i64(<vscale x 1 x float> [[DST:%.*]], <vscale x 1 x float> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x float> @llvm.riscv.vslideup.mask.nxv1f32.i64(<vscale x 1 x float> [[DST:%.*]], <vscale x 1 x float> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x float> [[TMP0]]
 //
 vfloat32mf2_t test_vslideup_vx_f32mf2_m(vbool64_t mask, vfloat32mf2_t dst,
@@ -1108,7 +1108,7 @@ vfloat32mf2_t test_vslideup_vx_f32mf2_m(vbool64_t mask, vfloat32mf2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vslideup.mask.nxv2f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 2 x float> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x float> @llvm.riscv.vslideup.mask.nxv2f32.i64(<vscale x 2 x float> [[DST:%.*]], <vscale x 2 x float> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x float> [[TMP0]]
 //
 vfloat32m1_t test_vslideup_vx_f32m1_m(vbool32_t mask, vfloat32m1_t dst,
@@ -1120,7 +1120,7 @@ vfloat32m1_t test_vslideup_vx_f32m1_m(vbool32_t mask, vfloat32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vslideup.mask.nxv4f32.i64(<vscale x 4 x float> [[DST:%.*]], <vscale x 4 x float> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x float> @llvm.riscv.vslideup.mask.nxv4f32.i64(<vscale x 4 x float> [[DST:%.*]], <vscale x 4 x float> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x float> [[TMP0]]
 //
 vfloat32m2_t test_vslideup_vx_f32m2_m(vbool16_t mask, vfloat32m2_t dst,
@@ -1132,7 +1132,7 @@ vfloat32m2_t test_vslideup_vx_f32m2_m(vbool16_t mask, vfloat32m2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vslideup.mask.nxv8f32.i64(<vscale x 8 x float> [[DST:%.*]], <vscale x 8 x float> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x float> @llvm.riscv.vslideup.mask.nxv8f32.i64(<vscale x 8 x float> [[DST:%.*]], <vscale x 8 x float> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x float> [[TMP0]]
 //
 vfloat32m4_t test_vslideup_vx_f32m4_m(vbool8_t mask, vfloat32m4_t dst,
@@ -1144,7 +1144,7 @@ vfloat32m4_t test_vslideup_vx_f32m4_m(vbool8_t mask, vfloat32m4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vslideup.mask.nxv16f32.i64(<vscale x 16 x float> [[DST:%.*]], <vscale x 16 x float> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x float> @llvm.riscv.vslideup.mask.nxv16f32.i64(<vscale x 16 x float> [[DST:%.*]], <vscale x 16 x float> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x float> [[TMP0]]
 //
 vfloat32m8_t test_vslideup_vx_f32m8_m(vbool4_t mask, vfloat32m8_t dst,
@@ -1156,7 +1156,7 @@ vfloat32m8_t test_vslideup_vx_f32m8_m(vbool4_t mask, vfloat32m8_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vslideup.mask.nxv1f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 1 x double> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x double> @llvm.riscv.vslideup.mask.nxv1f64.i64(<vscale x 1 x double> [[DST:%.*]], <vscale x 1 x double> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x double> [[TMP0]]
 //
 vfloat64m1_t test_vslideup_vx_f64m1_m(vbool64_t mask, vfloat64m1_t dst,
@@ -1168,7 +1168,7 @@ vfloat64m1_t test_vslideup_vx_f64m1_m(vbool64_t mask, vfloat64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vslideup.mask.nxv2f64.i64(<vscale x 2 x double> [[DST:%.*]], <vscale x 2 x double> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x double> @llvm.riscv.vslideup.mask.nxv2f64.i64(<vscale x 2 x double> [[DST:%.*]], <vscale x 2 x double> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x double> [[TMP0]]
 //
 vfloat64m2_t test_vslideup_vx_f64m2_m(vbool32_t mask, vfloat64m2_t dst,
@@ -1180,7 +1180,7 @@ vfloat64m2_t test_vslideup_vx_f64m2_m(vbool32_t mask, vfloat64m2_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vslideup.mask.nxv4f64.i64(<vscale x 4 x double> [[DST:%.*]], <vscale x 4 x double> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x double> @llvm.riscv.vslideup.mask.nxv4f64.i64(<vscale x 4 x double> [[DST:%.*]], <vscale x 4 x double> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x double> [[TMP0]]
 //
 vfloat64m4_t test_vslideup_vx_f64m4_m(vbool16_t mask, vfloat64m4_t dst,
@@ -1192,7 +1192,7 @@ vfloat64m4_t test_vslideup_vx_f64m4_m(vbool16_t mask, vfloat64m4_t dst,
 //
 // CHECK-RV64-LABEL: @test_vslideup_vx_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vslideup.mask.nxv8f64.i64(<vscale x 8 x double> [[DST:%.*]], <vscale x 8 x double> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x double> @llvm.riscv.vslideup.mask.nxv8f64.i64(<vscale x 8 x double> [[DST:%.*]], <vscale x 8 x double> [[SRC:%.*]], i64 [[OFFSET:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP0]]
 //
 vfloat64m8_t test_vslideup_vx_f64m8_m(vbool8_t mask, vfloat64m8_t dst,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsll.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsll.c
index f37c2de..532ac398 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsll.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsll.c
@@ -887,7 +887,7 @@ vuint64m8_t test_vsll_vx_u64m8(vuint64m8_t op1, size_t shift, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsll.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsll.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vsll_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, vuint8mf8_t shift, size_t vl) {
@@ -897,7 +897,7 @@ vint8mf8_t test_vsll_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsll.mask.nxv1i8.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsll.mask.nxv1i8.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vsll_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, size_t shift, size_t vl) {
@@ -907,7 +907,7 @@ vint8mf8_t test_vsll_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsll.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsll.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vsll_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, vuint8mf4_t shift, size_t vl) {
@@ -917,7 +917,7 @@ vint8mf4_t test_vsll_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsll.mask.nxv2i8.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsll.mask.nxv2i8.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vsll_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, size_t shift, size_t vl) {
@@ -927,7 +927,7 @@ vint8mf4_t test_vsll_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsll.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsll.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vsll_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, vuint8mf2_t shift, size_t vl) {
@@ -937,7 +937,7 @@ vint8mf2_t test_vsll_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsll.mask.nxv4i8.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsll.mask.nxv4i8.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vsll_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, size_t shift, size_t vl) {
@@ -947,7 +947,7 @@ vint8mf2_t test_vsll_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsll.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsll.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vsll_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, vuint8m1_t shift, size_t vl) {
@@ -957,7 +957,7 @@ vint8m1_t test_vsll_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsll.mask.nxv8i8.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsll.mask.nxv8i8.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vsll_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, size_t shift, size_t vl) {
@@ -967,7 +967,7 @@ vint8m1_t test_vsll_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsll.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsll.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vsll_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, vuint8m2_t shift, size_t vl) {
@@ -977,7 +977,7 @@ vint8m2_t test_vsll_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsll.mask.nxv16i8.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsll.mask.nxv16i8.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vsll_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, size_t shift, size_t vl) {
@@ -987,7 +987,7 @@ vint8m2_t test_vsll_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsll.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsll.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vsll_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, vuint8m4_t shift, size_t vl) {
@@ -997,7 +997,7 @@ vint8m4_t test_vsll_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsll.mask.nxv32i8.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsll.mask.nxv32i8.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vsll_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, size_t shift, size_t vl) {
@@ -1007,7 +1007,7 @@ vint8m4_t test_vsll_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsll.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsll.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vsll_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, vuint8m8_t shift, size_t vl) {
@@ -1017,7 +1017,7 @@ vint8m8_t test_vsll_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsll.mask.nxv64i8.i64.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsll.mask.nxv64i8.i64.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vsll_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, size_t shift, size_t vl) {
@@ -1027,7 +1027,7 @@ vint8m8_t test_vsll_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsll.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsll.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vsll_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, vuint16mf4_t shift, size_t vl) {
@@ -1037,7 +1037,7 @@ vint16mf4_t test_vsll_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsll.mask.nxv1i16.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsll.mask.nxv1i16.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vsll_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, size_t shift, size_t vl) {
@@ -1047,7 +1047,7 @@ vint16mf4_t test_vsll_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsll.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsll.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vsll_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, vuint16mf2_t shift, size_t vl) {
@@ -1057,7 +1057,7 @@ vint16mf2_t test_vsll_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsll.mask.nxv2i16.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsll.mask.nxv2i16.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vsll_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, size_t shift, size_t vl) {
@@ -1067,7 +1067,7 @@ vint16mf2_t test_vsll_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsll.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsll.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vsll_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, vuint16m1_t shift, size_t vl) {
@@ -1077,7 +1077,7 @@ vint16m1_t test_vsll_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsll.mask.nxv4i16.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsll.mask.nxv4i16.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vsll_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, size_t shift, size_t vl) {
@@ -1087,7 +1087,7 @@ vint16m1_t test_vsll_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsll.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsll.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vsll_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, vuint16m2_t shift, size_t vl) {
@@ -1097,7 +1097,7 @@ vint16m2_t test_vsll_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsll.mask.nxv8i16.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsll.mask.nxv8i16.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vsll_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, size_t shift, size_t vl) {
@@ -1107,7 +1107,7 @@ vint16m2_t test_vsll_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsll.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsll.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vsll_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, vuint16m4_t shift, size_t vl) {
@@ -1117,7 +1117,7 @@ vint16m4_t test_vsll_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsll.mask.nxv16i16.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsll.mask.nxv16i16.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vsll_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, size_t shift, size_t vl) {
@@ -1127,7 +1127,7 @@ vint16m4_t test_vsll_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsll.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsll.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vsll_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, vuint16m8_t shift, size_t vl) {
@@ -1137,7 +1137,7 @@ vint16m8_t test_vsll_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsll.mask.nxv32i16.i64.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsll.mask.nxv32i16.i64.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vsll_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, size_t shift, size_t vl) {
@@ -1147,7 +1147,7 @@ vint16m8_t test_vsll_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsll.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsll.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vsll_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, vuint32mf2_t shift, size_t vl) {
@@ -1157,7 +1157,7 @@ vint32mf2_t test_vsll_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsll.mask.nxv1i32.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsll.mask.nxv1i32.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vsll_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, size_t shift, size_t vl) {
@@ -1167,7 +1167,7 @@ vint32mf2_t test_vsll_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsll.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsll.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vsll_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, vuint32m1_t shift, size_t vl) {
@@ -1177,7 +1177,7 @@ vint32m1_t test_vsll_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsll.mask.nxv2i32.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsll.mask.nxv2i32.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vsll_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, size_t shift, size_t vl) {
@@ -1187,7 +1187,7 @@ vint32m1_t test_vsll_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsll.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsll.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vsll_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, vuint32m2_t shift, size_t vl) {
@@ -1197,7 +1197,7 @@ vint32m2_t test_vsll_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsll.mask.nxv4i32.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsll.mask.nxv4i32.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vsll_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, size_t shift, size_t vl) {
@@ -1207,7 +1207,7 @@ vint32m2_t test_vsll_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsll.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsll.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vsll_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, vuint32m4_t shift, size_t vl) {
@@ -1217,7 +1217,7 @@ vint32m4_t test_vsll_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsll.mask.nxv8i32.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsll.mask.nxv8i32.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vsll_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, size_t shift, size_t vl) {
@@ -1227,7 +1227,7 @@ vint32m4_t test_vsll_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsll.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsll.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vsll_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, vuint32m8_t shift, size_t vl) {
@@ -1237,7 +1237,7 @@ vint32m8_t test_vsll_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsll.mask.nxv16i32.i64.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsll.mask.nxv16i32.i64.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vsll_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, size_t shift, size_t vl) {
@@ -1247,7 +1247,7 @@ vint32m8_t test_vsll_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsll.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsll.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vsll_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, vuint64m1_t shift, size_t vl) {
@@ -1257,7 +1257,7 @@ vint64m1_t test_vsll_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsll.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsll.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vsll_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, size_t shift, size_t vl) {
@@ -1267,7 +1267,7 @@ vint64m1_t test_vsll_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsll.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsll.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vsll_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, vuint64m2_t shift, size_t vl) {
@@ -1277,7 +1277,7 @@ vint64m2_t test_vsll_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsll.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsll.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vsll_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, size_t shift, size_t vl) {
@@ -1287,7 +1287,7 @@ vint64m2_t test_vsll_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsll.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsll.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vsll_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, vuint64m4_t shift, size_t vl) {
@@ -1297,7 +1297,7 @@ vint64m4_t test_vsll_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsll.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsll.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vsll_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, size_t shift, size_t vl) {
@@ -1307,7 +1307,7 @@ vint64m4_t test_vsll_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsll.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsll.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vsll_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, vuint64m8_t shift, size_t vl) {
@@ -1317,7 +1317,7 @@ vint64m8_t test_vsll_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsll.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsll.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vsll_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, size_t shift, size_t vl) {
@@ -1327,7 +1327,7 @@ vint64m8_t test_vsll_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsll.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsll.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vsll_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, vuint8mf8_t shift, size_t vl) {
@@ -1337,7 +1337,7 @@ vuint8mf8_t test_vsll_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsll.mask.nxv1i8.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsll.mask.nxv1i8.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vsll_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, size_t shift, size_t vl) {
@@ -1347,7 +1347,7 @@ vuint8mf8_t test_vsll_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsll.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsll.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vsll_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, vuint8mf4_t shift, size_t vl) {
@@ -1357,7 +1357,7 @@ vuint8mf4_t test_vsll_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsll.mask.nxv2i8.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsll.mask.nxv2i8.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vsll_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, size_t shift, size_t vl) {
@@ -1367,7 +1367,7 @@ vuint8mf4_t test_vsll_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsll.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsll.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vsll_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, vuint8mf2_t shift, size_t vl) {
@@ -1377,7 +1377,7 @@ vuint8mf2_t test_vsll_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsll.mask.nxv4i8.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsll.mask.nxv4i8.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vsll_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, size_t shift, size_t vl) {
@@ -1387,7 +1387,7 @@ vuint8mf2_t test_vsll_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsll.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsll.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vsll_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, vuint8m1_t shift, size_t vl) {
@@ -1397,7 +1397,7 @@ vuint8m1_t test_vsll_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t o
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsll.mask.nxv8i8.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsll.mask.nxv8i8.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vsll_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, size_t shift, size_t vl) {
@@ -1407,7 +1407,7 @@ vuint8m1_t test_vsll_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t o
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsll.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsll.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vsll_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, vuint8m2_t shift, size_t vl) {
@@ -1417,7 +1417,7 @@ vuint8m2_t test_vsll_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t o
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsll.mask.nxv16i8.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsll.mask.nxv16i8.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vsll_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, size_t shift, size_t vl) {
@@ -1427,7 +1427,7 @@ vuint8m2_t test_vsll_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t o
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsll.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsll.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vsll_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, vuint8m4_t shift, size_t vl) {
@@ -1437,7 +1437,7 @@ vuint8m4_t test_vsll_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t o
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsll.mask.nxv32i8.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsll.mask.nxv32i8.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vsll_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, size_t shift, size_t vl) {
@@ -1447,7 +1447,7 @@ vuint8m4_t test_vsll_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t o
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsll.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsll.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vsll_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, vuint8m8_t shift, size_t vl) {
@@ -1457,7 +1457,7 @@ vuint8m8_t test_vsll_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t o
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsll.mask.nxv64i8.i64.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsll.mask.nxv64i8.i64.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vsll_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, size_t shift, size_t vl) {
@@ -1467,7 +1467,7 @@ vuint8m8_t test_vsll_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t o
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsll.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsll.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vsll_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, vuint16mf4_t shift, size_t vl) {
@@ -1477,7 +1477,7 @@ vuint16mf4_t test_vsll_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsll.mask.nxv1i16.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsll.mask.nxv1i16.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vsll_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, size_t shift, size_t vl) {
@@ -1487,7 +1487,7 @@ vuint16mf4_t test_vsll_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsll.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsll.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vsll_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, vuint16mf2_t shift, size_t vl) {
@@ -1497,7 +1497,7 @@ vuint16mf2_t test_vsll_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsll.mask.nxv2i16.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsll.mask.nxv2i16.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vsll_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, size_t shift, size_t vl) {
@@ -1507,7 +1507,7 @@ vuint16mf2_t test_vsll_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsll.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsll.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vsll_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, vuint16m1_t shift, size_t vl) {
@@ -1517,7 +1517,7 @@ vuint16m1_t test_vsll_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsll.mask.nxv4i16.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsll.mask.nxv4i16.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vsll_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, size_t shift, size_t vl) {
@@ -1527,7 +1527,7 @@ vuint16m1_t test_vsll_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsll.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsll.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vsll_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, vuint16m2_t shift, size_t vl) {
@@ -1537,7 +1537,7 @@ vuint16m2_t test_vsll_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsll.mask.nxv8i16.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsll.mask.nxv8i16.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vsll_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, size_t shift, size_t vl) {
@@ -1547,7 +1547,7 @@ vuint16m2_t test_vsll_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsll.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsll.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vsll_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, vuint16m4_t shift, size_t vl) {
@@ -1557,7 +1557,7 @@ vuint16m4_t test_vsll_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsll.mask.nxv16i16.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsll.mask.nxv16i16.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vsll_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, size_t shift, size_t vl) {
@@ -1567,7 +1567,7 @@ vuint16m4_t test_vsll_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsll.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsll.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vsll_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, vuint16m8_t shift, size_t vl) {
@@ -1577,7 +1577,7 @@ vuint16m8_t test_vsll_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsll.mask.nxv32i16.i64.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsll.mask.nxv32i16.i64.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vsll_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, size_t shift, size_t vl) {
@@ -1587,7 +1587,7 @@ vuint16m8_t test_vsll_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsll.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsll.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vsll_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, vuint32mf2_t shift, size_t vl) {
@@ -1597,7 +1597,7 @@ vuint32mf2_t test_vsll_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsll.mask.nxv1i32.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsll.mask.nxv1i32.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vsll_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, size_t shift, size_t vl) {
@@ -1607,7 +1607,7 @@ vuint32mf2_t test_vsll_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsll.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsll.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vsll_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, vuint32m1_t shift, size_t vl) {
@@ -1617,7 +1617,7 @@ vuint32m1_t test_vsll_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsll.mask.nxv2i32.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsll.mask.nxv2i32.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vsll_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, size_t shift, size_t vl) {
@@ -1627,7 +1627,7 @@ vuint32m1_t test_vsll_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsll.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsll.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vsll_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, vuint32m2_t shift, size_t vl) {
@@ -1637,7 +1637,7 @@ vuint32m2_t test_vsll_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsll.mask.nxv4i32.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsll.mask.nxv4i32.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vsll_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, size_t shift, size_t vl) {
@@ -1647,7 +1647,7 @@ vuint32m2_t test_vsll_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsll.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsll.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vsll_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, vuint32m4_t shift, size_t vl) {
@@ -1657,7 +1657,7 @@ vuint32m4_t test_vsll_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsll.mask.nxv8i32.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsll.mask.nxv8i32.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vsll_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, size_t shift, size_t vl) {
@@ -1667,7 +1667,7 @@ vuint32m4_t test_vsll_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsll.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsll.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vsll_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, vuint32m8_t shift, size_t vl) {
@@ -1677,7 +1677,7 @@ vuint32m8_t test_vsll_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsll.mask.nxv16i32.i64.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsll.mask.nxv16i32.i64.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vsll_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, size_t shift, size_t vl) {
@@ -1687,7 +1687,7 @@ vuint32m8_t test_vsll_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsll.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsll.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vsll_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, vuint64m1_t shift, size_t vl) {
@@ -1697,7 +1697,7 @@ vuint64m1_t test_vsll_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsll.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsll.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vsll_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, size_t shift, size_t vl) {
@@ -1707,7 +1707,7 @@ vuint64m1_t test_vsll_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsll.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsll.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vsll_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, vuint64m2_t shift, size_t vl) {
@@ -1717,7 +1717,7 @@ vuint64m2_t test_vsll_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsll.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsll.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vsll_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, size_t shift, size_t vl) {
@@ -1727,7 +1727,7 @@ vuint64m2_t test_vsll_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsll.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsll.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vsll_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, vuint64m4_t shift, size_t vl) {
@@ -1737,7 +1737,7 @@ vuint64m4_t test_vsll_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsll.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsll.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vsll_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, size_t shift, size_t vl) {
@@ -1747,7 +1747,7 @@ vuint64m4_t test_vsll_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vsll_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsll.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsll.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vsll_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, vuint64m8_t shift, size_t vl) {
@@ -1757,7 +1757,7 @@ vuint64m8_t test_vsll_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8
 //
 // CHECK-RV64-LABEL: @test_vsll_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsll.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsll.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vsll_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, size_t shift, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsmul.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsmul.c
index 347a44a..e2ae796 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsmul.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsmul.c
@@ -447,7 +447,7 @@ vint64m8_t test_vsmul_vx_i64m8(vint64m8_t op1, int64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsmul.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsmul.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vsmul_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
@@ -458,7 +458,7 @@ vint8mf8_t test_vsmul_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsmul.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsmul.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vsmul_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
@@ -469,7 +469,7 @@ vint8mf8_t test_vsmul_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsmul.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsmul.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vsmul_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
@@ -480,7 +480,7 @@ vint8mf4_t test_vsmul_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsmul.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsmul.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vsmul_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
@@ -491,7 +491,7 @@ vint8mf4_t test_vsmul_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsmul.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsmul.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vsmul_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
@@ -502,7 +502,7 @@ vint8mf2_t test_vsmul_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsmul.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsmul.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vsmul_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
@@ -513,7 +513,7 @@ vint8mf2_t test_vsmul_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsmul.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsmul.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vsmul_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
@@ -524,7 +524,7 @@ vint8m1_t test_vsmul_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsmul.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsmul.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vsmul_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
@@ -535,7 +535,7 @@ vint8m1_t test_vsmul_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsmul.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsmul.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vsmul_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
@@ -546,7 +546,7 @@ vint8m2_t test_vsmul_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsmul.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsmul.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vsmul_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
@@ -557,7 +557,7 @@ vint8m2_t test_vsmul_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsmul.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsmul.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vsmul_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
@@ -568,7 +568,7 @@ vint8m4_t test_vsmul_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsmul.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsmul.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vsmul_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
@@ -579,7 +579,7 @@ vint8m4_t test_vsmul_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsmul.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsmul.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vsmul_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
@@ -590,7 +590,7 @@ vint8m8_t test_vsmul_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsmul.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsmul.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vsmul_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
@@ -601,7 +601,7 @@ vint8m8_t test_vsmul_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsmul.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsmul.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vsmul_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -613,7 +613,7 @@ vint16mf4_t test_vsmul_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsmul.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsmul.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vsmul_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -624,7 +624,7 @@ vint16mf4_t test_vsmul_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsmul.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsmul.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vsmul_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -636,7 +636,7 @@ vint16mf2_t test_vsmul_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsmul.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsmul.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vsmul_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -647,7 +647,7 @@ vint16mf2_t test_vsmul_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsmul.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsmul.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vsmul_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -658,7 +658,7 @@ vint16m1_t test_vsmul_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsmul.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsmul.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vsmul_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -669,7 +669,7 @@ vint16m1_t test_vsmul_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsmul.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsmul.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vsmul_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -680,7 +680,7 @@ vint16m2_t test_vsmul_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsmul.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsmul.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vsmul_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -691,7 +691,7 @@ vint16m2_t test_vsmul_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsmul.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsmul.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vsmul_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -702,7 +702,7 @@ vint16m4_t test_vsmul_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsmul.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsmul.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vsmul_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -713,7 +713,7 @@ vint16m4_t test_vsmul_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsmul.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsmul.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vsmul_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -724,7 +724,7 @@ vint16m8_t test_vsmul_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsmul.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsmul.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vsmul_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -735,7 +735,7 @@ vint16m8_t test_vsmul_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsmul.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsmul.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vsmul_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -747,7 +747,7 @@ vint32mf2_t test_vsmul_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsmul.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsmul.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vsmul_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -758,7 +758,7 @@ vint32mf2_t test_vsmul_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsmul.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsmul.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vsmul_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -769,7 +769,7 @@ vint32m1_t test_vsmul_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsmul.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsmul.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vsmul_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -780,7 +780,7 @@ vint32m1_t test_vsmul_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsmul.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsmul.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vsmul_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -791,7 +791,7 @@ vint32m2_t test_vsmul_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsmul.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsmul.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vsmul_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -802,7 +802,7 @@ vint32m2_t test_vsmul_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsmul.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsmul.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vsmul_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -813,7 +813,7 @@ vint32m4_t test_vsmul_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsmul.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsmul.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vsmul_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -824,7 +824,7 @@ vint32m4_t test_vsmul_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsmul.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsmul.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vsmul_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -835,7 +835,7 @@ vint32m8_t test_vsmul_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsmul.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsmul.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vsmul_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -846,7 +846,7 @@ vint32m8_t test_vsmul_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsmul.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsmul.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vsmul_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -857,7 +857,7 @@ vint64m1_t test_vsmul_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsmul.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsmul.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vsmul_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -868,7 +868,7 @@ vint64m1_t test_vsmul_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsmul.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsmul.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vsmul_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -879,7 +879,7 @@ vint64m2_t test_vsmul_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsmul.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsmul.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vsmul_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -890,7 +890,7 @@ vint64m2_t test_vsmul_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsmul.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsmul.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vsmul_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -901,7 +901,7 @@ vint64m4_t test_vsmul_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsmul.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsmul.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vsmul_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -912,7 +912,7 @@ vint64m4_t test_vsmul_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsmul.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsmul.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vsmul_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -923,7 +923,7 @@ vint64m8_t test_vsmul_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vsmul_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsmul.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsmul.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vsmul_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsoxei.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsoxei.c
index 377d3e7..442e107 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsoxei.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsoxei.c
@@ -9,7 +9,7 @@
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8:#.*]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i8mf8(int8_t *base, vuint8mf8_t bindex, vint8mf8_t value,
@@ -21,7 +21,7 @@ void test_vsoxei8_v_i8mf8(int8_t *base, vuint8mf8_t bindex, vint8mf8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i8mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i8mf4(int8_t *base, vuint8mf4_t bindex, vint8mf4_t value,
@@ -33,7 +33,7 @@ void test_vsoxei8_v_i8mf4(int8_t *base, vuint8mf4_t bindex, vint8mf4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i8mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i8mf2(int8_t *base, vuint8mf2_t bindex, vint8mf2_t value,
@@ -45,7 +45,7 @@ void test_vsoxei8_v_i8mf2(int8_t *base, vuint8mf2_t bindex, vint8mf2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i8m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i8m1(int8_t *base, vuint8m1_t bindex, vint8m1_t value,
@@ -57,7 +57,7 @@ void test_vsoxei8_v_i8m1(int8_t *base, vuint8m1_t bindex, vint8m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i8m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i8m2(int8_t *base, vuint8m2_t bindex, vint8m2_t value,
@@ -69,7 +69,7 @@ void test_vsoxei8_v_i8m2(int8_t *base, vuint8m2_t bindex, vint8m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i8m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i8m4(int8_t *base, vuint8m4_t bindex, vint8m4_t value,
@@ -81,7 +81,7 @@ void test_vsoxei8_v_i8m4(int8_t *base, vuint8m4_t bindex, vint8m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i8m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i8m8(int8_t *base, vuint8m8_t bindex, vint8m8_t value,
@@ -93,7 +93,7 @@ void test_vsoxei8_v_i8m8(int8_t *base, vuint8m8_t bindex, vint8m8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i8mf8(int8_t *base, vuint16mf4_t bindex, vint8mf8_t value,
@@ -105,7 +105,7 @@ void test_vsoxei16_v_i8mf8(int8_t *base, vuint16mf4_t bindex, vint8mf8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i8mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i8mf4(int8_t *base, vuint16mf2_t bindex, vint8mf4_t value,
@@ -117,7 +117,7 @@ void test_vsoxei16_v_i8mf4(int8_t *base, vuint16mf2_t bindex, vint8mf4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i8mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i8mf2(int8_t *base, vuint16m1_t bindex, vint8mf2_t value,
@@ -129,7 +129,7 @@ void test_vsoxei16_v_i8mf2(int8_t *base, vuint16m1_t bindex, vint8mf2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i8m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i8m1(int8_t *base, vuint16m2_t bindex, vint8m1_t value,
@@ -141,7 +141,7 @@ void test_vsoxei16_v_i8m1(int8_t *base, vuint16m2_t bindex, vint8m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i8m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i8m2(int8_t *base, vuint16m4_t bindex, vint8m2_t value,
@@ -153,7 +153,7 @@ void test_vsoxei16_v_i8m2(int8_t *base, vuint16m4_t bindex, vint8m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i8m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i8m4(int8_t *base, vuint16m8_t bindex, vint8m4_t value,
@@ -165,7 +165,7 @@ void test_vsoxei16_v_i8m4(int8_t *base, vuint16m8_t bindex, vint8m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i8mf8(int8_t *base, vuint32mf2_t bindex, vint8mf8_t value,
@@ -177,7 +177,7 @@ void test_vsoxei32_v_i8mf8(int8_t *base, vuint32mf2_t bindex, vint8mf8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i8mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i8mf4(int8_t *base, vuint32m1_t bindex, vint8mf4_t value,
@@ -189,7 +189,7 @@ void test_vsoxei32_v_i8mf4(int8_t *base, vuint32m1_t bindex, vint8mf4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i8mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i8mf2(int8_t *base, vuint32m2_t bindex, vint8mf2_t value,
@@ -201,7 +201,7 @@ void test_vsoxei32_v_i8mf2(int8_t *base, vuint32m2_t bindex, vint8mf2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i8m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i8m1(int8_t *base, vuint32m4_t bindex, vint8m1_t value,
@@ -213,7 +213,7 @@ void test_vsoxei32_v_i8m1(int8_t *base, vuint32m4_t bindex, vint8m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i8m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i8m2(int8_t *base, vuint32m8_t bindex, vint8m2_t value,
@@ -225,7 +225,7 @@ void test_vsoxei32_v_i8m2(int8_t *base, vuint32m8_t bindex, vint8m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i8mf8(int8_t *base, vuint64m1_t bindex, vint8mf8_t value,
@@ -237,7 +237,7 @@ void test_vsoxei64_v_i8mf8(int8_t *base, vuint64m1_t bindex, vint8mf8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i8mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i8mf4(int8_t *base, vuint64m2_t bindex, vint8mf4_t value,
@@ -249,7 +249,7 @@ void test_vsoxei64_v_i8mf4(int8_t *base, vuint64m2_t bindex, vint8mf4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i8mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i8mf2(int8_t *base, vuint64m4_t bindex, vint8mf2_t value,
@@ -261,7 +261,7 @@ void test_vsoxei64_v_i8mf2(int8_t *base, vuint64m4_t bindex, vint8mf2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i8m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i8m1(int8_t *base, vuint64m8_t bindex, vint8m1_t value,
@@ -273,7 +273,7 @@ void test_vsoxei64_v_i8m1(int8_t *base, vuint64m8_t bindex, vint8m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i16mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i16mf4(int16_t *base, vuint8mf8_t bindex, vint16mf4_t value,
@@ -285,7 +285,7 @@ void test_vsoxei8_v_i16mf4(int16_t *base, vuint8mf8_t bindex, vint16mf4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i16mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i16mf2(int16_t *base, vuint8mf4_t bindex, vint16mf2_t value,
@@ -297,7 +297,7 @@ void test_vsoxei8_v_i16mf2(int16_t *base, vuint8mf4_t bindex, vint16mf2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i16m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i16m1(int16_t *base, vuint8mf2_t bindex, vint16m1_t value,
@@ -309,7 +309,7 @@ void test_vsoxei8_v_i16m1(int16_t *base, vuint8mf2_t bindex, vint16m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i16m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i16m2(int16_t *base, vuint8m1_t bindex, vint16m2_t value,
@@ -321,7 +321,7 @@ void test_vsoxei8_v_i16m2(int16_t *base, vuint8m1_t bindex, vint16m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i16m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i16m4(int16_t *base, vuint8m2_t bindex, vint16m4_t value,
@@ -333,7 +333,7 @@ void test_vsoxei8_v_i16m4(int16_t *base, vuint8m2_t bindex, vint16m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i16m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i16m8(int16_t *base, vuint8m4_t bindex, vint16m8_t value,
@@ -345,7 +345,7 @@ void test_vsoxei8_v_i16m8(int16_t *base, vuint8m4_t bindex, vint16m8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i16mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i16mf4(int16_t *base, vuint16mf4_t bindex,
@@ -357,7 +357,7 @@ void test_vsoxei16_v_i16mf4(int16_t *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i16mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i16mf2(int16_t *base, vuint16mf2_t bindex,
@@ -369,7 +369,7 @@ void test_vsoxei16_v_i16mf2(int16_t *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i16m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i16m1(int16_t *base, vuint16m1_t bindex, vint16m1_t value,
@@ -381,7 +381,7 @@ void test_vsoxei16_v_i16m1(int16_t *base, vuint16m1_t bindex, vint16m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i16m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i16m2(int16_t *base, vuint16m2_t bindex, vint16m2_t value,
@@ -393,7 +393,7 @@ void test_vsoxei16_v_i16m2(int16_t *base, vuint16m2_t bindex, vint16m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i16m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i16m4(int16_t *base, vuint16m4_t bindex, vint16m4_t value,
@@ -405,7 +405,7 @@ void test_vsoxei16_v_i16m4(int16_t *base, vuint16m4_t bindex, vint16m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i16m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i16m8(int16_t *base, vuint16m8_t bindex, vint16m8_t value,
@@ -417,7 +417,7 @@ void test_vsoxei16_v_i16m8(int16_t *base, vuint16m8_t bindex, vint16m8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i16mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i16mf4(int16_t *base, vuint32mf2_t bindex,
@@ -429,7 +429,7 @@ void test_vsoxei32_v_i16mf4(int16_t *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i16mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i16mf2(int16_t *base, vuint32m1_t bindex,
@@ -441,7 +441,7 @@ void test_vsoxei32_v_i16mf2(int16_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i16m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i16m1(int16_t *base, vuint32m2_t bindex, vint16m1_t value,
@@ -453,7 +453,7 @@ void test_vsoxei32_v_i16m1(int16_t *base, vuint32m2_t bindex, vint16m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i16m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i16m2(int16_t *base, vuint32m4_t bindex, vint16m2_t value,
@@ -465,7 +465,7 @@ void test_vsoxei32_v_i16m2(int16_t *base, vuint32m4_t bindex, vint16m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i16m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i16m4(int16_t *base, vuint32m8_t bindex, vint16m4_t value,
@@ -477,7 +477,7 @@ void test_vsoxei32_v_i16m4(int16_t *base, vuint32m8_t bindex, vint16m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i16mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i16mf4(int16_t *base, vuint64m1_t bindex,
@@ -489,7 +489,7 @@ void test_vsoxei64_v_i16mf4(int16_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i16mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i16mf2(int16_t *base, vuint64m2_t bindex,
@@ -501,7 +501,7 @@ void test_vsoxei64_v_i16mf2(int16_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i16m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i16m1(int16_t *base, vuint64m4_t bindex, vint16m1_t value,
@@ -513,7 +513,7 @@ void test_vsoxei64_v_i16m1(int16_t *base, vuint64m4_t bindex, vint16m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i16m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i16m2(int16_t *base, vuint64m8_t bindex, vint16m2_t value,
@@ -525,7 +525,7 @@ void test_vsoxei64_v_i16m2(int16_t *base, vuint64m8_t bindex, vint16m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i32mf2(int32_t *base, vuint8mf8_t bindex, vint32mf2_t value,
@@ -537,7 +537,7 @@ void test_vsoxei8_v_i32mf2(int32_t *base, vuint8mf8_t bindex, vint32mf2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i32m1(int32_t *base, vuint8mf4_t bindex, vint32m1_t value,
@@ -549,7 +549,7 @@ void test_vsoxei8_v_i32m1(int32_t *base, vuint8mf4_t bindex, vint32m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i32m2(int32_t *base, vuint8mf2_t bindex, vint32m2_t value,
@@ -561,7 +561,7 @@ void test_vsoxei8_v_i32m2(int32_t *base, vuint8mf2_t bindex, vint32m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i32m4(int32_t *base, vuint8m1_t bindex, vint32m4_t value,
@@ -573,7 +573,7 @@ void test_vsoxei8_v_i32m4(int32_t *base, vuint8m1_t bindex, vint32m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i32m8(int32_t *base, vuint8m2_t bindex, vint32m8_t value,
@@ -585,7 +585,7 @@ void test_vsoxei8_v_i32m8(int32_t *base, vuint8m2_t bindex, vint32m8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i32mf2(int32_t *base, vuint16mf4_t bindex,
@@ -597,7 +597,7 @@ void test_vsoxei16_v_i32mf2(int32_t *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i32m1(int32_t *base, vuint16mf2_t bindex, vint32m1_t value,
@@ -609,7 +609,7 @@ void test_vsoxei16_v_i32m1(int32_t *base, vuint16mf2_t bindex, vint32m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i32m2(int32_t *base, vuint16m1_t bindex, vint32m2_t value,
@@ -621,7 +621,7 @@ void test_vsoxei16_v_i32m2(int32_t *base, vuint16m1_t bindex, vint32m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i32m4(int32_t *base, vuint16m2_t bindex, vint32m4_t value,
@@ -633,7 +633,7 @@ void test_vsoxei16_v_i32m4(int32_t *base, vuint16m2_t bindex, vint32m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i32m8(int32_t *base, vuint16m4_t bindex, vint32m8_t value,
@@ -645,7 +645,7 @@ void test_vsoxei16_v_i32m8(int32_t *base, vuint16m4_t bindex, vint32m8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i32mf2(int32_t *base, vuint32mf2_t bindex,
@@ -657,7 +657,7 @@ void test_vsoxei32_v_i32mf2(int32_t *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i32m1(int32_t *base, vuint32m1_t bindex, vint32m1_t value,
@@ -669,7 +669,7 @@ void test_vsoxei32_v_i32m1(int32_t *base, vuint32m1_t bindex, vint32m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i32m2(int32_t *base, vuint32m2_t bindex, vint32m2_t value,
@@ -681,7 +681,7 @@ void test_vsoxei32_v_i32m2(int32_t *base, vuint32m2_t bindex, vint32m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i32m4(int32_t *base, vuint32m4_t bindex, vint32m4_t value,
@@ -693,7 +693,7 @@ void test_vsoxei32_v_i32m4(int32_t *base, vuint32m4_t bindex, vint32m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i32m8(int32_t *base, vuint32m8_t bindex, vint32m8_t value,
@@ -705,7 +705,7 @@ void test_vsoxei32_v_i32m8(int32_t *base, vuint32m8_t bindex, vint32m8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i32mf2(int32_t *base, vuint64m1_t bindex,
@@ -717,7 +717,7 @@ void test_vsoxei64_v_i32mf2(int32_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i32m1(int32_t *base, vuint64m2_t bindex, vint32m1_t value,
@@ -729,7 +729,7 @@ void test_vsoxei64_v_i32m1(int32_t *base, vuint64m2_t bindex, vint32m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i32m2(int32_t *base, vuint64m4_t bindex, vint32m2_t value,
@@ -741,7 +741,7 @@ void test_vsoxei64_v_i32m2(int32_t *base, vuint64m4_t bindex, vint32m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i32m4(int32_t *base, vuint64m8_t bindex, vint32m4_t value,
@@ -753,7 +753,7 @@ void test_vsoxei64_v_i32m4(int32_t *base, vuint64m8_t bindex, vint32m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i64m1(int64_t *base, vuint8mf8_t bindex, vint64m1_t value,
@@ -765,7 +765,7 @@ void test_vsoxei8_v_i64m1(int64_t *base, vuint8mf8_t bindex, vint64m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i64m2(int64_t *base, vuint8mf4_t bindex, vint64m2_t value,
@@ -777,7 +777,7 @@ void test_vsoxei8_v_i64m2(int64_t *base, vuint8mf4_t bindex, vint64m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i64m4(int64_t *base, vuint8mf2_t bindex, vint64m4_t value,
@@ -789,7 +789,7 @@ void test_vsoxei8_v_i64m4(int64_t *base, vuint8mf2_t bindex, vint64m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i64m8(int64_t *base, vuint8m1_t bindex, vint64m8_t value,
@@ -801,7 +801,7 @@ void test_vsoxei8_v_i64m8(int64_t *base, vuint8m1_t bindex, vint64m8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i64m1(int64_t *base, vuint16mf4_t bindex, vint64m1_t value,
@@ -813,7 +813,7 @@ void test_vsoxei16_v_i64m1(int64_t *base, vuint16mf4_t bindex, vint64m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i64m2(int64_t *base, vuint16mf2_t bindex, vint64m2_t value,
@@ -825,7 +825,7 @@ void test_vsoxei16_v_i64m2(int64_t *base, vuint16mf2_t bindex, vint64m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i64m4(int64_t *base, vuint16m1_t bindex, vint64m4_t value,
@@ -837,7 +837,7 @@ void test_vsoxei16_v_i64m4(int64_t *base, vuint16m1_t bindex, vint64m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i64m8(int64_t *base, vuint16m2_t bindex, vint64m8_t value,
@@ -849,7 +849,7 @@ void test_vsoxei16_v_i64m8(int64_t *base, vuint16m2_t bindex, vint64m8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i64m1(int64_t *base, vuint32mf2_t bindex, vint64m1_t value,
@@ -861,7 +861,7 @@ void test_vsoxei32_v_i64m1(int64_t *base, vuint32mf2_t bindex, vint64m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i64m2(int64_t *base, vuint32m1_t bindex, vint64m2_t value,
@@ -873,7 +873,7 @@ void test_vsoxei32_v_i64m2(int64_t *base, vuint32m1_t bindex, vint64m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i64m4(int64_t *base, vuint32m2_t bindex, vint64m4_t value,
@@ -885,7 +885,7 @@ void test_vsoxei32_v_i64m4(int64_t *base, vuint32m2_t bindex, vint64m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i64m8(int64_t *base, vuint32m4_t bindex, vint64m8_t value,
@@ -897,7 +897,7 @@ void test_vsoxei32_v_i64m8(int64_t *base, vuint32m4_t bindex, vint64m8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i64m1(int64_t *base, vuint64m1_t bindex, vint64m1_t value,
@@ -909,7 +909,7 @@ void test_vsoxei64_v_i64m1(int64_t *base, vuint64m1_t bindex, vint64m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i64m2(int64_t *base, vuint64m2_t bindex, vint64m2_t value,
@@ -921,7 +921,7 @@ void test_vsoxei64_v_i64m2(int64_t *base, vuint64m2_t bindex, vint64m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i64m4(int64_t *base, vuint64m4_t bindex, vint64m4_t value,
@@ -933,7 +933,7 @@ void test_vsoxei64_v_i64m4(int64_t *base, vuint64m4_t bindex, vint64m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i64m8(int64_t *base, vuint64m8_t bindex, vint64m8_t value,
@@ -945,7 +945,7 @@ void test_vsoxei64_v_i64m8(int64_t *base, vuint64m8_t bindex, vint64m8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u8mf8(uint8_t *base, vuint8mf8_t bindex, vuint8mf8_t value,
@@ -957,7 +957,7 @@ void test_vsoxei8_v_u8mf8(uint8_t *base, vuint8mf8_t bindex, vuint8mf8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u8mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u8mf4(uint8_t *base, vuint8mf4_t bindex, vuint8mf4_t value,
@@ -969,7 +969,7 @@ void test_vsoxei8_v_u8mf4(uint8_t *base, vuint8mf4_t bindex, vuint8mf4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u8mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u8mf2(uint8_t *base, vuint8mf2_t bindex, vuint8mf2_t value,
@@ -981,7 +981,7 @@ void test_vsoxei8_v_u8mf2(uint8_t *base, vuint8mf2_t bindex, vuint8mf2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u8m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u8m1(uint8_t *base, vuint8m1_t bindex, vuint8m1_t value,
@@ -993,7 +993,7 @@ void test_vsoxei8_v_u8m1(uint8_t *base, vuint8m1_t bindex, vuint8m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u8m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u8m2(uint8_t *base, vuint8m2_t bindex, vuint8m2_t value,
@@ -1005,7 +1005,7 @@ void test_vsoxei8_v_u8m2(uint8_t *base, vuint8m2_t bindex, vuint8m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u8m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u8m4(uint8_t *base, vuint8m4_t bindex, vuint8m4_t value,
@@ -1017,7 +1017,7 @@ void test_vsoxei8_v_u8m4(uint8_t *base, vuint8m4_t bindex, vuint8m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u8m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u8m8(uint8_t *base, vuint8m8_t bindex, vuint8m8_t value,
@@ -1029,7 +1029,7 @@ void test_vsoxei8_v_u8m8(uint8_t *base, vuint8m8_t bindex, vuint8m8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u8mf8(uint8_t *base, vuint16mf4_t bindex,
@@ -1041,7 +1041,7 @@ void test_vsoxei16_v_u8mf8(uint8_t *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u8mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u8mf4(uint8_t *base, vuint16mf2_t bindex,
@@ -1053,7 +1053,7 @@ void test_vsoxei16_v_u8mf4(uint8_t *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u8mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u8mf2(uint8_t *base, vuint16m1_t bindex, vuint8mf2_t value,
@@ -1065,7 +1065,7 @@ void test_vsoxei16_v_u8mf2(uint8_t *base, vuint16m1_t bindex, vuint8mf2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u8m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u8m1(uint8_t *base, vuint16m2_t bindex, vuint8m1_t value,
@@ -1077,7 +1077,7 @@ void test_vsoxei16_v_u8m1(uint8_t *base, vuint16m2_t bindex, vuint8m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u8m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u8m2(uint8_t *base, vuint16m4_t bindex, vuint8m2_t value,
@@ -1089,7 +1089,7 @@ void test_vsoxei16_v_u8m2(uint8_t *base, vuint16m4_t bindex, vuint8m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u8m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u8m4(uint8_t *base, vuint16m8_t bindex, vuint8m4_t value,
@@ -1101,7 +1101,7 @@ void test_vsoxei16_v_u8m4(uint8_t *base, vuint16m8_t bindex, vuint8m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u8mf8(uint8_t *base, vuint32mf2_t bindex,
@@ -1113,7 +1113,7 @@ void test_vsoxei32_v_u8mf8(uint8_t *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u8mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u8mf4(uint8_t *base, vuint32m1_t bindex, vuint8mf4_t value,
@@ -1125,7 +1125,7 @@ void test_vsoxei32_v_u8mf4(uint8_t *base, vuint32m1_t bindex, vuint8mf4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u8mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u8mf2(uint8_t *base, vuint32m2_t bindex, vuint8mf2_t value,
@@ -1137,7 +1137,7 @@ void test_vsoxei32_v_u8mf2(uint8_t *base, vuint32m2_t bindex, vuint8mf2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u8m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u8m1(uint8_t *base, vuint32m4_t bindex, vuint8m1_t value,
@@ -1149,7 +1149,7 @@ void test_vsoxei32_v_u8m1(uint8_t *base, vuint32m4_t bindex, vuint8m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u8m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u8m2(uint8_t *base, vuint32m8_t bindex, vuint8m2_t value,
@@ -1161,7 +1161,7 @@ void test_vsoxei32_v_u8m2(uint8_t *base, vuint32m8_t bindex, vuint8m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u8mf8(uint8_t *base, vuint64m1_t bindex, vuint8mf8_t value,
@@ -1173,7 +1173,7 @@ void test_vsoxei64_v_u8mf8(uint8_t *base, vuint64m1_t bindex, vuint8mf8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u8mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u8mf4(uint8_t *base, vuint64m2_t bindex, vuint8mf4_t value,
@@ -1185,7 +1185,7 @@ void test_vsoxei64_v_u8mf4(uint8_t *base, vuint64m2_t bindex, vuint8mf4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u8mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u8mf2(uint8_t *base, vuint64m4_t bindex, vuint8mf2_t value,
@@ -1197,7 +1197,7 @@ void test_vsoxei64_v_u8mf2(uint8_t *base, vuint64m4_t bindex, vuint8mf2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u8m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u8m1(uint8_t *base, vuint64m8_t bindex, vuint8m1_t value,
@@ -1209,7 +1209,7 @@ void test_vsoxei64_v_u8m1(uint8_t *base, vuint64m8_t bindex, vuint8m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u16mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u16mf4(uint16_t *base, vuint8mf8_t bindex,
@@ -1221,7 +1221,7 @@ void test_vsoxei8_v_u16mf4(uint16_t *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u16mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u16mf2(uint16_t *base, vuint8mf4_t bindex,
@@ -1233,7 +1233,7 @@ void test_vsoxei8_v_u16mf2(uint16_t *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u16m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u16m1(uint16_t *base, vuint8mf2_t bindex, vuint16m1_t value,
@@ -1245,7 +1245,7 @@ void test_vsoxei8_v_u16m1(uint16_t *base, vuint8mf2_t bindex, vuint16m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u16m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u16m2(uint16_t *base, vuint8m1_t bindex, vuint16m2_t value,
@@ -1257,7 +1257,7 @@ void test_vsoxei8_v_u16m2(uint16_t *base, vuint8m1_t bindex, vuint16m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u16m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u16m4(uint16_t *base, vuint8m2_t bindex, vuint16m4_t value,
@@ -1269,7 +1269,7 @@ void test_vsoxei8_v_u16m4(uint16_t *base, vuint8m2_t bindex, vuint16m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u16m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u16m8(uint16_t *base, vuint8m4_t bindex, vuint16m8_t value,
@@ -1281,7 +1281,7 @@ void test_vsoxei8_v_u16m8(uint16_t *base, vuint8m4_t bindex, vuint16m8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u16mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u16mf4(uint16_t *base, vuint16mf4_t bindex,
@@ -1293,7 +1293,7 @@ void test_vsoxei16_v_u16mf4(uint16_t *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u16mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u16mf2(uint16_t *base, vuint16mf2_t bindex,
@@ -1305,7 +1305,7 @@ void test_vsoxei16_v_u16mf2(uint16_t *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u16m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u16m1(uint16_t *base, vuint16m1_t bindex,
@@ -1317,7 +1317,7 @@ void test_vsoxei16_v_u16m1(uint16_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u16m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u16m2(uint16_t *base, vuint16m2_t bindex,
@@ -1329,7 +1329,7 @@ void test_vsoxei16_v_u16m2(uint16_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u16m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u16m4(uint16_t *base, vuint16m4_t bindex,
@@ -1341,7 +1341,7 @@ void test_vsoxei16_v_u16m4(uint16_t *base, vuint16m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u16m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u16m8(uint16_t *base, vuint16m8_t bindex,
@@ -1353,7 +1353,7 @@ void test_vsoxei16_v_u16m8(uint16_t *base, vuint16m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u16mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u16mf4(uint16_t *base, vuint32mf2_t bindex,
@@ -1365,7 +1365,7 @@ void test_vsoxei32_v_u16mf4(uint16_t *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u16mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u16mf2(uint16_t *base, vuint32m1_t bindex,
@@ -1377,7 +1377,7 @@ void test_vsoxei32_v_u16mf2(uint16_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u16m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u16m1(uint16_t *base, vuint32m2_t bindex,
@@ -1389,7 +1389,7 @@ void test_vsoxei32_v_u16m1(uint16_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u16m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u16m2(uint16_t *base, vuint32m4_t bindex,
@@ -1401,7 +1401,7 @@ void test_vsoxei32_v_u16m2(uint16_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u16m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u16m4(uint16_t *base, vuint32m8_t bindex,
@@ -1413,7 +1413,7 @@ void test_vsoxei32_v_u16m4(uint16_t *base, vuint32m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u16mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u16mf4(uint16_t *base, vuint64m1_t bindex,
@@ -1425,7 +1425,7 @@ void test_vsoxei64_v_u16mf4(uint16_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u16mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u16mf2(uint16_t *base, vuint64m2_t bindex,
@@ -1437,7 +1437,7 @@ void test_vsoxei64_v_u16mf2(uint16_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u16m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u16m1(uint16_t *base, vuint64m4_t bindex,
@@ -1449,7 +1449,7 @@ void test_vsoxei64_v_u16m1(uint16_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u16m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u16m2(uint16_t *base, vuint64m8_t bindex,
@@ -1461,7 +1461,7 @@ void test_vsoxei64_v_u16m2(uint16_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u32mf2(uint32_t *base, vuint8mf8_t bindex,
@@ -1473,7 +1473,7 @@ void test_vsoxei8_v_u32mf2(uint32_t *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u32m1(uint32_t *base, vuint8mf4_t bindex, vuint32m1_t value,
@@ -1485,7 +1485,7 @@ void test_vsoxei8_v_u32m1(uint32_t *base, vuint8mf4_t bindex, vuint32m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u32m2(uint32_t *base, vuint8mf2_t bindex, vuint32m2_t value,
@@ -1497,7 +1497,7 @@ void test_vsoxei8_v_u32m2(uint32_t *base, vuint8mf2_t bindex, vuint32m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u32m4(uint32_t *base, vuint8m1_t bindex, vuint32m4_t value,
@@ -1509,7 +1509,7 @@ void test_vsoxei8_v_u32m4(uint32_t *base, vuint8m1_t bindex, vuint32m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u32m8(uint32_t *base, vuint8m2_t bindex, vuint32m8_t value,
@@ -1521,7 +1521,7 @@ void test_vsoxei8_v_u32m8(uint32_t *base, vuint8m2_t bindex, vuint32m8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u32mf2(uint32_t *base, vuint16mf4_t bindex,
@@ -1533,7 +1533,7 @@ void test_vsoxei16_v_u32mf2(uint32_t *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u32m1(uint32_t *base, vuint16mf2_t bindex,
@@ -1545,7 +1545,7 @@ void test_vsoxei16_v_u32m1(uint32_t *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u32m2(uint32_t *base, vuint16m1_t bindex,
@@ -1557,7 +1557,7 @@ void test_vsoxei16_v_u32m2(uint32_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u32m4(uint32_t *base, vuint16m2_t bindex,
@@ -1569,7 +1569,7 @@ void test_vsoxei16_v_u32m4(uint32_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u32m8(uint32_t *base, vuint16m4_t bindex,
@@ -1581,7 +1581,7 @@ void test_vsoxei16_v_u32m8(uint32_t *base, vuint16m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u32mf2(uint32_t *base, vuint32mf2_t bindex,
@@ -1593,7 +1593,7 @@ void test_vsoxei32_v_u32mf2(uint32_t *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u32m1(uint32_t *base, vuint32m1_t bindex,
@@ -1605,7 +1605,7 @@ void test_vsoxei32_v_u32m1(uint32_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u32m2(uint32_t *base, vuint32m2_t bindex,
@@ -1617,7 +1617,7 @@ void test_vsoxei32_v_u32m2(uint32_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u32m4(uint32_t *base, vuint32m4_t bindex,
@@ -1629,7 +1629,7 @@ void test_vsoxei32_v_u32m4(uint32_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u32m8(uint32_t *base, vuint32m8_t bindex,
@@ -1641,7 +1641,7 @@ void test_vsoxei32_v_u32m8(uint32_t *base, vuint32m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u32mf2(uint32_t *base, vuint64m1_t bindex,
@@ -1653,7 +1653,7 @@ void test_vsoxei64_v_u32mf2(uint32_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u32m1(uint32_t *base, vuint64m2_t bindex,
@@ -1665,7 +1665,7 @@ void test_vsoxei64_v_u32m1(uint32_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u32m2(uint32_t *base, vuint64m4_t bindex,
@@ -1677,7 +1677,7 @@ void test_vsoxei64_v_u32m2(uint32_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u32m4(uint32_t *base, vuint64m8_t bindex,
@@ -1689,7 +1689,7 @@ void test_vsoxei64_v_u32m4(uint32_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u64m1(uint64_t *base, vuint8mf8_t bindex, vuint64m1_t value,
@@ -1701,7 +1701,7 @@ void test_vsoxei8_v_u64m1(uint64_t *base, vuint8mf8_t bindex, vuint64m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u64m2(uint64_t *base, vuint8mf4_t bindex, vuint64m2_t value,
@@ -1713,7 +1713,7 @@ void test_vsoxei8_v_u64m2(uint64_t *base, vuint8mf4_t bindex, vuint64m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u64m4(uint64_t *base, vuint8mf2_t bindex, vuint64m4_t value,
@@ -1725,7 +1725,7 @@ void test_vsoxei8_v_u64m4(uint64_t *base, vuint8mf2_t bindex, vuint64m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u64m8(uint64_t *base, vuint8m1_t bindex, vuint64m8_t value,
@@ -1737,7 +1737,7 @@ void test_vsoxei8_v_u64m8(uint64_t *base, vuint8m1_t bindex, vuint64m8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u64m1(uint64_t *base, vuint16mf4_t bindex,
@@ -1749,7 +1749,7 @@ void test_vsoxei16_v_u64m1(uint64_t *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u64m2(uint64_t *base, vuint16mf2_t bindex,
@@ -1761,7 +1761,7 @@ void test_vsoxei16_v_u64m2(uint64_t *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u64m4(uint64_t *base, vuint16m1_t bindex,
@@ -1773,7 +1773,7 @@ void test_vsoxei16_v_u64m4(uint64_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u64m8(uint64_t *base, vuint16m2_t bindex,
@@ -1785,7 +1785,7 @@ void test_vsoxei16_v_u64m8(uint64_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u64m1(uint64_t *base, vuint32mf2_t bindex,
@@ -1797,7 +1797,7 @@ void test_vsoxei32_v_u64m1(uint64_t *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u64m2(uint64_t *base, vuint32m1_t bindex,
@@ -1809,7 +1809,7 @@ void test_vsoxei32_v_u64m2(uint64_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u64m4(uint64_t *base, vuint32m2_t bindex,
@@ -1821,7 +1821,7 @@ void test_vsoxei32_v_u64m4(uint64_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u64m8(uint64_t *base, vuint32m4_t bindex,
@@ -1833,7 +1833,7 @@ void test_vsoxei32_v_u64m8(uint64_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u64m1(uint64_t *base, vuint64m1_t bindex,
@@ -1845,7 +1845,7 @@ void test_vsoxei64_v_u64m1(uint64_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u64m2(uint64_t *base, vuint64m2_t bindex,
@@ -1857,7 +1857,7 @@ void test_vsoxei64_v_u64m2(uint64_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u64m4(uint64_t *base, vuint64m4_t bindex,
@@ -1869,7 +1869,7 @@ void test_vsoxei64_v_u64m4(uint64_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u64m8(uint64_t *base, vuint64m8_t bindex,
@@ -1881,7 +1881,7 @@ void test_vsoxei64_v_u64m8(uint64_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_f32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1f32.nxv1i8.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1f32.nxv1i8.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_f32mf2(float *base, vuint8mf8_t bindex, vfloat32mf2_t value,
@@ -1893,7 +1893,7 @@ void test_vsoxei8_v_f32mf2(float *base, vuint8mf8_t bindex, vfloat32mf2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_f32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2f32.nxv2i8.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2f32.nxv2i8.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_f32m1(float *base, vuint8mf4_t bindex, vfloat32m1_t value,
@@ -1905,7 +1905,7 @@ void test_vsoxei8_v_f32m1(float *base, vuint8mf4_t bindex, vfloat32m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_f32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4f32.nxv4i8.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4f32.nxv4i8.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_f32m2(float *base, vuint8mf2_t bindex, vfloat32m2_t value,
@@ -1917,7 +1917,7 @@ void test_vsoxei8_v_f32m2(float *base, vuint8mf2_t bindex, vfloat32m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_f32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8f32.nxv8i8.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8f32.nxv8i8.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_f32m4(float *base, vuint8m1_t bindex, vfloat32m4_t value,
@@ -1929,7 +1929,7 @@ void test_vsoxei8_v_f32m4(float *base, vuint8m1_t bindex, vfloat32m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_f32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16f32.nxv16i8.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16f32.nxv16i8.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_f32m8(float *base, vuint8m2_t bindex, vfloat32m8_t value,
@@ -1941,7 +1941,7 @@ void test_vsoxei8_v_f32m8(float *base, vuint8m2_t bindex, vfloat32m8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_f32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1f32.nxv1i16.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1f32.nxv1i16.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_f32mf2(float *base, vuint16mf4_t bindex,
@@ -1953,7 +1953,7 @@ void test_vsoxei16_v_f32mf2(float *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_f32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2f32.nxv2i16.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2f32.nxv2i16.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_f32m1(float *base, vuint16mf2_t bindex, vfloat32m1_t value,
@@ -1965,7 +1965,7 @@ void test_vsoxei16_v_f32m1(float *base, vuint16mf2_t bindex, vfloat32m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_f32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4f32.nxv4i16.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4f32.nxv4i16.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_f32m2(float *base, vuint16m1_t bindex, vfloat32m2_t value,
@@ -1977,7 +1977,7 @@ void test_vsoxei16_v_f32m2(float *base, vuint16m1_t bindex, vfloat32m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_f32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8f32.nxv8i16.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8f32.nxv8i16.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_f32m4(float *base, vuint16m2_t bindex, vfloat32m4_t value,
@@ -1989,7 +1989,7 @@ void test_vsoxei16_v_f32m4(float *base, vuint16m2_t bindex, vfloat32m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_f32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16f32.nxv16i16.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16f32.nxv16i16.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_f32m8(float *base, vuint16m4_t bindex, vfloat32m8_t value,
@@ -2001,7 +2001,7 @@ void test_vsoxei16_v_f32m8(float *base, vuint16m4_t bindex, vfloat32m8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_f32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1f32.nxv1i32.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1f32.nxv1i32.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_f32mf2(float *base, vuint32mf2_t bindex,
@@ -2013,7 +2013,7 @@ void test_vsoxei32_v_f32mf2(float *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_f32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2f32.nxv2i32.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2f32.nxv2i32.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_f32m1(float *base, vuint32m1_t bindex, vfloat32m1_t value,
@@ -2025,7 +2025,7 @@ void test_vsoxei32_v_f32m1(float *base, vuint32m1_t bindex, vfloat32m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_f32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4f32.nxv4i32.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4f32.nxv4i32.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_f32m2(float *base, vuint32m2_t bindex, vfloat32m2_t value,
@@ -2037,7 +2037,7 @@ void test_vsoxei32_v_f32m2(float *base, vuint32m2_t bindex, vfloat32m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_f32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8f32.nxv8i32.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8f32.nxv8i32.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_f32m4(float *base, vuint32m4_t bindex, vfloat32m4_t value,
@@ -2049,7 +2049,7 @@ void test_vsoxei32_v_f32m4(float *base, vuint32m4_t bindex, vfloat32m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_f32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16f32.nxv16i32.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv16f32.nxv16i32.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_f32m8(float *base, vuint32m8_t bindex, vfloat32m8_t value,
@@ -2061,7 +2061,7 @@ void test_vsoxei32_v_f32m8(float *base, vuint32m8_t bindex, vfloat32m8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_f32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1f32.nxv1i64.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1f32.nxv1i64.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_f32mf2(float *base, vuint64m1_t bindex,
@@ -2073,7 +2073,7 @@ void test_vsoxei64_v_f32mf2(float *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_f32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2f32.nxv2i64.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2f32.nxv2i64.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_f32m1(float *base, vuint64m2_t bindex, vfloat32m1_t value,
@@ -2085,7 +2085,7 @@ void test_vsoxei64_v_f32m1(float *base, vuint64m2_t bindex, vfloat32m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_f32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4f32.nxv4i64.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4f32.nxv4i64.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_f32m2(float *base, vuint64m4_t bindex, vfloat32m2_t value,
@@ -2097,7 +2097,7 @@ void test_vsoxei64_v_f32m2(float *base, vuint64m4_t bindex, vfloat32m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_f32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8f32.nxv8i64.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8f32.nxv8i64.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_f32m4(float *base, vuint64m8_t bindex, vfloat32m4_t value,
@@ -2109,7 +2109,7 @@ void test_vsoxei64_v_f32m4(float *base, vuint64m8_t bindex, vfloat32m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_f64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1f64.nxv1i8.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1f64.nxv1i8.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_f64m1(double *base, vuint8mf8_t bindex, vfloat64m1_t value,
@@ -2121,7 +2121,7 @@ void test_vsoxei8_v_f64m1(double *base, vuint8mf8_t bindex, vfloat64m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_f64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2f64.nxv2i8.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2f64.nxv2i8.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_f64m2(double *base, vuint8mf4_t bindex, vfloat64m2_t value,
@@ -2133,7 +2133,7 @@ void test_vsoxei8_v_f64m2(double *base, vuint8mf4_t bindex, vfloat64m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_f64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4f64.nxv4i8.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4f64.nxv4i8.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_f64m4(double *base, vuint8mf2_t bindex, vfloat64m4_t value,
@@ -2145,7 +2145,7 @@ void test_vsoxei8_v_f64m4(double *base, vuint8mf2_t bindex, vfloat64m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_f64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8f64.nxv8i8.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8f64.nxv8i8.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_f64m8(double *base, vuint8m1_t bindex, vfloat64m8_t value,
@@ -2157,7 +2157,7 @@ void test_vsoxei8_v_f64m8(double *base, vuint8m1_t bindex, vfloat64m8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_f64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1f64.nxv1i16.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1f64.nxv1i16.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_f64m1(double *base, vuint16mf4_t bindex,
@@ -2169,7 +2169,7 @@ void test_vsoxei16_v_f64m1(double *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_f64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2f64.nxv2i16.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2f64.nxv2i16.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_f64m2(double *base, vuint16mf2_t bindex,
@@ -2181,7 +2181,7 @@ void test_vsoxei16_v_f64m2(double *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_f64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4f64.nxv4i16.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4f64.nxv4i16.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_f64m4(double *base, vuint16m1_t bindex, vfloat64m4_t value,
@@ -2193,7 +2193,7 @@ void test_vsoxei16_v_f64m4(double *base, vuint16m1_t bindex, vfloat64m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_f64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8f64.nxv8i16.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8f64.nxv8i16.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_f64m8(double *base, vuint16m2_t bindex, vfloat64m8_t value,
@@ -2205,7 +2205,7 @@ void test_vsoxei16_v_f64m8(double *base, vuint16m2_t bindex, vfloat64m8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_f64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1f64.nxv1i32.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1f64.nxv1i32.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_f64m1(double *base, vuint32mf2_t bindex,
@@ -2217,7 +2217,7 @@ void test_vsoxei32_v_f64m1(double *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_f64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2f64.nxv2i32.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2f64.nxv2i32.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_f64m2(double *base, vuint32m1_t bindex, vfloat64m2_t value,
@@ -2229,7 +2229,7 @@ void test_vsoxei32_v_f64m2(double *base, vuint32m1_t bindex, vfloat64m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_f64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4f64.nxv4i32.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4f64.nxv4i32.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_f64m4(double *base, vuint32m2_t bindex, vfloat64m4_t value,
@@ -2241,7 +2241,7 @@ void test_vsoxei32_v_f64m4(double *base, vuint32m2_t bindex, vfloat64m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_f64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8f64.nxv8i32.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8f64.nxv8i32.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_f64m8(double *base, vuint32m4_t bindex, vfloat64m8_t value,
@@ -2253,7 +2253,7 @@ void test_vsoxei32_v_f64m8(double *base, vuint32m4_t bindex, vfloat64m8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_f64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1f64.nxv1i64.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv1f64.nxv1i64.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_f64m1(double *base, vuint64m1_t bindex, vfloat64m1_t value,
@@ -2265,7 +2265,7 @@ void test_vsoxei64_v_f64m1(double *base, vuint64m1_t bindex, vfloat64m1_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_f64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2f64.nxv2i64.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv2f64.nxv2i64.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_f64m2(double *base, vuint64m2_t bindex, vfloat64m2_t value,
@@ -2277,7 +2277,7 @@ void test_vsoxei64_v_f64m2(double *base, vuint64m2_t bindex, vfloat64m2_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_f64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4f64.nxv4i64.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv4f64.nxv4i64.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_f64m4(double *base, vuint64m4_t bindex, vfloat64m4_t value,
@@ -2289,7 +2289,7 @@ void test_vsoxei64_v_f64m4(double *base, vuint64m4_t bindex, vfloat64m4_t value,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_f64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8f64.nxv8i64.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.nxv8f64.nxv8i64.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_f64m8(double *base, vuint64m8_t bindex, vfloat64m8_t value,
@@ -2301,7 +2301,7 @@ void test_vsoxei64_v_f64m8(double *base, vuint64m8_t bindex, vfloat64m8_t value,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i8mf8_m(vbool64_t mask, int8_t *base, vuint8mf8_t bindex,
@@ -2313,7 +2313,7 @@ void test_vsoxei8_v_i8mf8_m(vbool64_t mask, int8_t *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i8mf4_m(vbool32_t mask, int8_t *base, vuint8mf4_t bindex,
@@ -2325,7 +2325,7 @@ void test_vsoxei8_v_i8mf4_m(vbool32_t mask, int8_t *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i8mf2_m(vbool16_t mask, int8_t *base, vuint8mf2_t bindex,
@@ -2337,7 +2337,7 @@ void test_vsoxei8_v_i8mf2_m(vbool16_t mask, int8_t *base, vuint8mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i8m1_m(vbool8_t mask, int8_t *base, vuint8m1_t bindex,
@@ -2349,7 +2349,7 @@ void test_vsoxei8_v_i8m1_m(vbool8_t mask, int8_t *base, vuint8m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i8m2_m(vbool4_t mask, int8_t *base, vuint8m2_t bindex,
@@ -2361,7 +2361,7 @@ void test_vsoxei8_v_i8m2_m(vbool4_t mask, int8_t *base, vuint8m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i8m4_m(vbool2_t mask, int8_t *base, vuint8m4_t bindex,
@@ -2373,7 +2373,7 @@ void test_vsoxei8_v_i8m4_m(vbool2_t mask, int8_t *base, vuint8m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i8m8_m(vbool1_t mask, int8_t *base, vuint8m8_t bindex,
@@ -2385,7 +2385,7 @@ void test_vsoxei8_v_i8m8_m(vbool1_t mask, int8_t *base, vuint8m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i8mf8_m(vbool64_t mask, int8_t *base, vuint16mf4_t bindex,
@@ -2397,7 +2397,7 @@ void test_vsoxei16_v_i8mf8_m(vbool64_t mask, int8_t *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i8mf4_m(vbool32_t mask, int8_t *base, vuint16mf2_t bindex,
@@ -2409,7 +2409,7 @@ void test_vsoxei16_v_i8mf4_m(vbool32_t mask, int8_t *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i8mf2_m(vbool16_t mask, int8_t *base, vuint16m1_t bindex,
@@ -2421,7 +2421,7 @@ void test_vsoxei16_v_i8mf2_m(vbool16_t mask, int8_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i8m1_m(vbool8_t mask, int8_t *base, vuint16m2_t bindex,
@@ -2433,7 +2433,7 @@ void test_vsoxei16_v_i8m1_m(vbool8_t mask, int8_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i8m2_m(vbool4_t mask, int8_t *base, vuint16m4_t bindex,
@@ -2445,7 +2445,7 @@ void test_vsoxei16_v_i8m2_m(vbool4_t mask, int8_t *base, vuint16m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i8m4_m(vbool2_t mask, int8_t *base, vuint16m8_t bindex,
@@ -2457,7 +2457,7 @@ void test_vsoxei16_v_i8m4_m(vbool2_t mask, int8_t *base, vuint16m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i8mf8_m(vbool64_t mask, int8_t *base, vuint32mf2_t bindex,
@@ -2469,7 +2469,7 @@ void test_vsoxei32_v_i8mf8_m(vbool64_t mask, int8_t *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i8mf4_m(vbool32_t mask, int8_t *base, vuint32m1_t bindex,
@@ -2481,7 +2481,7 @@ void test_vsoxei32_v_i8mf4_m(vbool32_t mask, int8_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i8mf2_m(vbool16_t mask, int8_t *base, vuint32m2_t bindex,
@@ -2493,7 +2493,7 @@ void test_vsoxei32_v_i8mf2_m(vbool16_t mask, int8_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i8m1_m(vbool8_t mask, int8_t *base, vuint32m4_t bindex,
@@ -2505,7 +2505,7 @@ void test_vsoxei32_v_i8m1_m(vbool8_t mask, int8_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i8m2_m(vbool4_t mask, int8_t *base, vuint32m8_t bindex,
@@ -2517,7 +2517,7 @@ void test_vsoxei32_v_i8m2_m(vbool4_t mask, int8_t *base, vuint32m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i8mf8_m(vbool64_t mask, int8_t *base, vuint64m1_t bindex,
@@ -2529,7 +2529,7 @@ void test_vsoxei64_v_i8mf8_m(vbool64_t mask, int8_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i8mf4_m(vbool32_t mask, int8_t *base, vuint64m2_t bindex,
@@ -2541,7 +2541,7 @@ void test_vsoxei64_v_i8mf4_m(vbool32_t mask, int8_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i8mf2_m(vbool16_t mask, int8_t *base, vuint64m4_t bindex,
@@ -2553,7 +2553,7 @@ void test_vsoxei64_v_i8mf2_m(vbool16_t mask, int8_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i8m1_m(vbool8_t mask, int8_t *base, vuint64m8_t bindex,
@@ -2565,7 +2565,7 @@ void test_vsoxei64_v_i8m1_m(vbool8_t mask, int8_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i16mf4_m(vbool64_t mask, int16_t *base, vuint8mf8_t bindex,
@@ -2577,7 +2577,7 @@ void test_vsoxei8_v_i16mf4_m(vbool64_t mask, int16_t *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i16mf2_m(vbool32_t mask, int16_t *base, vuint8mf4_t bindex,
@@ -2589,7 +2589,7 @@ void test_vsoxei8_v_i16mf2_m(vbool32_t mask, int16_t *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i16m1_m(vbool16_t mask, int16_t *base, vuint8mf2_t bindex,
@@ -2601,7 +2601,7 @@ void test_vsoxei8_v_i16m1_m(vbool16_t mask, int16_t *base, vuint8mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i16m2_m(vbool8_t mask, int16_t *base, vuint8m1_t bindex,
@@ -2613,7 +2613,7 @@ void test_vsoxei8_v_i16m2_m(vbool8_t mask, int16_t *base, vuint8m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i16m4_m(vbool4_t mask, int16_t *base, vuint8m2_t bindex,
@@ -2625,7 +2625,7 @@ void test_vsoxei8_v_i16m4_m(vbool4_t mask, int16_t *base, vuint8m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i16m8_m(vbool2_t mask, int16_t *base, vuint8m4_t bindex,
@@ -2637,7 +2637,7 @@ void test_vsoxei8_v_i16m8_m(vbool2_t mask, int16_t *base, vuint8m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i16mf4_m(vbool64_t mask, int16_t *base,
@@ -2650,7 +2650,7 @@ void test_vsoxei16_v_i16mf4_m(vbool64_t mask, int16_t *base,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i16mf2_m(vbool32_t mask, int16_t *base,
@@ -2663,7 +2663,7 @@ void test_vsoxei16_v_i16mf2_m(vbool32_t mask, int16_t *base,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i16m1_m(vbool16_t mask, int16_t *base, vuint16m1_t bindex,
@@ -2675,7 +2675,7 @@ void test_vsoxei16_v_i16m1_m(vbool16_t mask, int16_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i16m2_m(vbool8_t mask, int16_t *base, vuint16m2_t bindex,
@@ -2687,7 +2687,7 @@ void test_vsoxei16_v_i16m2_m(vbool8_t mask, int16_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i16m4_m(vbool4_t mask, int16_t *base, vuint16m4_t bindex,
@@ -2699,7 +2699,7 @@ void test_vsoxei16_v_i16m4_m(vbool4_t mask, int16_t *base, vuint16m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i16m8_m(vbool2_t mask, int16_t *base, vuint16m8_t bindex,
@@ -2711,7 +2711,7 @@ void test_vsoxei16_v_i16m8_m(vbool2_t mask, int16_t *base, vuint16m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i16mf4_m(vbool64_t mask, int16_t *base,
@@ -2724,7 +2724,7 @@ void test_vsoxei32_v_i16mf4_m(vbool64_t mask, int16_t *base,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i16mf2_m(vbool32_t mask, int16_t *base, vuint32m1_t bindex,
@@ -2736,7 +2736,7 @@ void test_vsoxei32_v_i16mf2_m(vbool32_t mask, int16_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i16m1_m(vbool16_t mask, int16_t *base, vuint32m2_t bindex,
@@ -2748,7 +2748,7 @@ void test_vsoxei32_v_i16m1_m(vbool16_t mask, int16_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i16m2_m(vbool8_t mask, int16_t *base, vuint32m4_t bindex,
@@ -2760,7 +2760,7 @@ void test_vsoxei32_v_i16m2_m(vbool8_t mask, int16_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i16m4_m(vbool4_t mask, int16_t *base, vuint32m8_t bindex,
@@ -2772,7 +2772,7 @@ void test_vsoxei32_v_i16m4_m(vbool4_t mask, int16_t *base, vuint32m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i16mf4_m(vbool64_t mask, int16_t *base, vuint64m1_t bindex,
@@ -2784,7 +2784,7 @@ void test_vsoxei64_v_i16mf4_m(vbool64_t mask, int16_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i16mf2_m(vbool32_t mask, int16_t *base, vuint64m2_t bindex,
@@ -2796,7 +2796,7 @@ void test_vsoxei64_v_i16mf2_m(vbool32_t mask, int16_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i16m1_m(vbool16_t mask, int16_t *base, vuint64m4_t bindex,
@@ -2808,7 +2808,7 @@ void test_vsoxei64_v_i16m1_m(vbool16_t mask, int16_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i16m2_m(vbool8_t mask, int16_t *base, vuint64m8_t bindex,
@@ -2820,7 +2820,7 @@ void test_vsoxei64_v_i16m2_m(vbool8_t mask, int16_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i32mf2_m(vbool64_t mask, int32_t *base, vuint8mf8_t bindex,
@@ -2832,7 +2832,7 @@ void test_vsoxei8_v_i32mf2_m(vbool64_t mask, int32_t *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i32m1_m(vbool32_t mask, int32_t *base, vuint8mf4_t bindex,
@@ -2844,7 +2844,7 @@ void test_vsoxei8_v_i32m1_m(vbool32_t mask, int32_t *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i32m2_m(vbool16_t mask, int32_t *base, vuint8mf2_t bindex,
@@ -2856,7 +2856,7 @@ void test_vsoxei8_v_i32m2_m(vbool16_t mask, int32_t *base, vuint8mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i32m4_m(vbool8_t mask, int32_t *base, vuint8m1_t bindex,
@@ -2868,7 +2868,7 @@ void test_vsoxei8_v_i32m4_m(vbool8_t mask, int32_t *base, vuint8m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i32m8_m(vbool4_t mask, int32_t *base, vuint8m2_t bindex,
@@ -2880,7 +2880,7 @@ void test_vsoxei8_v_i32m8_m(vbool4_t mask, int32_t *base, vuint8m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i32mf2_m(vbool64_t mask, int32_t *base,
@@ -2893,7 +2893,7 @@ void test_vsoxei16_v_i32mf2_m(vbool64_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i32m1_m(vbool32_t mask, int32_t *base, vuint16mf2_t bindex,
@@ -2905,7 +2905,7 @@ void test_vsoxei16_v_i32m1_m(vbool32_t mask, int32_t *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i32m2_m(vbool16_t mask, int32_t *base, vuint16m1_t bindex,
@@ -2917,7 +2917,7 @@ void test_vsoxei16_v_i32m2_m(vbool16_t mask, int32_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i32m4_m(vbool8_t mask, int32_t *base, vuint16m2_t bindex,
@@ -2929,7 +2929,7 @@ void test_vsoxei16_v_i32m4_m(vbool8_t mask, int32_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i32m8_m(vbool4_t mask, int32_t *base, vuint16m4_t bindex,
@@ -2941,7 +2941,7 @@ void test_vsoxei16_v_i32m8_m(vbool4_t mask, int32_t *base, vuint16m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i32mf2_m(vbool64_t mask, int32_t *base,
@@ -2954,7 +2954,7 @@ void test_vsoxei32_v_i32mf2_m(vbool64_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i32m1_m(vbool32_t mask, int32_t *base, vuint32m1_t bindex,
@@ -2966,7 +2966,7 @@ void test_vsoxei32_v_i32m1_m(vbool32_t mask, int32_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i32m2_m(vbool16_t mask, int32_t *base, vuint32m2_t bindex,
@@ -2978,7 +2978,7 @@ void test_vsoxei32_v_i32m2_m(vbool16_t mask, int32_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i32m4_m(vbool8_t mask, int32_t *base, vuint32m4_t bindex,
@@ -2990,7 +2990,7 @@ void test_vsoxei32_v_i32m4_m(vbool8_t mask, int32_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i32m8_m(vbool4_t mask, int32_t *base, vuint32m8_t bindex,
@@ -3002,7 +3002,7 @@ void test_vsoxei32_v_i32m8_m(vbool4_t mask, int32_t *base, vuint32m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i32mf2_m(vbool64_t mask, int32_t *base, vuint64m1_t bindex,
@@ -3014,7 +3014,7 @@ void test_vsoxei64_v_i32mf2_m(vbool64_t mask, int32_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i32m1_m(vbool32_t mask, int32_t *base, vuint64m2_t bindex,
@@ -3026,7 +3026,7 @@ void test_vsoxei64_v_i32m1_m(vbool32_t mask, int32_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i32m2_m(vbool16_t mask, int32_t *base, vuint64m4_t bindex,
@@ -3038,7 +3038,7 @@ void test_vsoxei64_v_i32m2_m(vbool16_t mask, int32_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i32m4_m(vbool8_t mask, int32_t *base, vuint64m8_t bindex,
@@ -3050,7 +3050,7 @@ void test_vsoxei64_v_i32m4_m(vbool8_t mask, int32_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i64m1_m(vbool64_t mask, int64_t *base, vuint8mf8_t bindex,
@@ -3062,7 +3062,7 @@ void test_vsoxei8_v_i64m1_m(vbool64_t mask, int64_t *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i64m2_m(vbool32_t mask, int64_t *base, vuint8mf4_t bindex,
@@ -3074,7 +3074,7 @@ void test_vsoxei8_v_i64m2_m(vbool32_t mask, int64_t *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i64m4_m(vbool16_t mask, int64_t *base, vuint8mf2_t bindex,
@@ -3086,7 +3086,7 @@ void test_vsoxei8_v_i64m4_m(vbool16_t mask, int64_t *base, vuint8mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_i64m8_m(vbool8_t mask, int64_t *base, vuint8m1_t bindex,
@@ -3098,7 +3098,7 @@ void test_vsoxei8_v_i64m8_m(vbool8_t mask, int64_t *base, vuint8m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i64m1_m(vbool64_t mask, int64_t *base, vuint16mf4_t bindex,
@@ -3110,7 +3110,7 @@ void test_vsoxei16_v_i64m1_m(vbool64_t mask, int64_t *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i64m2_m(vbool32_t mask, int64_t *base, vuint16mf2_t bindex,
@@ -3122,7 +3122,7 @@ void test_vsoxei16_v_i64m2_m(vbool32_t mask, int64_t *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i64m4_m(vbool16_t mask, int64_t *base, vuint16m1_t bindex,
@@ -3134,7 +3134,7 @@ void test_vsoxei16_v_i64m4_m(vbool16_t mask, int64_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_i64m8_m(vbool8_t mask, int64_t *base, vuint16m2_t bindex,
@@ -3146,7 +3146,7 @@ void test_vsoxei16_v_i64m8_m(vbool8_t mask, int64_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i64m1_m(vbool64_t mask, int64_t *base, vuint32mf2_t bindex,
@@ -3158,7 +3158,7 @@ void test_vsoxei32_v_i64m1_m(vbool64_t mask, int64_t *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i64m2_m(vbool32_t mask, int64_t *base, vuint32m1_t bindex,
@@ -3170,7 +3170,7 @@ void test_vsoxei32_v_i64m2_m(vbool32_t mask, int64_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i64m4_m(vbool16_t mask, int64_t *base, vuint32m2_t bindex,
@@ -3182,7 +3182,7 @@ void test_vsoxei32_v_i64m4_m(vbool16_t mask, int64_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_i64m8_m(vbool8_t mask, int64_t *base, vuint32m4_t bindex,
@@ -3194,7 +3194,7 @@ void test_vsoxei32_v_i64m8_m(vbool8_t mask, int64_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i64m1_m(vbool64_t mask, int64_t *base, vuint64m1_t bindex,
@@ -3206,7 +3206,7 @@ void test_vsoxei64_v_i64m1_m(vbool64_t mask, int64_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i64m2_m(vbool32_t mask, int64_t *base, vuint64m2_t bindex,
@@ -3218,7 +3218,7 @@ void test_vsoxei64_v_i64m2_m(vbool32_t mask, int64_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i64m4_m(vbool16_t mask, int64_t *base, vuint64m4_t bindex,
@@ -3230,7 +3230,7 @@ void test_vsoxei64_v_i64m4_m(vbool16_t mask, int64_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_i64m8_m(vbool8_t mask, int64_t *base, vuint64m8_t bindex,
@@ -3242,7 +3242,7 @@ void test_vsoxei64_v_i64m8_m(vbool8_t mask, int64_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u8mf8_m(vbool64_t mask, uint8_t *base, vuint8mf8_t bindex,
@@ -3254,7 +3254,7 @@ void test_vsoxei8_v_u8mf8_m(vbool64_t mask, uint8_t *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u8mf4_m(vbool32_t mask, uint8_t *base, vuint8mf4_t bindex,
@@ -3266,7 +3266,7 @@ void test_vsoxei8_v_u8mf4_m(vbool32_t mask, uint8_t *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u8mf2_m(vbool16_t mask, uint8_t *base, vuint8mf2_t bindex,
@@ -3278,7 +3278,7 @@ void test_vsoxei8_v_u8mf2_m(vbool16_t mask, uint8_t *base, vuint8mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u8m1_m(vbool8_t mask, uint8_t *base, vuint8m1_t bindex,
@@ -3290,7 +3290,7 @@ void test_vsoxei8_v_u8m1_m(vbool8_t mask, uint8_t *base, vuint8m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u8m2_m(vbool4_t mask, uint8_t *base, vuint8m2_t bindex,
@@ -3302,7 +3302,7 @@ void test_vsoxei8_v_u8m2_m(vbool4_t mask, uint8_t *base, vuint8m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u8m4_m(vbool2_t mask, uint8_t *base, vuint8m4_t bindex,
@@ -3314,7 +3314,7 @@ void test_vsoxei8_v_u8m4_m(vbool2_t mask, uint8_t *base, vuint8m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u8m8_m(vbool1_t mask, uint8_t *base, vuint8m8_t bindex,
@@ -3326,7 +3326,7 @@ void test_vsoxei8_v_u8m8_m(vbool1_t mask, uint8_t *base, vuint8m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u8mf8_m(vbool64_t mask, uint8_t *base, vuint16mf4_t bindex,
@@ -3338,7 +3338,7 @@ void test_vsoxei16_v_u8mf8_m(vbool64_t mask, uint8_t *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u8mf4_m(vbool32_t mask, uint8_t *base, vuint16mf2_t bindex,
@@ -3350,7 +3350,7 @@ void test_vsoxei16_v_u8mf4_m(vbool32_t mask, uint8_t *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u8mf2_m(vbool16_t mask, uint8_t *base, vuint16m1_t bindex,
@@ -3362,7 +3362,7 @@ void test_vsoxei16_v_u8mf2_m(vbool16_t mask, uint8_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u8m1_m(vbool8_t mask, uint8_t *base, vuint16m2_t bindex,
@@ -3374,7 +3374,7 @@ void test_vsoxei16_v_u8m1_m(vbool8_t mask, uint8_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u8m2_m(vbool4_t mask, uint8_t *base, vuint16m4_t bindex,
@@ -3386,7 +3386,7 @@ void test_vsoxei16_v_u8m2_m(vbool4_t mask, uint8_t *base, vuint16m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u8m4_m(vbool2_t mask, uint8_t *base, vuint16m8_t bindex,
@@ -3398,7 +3398,7 @@ void test_vsoxei16_v_u8m4_m(vbool2_t mask, uint8_t *base, vuint16m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u8mf8_m(vbool64_t mask, uint8_t *base, vuint32mf2_t bindex,
@@ -3410,7 +3410,7 @@ void test_vsoxei32_v_u8mf8_m(vbool64_t mask, uint8_t *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u8mf4_m(vbool32_t mask, uint8_t *base, vuint32m1_t bindex,
@@ -3422,7 +3422,7 @@ void test_vsoxei32_v_u8mf4_m(vbool32_t mask, uint8_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u8mf2_m(vbool16_t mask, uint8_t *base, vuint32m2_t bindex,
@@ -3434,7 +3434,7 @@ void test_vsoxei32_v_u8mf2_m(vbool16_t mask, uint8_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u8m1_m(vbool8_t mask, uint8_t *base, vuint32m4_t bindex,
@@ -3446,7 +3446,7 @@ void test_vsoxei32_v_u8m1_m(vbool8_t mask, uint8_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u8m2_m(vbool4_t mask, uint8_t *base, vuint32m8_t bindex,
@@ -3458,7 +3458,7 @@ void test_vsoxei32_v_u8m2_m(vbool4_t mask, uint8_t *base, vuint32m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u8mf8_m(vbool64_t mask, uint8_t *base, vuint64m1_t bindex,
@@ -3470,7 +3470,7 @@ void test_vsoxei64_v_u8mf8_m(vbool64_t mask, uint8_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u8mf4_m(vbool32_t mask, uint8_t *base, vuint64m2_t bindex,
@@ -3482,7 +3482,7 @@ void test_vsoxei64_v_u8mf4_m(vbool32_t mask, uint8_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u8mf2_m(vbool16_t mask, uint8_t *base, vuint64m4_t bindex,
@@ -3494,7 +3494,7 @@ void test_vsoxei64_v_u8mf2_m(vbool16_t mask, uint8_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u8m1_m(vbool8_t mask, uint8_t *base, vuint64m8_t bindex,
@@ -3506,7 +3506,7 @@ void test_vsoxei64_v_u8m1_m(vbool8_t mask, uint8_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u16mf4_m(vbool64_t mask, uint16_t *base, vuint8mf8_t bindex,
@@ -3518,7 +3518,7 @@ void test_vsoxei8_v_u16mf4_m(vbool64_t mask, uint16_t *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u16mf2_m(vbool32_t mask, uint16_t *base, vuint8mf4_t bindex,
@@ -3530,7 +3530,7 @@ void test_vsoxei8_v_u16mf2_m(vbool32_t mask, uint16_t *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u16m1_m(vbool16_t mask, uint16_t *base, vuint8mf2_t bindex,
@@ -3542,7 +3542,7 @@ void test_vsoxei8_v_u16m1_m(vbool16_t mask, uint16_t *base, vuint8mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u16m2_m(vbool8_t mask, uint16_t *base, vuint8m1_t bindex,
@@ -3554,7 +3554,7 @@ void test_vsoxei8_v_u16m2_m(vbool8_t mask, uint16_t *base, vuint8m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u16m4_m(vbool4_t mask, uint16_t *base, vuint8m2_t bindex,
@@ -3566,7 +3566,7 @@ void test_vsoxei8_v_u16m4_m(vbool4_t mask, uint16_t *base, vuint8m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u16m8_m(vbool2_t mask, uint16_t *base, vuint8m4_t bindex,
@@ -3578,7 +3578,7 @@ void test_vsoxei8_v_u16m8_m(vbool2_t mask, uint16_t *base, vuint8m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u16mf4_m(vbool64_t mask, uint16_t *base,
@@ -3591,7 +3591,7 @@ void test_vsoxei16_v_u16mf4_m(vbool64_t mask, uint16_t *base,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u16mf2_m(vbool32_t mask, uint16_t *base,
@@ -3604,7 +3604,7 @@ void test_vsoxei16_v_u16mf2_m(vbool32_t mask, uint16_t *base,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u16m1_m(vbool16_t mask, uint16_t *base, vuint16m1_t bindex,
@@ -3616,7 +3616,7 @@ void test_vsoxei16_v_u16m1_m(vbool16_t mask, uint16_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u16m2_m(vbool8_t mask, uint16_t *base, vuint16m2_t bindex,
@@ -3628,7 +3628,7 @@ void test_vsoxei16_v_u16m2_m(vbool8_t mask, uint16_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u16m4_m(vbool4_t mask, uint16_t *base, vuint16m4_t bindex,
@@ -3640,7 +3640,7 @@ void test_vsoxei16_v_u16m4_m(vbool4_t mask, uint16_t *base, vuint16m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u16m8_m(vbool2_t mask, uint16_t *base, vuint16m8_t bindex,
@@ -3652,7 +3652,7 @@ void test_vsoxei16_v_u16m8_m(vbool2_t mask, uint16_t *base, vuint16m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u16mf4_m(vbool64_t mask, uint16_t *base,
@@ -3665,7 +3665,7 @@ void test_vsoxei32_v_u16mf4_m(vbool64_t mask, uint16_t *base,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u16mf2_m(vbool32_t mask, uint16_t *base,
@@ -3678,7 +3678,7 @@ void test_vsoxei32_v_u16mf2_m(vbool32_t mask, uint16_t *base,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u16m1_m(vbool16_t mask, uint16_t *base, vuint32m2_t bindex,
@@ -3690,7 +3690,7 @@ void test_vsoxei32_v_u16m1_m(vbool16_t mask, uint16_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u16m2_m(vbool8_t mask, uint16_t *base, vuint32m4_t bindex,
@@ -3702,7 +3702,7 @@ void test_vsoxei32_v_u16m2_m(vbool8_t mask, uint16_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u16m4_m(vbool4_t mask, uint16_t *base, vuint32m8_t bindex,
@@ -3714,7 +3714,7 @@ void test_vsoxei32_v_u16m4_m(vbool4_t mask, uint16_t *base, vuint32m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u16mf4_m(vbool64_t mask, uint16_t *base,
@@ -3727,7 +3727,7 @@ void test_vsoxei64_v_u16mf4_m(vbool64_t mask, uint16_t *base,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u16mf2_m(vbool32_t mask, uint16_t *base,
@@ -3740,7 +3740,7 @@ void test_vsoxei64_v_u16mf2_m(vbool32_t mask, uint16_t *base,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u16m1_m(vbool16_t mask, uint16_t *base, vuint64m4_t bindex,
@@ -3752,7 +3752,7 @@ void test_vsoxei64_v_u16m1_m(vbool16_t mask, uint16_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u16m2_m(vbool8_t mask, uint16_t *base, vuint64m8_t bindex,
@@ -3764,7 +3764,7 @@ void test_vsoxei64_v_u16m2_m(vbool8_t mask, uint16_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u32mf2_m(vbool64_t mask, uint32_t *base, vuint8mf8_t bindex,
@@ -3776,7 +3776,7 @@ void test_vsoxei8_v_u32mf2_m(vbool64_t mask, uint32_t *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u32m1_m(vbool32_t mask, uint32_t *base, vuint8mf4_t bindex,
@@ -3788,7 +3788,7 @@ void test_vsoxei8_v_u32m1_m(vbool32_t mask, uint32_t *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u32m2_m(vbool16_t mask, uint32_t *base, vuint8mf2_t bindex,
@@ -3800,7 +3800,7 @@ void test_vsoxei8_v_u32m2_m(vbool16_t mask, uint32_t *base, vuint8mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u32m4_m(vbool8_t mask, uint32_t *base, vuint8m1_t bindex,
@@ -3812,7 +3812,7 @@ void test_vsoxei8_v_u32m4_m(vbool8_t mask, uint32_t *base, vuint8m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u32m8_m(vbool4_t mask, uint32_t *base, vuint8m2_t bindex,
@@ -3824,7 +3824,7 @@ void test_vsoxei8_v_u32m8_m(vbool4_t mask, uint32_t *base, vuint8m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u32mf2_m(vbool64_t mask, uint32_t *base,
@@ -3837,7 +3837,7 @@ void test_vsoxei16_v_u32mf2_m(vbool64_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u32m1_m(vbool32_t mask, uint32_t *base,
@@ -3850,7 +3850,7 @@ void test_vsoxei16_v_u32m1_m(vbool32_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u32m2_m(vbool16_t mask, uint32_t *base, vuint16m1_t bindex,
@@ -3862,7 +3862,7 @@ void test_vsoxei16_v_u32m2_m(vbool16_t mask, uint32_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u32m4_m(vbool8_t mask, uint32_t *base, vuint16m2_t bindex,
@@ -3874,7 +3874,7 @@ void test_vsoxei16_v_u32m4_m(vbool8_t mask, uint32_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u32m8_m(vbool4_t mask, uint32_t *base, vuint16m4_t bindex,
@@ -3886,7 +3886,7 @@ void test_vsoxei16_v_u32m8_m(vbool4_t mask, uint32_t *base, vuint16m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u32mf2_m(vbool64_t mask, uint32_t *base,
@@ -3899,7 +3899,7 @@ void test_vsoxei32_v_u32mf2_m(vbool64_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u32m1_m(vbool32_t mask, uint32_t *base, vuint32m1_t bindex,
@@ -3911,7 +3911,7 @@ void test_vsoxei32_v_u32m1_m(vbool32_t mask, uint32_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u32m2_m(vbool16_t mask, uint32_t *base, vuint32m2_t bindex,
@@ -3923,7 +3923,7 @@ void test_vsoxei32_v_u32m2_m(vbool16_t mask, uint32_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u32m4_m(vbool8_t mask, uint32_t *base, vuint32m4_t bindex,
@@ -3935,7 +3935,7 @@ void test_vsoxei32_v_u32m4_m(vbool8_t mask, uint32_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u32m8_m(vbool4_t mask, uint32_t *base, vuint32m8_t bindex,
@@ -3947,7 +3947,7 @@ void test_vsoxei32_v_u32m8_m(vbool4_t mask, uint32_t *base, vuint32m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u32mf2_m(vbool64_t mask, uint32_t *base,
@@ -3960,7 +3960,7 @@ void test_vsoxei64_v_u32mf2_m(vbool64_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u32m1_m(vbool32_t mask, uint32_t *base, vuint64m2_t bindex,
@@ -3972,7 +3972,7 @@ void test_vsoxei64_v_u32m1_m(vbool32_t mask, uint32_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u32m2_m(vbool16_t mask, uint32_t *base, vuint64m4_t bindex,
@@ -3984,7 +3984,7 @@ void test_vsoxei64_v_u32m2_m(vbool16_t mask, uint32_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u32m4_m(vbool8_t mask, uint32_t *base, vuint64m8_t bindex,
@@ -3996,7 +3996,7 @@ void test_vsoxei64_v_u32m4_m(vbool8_t mask, uint32_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u64m1_m(vbool64_t mask, uint64_t *base, vuint8mf8_t bindex,
@@ -4008,7 +4008,7 @@ void test_vsoxei8_v_u64m1_m(vbool64_t mask, uint64_t *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u64m2_m(vbool32_t mask, uint64_t *base, vuint8mf4_t bindex,
@@ -4020,7 +4020,7 @@ void test_vsoxei8_v_u64m2_m(vbool32_t mask, uint64_t *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u64m4_m(vbool16_t mask, uint64_t *base, vuint8mf2_t bindex,
@@ -4032,7 +4032,7 @@ void test_vsoxei8_v_u64m4_m(vbool16_t mask, uint64_t *base, vuint8mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_u64m8_m(vbool8_t mask, uint64_t *base, vuint8m1_t bindex,
@@ -4044,7 +4044,7 @@ void test_vsoxei8_v_u64m8_m(vbool8_t mask, uint64_t *base, vuint8m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u64m1_m(vbool64_t mask, uint64_t *base,
@@ -4057,7 +4057,7 @@ void test_vsoxei16_v_u64m1_m(vbool64_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u64m2_m(vbool32_t mask, uint64_t *base,
@@ -4070,7 +4070,7 @@ void test_vsoxei16_v_u64m2_m(vbool32_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u64m4_m(vbool16_t mask, uint64_t *base, vuint16m1_t bindex,
@@ -4082,7 +4082,7 @@ void test_vsoxei16_v_u64m4_m(vbool16_t mask, uint64_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_u64m8_m(vbool8_t mask, uint64_t *base, vuint16m2_t bindex,
@@ -4094,7 +4094,7 @@ void test_vsoxei16_v_u64m8_m(vbool8_t mask, uint64_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u64m1_m(vbool64_t mask, uint64_t *base,
@@ -4107,7 +4107,7 @@ void test_vsoxei32_v_u64m1_m(vbool64_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u64m2_m(vbool32_t mask, uint64_t *base, vuint32m1_t bindex,
@@ -4119,7 +4119,7 @@ void test_vsoxei32_v_u64m2_m(vbool32_t mask, uint64_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u64m4_m(vbool16_t mask, uint64_t *base, vuint32m2_t bindex,
@@ -4131,7 +4131,7 @@ void test_vsoxei32_v_u64m4_m(vbool16_t mask, uint64_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_u64m8_m(vbool8_t mask, uint64_t *base, vuint32m4_t bindex,
@@ -4143,7 +4143,7 @@ void test_vsoxei32_v_u64m8_m(vbool8_t mask, uint64_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u64m1_m(vbool64_t mask, uint64_t *base, vuint64m1_t bindex,
@@ -4155,7 +4155,7 @@ void test_vsoxei64_v_u64m1_m(vbool64_t mask, uint64_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u64m2_m(vbool32_t mask, uint64_t *base, vuint64m2_t bindex,
@@ -4167,7 +4167,7 @@ void test_vsoxei64_v_u64m2_m(vbool32_t mask, uint64_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u64m4_m(vbool16_t mask, uint64_t *base, vuint64m4_t bindex,
@@ -4179,7 +4179,7 @@ void test_vsoxei64_v_u64m4_m(vbool16_t mask, uint64_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_u64m8_m(vbool8_t mask, uint64_t *base, vuint64m8_t bindex,
@@ -4191,7 +4191,7 @@ void test_vsoxei64_v_u64m8_m(vbool8_t mask, uint64_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1f32.nxv1i8.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1f32.nxv1i8.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_f32mf2_m(vbool64_t mask, float *base, vuint8mf8_t bindex,
@@ -4203,7 +4203,7 @@ void test_vsoxei8_v_f32mf2_m(vbool64_t mask, float *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2f32.nxv2i8.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2f32.nxv2i8.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_f32m1_m(vbool32_t mask, float *base, vuint8mf4_t bindex,
@@ -4215,7 +4215,7 @@ void test_vsoxei8_v_f32m1_m(vbool32_t mask, float *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4f32.nxv4i8.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4f32.nxv4i8.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_f32m2_m(vbool16_t mask, float *base, vuint8mf2_t bindex,
@@ -4227,7 +4227,7 @@ void test_vsoxei8_v_f32m2_m(vbool16_t mask, float *base, vuint8mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8f32.nxv8i8.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8f32.nxv8i8.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_f32m4_m(vbool8_t mask, float *base, vuint8m1_t bindex,
@@ -4239,7 +4239,7 @@ void test_vsoxei8_v_f32m4_m(vbool8_t mask, float *base, vuint8m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16f32.nxv16i8.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16f32.nxv16i8.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_f32m8_m(vbool4_t mask, float *base, vuint8m2_t bindex,
@@ -4251,7 +4251,7 @@ void test_vsoxei8_v_f32m8_m(vbool4_t mask, float *base, vuint8m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1f32.nxv1i16.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1f32.nxv1i16.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_f32mf2_m(vbool64_t mask, float *base, vuint16mf4_t bindex,
@@ -4263,7 +4263,7 @@ void test_vsoxei16_v_f32mf2_m(vbool64_t mask, float *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2f32.nxv2i16.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2f32.nxv2i16.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_f32m1_m(vbool32_t mask, float *base, vuint16mf2_t bindex,
@@ -4275,7 +4275,7 @@ void test_vsoxei16_v_f32m1_m(vbool32_t mask, float *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4f32.nxv4i16.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4f32.nxv4i16.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_f32m2_m(vbool16_t mask, float *base, vuint16m1_t bindex,
@@ -4287,7 +4287,7 @@ void test_vsoxei16_v_f32m2_m(vbool16_t mask, float *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8f32.nxv8i16.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8f32.nxv8i16.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_f32m4_m(vbool8_t mask, float *base, vuint16m2_t bindex,
@@ -4299,7 +4299,7 @@ void test_vsoxei16_v_f32m4_m(vbool8_t mask, float *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16f32.nxv16i16.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16f32.nxv16i16.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_f32m8_m(vbool4_t mask, float *base, vuint16m4_t bindex,
@@ -4311,7 +4311,7 @@ void test_vsoxei16_v_f32m8_m(vbool4_t mask, float *base, vuint16m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1f32.nxv1i32.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1f32.nxv1i32.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_f32mf2_m(vbool64_t mask, float *base, vuint32mf2_t bindex,
@@ -4323,7 +4323,7 @@ void test_vsoxei32_v_f32mf2_m(vbool64_t mask, float *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2f32.nxv2i32.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2f32.nxv2i32.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_f32m1_m(vbool32_t mask, float *base, vuint32m1_t bindex,
@@ -4335,7 +4335,7 @@ void test_vsoxei32_v_f32m1_m(vbool32_t mask, float *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4f32.nxv4i32.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4f32.nxv4i32.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_f32m2_m(vbool16_t mask, float *base, vuint32m2_t bindex,
@@ -4347,7 +4347,7 @@ void test_vsoxei32_v_f32m2_m(vbool16_t mask, float *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8f32.nxv8i32.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8f32.nxv8i32.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_f32m4_m(vbool8_t mask, float *base, vuint32m4_t bindex,
@@ -4359,7 +4359,7 @@ void test_vsoxei32_v_f32m4_m(vbool8_t mask, float *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16f32.nxv16i32.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv16f32.nxv16i32.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_f32m8_m(vbool4_t mask, float *base, vuint32m8_t bindex,
@@ -4371,7 +4371,7 @@ void test_vsoxei32_v_f32m8_m(vbool4_t mask, float *base, vuint32m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1f32.nxv1i64.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1f32.nxv1i64.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_f32mf2_m(vbool64_t mask, float *base, vuint64m1_t bindex,
@@ -4383,7 +4383,7 @@ void test_vsoxei64_v_f32mf2_m(vbool64_t mask, float *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2f32.nxv2i64.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2f32.nxv2i64.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_f32m1_m(vbool32_t mask, float *base, vuint64m2_t bindex,
@@ -4395,7 +4395,7 @@ void test_vsoxei64_v_f32m1_m(vbool32_t mask, float *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4f32.nxv4i64.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4f32.nxv4i64.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_f32m2_m(vbool16_t mask, float *base, vuint64m4_t bindex,
@@ -4407,7 +4407,7 @@ void test_vsoxei64_v_f32m2_m(vbool16_t mask, float *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8f32.nxv8i64.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8f32.nxv8i64.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_f32m4_m(vbool8_t mask, float *base, vuint64m8_t bindex,
@@ -4419,7 +4419,7 @@ void test_vsoxei64_v_f32m4_m(vbool8_t mask, float *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1f64.nxv1i8.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1f64.nxv1i8.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_f64m1_m(vbool64_t mask, double *base, vuint8mf8_t bindex,
@@ -4431,7 +4431,7 @@ void test_vsoxei8_v_f64m1_m(vbool64_t mask, double *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2f64.nxv2i8.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2f64.nxv2i8.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_f64m2_m(vbool32_t mask, double *base, vuint8mf4_t bindex,
@@ -4443,7 +4443,7 @@ void test_vsoxei8_v_f64m2_m(vbool32_t mask, double *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4f64.nxv4i8.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4f64.nxv4i8.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_f64m4_m(vbool16_t mask, double *base, vuint8mf2_t bindex,
@@ -4455,7 +4455,7 @@ void test_vsoxei8_v_f64m4_m(vbool16_t mask, double *base, vuint8mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei8_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8f64.nxv8i8.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8f64.nxv8i8.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei8_v_f64m8_m(vbool8_t mask, double *base, vuint8m1_t bindex,
@@ -4467,7 +4467,7 @@ void test_vsoxei8_v_f64m8_m(vbool8_t mask, double *base, vuint8m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1f64.nxv1i16.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1f64.nxv1i16.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_f64m1_m(vbool64_t mask, double *base, vuint16mf4_t bindex,
@@ -4479,7 +4479,7 @@ void test_vsoxei16_v_f64m1_m(vbool64_t mask, double *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2f64.nxv2i16.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2f64.nxv2i16.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_f64m2_m(vbool32_t mask, double *base, vuint16mf2_t bindex,
@@ -4491,7 +4491,7 @@ void test_vsoxei16_v_f64m2_m(vbool32_t mask, double *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4f64.nxv4i16.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4f64.nxv4i16.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_f64m4_m(vbool16_t mask, double *base, vuint16m1_t bindex,
@@ -4503,7 +4503,7 @@ void test_vsoxei16_v_f64m4_m(vbool16_t mask, double *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei16_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8f64.nxv8i16.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8f64.nxv8i16.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei16_v_f64m8_m(vbool8_t mask, double *base, vuint16m2_t bindex,
@@ -4515,7 +4515,7 @@ void test_vsoxei16_v_f64m8_m(vbool8_t mask, double *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1f64.nxv1i32.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1f64.nxv1i32.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_f64m1_m(vbool64_t mask, double *base, vuint32mf2_t bindex,
@@ -4527,7 +4527,7 @@ void test_vsoxei32_v_f64m1_m(vbool64_t mask, double *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2f64.nxv2i32.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2f64.nxv2i32.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_f64m2_m(vbool32_t mask, double *base, vuint32m1_t bindex,
@@ -4539,7 +4539,7 @@ void test_vsoxei32_v_f64m2_m(vbool32_t mask, double *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4f64.nxv4i32.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4f64.nxv4i32.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_f64m4_m(vbool16_t mask, double *base, vuint32m2_t bindex,
@@ -4551,7 +4551,7 @@ void test_vsoxei32_v_f64m4_m(vbool16_t mask, double *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei32_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8f64.nxv8i32.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8f64.nxv8i32.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei32_v_f64m8_m(vbool8_t mask, double *base, vuint32m4_t bindex,
@@ -4563,7 +4563,7 @@ void test_vsoxei32_v_f64m8_m(vbool8_t mask, double *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1f64.nxv1i64.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv1f64.nxv1i64.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_f64m1_m(vbool64_t mask, double *base, vuint64m1_t bindex,
@@ -4575,7 +4575,7 @@ void test_vsoxei64_v_f64m1_m(vbool64_t mask, double *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2f64.nxv2i64.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv2f64.nxv2i64.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_f64m2_m(vbool32_t mask, double *base, vuint64m2_t bindex,
@@ -4587,7 +4587,7 @@ void test_vsoxei64_v_f64m2_m(vbool32_t mask, double *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4f64.nxv4i64.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv4f64.nxv4i64.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_f64m4_m(vbool16_t mask, double *base, vuint64m4_t bindex,
@@ -4599,7 +4599,7 @@ void test_vsoxei64_v_f64m4_m(vbool16_t mask, double *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsoxei64_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8f64.nxv8i64.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsoxei.mask.nxv8f64.nxv8i64.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsoxei64_v_f64m8_m(vbool8_t mask, double *base, vuint64m8_t bindex,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsra.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsra.c
index c88d491..389f4b0 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsra.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsra.c
@@ -447,7 +447,7 @@ vint64m8_t test_vsra_vx_i64m8(vint64m8_t op1, size_t shift, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsra.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsra.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vsra_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, vuint8mf8_t shift, size_t vl) {
@@ -457,7 +457,7 @@ vint8mf8_t test_vsra_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsra.mask.nxv1i8.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsra.mask.nxv1i8.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vsra_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, size_t shift, size_t vl) {
@@ -467,7 +467,7 @@ vint8mf8_t test_vsra_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsra.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsra.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vsra_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, vuint8mf4_t shift, size_t vl) {
@@ -477,7 +477,7 @@ vint8mf4_t test_vsra_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsra.mask.nxv2i8.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsra.mask.nxv2i8.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vsra_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, size_t shift, size_t vl) {
@@ -487,7 +487,7 @@ vint8mf4_t test_vsra_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsra.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsra.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vsra_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, vuint8mf2_t shift, size_t vl) {
@@ -497,7 +497,7 @@ vint8mf2_t test_vsra_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsra.mask.nxv4i8.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsra.mask.nxv4i8.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vsra_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, size_t shift, size_t vl) {
@@ -507,7 +507,7 @@ vint8mf2_t test_vsra_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsra.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsra.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vsra_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, vuint8m1_t shift, size_t vl) {
@@ -517,7 +517,7 @@ vint8m1_t test_vsra_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsra.mask.nxv8i8.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsra.mask.nxv8i8.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vsra_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, size_t shift, size_t vl) {
@@ -527,7 +527,7 @@ vint8m1_t test_vsra_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsra.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsra.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vsra_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, vuint8m2_t shift, size_t vl) {
@@ -537,7 +537,7 @@ vint8m2_t test_vsra_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsra.mask.nxv16i8.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsra.mask.nxv16i8.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vsra_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, size_t shift, size_t vl) {
@@ -547,7 +547,7 @@ vint8m2_t test_vsra_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsra.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsra.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vsra_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, vuint8m4_t shift, size_t vl) {
@@ -557,7 +557,7 @@ vint8m4_t test_vsra_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsra.mask.nxv32i8.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsra.mask.nxv32i8.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vsra_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, size_t shift, size_t vl) {
@@ -567,7 +567,7 @@ vint8m4_t test_vsra_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsra.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsra.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vsra_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, vuint8m8_t shift, size_t vl) {
@@ -577,7 +577,7 @@ vint8m8_t test_vsra_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsra.mask.nxv64i8.i64.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsra.mask.nxv64i8.i64.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vsra_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, size_t shift, size_t vl) {
@@ -587,7 +587,7 @@ vint8m8_t test_vsra_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsra.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsra.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vsra_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, vuint16mf4_t shift, size_t vl) {
@@ -597,7 +597,7 @@ vint16mf4_t test_vsra_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsra.mask.nxv1i16.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsra.mask.nxv1i16.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vsra_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, size_t shift, size_t vl) {
@@ -607,7 +607,7 @@ vint16mf4_t test_vsra_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsra.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsra.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vsra_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, vuint16mf2_t shift, size_t vl) {
@@ -617,7 +617,7 @@ vint16mf2_t test_vsra_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsra.mask.nxv2i16.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsra.mask.nxv2i16.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vsra_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, size_t shift, size_t vl) {
@@ -627,7 +627,7 @@ vint16mf2_t test_vsra_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsra.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsra.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vsra_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, vuint16m1_t shift, size_t vl) {
@@ -637,7 +637,7 @@ vint16m1_t test_vsra_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsra.mask.nxv4i16.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsra.mask.nxv4i16.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vsra_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, size_t shift, size_t vl) {
@@ -647,7 +647,7 @@ vint16m1_t test_vsra_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsra.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsra.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vsra_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, vuint16m2_t shift, size_t vl) {
@@ -657,7 +657,7 @@ vint16m2_t test_vsra_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsra.mask.nxv8i16.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsra.mask.nxv8i16.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vsra_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, size_t shift, size_t vl) {
@@ -667,7 +667,7 @@ vint16m2_t test_vsra_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsra.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsra.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vsra_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, vuint16m4_t shift, size_t vl) {
@@ -677,7 +677,7 @@ vint16m4_t test_vsra_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsra.mask.nxv16i16.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsra.mask.nxv16i16.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vsra_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, size_t shift, size_t vl) {
@@ -687,7 +687,7 @@ vint16m4_t test_vsra_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsra.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsra.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vsra_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, vuint16m8_t shift, size_t vl) {
@@ -697,7 +697,7 @@ vint16m8_t test_vsra_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsra.mask.nxv32i16.i64.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsra.mask.nxv32i16.i64.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vsra_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, size_t shift, size_t vl) {
@@ -707,7 +707,7 @@ vint16m8_t test_vsra_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsra.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsra.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vsra_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, vuint32mf2_t shift, size_t vl) {
@@ -717,7 +717,7 @@ vint32mf2_t test_vsra_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsra.mask.nxv1i32.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsra.mask.nxv1i32.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vsra_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, size_t shift, size_t vl) {
@@ -727,7 +727,7 @@ vint32mf2_t test_vsra_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsra.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsra.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vsra_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, vuint32m1_t shift, size_t vl) {
@@ -737,7 +737,7 @@ vint32m1_t test_vsra_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsra.mask.nxv2i32.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsra.mask.nxv2i32.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vsra_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, size_t shift, size_t vl) {
@@ -747,7 +747,7 @@ vint32m1_t test_vsra_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsra.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsra.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vsra_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, vuint32m2_t shift, size_t vl) {
@@ -757,7 +757,7 @@ vint32m2_t test_vsra_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsra.mask.nxv4i32.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsra.mask.nxv4i32.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vsra_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, size_t shift, size_t vl) {
@@ -767,7 +767,7 @@ vint32m2_t test_vsra_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsra.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsra.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vsra_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, vuint32m4_t shift, size_t vl) {
@@ -777,7 +777,7 @@ vint32m4_t test_vsra_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsra.mask.nxv8i32.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsra.mask.nxv8i32.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vsra_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, size_t shift, size_t vl) {
@@ -787,7 +787,7 @@ vint32m4_t test_vsra_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsra.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsra.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vsra_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, vuint32m8_t shift, size_t vl) {
@@ -797,7 +797,7 @@ vint32m8_t test_vsra_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsra.mask.nxv16i32.i64.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsra.mask.nxv16i32.i64.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vsra_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, size_t shift, size_t vl) {
@@ -807,7 +807,7 @@ vint32m8_t test_vsra_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsra.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsra.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vsra_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, vuint64m1_t shift, size_t vl) {
@@ -817,7 +817,7 @@ vint64m1_t test_vsra_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsra.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsra.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vsra_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, size_t shift, size_t vl) {
@@ -827,7 +827,7 @@ vint64m1_t test_vsra_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsra.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsra.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vsra_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, vuint64m2_t shift, size_t vl) {
@@ -837,7 +837,7 @@ vint64m2_t test_vsra_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsra.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsra.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vsra_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, size_t shift, size_t vl) {
@@ -847,7 +847,7 @@ vint64m2_t test_vsra_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsra.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsra.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vsra_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, vuint64m4_t shift, size_t vl) {
@@ -857,7 +857,7 @@ vint64m4_t test_vsra_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsra.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsra.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vsra_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, size_t shift, size_t vl) {
@@ -867,7 +867,7 @@ vint64m4_t test_vsra_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsra.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsra.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vsra_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, vuint64m8_t shift, size_t vl) {
@@ -877,7 +877,7 @@ vint64m8_t test_vsra_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vsra_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsra.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsra.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vsra_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, size_t shift, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsrl.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsrl.c
index 71c9973..afc25dc 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsrl.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsrl.c
@@ -447,7 +447,7 @@ vuint64m8_t test_vsrl_vx_u64m8(vuint64m8_t op1, size_t shift, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsrl.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsrl.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vsrl_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, vuint8mf8_t shift, size_t vl) {
@@ -457,7 +457,7 @@ vuint8mf8_t test_vsrl_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsrl.mask.nxv1i8.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsrl.mask.nxv1i8.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vsrl_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, size_t shift, size_t vl) {
@@ -467,7 +467,7 @@ vuint8mf8_t test_vsrl_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsrl.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsrl.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vsrl_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, vuint8mf4_t shift, size_t vl) {
@@ -477,7 +477,7 @@ vuint8mf4_t test_vsrl_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsrl.mask.nxv2i8.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsrl.mask.nxv2i8.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vsrl_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, size_t shift, size_t vl) {
@@ -487,7 +487,7 @@ vuint8mf4_t test_vsrl_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsrl.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsrl.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vsrl_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, vuint8mf2_t shift, size_t vl) {
@@ -497,7 +497,7 @@ vuint8mf2_t test_vsrl_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsrl.mask.nxv4i8.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsrl.mask.nxv4i8.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vsrl_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, size_t shift, size_t vl) {
@@ -507,7 +507,7 @@ vuint8mf2_t test_vsrl_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsrl.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsrl.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vsrl_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, vuint8m1_t shift, size_t vl) {
@@ -517,7 +517,7 @@ vuint8m1_t test_vsrl_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t o
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsrl.mask.nxv8i8.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsrl.mask.nxv8i8.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vsrl_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, size_t shift, size_t vl) {
@@ -527,7 +527,7 @@ vuint8m1_t test_vsrl_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t o
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsrl.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsrl.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vsrl_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, vuint8m2_t shift, size_t vl) {
@@ -537,7 +537,7 @@ vuint8m2_t test_vsrl_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t o
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsrl.mask.nxv16i8.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsrl.mask.nxv16i8.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vsrl_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, size_t shift, size_t vl) {
@@ -547,7 +547,7 @@ vuint8m2_t test_vsrl_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t o
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsrl.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsrl.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vsrl_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, vuint8m4_t shift, size_t vl) {
@@ -557,7 +557,7 @@ vuint8m4_t test_vsrl_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t o
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsrl.mask.nxv32i8.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsrl.mask.nxv32i8.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vsrl_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, size_t shift, size_t vl) {
@@ -567,7 +567,7 @@ vuint8m4_t test_vsrl_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t o
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsrl.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsrl.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vsrl_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, vuint8m8_t shift, size_t vl) {
@@ -577,7 +577,7 @@ vuint8m8_t test_vsrl_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t o
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsrl.mask.nxv64i8.i64.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsrl.mask.nxv64i8.i64.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vsrl_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, size_t shift, size_t vl) {
@@ -587,7 +587,7 @@ vuint8m8_t test_vsrl_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t o
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsrl.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsrl.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vsrl_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, vuint16mf4_t shift, size_t vl) {
@@ -597,7 +597,7 @@ vuint16mf4_t test_vsrl_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsrl.mask.nxv1i16.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsrl.mask.nxv1i16.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vsrl_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, size_t shift, size_t vl) {
@@ -607,7 +607,7 @@ vuint16mf4_t test_vsrl_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsrl.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsrl.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vsrl_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, vuint16mf2_t shift, size_t vl) {
@@ -617,7 +617,7 @@ vuint16mf2_t test_vsrl_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsrl.mask.nxv2i16.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsrl.mask.nxv2i16.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vsrl_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, size_t shift, size_t vl) {
@@ -627,7 +627,7 @@ vuint16mf2_t test_vsrl_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsrl.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsrl.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vsrl_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, vuint16m1_t shift, size_t vl) {
@@ -637,7 +637,7 @@ vuint16m1_t test_vsrl_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsrl.mask.nxv4i16.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsrl.mask.nxv4i16.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vsrl_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, size_t shift, size_t vl) {
@@ -647,7 +647,7 @@ vuint16m1_t test_vsrl_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsrl.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsrl.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vsrl_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, vuint16m2_t shift, size_t vl) {
@@ -657,7 +657,7 @@ vuint16m2_t test_vsrl_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsrl.mask.nxv8i16.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsrl.mask.nxv8i16.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vsrl_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, size_t shift, size_t vl) {
@@ -667,7 +667,7 @@ vuint16m2_t test_vsrl_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsrl.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsrl.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vsrl_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, vuint16m4_t shift, size_t vl) {
@@ -677,7 +677,7 @@ vuint16m4_t test_vsrl_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsrl.mask.nxv16i16.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsrl.mask.nxv16i16.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vsrl_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, size_t shift, size_t vl) {
@@ -687,7 +687,7 @@ vuint16m4_t test_vsrl_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsrl.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsrl.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vsrl_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, vuint16m8_t shift, size_t vl) {
@@ -697,7 +697,7 @@ vuint16m8_t test_vsrl_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsrl.mask.nxv32i16.i64.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsrl.mask.nxv32i16.i64.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vsrl_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, size_t shift, size_t vl) {
@@ -707,7 +707,7 @@ vuint16m8_t test_vsrl_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsrl.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsrl.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vsrl_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, vuint32mf2_t shift, size_t vl) {
@@ -717,7 +717,7 @@ vuint32mf2_t test_vsrl_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsrl.mask.nxv1i32.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsrl.mask.nxv1i32.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vsrl_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, size_t shift, size_t vl) {
@@ -727,7 +727,7 @@ vuint32mf2_t test_vsrl_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsrl.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsrl.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vsrl_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, vuint32m1_t shift, size_t vl) {
@@ -737,7 +737,7 @@ vuint32m1_t test_vsrl_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsrl.mask.nxv2i32.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsrl.mask.nxv2i32.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vsrl_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, size_t shift, size_t vl) {
@@ -747,7 +747,7 @@ vuint32m1_t test_vsrl_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsrl.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsrl.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vsrl_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, vuint32m2_t shift, size_t vl) {
@@ -757,7 +757,7 @@ vuint32m2_t test_vsrl_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsrl.mask.nxv4i32.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsrl.mask.nxv4i32.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vsrl_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, size_t shift, size_t vl) {
@@ -767,7 +767,7 @@ vuint32m2_t test_vsrl_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsrl.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsrl.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vsrl_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, vuint32m4_t shift, size_t vl) {
@@ -777,7 +777,7 @@ vuint32m4_t test_vsrl_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsrl.mask.nxv8i32.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsrl.mask.nxv8i32.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vsrl_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, size_t shift, size_t vl) {
@@ -787,7 +787,7 @@ vuint32m4_t test_vsrl_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsrl.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsrl.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vsrl_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, vuint32m8_t shift, size_t vl) {
@@ -797,7 +797,7 @@ vuint32m8_t test_vsrl_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsrl.mask.nxv16i32.i64.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsrl.mask.nxv16i32.i64.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vsrl_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, size_t shift, size_t vl) {
@@ -807,7 +807,7 @@ vuint32m8_t test_vsrl_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsrl.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsrl.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vsrl_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, vuint64m1_t shift, size_t vl) {
@@ -817,7 +817,7 @@ vuint64m1_t test_vsrl_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsrl.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsrl.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vsrl_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, size_t shift, size_t vl) {
@@ -827,7 +827,7 @@ vuint64m1_t test_vsrl_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsrl.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsrl.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vsrl_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, vuint64m2_t shift, size_t vl) {
@@ -837,7 +837,7 @@ vuint64m2_t test_vsrl_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsrl.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsrl.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vsrl_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, size_t shift, size_t vl) {
@@ -847,7 +847,7 @@ vuint64m2_t test_vsrl_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsrl.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsrl.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vsrl_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, vuint64m4_t shift, size_t vl) {
@@ -857,7 +857,7 @@ vuint64m4_t test_vsrl_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsrl.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsrl.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vsrl_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, size_t shift, size_t vl) {
@@ -867,7 +867,7 @@ vuint64m4_t test_vsrl_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vsrl_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsrl.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsrl.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vsrl_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, vuint64m8_t shift, size_t vl) {
@@ -877,7 +877,7 @@ vuint64m8_t test_vsrl_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8
 //
 // CHECK-RV64-LABEL: @test_vsrl_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsrl.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsrl.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vsrl_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, size_t shift, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsse.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsse.c
index 2073ff6..f64b6d1 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsse.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsse.c
@@ -9,7 +9,7 @@
 // CHECK-RV64-LABEL: @test_vsse8_v_i8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8:#.*]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_i8mf8(int8_t *base, ptrdiff_t bstride, vint8mf8_t value,
@@ -21,7 +21,7 @@ void test_vsse8_v_i8mf8(int8_t *base, ptrdiff_t bstride, vint8mf8_t value,
 // CHECK-RV64-LABEL: @test_vsse8_v_i8mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_i8mf4(int8_t *base, ptrdiff_t bstride, vint8mf4_t value,
@@ -33,7 +33,7 @@ void test_vsse8_v_i8mf4(int8_t *base, ptrdiff_t bstride, vint8mf4_t value,
 // CHECK-RV64-LABEL: @test_vsse8_v_i8mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_i8mf2(int8_t *base, ptrdiff_t bstride, vint8mf2_t value,
@@ -45,7 +45,7 @@ void test_vsse8_v_i8mf2(int8_t *base, ptrdiff_t bstride, vint8mf2_t value,
 // CHECK-RV64-LABEL: @test_vsse8_v_i8m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_i8m1(int8_t *base, ptrdiff_t bstride, vint8m1_t value,
@@ -57,7 +57,7 @@ void test_vsse8_v_i8m1(int8_t *base, ptrdiff_t bstride, vint8m1_t value,
 // CHECK-RV64-LABEL: @test_vsse8_v_i8m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_i8m2(int8_t *base, ptrdiff_t bstride, vint8m2_t value,
@@ -69,7 +69,7 @@ void test_vsse8_v_i8m2(int8_t *base, ptrdiff_t bstride, vint8m2_t value,
 // CHECK-RV64-LABEL: @test_vsse8_v_i8m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_i8m4(int8_t *base, ptrdiff_t bstride, vint8m4_t value,
@@ -81,7 +81,7 @@ void test_vsse8_v_i8m4(int8_t *base, ptrdiff_t bstride, vint8m4_t value,
 // CHECK-RV64-LABEL: @test_vsse8_v_i8m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_i8m8(int8_t *base, ptrdiff_t bstride, vint8m8_t value,
@@ -93,7 +93,7 @@ void test_vsse8_v_i8m8(int8_t *base, ptrdiff_t bstride, vint8m8_t value,
 // CHECK-RV64-LABEL: @test_vsse16_v_i16mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_i16mf4(int16_t *base, ptrdiff_t bstride, vint16mf4_t value,
@@ -105,7 +105,7 @@ void test_vsse16_v_i16mf4(int16_t *base, ptrdiff_t bstride, vint16mf4_t value,
 // CHECK-RV64-LABEL: @test_vsse16_v_i16mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_i16mf2(int16_t *base, ptrdiff_t bstride, vint16mf2_t value,
@@ -117,7 +117,7 @@ void test_vsse16_v_i16mf2(int16_t *base, ptrdiff_t bstride, vint16mf2_t value,
 // CHECK-RV64-LABEL: @test_vsse16_v_i16m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_i16m1(int16_t *base, ptrdiff_t bstride, vint16m1_t value,
@@ -129,7 +129,7 @@ void test_vsse16_v_i16m1(int16_t *base, ptrdiff_t bstride, vint16m1_t value,
 // CHECK-RV64-LABEL: @test_vsse16_v_i16m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_i16m2(int16_t *base, ptrdiff_t bstride, vint16m2_t value,
@@ -141,7 +141,7 @@ void test_vsse16_v_i16m2(int16_t *base, ptrdiff_t bstride, vint16m2_t value,
 // CHECK-RV64-LABEL: @test_vsse16_v_i16m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_i16m4(int16_t *base, ptrdiff_t bstride, vint16m4_t value,
@@ -153,7 +153,7 @@ void test_vsse16_v_i16m4(int16_t *base, ptrdiff_t bstride, vint16m4_t value,
 // CHECK-RV64-LABEL: @test_vsse16_v_i16m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_i16m8(int16_t *base, ptrdiff_t bstride, vint16m8_t value,
@@ -165,7 +165,7 @@ void test_vsse16_v_i16m8(int16_t *base, ptrdiff_t bstride, vint16m8_t value,
 // CHECK-RV64-LABEL: @test_vsse32_v_i32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_i32mf2(int32_t *base, ptrdiff_t bstride, vint32mf2_t value,
@@ -177,7 +177,7 @@ void test_vsse32_v_i32mf2(int32_t *base, ptrdiff_t bstride, vint32mf2_t value,
 // CHECK-RV64-LABEL: @test_vsse32_v_i32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_i32m1(int32_t *base, ptrdiff_t bstride, vint32m1_t value,
@@ -189,7 +189,7 @@ void test_vsse32_v_i32m1(int32_t *base, ptrdiff_t bstride, vint32m1_t value,
 // CHECK-RV64-LABEL: @test_vsse32_v_i32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_i32m2(int32_t *base, ptrdiff_t bstride, vint32m2_t value,
@@ -201,7 +201,7 @@ void test_vsse32_v_i32m2(int32_t *base, ptrdiff_t bstride, vint32m2_t value,
 // CHECK-RV64-LABEL: @test_vsse32_v_i32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_i32m4(int32_t *base, ptrdiff_t bstride, vint32m4_t value,
@@ -213,7 +213,7 @@ void test_vsse32_v_i32m4(int32_t *base, ptrdiff_t bstride, vint32m4_t value,
 // CHECK-RV64-LABEL: @test_vsse32_v_i32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_i32m8(int32_t *base, ptrdiff_t bstride, vint32m8_t value,
@@ -225,7 +225,7 @@ void test_vsse32_v_i32m8(int32_t *base, ptrdiff_t bstride, vint32m8_t value,
 // CHECK-RV64-LABEL: @test_vsse64_v_i64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_i64m1(int64_t *base, ptrdiff_t bstride, vint64m1_t value,
@@ -237,7 +237,7 @@ void test_vsse64_v_i64m1(int64_t *base, ptrdiff_t bstride, vint64m1_t value,
 // CHECK-RV64-LABEL: @test_vsse64_v_i64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_i64m2(int64_t *base, ptrdiff_t bstride, vint64m2_t value,
@@ -249,7 +249,7 @@ void test_vsse64_v_i64m2(int64_t *base, ptrdiff_t bstride, vint64m2_t value,
 // CHECK-RV64-LABEL: @test_vsse64_v_i64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_i64m4(int64_t *base, ptrdiff_t bstride, vint64m4_t value,
@@ -261,7 +261,7 @@ void test_vsse64_v_i64m4(int64_t *base, ptrdiff_t bstride, vint64m4_t value,
 // CHECK-RV64-LABEL: @test_vsse64_v_i64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_i64m8(int64_t *base, ptrdiff_t bstride, vint64m8_t value,
@@ -273,7 +273,7 @@ void test_vsse64_v_i64m8(int64_t *base, ptrdiff_t bstride, vint64m8_t value,
 // CHECK-RV64-LABEL: @test_vsse8_v_u8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_u8mf8(uint8_t *base, ptrdiff_t bstride, vuint8mf8_t value,
@@ -285,7 +285,7 @@ void test_vsse8_v_u8mf8(uint8_t *base, ptrdiff_t bstride, vuint8mf8_t value,
 // CHECK-RV64-LABEL: @test_vsse8_v_u8mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_u8mf4(uint8_t *base, ptrdiff_t bstride, vuint8mf4_t value,
@@ -297,7 +297,7 @@ void test_vsse8_v_u8mf4(uint8_t *base, ptrdiff_t bstride, vuint8mf4_t value,
 // CHECK-RV64-LABEL: @test_vsse8_v_u8mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_u8mf2(uint8_t *base, ptrdiff_t bstride, vuint8mf2_t value,
@@ -309,7 +309,7 @@ void test_vsse8_v_u8mf2(uint8_t *base, ptrdiff_t bstride, vuint8mf2_t value,
 // CHECK-RV64-LABEL: @test_vsse8_v_u8m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_u8m1(uint8_t *base, ptrdiff_t bstride, vuint8m1_t value,
@@ -321,7 +321,7 @@ void test_vsse8_v_u8m1(uint8_t *base, ptrdiff_t bstride, vuint8m1_t value,
 // CHECK-RV64-LABEL: @test_vsse8_v_u8m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_u8m2(uint8_t *base, ptrdiff_t bstride, vuint8m2_t value,
@@ -333,7 +333,7 @@ void test_vsse8_v_u8m2(uint8_t *base, ptrdiff_t bstride, vuint8m2_t value,
 // CHECK-RV64-LABEL: @test_vsse8_v_u8m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_u8m4(uint8_t *base, ptrdiff_t bstride, vuint8m4_t value,
@@ -345,7 +345,7 @@ void test_vsse8_v_u8m4(uint8_t *base, ptrdiff_t bstride, vuint8m4_t value,
 // CHECK-RV64-LABEL: @test_vsse8_v_u8m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_u8m8(uint8_t *base, ptrdiff_t bstride, vuint8m8_t value,
@@ -357,7 +357,7 @@ void test_vsse8_v_u8m8(uint8_t *base, ptrdiff_t bstride, vuint8m8_t value,
 // CHECK-RV64-LABEL: @test_vsse16_v_u16mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_u16mf4(uint16_t *base, ptrdiff_t bstride, vuint16mf4_t value,
@@ -369,7 +369,7 @@ void test_vsse16_v_u16mf4(uint16_t *base, ptrdiff_t bstride, vuint16mf4_t value,
 // CHECK-RV64-LABEL: @test_vsse16_v_u16mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_u16mf2(uint16_t *base, ptrdiff_t bstride, vuint16mf2_t value,
@@ -381,7 +381,7 @@ void test_vsse16_v_u16mf2(uint16_t *base, ptrdiff_t bstride, vuint16mf2_t value,
 // CHECK-RV64-LABEL: @test_vsse16_v_u16m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_u16m1(uint16_t *base, ptrdiff_t bstride, vuint16m1_t value,
@@ -393,7 +393,7 @@ void test_vsse16_v_u16m1(uint16_t *base, ptrdiff_t bstride, vuint16m1_t value,
 // CHECK-RV64-LABEL: @test_vsse16_v_u16m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_u16m2(uint16_t *base, ptrdiff_t bstride, vuint16m2_t value,
@@ -405,7 +405,7 @@ void test_vsse16_v_u16m2(uint16_t *base, ptrdiff_t bstride, vuint16m2_t value,
 // CHECK-RV64-LABEL: @test_vsse16_v_u16m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_u16m4(uint16_t *base, ptrdiff_t bstride, vuint16m4_t value,
@@ -417,7 +417,7 @@ void test_vsse16_v_u16m4(uint16_t *base, ptrdiff_t bstride, vuint16m4_t value,
 // CHECK-RV64-LABEL: @test_vsse16_v_u16m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_u16m8(uint16_t *base, ptrdiff_t bstride, vuint16m8_t value,
@@ -429,7 +429,7 @@ void test_vsse16_v_u16m8(uint16_t *base, ptrdiff_t bstride, vuint16m8_t value,
 // CHECK-RV64-LABEL: @test_vsse32_v_u32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_u32mf2(uint32_t *base, ptrdiff_t bstride, vuint32mf2_t value,
@@ -441,7 +441,7 @@ void test_vsse32_v_u32mf2(uint32_t *base, ptrdiff_t bstride, vuint32mf2_t value,
 // CHECK-RV64-LABEL: @test_vsse32_v_u32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_u32m1(uint32_t *base, ptrdiff_t bstride, vuint32m1_t value,
@@ -453,7 +453,7 @@ void test_vsse32_v_u32m1(uint32_t *base, ptrdiff_t bstride, vuint32m1_t value,
 // CHECK-RV64-LABEL: @test_vsse32_v_u32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_u32m2(uint32_t *base, ptrdiff_t bstride, vuint32m2_t value,
@@ -465,7 +465,7 @@ void test_vsse32_v_u32m2(uint32_t *base, ptrdiff_t bstride, vuint32m2_t value,
 // CHECK-RV64-LABEL: @test_vsse32_v_u32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_u32m4(uint32_t *base, ptrdiff_t bstride, vuint32m4_t value,
@@ -477,7 +477,7 @@ void test_vsse32_v_u32m4(uint32_t *base, ptrdiff_t bstride, vuint32m4_t value,
 // CHECK-RV64-LABEL: @test_vsse32_v_u32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_u32m8(uint32_t *base, ptrdiff_t bstride, vuint32m8_t value,
@@ -489,7 +489,7 @@ void test_vsse32_v_u32m8(uint32_t *base, ptrdiff_t bstride, vuint32m8_t value,
 // CHECK-RV64-LABEL: @test_vsse64_v_u64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_u64m1(uint64_t *base, ptrdiff_t bstride, vuint64m1_t value,
@@ -501,7 +501,7 @@ void test_vsse64_v_u64m1(uint64_t *base, ptrdiff_t bstride, vuint64m1_t value,
 // CHECK-RV64-LABEL: @test_vsse64_v_u64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_u64m2(uint64_t *base, ptrdiff_t bstride, vuint64m2_t value,
@@ -513,7 +513,7 @@ void test_vsse64_v_u64m2(uint64_t *base, ptrdiff_t bstride, vuint64m2_t value,
 // CHECK-RV64-LABEL: @test_vsse64_v_u64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_u64m4(uint64_t *base, ptrdiff_t bstride, vuint64m4_t value,
@@ -525,7 +525,7 @@ void test_vsse64_v_u64m4(uint64_t *base, ptrdiff_t bstride, vuint64m4_t value,
 // CHECK-RV64-LABEL: @test_vsse64_v_u64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_u64m8(uint64_t *base, ptrdiff_t bstride, vuint64m8_t value,
@@ -537,7 +537,7 @@ void test_vsse64_v_u64m8(uint64_t *base, ptrdiff_t bstride, vuint64m8_t value,
 // CHECK-RV64-LABEL: @test_vsse32_v_f32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1f32.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1f32.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_f32mf2(float *base, ptrdiff_t bstride, vfloat32mf2_t value,
@@ -549,7 +549,7 @@ void test_vsse32_v_f32mf2(float *base, ptrdiff_t bstride, vfloat32mf2_t value,
 // CHECK-RV64-LABEL: @test_vsse32_v_f32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv2f32.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv2f32.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_f32m1(float *base, ptrdiff_t bstride, vfloat32m1_t value,
@@ -561,7 +561,7 @@ void test_vsse32_v_f32m1(float *base, ptrdiff_t bstride, vfloat32m1_t value,
 // CHECK-RV64-LABEL: @test_vsse32_v_f32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv4f32.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv4f32.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_f32m2(float *base, ptrdiff_t bstride, vfloat32m2_t value,
@@ -573,7 +573,7 @@ void test_vsse32_v_f32m2(float *base, ptrdiff_t bstride, vfloat32m2_t value,
 // CHECK-RV64-LABEL: @test_vsse32_v_f32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv8f32.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv8f32.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_f32m4(float *base, ptrdiff_t bstride, vfloat32m4_t value,
@@ -585,7 +585,7 @@ void test_vsse32_v_f32m4(float *base, ptrdiff_t bstride, vfloat32m4_t value,
 // CHECK-RV64-LABEL: @test_vsse32_v_f32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv16f32.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv16f32.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_f32m8(float *base, ptrdiff_t bstride, vfloat32m8_t value,
@@ -597,7 +597,7 @@ void test_vsse32_v_f32m8(float *base, ptrdiff_t bstride, vfloat32m8_t value,
 // CHECK-RV64-LABEL: @test_vsse64_v_f64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1f64.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1f64.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_f64m1(double *base, ptrdiff_t bstride, vfloat64m1_t value,
@@ -609,7 +609,7 @@ void test_vsse64_v_f64m1(double *base, ptrdiff_t bstride, vfloat64m1_t value,
 // CHECK-RV64-LABEL: @test_vsse64_v_f64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv2f64.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv2f64.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_f64m2(double *base, ptrdiff_t bstride, vfloat64m2_t value,
@@ -621,7 +621,7 @@ void test_vsse64_v_f64m2(double *base, ptrdiff_t bstride, vfloat64m2_t value,
 // CHECK-RV64-LABEL: @test_vsse64_v_f64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv4f64.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv4f64.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_f64m4(double *base, ptrdiff_t bstride, vfloat64m4_t value,
@@ -633,7 +633,7 @@ void test_vsse64_v_f64m4(double *base, ptrdiff_t bstride, vfloat64m4_t value,
 // CHECK-RV64-LABEL: @test_vsse64_v_f64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv8f64.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv8f64.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_f64m8(double *base, ptrdiff_t bstride, vfloat64m8_t value,
@@ -645,7 +645,7 @@ void test_vsse64_v_f64m8(double *base, ptrdiff_t bstride, vfloat64m8_t value,
 // CHECK-RV64-LABEL: @test_vsse8_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_i8mf8_m(vbool64_t mask, int8_t *base, ptrdiff_t bstride,
@@ -657,7 +657,7 @@ void test_vsse8_v_i8mf8_m(vbool64_t mask, int8_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse8_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_i8mf4_m(vbool32_t mask, int8_t *base, ptrdiff_t bstride,
@@ -669,7 +669,7 @@ void test_vsse8_v_i8mf4_m(vbool32_t mask, int8_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse8_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_i8mf2_m(vbool16_t mask, int8_t *base, ptrdiff_t bstride,
@@ -681,7 +681,7 @@ void test_vsse8_v_i8mf2_m(vbool16_t mask, int8_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse8_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_i8m1_m(vbool8_t mask, int8_t *base, ptrdiff_t bstride,
@@ -693,7 +693,7 @@ void test_vsse8_v_i8m1_m(vbool8_t mask, int8_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse8_v_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_i8m2_m(vbool4_t mask, int8_t *base, ptrdiff_t bstride,
@@ -705,7 +705,7 @@ void test_vsse8_v_i8m2_m(vbool4_t mask, int8_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse8_v_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_i8m4_m(vbool2_t mask, int8_t *base, ptrdiff_t bstride,
@@ -717,7 +717,7 @@ void test_vsse8_v_i8m4_m(vbool2_t mask, int8_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse8_v_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_i8m8_m(vbool1_t mask, int8_t *base, ptrdiff_t bstride,
@@ -729,7 +729,7 @@ void test_vsse8_v_i8m8_m(vbool1_t mask, int8_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse16_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_i16mf4_m(vbool64_t mask, int16_t *base, ptrdiff_t bstride,
@@ -741,7 +741,7 @@ void test_vsse16_v_i16mf4_m(vbool64_t mask, int16_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse16_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_i16mf2_m(vbool32_t mask, int16_t *base, ptrdiff_t bstride,
@@ -753,7 +753,7 @@ void test_vsse16_v_i16mf2_m(vbool32_t mask, int16_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse16_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_i16m1_m(vbool16_t mask, int16_t *base, ptrdiff_t bstride,
@@ -765,7 +765,7 @@ void test_vsse16_v_i16m1_m(vbool16_t mask, int16_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse16_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_i16m2_m(vbool8_t mask, int16_t *base, ptrdiff_t bstride,
@@ -777,7 +777,7 @@ void test_vsse16_v_i16m2_m(vbool8_t mask, int16_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse16_v_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_i16m4_m(vbool4_t mask, int16_t *base, ptrdiff_t bstride,
@@ -789,7 +789,7 @@ void test_vsse16_v_i16m4_m(vbool4_t mask, int16_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse16_v_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_i16m8_m(vbool2_t mask, int16_t *base, ptrdiff_t bstride,
@@ -801,7 +801,7 @@ void test_vsse16_v_i16m8_m(vbool2_t mask, int16_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse32_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_i32mf2_m(vbool64_t mask, int32_t *base, ptrdiff_t bstride,
@@ -813,7 +813,7 @@ void test_vsse32_v_i32mf2_m(vbool64_t mask, int32_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse32_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_i32m1_m(vbool32_t mask, int32_t *base, ptrdiff_t bstride,
@@ -825,7 +825,7 @@ void test_vsse32_v_i32m1_m(vbool32_t mask, int32_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse32_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_i32m2_m(vbool16_t mask, int32_t *base, ptrdiff_t bstride,
@@ -837,7 +837,7 @@ void test_vsse32_v_i32m2_m(vbool16_t mask, int32_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse32_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_i32m4_m(vbool8_t mask, int32_t *base, ptrdiff_t bstride,
@@ -849,7 +849,7 @@ void test_vsse32_v_i32m4_m(vbool8_t mask, int32_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse32_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_i32m8_m(vbool4_t mask, int32_t *base, ptrdiff_t bstride,
@@ -861,7 +861,7 @@ void test_vsse32_v_i32m8_m(vbool4_t mask, int32_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse64_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_i64m1_m(vbool64_t mask, int64_t *base, ptrdiff_t bstride,
@@ -873,7 +873,7 @@ void test_vsse64_v_i64m1_m(vbool64_t mask, int64_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse64_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_i64m2_m(vbool32_t mask, int64_t *base, ptrdiff_t bstride,
@@ -885,7 +885,7 @@ void test_vsse64_v_i64m2_m(vbool32_t mask, int64_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse64_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_i64m4_m(vbool16_t mask, int64_t *base, ptrdiff_t bstride,
@@ -897,7 +897,7 @@ void test_vsse64_v_i64m4_m(vbool16_t mask, int64_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse64_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_i64m8_m(vbool8_t mask, int64_t *base, ptrdiff_t bstride,
@@ -909,7 +909,7 @@ void test_vsse64_v_i64m8_m(vbool8_t mask, int64_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse8_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_u8mf8_m(vbool64_t mask, uint8_t *base, ptrdiff_t bstride,
@@ -921,7 +921,7 @@ void test_vsse8_v_u8mf8_m(vbool64_t mask, uint8_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse8_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_u8mf4_m(vbool32_t mask, uint8_t *base, ptrdiff_t bstride,
@@ -933,7 +933,7 @@ void test_vsse8_v_u8mf4_m(vbool32_t mask, uint8_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse8_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_u8mf2_m(vbool16_t mask, uint8_t *base, ptrdiff_t bstride,
@@ -945,7 +945,7 @@ void test_vsse8_v_u8mf2_m(vbool16_t mask, uint8_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse8_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_u8m1_m(vbool8_t mask, uint8_t *base, ptrdiff_t bstride,
@@ -957,7 +957,7 @@ void test_vsse8_v_u8m1_m(vbool8_t mask, uint8_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse8_v_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_u8m2_m(vbool4_t mask, uint8_t *base, ptrdiff_t bstride,
@@ -969,7 +969,7 @@ void test_vsse8_v_u8m2_m(vbool4_t mask, uint8_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse8_v_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_u8m4_m(vbool2_t mask, uint8_t *base, ptrdiff_t bstride,
@@ -981,7 +981,7 @@ void test_vsse8_v_u8m4_m(vbool2_t mask, uint8_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse8_v_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse8_v_u8m8_m(vbool1_t mask, uint8_t *base, ptrdiff_t bstride,
@@ -993,7 +993,7 @@ void test_vsse8_v_u8m8_m(vbool1_t mask, uint8_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse16_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_u16mf4_m(vbool64_t mask, uint16_t *base, ptrdiff_t bstride,
@@ -1005,7 +1005,7 @@ void test_vsse16_v_u16mf4_m(vbool64_t mask, uint16_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse16_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_u16mf2_m(vbool32_t mask, uint16_t *base, ptrdiff_t bstride,
@@ -1017,7 +1017,7 @@ void test_vsse16_v_u16mf2_m(vbool32_t mask, uint16_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse16_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_u16m1_m(vbool16_t mask, uint16_t *base, ptrdiff_t bstride,
@@ -1029,7 +1029,7 @@ void test_vsse16_v_u16m1_m(vbool16_t mask, uint16_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse16_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_u16m2_m(vbool8_t mask, uint16_t *base, ptrdiff_t bstride,
@@ -1041,7 +1041,7 @@ void test_vsse16_v_u16m2_m(vbool8_t mask, uint16_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse16_v_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_u16m4_m(vbool4_t mask, uint16_t *base, ptrdiff_t bstride,
@@ -1053,7 +1053,7 @@ void test_vsse16_v_u16m4_m(vbool4_t mask, uint16_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse16_v_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse16_v_u16m8_m(vbool2_t mask, uint16_t *base, ptrdiff_t bstride,
@@ -1065,7 +1065,7 @@ void test_vsse16_v_u16m8_m(vbool2_t mask, uint16_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse32_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_u32mf2_m(vbool64_t mask, uint32_t *base, ptrdiff_t bstride,
@@ -1077,7 +1077,7 @@ void test_vsse32_v_u32mf2_m(vbool64_t mask, uint32_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse32_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_u32m1_m(vbool32_t mask, uint32_t *base, ptrdiff_t bstride,
@@ -1089,7 +1089,7 @@ void test_vsse32_v_u32m1_m(vbool32_t mask, uint32_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse32_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_u32m2_m(vbool16_t mask, uint32_t *base, ptrdiff_t bstride,
@@ -1101,7 +1101,7 @@ void test_vsse32_v_u32m2_m(vbool16_t mask, uint32_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse32_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_u32m4_m(vbool8_t mask, uint32_t *base, ptrdiff_t bstride,
@@ -1113,7 +1113,7 @@ void test_vsse32_v_u32m4_m(vbool8_t mask, uint32_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse32_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_u32m8_m(vbool4_t mask, uint32_t *base, ptrdiff_t bstride,
@@ -1125,7 +1125,7 @@ void test_vsse32_v_u32m8_m(vbool4_t mask, uint32_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse64_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_u64m1_m(vbool64_t mask, uint64_t *base, ptrdiff_t bstride,
@@ -1137,7 +1137,7 @@ void test_vsse64_v_u64m1_m(vbool64_t mask, uint64_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse64_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_u64m2_m(vbool32_t mask, uint64_t *base, ptrdiff_t bstride,
@@ -1149,7 +1149,7 @@ void test_vsse64_v_u64m2_m(vbool32_t mask, uint64_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse64_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_u64m4_m(vbool16_t mask, uint64_t *base, ptrdiff_t bstride,
@@ -1161,7 +1161,7 @@ void test_vsse64_v_u64m4_m(vbool16_t mask, uint64_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse64_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_u64m8_m(vbool8_t mask, uint64_t *base, ptrdiff_t bstride,
@@ -1173,7 +1173,7 @@ void test_vsse64_v_u64m8_m(vbool8_t mask, uint64_t *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse32_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv1f32.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv1f32.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_f32mf2_m(vbool64_t mask, float *base, ptrdiff_t bstride,
@@ -1185,7 +1185,7 @@ void test_vsse32_v_f32mf2_m(vbool64_t mask, float *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse32_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv2f32.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv2f32.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_f32m1_m(vbool32_t mask, float *base, ptrdiff_t bstride,
@@ -1197,7 +1197,7 @@ void test_vsse32_v_f32m1_m(vbool32_t mask, float *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse32_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv4f32.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv4f32.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_f32m2_m(vbool16_t mask, float *base, ptrdiff_t bstride,
@@ -1209,7 +1209,7 @@ void test_vsse32_v_f32m2_m(vbool16_t mask, float *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse32_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8f32.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8f32.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_f32m4_m(vbool8_t mask, float *base, ptrdiff_t bstride,
@@ -1221,7 +1221,7 @@ void test_vsse32_v_f32m4_m(vbool8_t mask, float *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse32_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv16f32.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv16f32.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse32_v_f32m8_m(vbool4_t mask, float *base, ptrdiff_t bstride,
@@ -1233,7 +1233,7 @@ void test_vsse32_v_f32m8_m(vbool4_t mask, float *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse64_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv1f64.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv1f64.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_f64m1_m(vbool64_t mask, double *base, ptrdiff_t bstride,
@@ -1245,7 +1245,7 @@ void test_vsse64_v_f64m1_m(vbool64_t mask, double *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse64_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv2f64.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv2f64.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_f64m2_m(vbool32_t mask, double *base, ptrdiff_t bstride,
@@ -1257,7 +1257,7 @@ void test_vsse64_v_f64m2_m(vbool32_t mask, double *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse64_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv4f64.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv4f64.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_f64m4_m(vbool16_t mask, double *base, ptrdiff_t bstride,
@@ -1269,7 +1269,7 @@ void test_vsse64_v_f64m4_m(vbool16_t mask, double *base, ptrdiff_t bstride,
 // CHECK-RV64-LABEL: @test_vsse64_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8f64.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8f64.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_f64m8_m(vbool8_t mask, double *base, ptrdiff_t bstride,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vssra.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vssra.c
index 4fc3ae2..7b3599b 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vssra.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vssra.c
@@ -450,7 +450,7 @@ vint64m8_t test_vssra_vx_i64m8(vint64m8_t op1, size_t shift, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vssra.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vssra.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vssra_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
@@ -461,7 +461,7 @@ vint8mf8_t test_vssra_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vssra.mask.nxv1i8.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vssra.mask.nxv1i8.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vssra_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
@@ -472,7 +472,7 @@ vint8mf8_t test_vssra_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vssra.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vssra.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vssra_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
@@ -483,7 +483,7 @@ vint8mf4_t test_vssra_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vssra.mask.nxv2i8.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vssra.mask.nxv2i8.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vssra_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
@@ -494,7 +494,7 @@ vint8mf4_t test_vssra_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vssra.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vssra.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vssra_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
@@ -505,7 +505,7 @@ vint8mf2_t test_vssra_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vssra.mask.nxv4i8.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vssra.mask.nxv4i8.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vssra_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
@@ -516,7 +516,7 @@ vint8mf2_t test_vssra_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vssra.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vssra.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vssra_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
@@ -527,7 +527,7 @@ vint8m1_t test_vssra_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vssra.mask.nxv8i8.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vssra.mask.nxv8i8.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vssra_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
@@ -538,7 +538,7 @@ vint8m1_t test_vssra_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vssra.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vssra.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vssra_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
@@ -549,7 +549,7 @@ vint8m2_t test_vssra_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vssra.mask.nxv16i8.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vssra.mask.nxv16i8.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vssra_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
@@ -560,7 +560,7 @@ vint8m2_t test_vssra_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vssra.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vssra.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vssra_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
@@ -571,7 +571,7 @@ vint8m4_t test_vssra_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vssra.mask.nxv32i8.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vssra.mask.nxv32i8.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vssra_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
@@ -582,7 +582,7 @@ vint8m4_t test_vssra_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vssra.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vssra.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vssra_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
@@ -593,7 +593,7 @@ vint8m8_t test_vssra_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vssra.mask.nxv64i8.i64.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vssra.mask.nxv64i8.i64.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vssra_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
@@ -604,7 +604,7 @@ vint8m8_t test_vssra_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vssra.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vssra.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vssra_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -616,7 +616,7 @@ vint16mf4_t test_vssra_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vssra.mask.nxv1i16.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vssra.mask.nxv1i16.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vssra_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -627,7 +627,7 @@ vint16mf4_t test_vssra_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vssra.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vssra.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vssra_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -639,7 +639,7 @@ vint16mf2_t test_vssra_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vssra.mask.nxv2i16.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vssra.mask.nxv2i16.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vssra_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -650,7 +650,7 @@ vint16mf2_t test_vssra_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vssra.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vssra.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vssra_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -661,7 +661,7 @@ vint16m1_t test_vssra_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vssra.mask.nxv4i16.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vssra.mask.nxv4i16.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vssra_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -672,7 +672,7 @@ vint16m1_t test_vssra_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vssra.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vssra.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vssra_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -683,7 +683,7 @@ vint16m2_t test_vssra_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vssra.mask.nxv8i16.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vssra.mask.nxv8i16.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vssra_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -694,7 +694,7 @@ vint16m2_t test_vssra_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vssra.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vssra.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vssra_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -705,7 +705,7 @@ vint16m4_t test_vssra_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vssra.mask.nxv16i16.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vssra.mask.nxv16i16.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vssra_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -716,7 +716,7 @@ vint16m4_t test_vssra_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vssra.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vssra.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vssra_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -727,7 +727,7 @@ vint16m8_t test_vssra_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vssra.mask.nxv32i16.i64.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vssra.mask.nxv32i16.i64.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vssra_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -738,7 +738,7 @@ vint16m8_t test_vssra_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vssra.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vssra.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vssra_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -750,7 +750,7 @@ vint32mf2_t test_vssra_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vssra.mask.nxv1i32.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vssra.mask.nxv1i32.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vssra_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -761,7 +761,7 @@ vint32mf2_t test_vssra_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vssra.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vssra.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vssra_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -772,7 +772,7 @@ vint32m1_t test_vssra_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vssra.mask.nxv2i32.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vssra.mask.nxv2i32.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vssra_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -783,7 +783,7 @@ vint32m1_t test_vssra_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vssra.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vssra.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vssra_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -794,7 +794,7 @@ vint32m2_t test_vssra_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vssra.mask.nxv4i32.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vssra.mask.nxv4i32.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vssra_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -805,7 +805,7 @@ vint32m2_t test_vssra_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vssra.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vssra.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vssra_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -816,7 +816,7 @@ vint32m4_t test_vssra_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vssra.mask.nxv8i32.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vssra.mask.nxv8i32.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vssra_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -827,7 +827,7 @@ vint32m4_t test_vssra_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vssra.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vssra.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vssra_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -838,7 +838,7 @@ vint32m8_t test_vssra_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vssra.mask.nxv16i32.i64.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vssra.mask.nxv16i32.i64.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vssra_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -849,7 +849,7 @@ vint32m8_t test_vssra_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vssra.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vssra.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vssra_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -860,7 +860,7 @@ vint64m1_t test_vssra_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vssra.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vssra.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vssra_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -871,7 +871,7 @@ vint64m1_t test_vssra_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vssra.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vssra.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vssra_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -882,7 +882,7 @@ vint64m2_t test_vssra_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vssra.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vssra.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vssra_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -893,7 +893,7 @@ vint64m2_t test_vssra_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vssra.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vssra.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vssra_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -904,7 +904,7 @@ vint64m4_t test_vssra_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vssra.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vssra.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vssra_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -915,7 +915,7 @@ vint64m4_t test_vssra_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vssra.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vssra.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vssra_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -926,7 +926,7 @@ vint64m8_t test_vssra_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssra_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vssra.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vssra.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vssra_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vssrl.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vssrl.c
index 0428ce7..6c6b9d3 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vssrl.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vssrl.c
@@ -450,7 +450,7 @@ vuint64m8_t test_vssrl_vx_u64m8(vuint64m8_t op1, size_t shift, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vssrl.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vssrl.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vssrl_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
@@ -462,7 +462,7 @@ vuint8mf8_t test_vssrl_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vssrl.mask.nxv1i8.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vssrl.mask.nxv1i8.i64.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vssrl_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
@@ -473,7 +473,7 @@ vuint8mf8_t test_vssrl_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vssrl.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vssrl.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vssrl_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
@@ -485,7 +485,7 @@ vuint8mf4_t test_vssrl_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vssrl.mask.nxv2i8.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vssrl.mask.nxv2i8.i64.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vssrl_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
@@ -496,7 +496,7 @@ vuint8mf4_t test_vssrl_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vssrl.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vssrl.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vssrl_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
@@ -508,7 +508,7 @@ vuint8mf2_t test_vssrl_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vssrl.mask.nxv4i8.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vssrl.mask.nxv4i8.i64.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vssrl_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
@@ -519,7 +519,7 @@ vuint8mf2_t test_vssrl_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vssrl.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vssrl.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vssrl_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
@@ -530,7 +530,7 @@ vuint8m1_t test_vssrl_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vssrl.mask.nxv8i8.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vssrl.mask.nxv8i8.i64.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vssrl_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
@@ -541,7 +541,7 @@ vuint8m1_t test_vssrl_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vssrl.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vssrl.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vssrl_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
@@ -552,7 +552,7 @@ vuint8m2_t test_vssrl_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vssrl.mask.nxv16i8.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vssrl.mask.nxv16i8.i64.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vssrl_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
@@ -563,7 +563,7 @@ vuint8m2_t test_vssrl_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vssrl.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vssrl.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vssrl_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
@@ -574,7 +574,7 @@ vuint8m4_t test_vssrl_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vssrl.mask.nxv32i8.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vssrl.mask.nxv32i8.i64.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vssrl_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
@@ -585,7 +585,7 @@ vuint8m4_t test_vssrl_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vssrl.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vssrl.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vssrl_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
@@ -596,7 +596,7 @@ vuint8m8_t test_vssrl_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vssrl.mask.nxv64i8.i64.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vssrl.mask.nxv64i8.i64.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vssrl_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
@@ -607,7 +607,7 @@ vuint8m8_t test_vssrl_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vssrl.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vssrl.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vssrl_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -619,7 +619,7 @@ vuint16mf4_t test_vssrl_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vssrl.mask.nxv1i16.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vssrl.mask.nxv1i16.i64.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vssrl_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -630,7 +630,7 @@ vuint16mf4_t test_vssrl_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vssrl.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vssrl.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vssrl_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -642,7 +642,7 @@ vuint16mf2_t test_vssrl_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vssrl.mask.nxv2i16.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vssrl.mask.nxv2i16.i64.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vssrl_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -653,7 +653,7 @@ vuint16mf2_t test_vssrl_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vssrl.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vssrl.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vssrl_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -665,7 +665,7 @@ vuint16m1_t test_vssrl_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vssrl.mask.nxv4i16.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vssrl.mask.nxv4i16.i64.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vssrl_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -676,7 +676,7 @@ vuint16m1_t test_vssrl_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vssrl.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vssrl.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vssrl_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -688,7 +688,7 @@ vuint16m2_t test_vssrl_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vssrl.mask.nxv8i16.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vssrl.mask.nxv8i16.i64.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vssrl_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -699,7 +699,7 @@ vuint16m2_t test_vssrl_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vssrl.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vssrl.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vssrl_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -711,7 +711,7 @@ vuint16m4_t test_vssrl_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vssrl.mask.nxv16i16.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vssrl.mask.nxv16i16.i64.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vssrl_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -722,7 +722,7 @@ vuint16m4_t test_vssrl_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vssrl.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vssrl.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vssrl_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -734,7 +734,7 @@ vuint16m8_t test_vssrl_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vssrl.mask.nxv32i16.i64.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vssrl.mask.nxv32i16.i64.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vssrl_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -745,7 +745,7 @@ vuint16m8_t test_vssrl_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vssrl.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vssrl.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vssrl_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -757,7 +757,7 @@ vuint32mf2_t test_vssrl_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vssrl.mask.nxv1i32.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vssrl.mask.nxv1i32.i64.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vssrl_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -768,7 +768,7 @@ vuint32mf2_t test_vssrl_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vssrl.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vssrl.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vssrl_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -780,7 +780,7 @@ vuint32m1_t test_vssrl_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vssrl.mask.nxv2i32.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vssrl.mask.nxv2i32.i64.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vssrl_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -791,7 +791,7 @@ vuint32m1_t test_vssrl_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vssrl.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vssrl.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vssrl_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -803,7 +803,7 @@ vuint32m2_t test_vssrl_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vssrl.mask.nxv4i32.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vssrl.mask.nxv4i32.i64.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vssrl_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -814,7 +814,7 @@ vuint32m2_t test_vssrl_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vssrl.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vssrl.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vssrl_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -826,7 +826,7 @@ vuint32m4_t test_vssrl_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vssrl.mask.nxv8i32.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vssrl.mask.nxv8i32.i64.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vssrl_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -837,7 +837,7 @@ vuint32m4_t test_vssrl_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vssrl.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vssrl.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vssrl_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -849,7 +849,7 @@ vuint32m8_t test_vssrl_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vssrl.mask.nxv16i32.i64.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vssrl.mask.nxv16i32.i64.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vssrl_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -860,7 +860,7 @@ vuint32m8_t test_vssrl_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vssrl.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vssrl.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vssrl_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -872,7 +872,7 @@ vuint64m1_t test_vssrl_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vssrl.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vssrl.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vssrl_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -883,7 +883,7 @@ vuint64m1_t test_vssrl_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vssrl.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vssrl.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vssrl_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -895,7 +895,7 @@ vuint64m2_t test_vssrl_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vssrl.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vssrl.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vssrl_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -906,7 +906,7 @@ vuint64m2_t test_vssrl_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vssrl.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vssrl.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vssrl_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -918,7 +918,7 @@ vuint64m4_t test_vssrl_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vssrl.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vssrl.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vssrl_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -929,7 +929,7 @@ vuint64m4_t test_vssrl_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vssrl.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vssrl.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vssrl_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
@@ -941,7 +941,7 @@ vuint64m8_t test_vssrl_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssrl_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vssrl.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vssrl.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[SHIFT:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vssrl_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vssub.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vssub.c
index b0bdae2..02ec458 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vssub.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vssub.c
@@ -890,7 +890,7 @@ vuint64m8_t test_vssubu_vx_u64m8(vuint64m8_t op1, uint64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vssub.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vssub.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vssub_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
@@ -901,7 +901,7 @@ vint8mf8_t test_vssub_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vssub.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vssub.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vssub_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
@@ -912,7 +912,7 @@ vint8mf8_t test_vssub_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vssub.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vssub.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vssub_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
@@ -923,7 +923,7 @@ vint8mf4_t test_vssub_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vssub.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vssub.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vssub_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
@@ -934,7 +934,7 @@ vint8mf4_t test_vssub_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vssub.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vssub.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vssub_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
@@ -945,7 +945,7 @@ vint8mf2_t test_vssub_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vssub.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vssub.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vssub_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
@@ -956,7 +956,7 @@ vint8mf2_t test_vssub_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vssub.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vssub.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vssub_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
@@ -967,7 +967,7 @@ vint8m1_t test_vssub_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vssub.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vssub.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vssub_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
@@ -978,7 +978,7 @@ vint8m1_t test_vssub_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vssub.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vssub.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vssub_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
@@ -989,7 +989,7 @@ vint8m2_t test_vssub_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vssub.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vssub.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vssub_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
@@ -1000,7 +1000,7 @@ vint8m2_t test_vssub_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vssub.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vssub.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vssub_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
@@ -1011,7 +1011,7 @@ vint8m4_t test_vssub_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vssub.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vssub.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vssub_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
@@ -1022,7 +1022,7 @@ vint8m4_t test_vssub_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vssub.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vssub.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vssub_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
@@ -1033,7 +1033,7 @@ vint8m8_t test_vssub_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vssub.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vssub.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vssub_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
@@ -1044,7 +1044,7 @@ vint8m8_t test_vssub_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vssub.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vssub.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vssub_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -1056,7 +1056,7 @@ vint16mf4_t test_vssub_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vssub.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vssub.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vssub_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -1067,7 +1067,7 @@ vint16mf4_t test_vssub_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vssub.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vssub.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vssub_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -1079,7 +1079,7 @@ vint16mf2_t test_vssub_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vssub.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vssub.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vssub_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -1090,7 +1090,7 @@ vint16mf2_t test_vssub_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vssub.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vssub.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vssub_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -1101,7 +1101,7 @@ vint16m1_t test_vssub_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vssub.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vssub.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vssub_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -1112,7 +1112,7 @@ vint16m1_t test_vssub_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vssub.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vssub.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vssub_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -1123,7 +1123,7 @@ vint16m2_t test_vssub_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vssub.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vssub.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vssub_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -1134,7 +1134,7 @@ vint16m2_t test_vssub_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vssub.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vssub.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vssub_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -1145,7 +1145,7 @@ vint16m4_t test_vssub_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vssub.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vssub.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vssub_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -1156,7 +1156,7 @@ vint16m4_t test_vssub_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vssub.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vssub.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vssub_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -1167,7 +1167,7 @@ vint16m8_t test_vssub_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vssub.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vssub.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vssub_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -1178,7 +1178,7 @@ vint16m8_t test_vssub_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vssub.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vssub.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vssub_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -1190,7 +1190,7 @@ vint32mf2_t test_vssub_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vssub.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vssub.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vssub_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -1201,7 +1201,7 @@ vint32mf2_t test_vssub_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vssub.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vssub.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vssub_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -1212,7 +1212,7 @@ vint32m1_t test_vssub_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vssub.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vssub.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vssub_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -1223,7 +1223,7 @@ vint32m1_t test_vssub_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vssub.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vssub.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vssub_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -1234,7 +1234,7 @@ vint32m2_t test_vssub_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vssub.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vssub.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vssub_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -1245,7 +1245,7 @@ vint32m2_t test_vssub_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vssub.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vssub.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vssub_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -1256,7 +1256,7 @@ vint32m4_t test_vssub_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vssub.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vssub.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vssub_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -1267,7 +1267,7 @@ vint32m4_t test_vssub_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vssub.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vssub.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vssub_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -1278,7 +1278,7 @@ vint32m8_t test_vssub_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vssub.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vssub.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vssub_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -1289,7 +1289,7 @@ vint32m8_t test_vssub_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vssub.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vssub.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vssub_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -1300,7 +1300,7 @@ vint64m1_t test_vssub_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vssub.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vssub.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vssub_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -1311,7 +1311,7 @@ vint64m1_t test_vssub_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vssub.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vssub.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vssub_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -1322,7 +1322,7 @@ vint64m2_t test_vssub_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vssub.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vssub.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vssub_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -1333,7 +1333,7 @@ vint64m2_t test_vssub_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vssub.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vssub.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vssub_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -1344,7 +1344,7 @@ vint64m4_t test_vssub_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vssub.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vssub.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vssub_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -1355,7 +1355,7 @@ vint64m4_t test_vssub_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vssub.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vssub.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vssub_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -1366,7 +1366,7 @@ vint64m8_t test_vssub_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssub_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vssub.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vssub.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vssub_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -1377,7 +1377,7 @@ vint64m8_t test_vssub_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vssubu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vssubu.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vssubu_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
@@ -1389,7 +1389,7 @@ vuint8mf8_t test_vssubu_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vssubu.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vssubu.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vssubu_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
@@ -1400,7 +1400,7 @@ vuint8mf8_t test_vssubu_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vssubu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vssubu.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vssubu_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
@@ -1412,7 +1412,7 @@ vuint8mf4_t test_vssubu_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vssubu.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vssubu.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vssubu_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
@@ -1423,7 +1423,7 @@ vuint8mf4_t test_vssubu_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vssubu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vssubu.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vssubu_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
@@ -1435,7 +1435,7 @@ vuint8mf2_t test_vssubu_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vssubu.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vssubu.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vssubu_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
@@ -1446,7 +1446,7 @@ vuint8mf2_t test_vssubu_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vssubu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vssubu.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vssubu_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
@@ -1457,7 +1457,7 @@ vuint8m1_t test_vssubu_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vssubu.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vssubu.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vssubu_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
@@ -1468,7 +1468,7 @@ vuint8m1_t test_vssubu_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vssubu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vssubu.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vssubu_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
@@ -1479,7 +1479,7 @@ vuint8m2_t test_vssubu_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vssubu.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vssubu.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vssubu_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
@@ -1490,7 +1490,7 @@ vuint8m2_t test_vssubu_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vssubu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vssubu.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vssubu_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
@@ -1501,7 +1501,7 @@ vuint8m4_t test_vssubu_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vssubu.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vssubu.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vssubu_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
@@ -1512,7 +1512,7 @@ vuint8m4_t test_vssubu_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vssubu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vssubu.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vssubu_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
@@ -1523,7 +1523,7 @@ vuint8m8_t test_vssubu_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vssubu.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vssubu.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vssubu_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
@@ -1534,7 +1534,7 @@ vuint8m8_t test_vssubu_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vssubu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vssubu.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vssubu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -1546,7 +1546,7 @@ vuint16mf4_t test_vssubu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vssubu.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vssubu.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vssubu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -1558,7 +1558,7 @@ vuint16mf4_t test_vssubu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vssubu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vssubu.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vssubu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -1570,7 +1570,7 @@ vuint16mf2_t test_vssubu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vssubu.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vssubu.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vssubu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -1582,7 +1582,7 @@ vuint16mf2_t test_vssubu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vssubu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vssubu.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vssubu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -1594,7 +1594,7 @@ vuint16m1_t test_vssubu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vssubu.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vssubu.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vssubu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -1605,7 +1605,7 @@ vuint16m1_t test_vssubu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vssubu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vssubu.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vssubu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -1617,7 +1617,7 @@ vuint16m2_t test_vssubu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vssubu.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vssubu.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vssubu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -1628,7 +1628,7 @@ vuint16m2_t test_vssubu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vssubu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vssubu.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vssubu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -1640,7 +1640,7 @@ vuint16m4_t test_vssubu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vssubu.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vssubu.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vssubu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -1651,7 +1651,7 @@ vuint16m4_t test_vssubu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vssubu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vssubu.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vssubu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -1663,7 +1663,7 @@ vuint16m8_t test_vssubu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vssubu.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vssubu.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vssubu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -1674,7 +1674,7 @@ vuint16m8_t test_vssubu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vssubu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vssubu.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vssubu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -1686,7 +1686,7 @@ vuint32mf2_t test_vssubu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vssubu.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vssubu.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vssubu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -1698,7 +1698,7 @@ vuint32mf2_t test_vssubu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vssubu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vssubu.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vssubu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -1710,7 +1710,7 @@ vuint32m1_t test_vssubu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vssubu.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vssubu.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vssubu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -1721,7 +1721,7 @@ vuint32m1_t test_vssubu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vssubu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vssubu.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vssubu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -1733,7 +1733,7 @@ vuint32m2_t test_vssubu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vssubu.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vssubu.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vssubu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -1744,7 +1744,7 @@ vuint32m2_t test_vssubu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vssubu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vssubu.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vssubu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -1756,7 +1756,7 @@ vuint32m4_t test_vssubu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vssubu.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vssubu.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vssubu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -1767,7 +1767,7 @@ vuint32m4_t test_vssubu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vssubu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vssubu.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vssubu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -1779,7 +1779,7 @@ vuint32m8_t test_vssubu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vssubu.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vssubu.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vssubu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -1790,7 +1790,7 @@ vuint32m8_t test_vssubu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vssubu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vssubu.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vssubu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -1802,7 +1802,7 @@ vuint64m1_t test_vssubu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vssubu.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vssubu.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vssubu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -1813,7 +1813,7 @@ vuint64m1_t test_vssubu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vssubu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vssubu.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vssubu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -1825,7 +1825,7 @@ vuint64m2_t test_vssubu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vssubu.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vssubu.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vssubu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -1836,7 +1836,7 @@ vuint64m2_t test_vssubu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vssubu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vssubu.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vssubu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -1848,7 +1848,7 @@ vuint64m4_t test_vssubu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vssubu.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vssubu.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vssubu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -1859,7 +1859,7 @@ vuint64m4_t test_vssubu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vssubu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vssubu.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vssubu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
@@ -1871,7 +1871,7 @@ vuint64m8_t test_vssubu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vssubu_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vssubu.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR1]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vssubu.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vssubu_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsub.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsub.c
index 1ad5d00..51d9848 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsub.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsub.c
@@ -887,7 +887,7 @@ vuint64m8_t test_vsub_vx_u64m8(vuint64m8_t op1, uint64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsub.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsub.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vsub_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, vint8mf8_t op2, size_t vl) {
@@ -897,7 +897,7 @@ vint8mf8_t test_vsub_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsub.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsub.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vsub_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, int8_t op2, size_t vl) {
@@ -907,7 +907,7 @@ vint8mf8_t test_vsub_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsub.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsub.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vsub_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, vint8mf4_t op2, size_t vl) {
@@ -917,7 +917,7 @@ vint8mf4_t test_vsub_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsub.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsub.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vsub_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, int8_t op2, size_t vl) {
@@ -927,7 +927,7 @@ vint8mf4_t test_vsub_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsub.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsub.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vsub_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, vint8mf2_t op2, size_t vl) {
@@ -937,7 +937,7 @@ vint8mf2_t test_vsub_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsub.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsub.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vsub_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, int8_t op2, size_t vl) {
@@ -947,7 +947,7 @@ vint8mf2_t test_vsub_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsub.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsub.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vsub_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, vint8m1_t op2, size_t vl) {
@@ -957,7 +957,7 @@ vint8m1_t test_vsub_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsub.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsub.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vsub_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, int8_t op2, size_t vl) {
@@ -967,7 +967,7 @@ vint8m1_t test_vsub_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsub.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsub.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vsub_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, vint8m2_t op2, size_t vl) {
@@ -977,7 +977,7 @@ vint8m2_t test_vsub_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsub.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsub.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vsub_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, int8_t op2, size_t vl) {
@@ -987,7 +987,7 @@ vint8m2_t test_vsub_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsub.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsub.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vsub_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, vint8m4_t op2, size_t vl) {
@@ -997,7 +997,7 @@ vint8m4_t test_vsub_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsub.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsub.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vsub_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, int8_t op2, size_t vl) {
@@ -1007,7 +1007,7 @@ vint8m4_t test_vsub_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsub.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsub.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vsub_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, vint8m8_t op2, size_t vl) {
@@ -1017,7 +1017,7 @@ vint8m8_t test_vsub_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsub.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsub.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vsub_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, int8_t op2, size_t vl) {
@@ -1027,7 +1027,7 @@ vint8m8_t test_vsub_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsub.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsub.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vsub_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, vint16mf4_t op2, size_t vl) {
@@ -1037,7 +1037,7 @@ vint16mf4_t test_vsub_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsub.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsub.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vsub_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, int16_t op2, size_t vl) {
@@ -1047,7 +1047,7 @@ vint16mf4_t test_vsub_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsub.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsub.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vsub_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, vint16mf2_t op2, size_t vl) {
@@ -1057,7 +1057,7 @@ vint16mf2_t test_vsub_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsub.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsub.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vsub_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, int16_t op2, size_t vl) {
@@ -1067,7 +1067,7 @@ vint16mf2_t test_vsub_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsub.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsub.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vsub_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, vint16m1_t op2, size_t vl) {
@@ -1077,7 +1077,7 @@ vint16m1_t test_vsub_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsub.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsub.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vsub_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, int16_t op2, size_t vl) {
@@ -1087,7 +1087,7 @@ vint16m1_t test_vsub_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsub.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsub.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vsub_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, vint16m2_t op2, size_t vl) {
@@ -1097,7 +1097,7 @@ vint16m2_t test_vsub_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsub.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsub.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vsub_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, int16_t op2, size_t vl) {
@@ -1107,7 +1107,7 @@ vint16m2_t test_vsub_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsub.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsub.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vsub_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, vint16m4_t op2, size_t vl) {
@@ -1117,7 +1117,7 @@ vint16m4_t test_vsub_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsub.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsub.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vsub_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, int16_t op2, size_t vl) {
@@ -1127,7 +1127,7 @@ vint16m4_t test_vsub_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsub.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsub.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vsub_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, vint16m8_t op2, size_t vl) {
@@ -1137,7 +1137,7 @@ vint16m8_t test_vsub_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsub.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsub.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vsub_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, int16_t op2, size_t vl) {
@@ -1147,7 +1147,7 @@ vint16m8_t test_vsub_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsub.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsub.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vsub_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, vint32mf2_t op2, size_t vl) {
@@ -1157,7 +1157,7 @@ vint32mf2_t test_vsub_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsub.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsub.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vsub_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, int32_t op2, size_t vl) {
@@ -1167,7 +1167,7 @@ vint32mf2_t test_vsub_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsub.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsub.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vsub_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, vint32m1_t op2, size_t vl) {
@@ -1177,7 +1177,7 @@ vint32m1_t test_vsub_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsub.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsub.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vsub_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, int32_t op2, size_t vl) {
@@ -1187,7 +1187,7 @@ vint32m1_t test_vsub_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsub.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsub.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vsub_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, vint32m2_t op2, size_t vl) {
@@ -1197,7 +1197,7 @@ vint32m2_t test_vsub_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsub.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsub.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vsub_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, int32_t op2, size_t vl) {
@@ -1207,7 +1207,7 @@ vint32m2_t test_vsub_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsub.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsub.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vsub_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, vint32m4_t op2, size_t vl) {
@@ -1217,7 +1217,7 @@ vint32m4_t test_vsub_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsub.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsub.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vsub_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, int32_t op2, size_t vl) {
@@ -1227,7 +1227,7 @@ vint32m4_t test_vsub_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsub.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsub.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vsub_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, vint32m8_t op2, size_t vl) {
@@ -1237,7 +1237,7 @@ vint32m8_t test_vsub_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsub.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsub.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vsub_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, int32_t op2, size_t vl) {
@@ -1247,7 +1247,7 @@ vint32m8_t test_vsub_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsub.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsub.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vsub_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, vint64m1_t op2, size_t vl) {
@@ -1257,7 +1257,7 @@ vint64m1_t test_vsub_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsub.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsub.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vsub_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, int64_t op2, size_t vl) {
@@ -1267,7 +1267,7 @@ vint64m1_t test_vsub_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsub.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsub.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vsub_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, vint64m2_t op2, size_t vl) {
@@ -1277,7 +1277,7 @@ vint64m2_t test_vsub_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsub.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsub.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vsub_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, int64_t op2, size_t vl) {
@@ -1287,7 +1287,7 @@ vint64m2_t test_vsub_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsub.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsub.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vsub_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, vint64m4_t op2, size_t vl) {
@@ -1297,7 +1297,7 @@ vint64m4_t test_vsub_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsub.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsub.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vsub_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, int64_t op2, size_t vl) {
@@ -1307,7 +1307,7 @@ vint64m4_t test_vsub_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsub.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsub.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vsub_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, vint64m8_t op2, size_t vl) {
@@ -1317,7 +1317,7 @@ vint64m8_t test_vsub_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsub.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsub.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vsub_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, int64_t op2, size_t vl) {
@@ -1327,7 +1327,7 @@ vint64m8_t test_vsub_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsub.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsub.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vsub_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -1337,7 +1337,7 @@ vuint8mf8_t test_vsub_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsub.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vsub.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vsub_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, uint8_t op2, size_t vl) {
@@ -1347,7 +1347,7 @@ vuint8mf8_t test_vsub_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsub.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsub.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vsub_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, vuint8mf4_t op2, size_t vl) {
@@ -1357,7 +1357,7 @@ vuint8mf4_t test_vsub_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsub.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vsub.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vsub_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, uint8_t op2, size_t vl) {
@@ -1367,7 +1367,7 @@ vuint8mf4_t test_vsub_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsub.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsub.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vsub_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, vuint8mf2_t op2, size_t vl) {
@@ -1377,7 +1377,7 @@ vuint8mf2_t test_vsub_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsub.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vsub.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vsub_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, uint8_t op2, size_t vl) {
@@ -1387,7 +1387,7 @@ vuint8mf2_t test_vsub_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsub.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsub.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vsub_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, vuint8m1_t op2, size_t vl) {
@@ -1397,7 +1397,7 @@ vuint8m1_t test_vsub_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t o
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsub.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vsub.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vsub_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, uint8_t op2, size_t vl) {
@@ -1407,7 +1407,7 @@ vuint8m1_t test_vsub_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t o
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsub.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsub.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vsub_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, vuint8m2_t op2, size_t vl) {
@@ -1417,7 +1417,7 @@ vuint8m2_t test_vsub_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t o
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsub.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vsub.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vsub_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, uint8_t op2, size_t vl) {
@@ -1427,7 +1427,7 @@ vuint8m2_t test_vsub_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t o
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsub.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsub.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vsub_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, vuint8m4_t op2, size_t vl) {
@@ -1437,7 +1437,7 @@ vuint8m4_t test_vsub_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t o
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsub.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vsub.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vsub_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, uint8_t op2, size_t vl) {
@@ -1447,7 +1447,7 @@ vuint8m4_t test_vsub_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t o
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsub.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsub.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vsub_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, vuint8m8_t op2, size_t vl) {
@@ -1457,7 +1457,7 @@ vuint8m8_t test_vsub_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t o
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsub.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vsub.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vsub_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, uint8_t op2, size_t vl) {
@@ -1467,7 +1467,7 @@ vuint8m8_t test_vsub_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t o
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsub.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsub.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vsub_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, vuint16mf4_t op2, size_t vl) {
@@ -1477,7 +1477,7 @@ vuint16mf4_t test_vsub_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsub.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vsub.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vsub_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, uint16_t op2, size_t vl) {
@@ -1487,7 +1487,7 @@ vuint16mf4_t test_vsub_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsub.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsub.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vsub_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, vuint16mf2_t op2, size_t vl) {
@@ -1497,7 +1497,7 @@ vuint16mf2_t test_vsub_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsub.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vsub.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vsub_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, uint16_t op2, size_t vl) {
@@ -1507,7 +1507,7 @@ vuint16mf2_t test_vsub_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsub.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsub.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vsub_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, vuint16m1_t op2, size_t vl) {
@@ -1517,7 +1517,7 @@ vuint16m1_t test_vsub_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsub.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vsub.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vsub_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, uint16_t op2, size_t vl) {
@@ -1527,7 +1527,7 @@ vuint16m1_t test_vsub_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsub.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsub.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vsub_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, vuint16m2_t op2, size_t vl) {
@@ -1537,7 +1537,7 @@ vuint16m2_t test_vsub_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsub.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vsub.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vsub_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, uint16_t op2, size_t vl) {
@@ -1547,7 +1547,7 @@ vuint16m2_t test_vsub_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsub.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsub.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vsub_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, vuint16m4_t op2, size_t vl) {
@@ -1557,7 +1557,7 @@ vuint16m4_t test_vsub_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsub.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vsub.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vsub_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, uint16_t op2, size_t vl) {
@@ -1567,7 +1567,7 @@ vuint16m4_t test_vsub_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsub.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsub.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vsub_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, vuint16m8_t op2, size_t vl) {
@@ -1577,7 +1577,7 @@ vuint16m8_t test_vsub_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsub.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vsub.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vsub_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, uint16_t op2, size_t vl) {
@@ -1587,7 +1587,7 @@ vuint16m8_t test_vsub_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsub.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsub.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vsub_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, vuint32mf2_t op2, size_t vl) {
@@ -1597,7 +1597,7 @@ vuint32mf2_t test_vsub_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsub.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vsub.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vsub_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, uint32_t op2, size_t vl) {
@@ -1607,7 +1607,7 @@ vuint32mf2_t test_vsub_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsub.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsub.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vsub_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, vuint32m1_t op2, size_t vl) {
@@ -1617,7 +1617,7 @@ vuint32m1_t test_vsub_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsub.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vsub.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vsub_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, uint32_t op2, size_t vl) {
@@ -1627,7 +1627,7 @@ vuint32m1_t test_vsub_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsub.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsub.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vsub_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, vuint32m2_t op2, size_t vl) {
@@ -1637,7 +1637,7 @@ vuint32m2_t test_vsub_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsub.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vsub.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vsub_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, uint32_t op2, size_t vl) {
@@ -1647,7 +1647,7 @@ vuint32m2_t test_vsub_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsub.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsub.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vsub_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, vuint32m4_t op2, size_t vl) {
@@ -1657,7 +1657,7 @@ vuint32m4_t test_vsub_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsub.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vsub.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vsub_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, uint32_t op2, size_t vl) {
@@ -1667,7 +1667,7 @@ vuint32m4_t test_vsub_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsub.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsub.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vsub_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, vuint32m8_t op2, size_t vl) {
@@ -1677,7 +1677,7 @@ vuint32m8_t test_vsub_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsub.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vsub.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vsub_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, uint32_t op2, size_t vl) {
@@ -1687,7 +1687,7 @@ vuint32m8_t test_vsub_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsub.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsub.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vsub_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, vuint64m1_t op2, size_t vl) {
@@ -1697,7 +1697,7 @@ vuint64m1_t test_vsub_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsub.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vsub.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vsub_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, uint64_t op2, size_t vl) {
@@ -1707,7 +1707,7 @@ vuint64m1_t test_vsub_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsub.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsub.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vsub_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, vuint64m2_t op2, size_t vl) {
@@ -1717,7 +1717,7 @@ vuint64m2_t test_vsub_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsub.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vsub.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vsub_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, uint64_t op2, size_t vl) {
@@ -1727,7 +1727,7 @@ vuint64m2_t test_vsub_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsub.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsub.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vsub_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, vuint64m4_t op2, size_t vl) {
@@ -1737,7 +1737,7 @@ vuint64m4_t test_vsub_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsub.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vsub.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vsub_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, uint64_t op2, size_t vl) {
@@ -1747,7 +1747,7 @@ vuint64m4_t test_vsub_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vsub_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsub.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsub.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vsub_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, vuint64m8_t op2, size_t vl) {
@@ -1757,7 +1757,7 @@ vuint64m8_t test_vsub_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8
 //
 // CHECK-RV64-LABEL: @test_vsub_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsub.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vsub.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vsub_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, uint64_t op2, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsuxei.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsuxei.c
index 45f8dd3..6c829d7 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsuxei.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsuxei.c
@@ -9,7 +9,7 @@
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8:#.*]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i8mf8(int8_t *base, vuint8mf8_t bindex, vint8mf8_t value,
@@ -21,7 +21,7 @@ void test_vsuxei8_v_i8mf8(int8_t *base, vuint8mf8_t bindex, vint8mf8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i8mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i8mf4(int8_t *base, vuint8mf4_t bindex, vint8mf4_t value,
@@ -33,7 +33,7 @@ void test_vsuxei8_v_i8mf4(int8_t *base, vuint8mf4_t bindex, vint8mf4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i8mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i8mf2(int8_t *base, vuint8mf2_t bindex, vint8mf2_t value,
@@ -45,7 +45,7 @@ void test_vsuxei8_v_i8mf2(int8_t *base, vuint8mf2_t bindex, vint8mf2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i8m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i8m1(int8_t *base, vuint8m1_t bindex, vint8m1_t value,
@@ -57,7 +57,7 @@ void test_vsuxei8_v_i8m1(int8_t *base, vuint8m1_t bindex, vint8m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i8m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i8m2(int8_t *base, vuint8m2_t bindex, vint8m2_t value,
@@ -69,7 +69,7 @@ void test_vsuxei8_v_i8m2(int8_t *base, vuint8m2_t bindex, vint8m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i8m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i8m4(int8_t *base, vuint8m4_t bindex, vint8m4_t value,
@@ -81,7 +81,7 @@ void test_vsuxei8_v_i8m4(int8_t *base, vuint8m4_t bindex, vint8m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i8m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i8m8(int8_t *base, vuint8m8_t bindex, vint8m8_t value,
@@ -93,7 +93,7 @@ void test_vsuxei8_v_i8m8(int8_t *base, vuint8m8_t bindex, vint8m8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i8mf8(int8_t *base, vuint16mf4_t bindex, vint8mf8_t value,
@@ -105,7 +105,7 @@ void test_vsuxei16_v_i8mf8(int8_t *base, vuint16mf4_t bindex, vint8mf8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i8mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i8mf4(int8_t *base, vuint16mf2_t bindex, vint8mf4_t value,
@@ -117,7 +117,7 @@ void test_vsuxei16_v_i8mf4(int8_t *base, vuint16mf2_t bindex, vint8mf4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i8mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i8mf2(int8_t *base, vuint16m1_t bindex, vint8mf2_t value,
@@ -129,7 +129,7 @@ void test_vsuxei16_v_i8mf2(int8_t *base, vuint16m1_t bindex, vint8mf2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i8m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i8m1(int8_t *base, vuint16m2_t bindex, vint8m1_t value,
@@ -141,7 +141,7 @@ void test_vsuxei16_v_i8m1(int8_t *base, vuint16m2_t bindex, vint8m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i8m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i8m2(int8_t *base, vuint16m4_t bindex, vint8m2_t value,
@@ -153,7 +153,7 @@ void test_vsuxei16_v_i8m2(int8_t *base, vuint16m4_t bindex, vint8m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i8m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i8m4(int8_t *base, vuint16m8_t bindex, vint8m4_t value,
@@ -165,7 +165,7 @@ void test_vsuxei16_v_i8m4(int8_t *base, vuint16m8_t bindex, vint8m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i8mf8(int8_t *base, vuint32mf2_t bindex, vint8mf8_t value,
@@ -177,7 +177,7 @@ void test_vsuxei32_v_i8mf8(int8_t *base, vuint32mf2_t bindex, vint8mf8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i8mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i8mf4(int8_t *base, vuint32m1_t bindex, vint8mf4_t value,
@@ -189,7 +189,7 @@ void test_vsuxei32_v_i8mf4(int8_t *base, vuint32m1_t bindex, vint8mf4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i8mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i8mf2(int8_t *base, vuint32m2_t bindex, vint8mf2_t value,
@@ -201,7 +201,7 @@ void test_vsuxei32_v_i8mf2(int8_t *base, vuint32m2_t bindex, vint8mf2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i8m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i8m1(int8_t *base, vuint32m4_t bindex, vint8m1_t value,
@@ -213,7 +213,7 @@ void test_vsuxei32_v_i8m1(int8_t *base, vuint32m4_t bindex, vint8m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i8m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i8m2(int8_t *base, vuint32m8_t bindex, vint8m2_t value,
@@ -225,7 +225,7 @@ void test_vsuxei32_v_i8m2(int8_t *base, vuint32m8_t bindex, vint8m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i8mf8(int8_t *base, vuint64m1_t bindex, vint8mf8_t value,
@@ -237,7 +237,7 @@ void test_vsuxei64_v_i8mf8(int8_t *base, vuint64m1_t bindex, vint8mf8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i8mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i8mf4(int8_t *base, vuint64m2_t bindex, vint8mf4_t value,
@@ -249,7 +249,7 @@ void test_vsuxei64_v_i8mf4(int8_t *base, vuint64m2_t bindex, vint8mf4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i8mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i8mf2(int8_t *base, vuint64m4_t bindex, vint8mf2_t value,
@@ -261,7 +261,7 @@ void test_vsuxei64_v_i8mf2(int8_t *base, vuint64m4_t bindex, vint8mf2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i8m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i8m1(int8_t *base, vuint64m8_t bindex, vint8m1_t value,
@@ -273,7 +273,7 @@ void test_vsuxei64_v_i8m1(int8_t *base, vuint64m8_t bindex, vint8m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i16mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i16mf4(int16_t *base, vuint8mf8_t bindex, vint16mf4_t value,
@@ -285,7 +285,7 @@ void test_vsuxei8_v_i16mf4(int16_t *base, vuint8mf8_t bindex, vint16mf4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i16mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i16mf2(int16_t *base, vuint8mf4_t bindex, vint16mf2_t value,
@@ -297,7 +297,7 @@ void test_vsuxei8_v_i16mf2(int16_t *base, vuint8mf4_t bindex, vint16mf2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i16m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i16m1(int16_t *base, vuint8mf2_t bindex, vint16m1_t value,
@@ -309,7 +309,7 @@ void test_vsuxei8_v_i16m1(int16_t *base, vuint8mf2_t bindex, vint16m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i16m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i16m2(int16_t *base, vuint8m1_t bindex, vint16m2_t value,
@@ -321,7 +321,7 @@ void test_vsuxei8_v_i16m2(int16_t *base, vuint8m1_t bindex, vint16m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i16m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i16m4(int16_t *base, vuint8m2_t bindex, vint16m4_t value,
@@ -333,7 +333,7 @@ void test_vsuxei8_v_i16m4(int16_t *base, vuint8m2_t bindex, vint16m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i16m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i16m8(int16_t *base, vuint8m4_t bindex, vint16m8_t value,
@@ -345,7 +345,7 @@ void test_vsuxei8_v_i16m8(int16_t *base, vuint8m4_t bindex, vint16m8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i16mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i16mf4(int16_t *base, vuint16mf4_t bindex,
@@ -357,7 +357,7 @@ void test_vsuxei16_v_i16mf4(int16_t *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i16mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i16mf2(int16_t *base, vuint16mf2_t bindex,
@@ -369,7 +369,7 @@ void test_vsuxei16_v_i16mf2(int16_t *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i16m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i16m1(int16_t *base, vuint16m1_t bindex, vint16m1_t value,
@@ -381,7 +381,7 @@ void test_vsuxei16_v_i16m1(int16_t *base, vuint16m1_t bindex, vint16m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i16m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i16m2(int16_t *base, vuint16m2_t bindex, vint16m2_t value,
@@ -393,7 +393,7 @@ void test_vsuxei16_v_i16m2(int16_t *base, vuint16m2_t bindex, vint16m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i16m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i16m4(int16_t *base, vuint16m4_t bindex, vint16m4_t value,
@@ -405,7 +405,7 @@ void test_vsuxei16_v_i16m4(int16_t *base, vuint16m4_t bindex, vint16m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i16m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i16m8(int16_t *base, vuint16m8_t bindex, vint16m8_t value,
@@ -417,7 +417,7 @@ void test_vsuxei16_v_i16m8(int16_t *base, vuint16m8_t bindex, vint16m8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i16mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i16mf4(int16_t *base, vuint32mf2_t bindex,
@@ -429,7 +429,7 @@ void test_vsuxei32_v_i16mf4(int16_t *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i16mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i16mf2(int16_t *base, vuint32m1_t bindex,
@@ -441,7 +441,7 @@ void test_vsuxei32_v_i16mf2(int16_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i16m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i16m1(int16_t *base, vuint32m2_t bindex, vint16m1_t value,
@@ -453,7 +453,7 @@ void test_vsuxei32_v_i16m1(int16_t *base, vuint32m2_t bindex, vint16m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i16m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i16m2(int16_t *base, vuint32m4_t bindex, vint16m2_t value,
@@ -465,7 +465,7 @@ void test_vsuxei32_v_i16m2(int16_t *base, vuint32m4_t bindex, vint16m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i16m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i16m4(int16_t *base, vuint32m8_t bindex, vint16m4_t value,
@@ -477,7 +477,7 @@ void test_vsuxei32_v_i16m4(int16_t *base, vuint32m8_t bindex, vint16m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i16mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i16mf4(int16_t *base, vuint64m1_t bindex,
@@ -489,7 +489,7 @@ void test_vsuxei64_v_i16mf4(int16_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i16mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i16mf2(int16_t *base, vuint64m2_t bindex,
@@ -501,7 +501,7 @@ void test_vsuxei64_v_i16mf2(int16_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i16m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i16m1(int16_t *base, vuint64m4_t bindex, vint16m1_t value,
@@ -513,7 +513,7 @@ void test_vsuxei64_v_i16m1(int16_t *base, vuint64m4_t bindex, vint16m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i16m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i16m2(int16_t *base, vuint64m8_t bindex, vint16m2_t value,
@@ -525,7 +525,7 @@ void test_vsuxei64_v_i16m2(int16_t *base, vuint64m8_t bindex, vint16m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i32mf2(int32_t *base, vuint8mf8_t bindex, vint32mf2_t value,
@@ -537,7 +537,7 @@ void test_vsuxei8_v_i32mf2(int32_t *base, vuint8mf8_t bindex, vint32mf2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i32m1(int32_t *base, vuint8mf4_t bindex, vint32m1_t value,
@@ -549,7 +549,7 @@ void test_vsuxei8_v_i32m1(int32_t *base, vuint8mf4_t bindex, vint32m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i32m2(int32_t *base, vuint8mf2_t bindex, vint32m2_t value,
@@ -561,7 +561,7 @@ void test_vsuxei8_v_i32m2(int32_t *base, vuint8mf2_t bindex, vint32m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i32m4(int32_t *base, vuint8m1_t bindex, vint32m4_t value,
@@ -573,7 +573,7 @@ void test_vsuxei8_v_i32m4(int32_t *base, vuint8m1_t bindex, vint32m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i32m8(int32_t *base, vuint8m2_t bindex, vint32m8_t value,
@@ -585,7 +585,7 @@ void test_vsuxei8_v_i32m8(int32_t *base, vuint8m2_t bindex, vint32m8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i32mf2(int32_t *base, vuint16mf4_t bindex,
@@ -597,7 +597,7 @@ void test_vsuxei16_v_i32mf2(int32_t *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i32m1(int32_t *base, vuint16mf2_t bindex, vint32m1_t value,
@@ -609,7 +609,7 @@ void test_vsuxei16_v_i32m1(int32_t *base, vuint16mf2_t bindex, vint32m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i32m2(int32_t *base, vuint16m1_t bindex, vint32m2_t value,
@@ -621,7 +621,7 @@ void test_vsuxei16_v_i32m2(int32_t *base, vuint16m1_t bindex, vint32m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i32m4(int32_t *base, vuint16m2_t bindex, vint32m4_t value,
@@ -633,7 +633,7 @@ void test_vsuxei16_v_i32m4(int32_t *base, vuint16m2_t bindex, vint32m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i32m8(int32_t *base, vuint16m4_t bindex, vint32m8_t value,
@@ -645,7 +645,7 @@ void test_vsuxei16_v_i32m8(int32_t *base, vuint16m4_t bindex, vint32m8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i32mf2(int32_t *base, vuint32mf2_t bindex,
@@ -657,7 +657,7 @@ void test_vsuxei32_v_i32mf2(int32_t *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i32m1(int32_t *base, vuint32m1_t bindex, vint32m1_t value,
@@ -669,7 +669,7 @@ void test_vsuxei32_v_i32m1(int32_t *base, vuint32m1_t bindex, vint32m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i32m2(int32_t *base, vuint32m2_t bindex, vint32m2_t value,
@@ -681,7 +681,7 @@ void test_vsuxei32_v_i32m2(int32_t *base, vuint32m2_t bindex, vint32m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i32m4(int32_t *base, vuint32m4_t bindex, vint32m4_t value,
@@ -693,7 +693,7 @@ void test_vsuxei32_v_i32m4(int32_t *base, vuint32m4_t bindex, vint32m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i32m8(int32_t *base, vuint32m8_t bindex, vint32m8_t value,
@@ -705,7 +705,7 @@ void test_vsuxei32_v_i32m8(int32_t *base, vuint32m8_t bindex, vint32m8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i32mf2(int32_t *base, vuint64m1_t bindex,
@@ -717,7 +717,7 @@ void test_vsuxei64_v_i32mf2(int32_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i32m1(int32_t *base, vuint64m2_t bindex, vint32m1_t value,
@@ -729,7 +729,7 @@ void test_vsuxei64_v_i32m1(int32_t *base, vuint64m2_t bindex, vint32m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i32m2(int32_t *base, vuint64m4_t bindex, vint32m2_t value,
@@ -741,7 +741,7 @@ void test_vsuxei64_v_i32m2(int32_t *base, vuint64m4_t bindex, vint32m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i32m4(int32_t *base, vuint64m8_t bindex, vint32m4_t value,
@@ -753,7 +753,7 @@ void test_vsuxei64_v_i32m4(int32_t *base, vuint64m8_t bindex, vint32m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i64m1(int64_t *base, vuint8mf8_t bindex, vint64m1_t value,
@@ -765,7 +765,7 @@ void test_vsuxei8_v_i64m1(int64_t *base, vuint8mf8_t bindex, vint64m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i64m2(int64_t *base, vuint8mf4_t bindex, vint64m2_t value,
@@ -777,7 +777,7 @@ void test_vsuxei8_v_i64m2(int64_t *base, vuint8mf4_t bindex, vint64m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i64m4(int64_t *base, vuint8mf2_t bindex, vint64m4_t value,
@@ -789,7 +789,7 @@ void test_vsuxei8_v_i64m4(int64_t *base, vuint8mf2_t bindex, vint64m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i64m8(int64_t *base, vuint8m1_t bindex, vint64m8_t value,
@@ -801,7 +801,7 @@ void test_vsuxei8_v_i64m8(int64_t *base, vuint8m1_t bindex, vint64m8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i64m1(int64_t *base, vuint16mf4_t bindex, vint64m1_t value,
@@ -813,7 +813,7 @@ void test_vsuxei16_v_i64m1(int64_t *base, vuint16mf4_t bindex, vint64m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i64m2(int64_t *base, vuint16mf2_t bindex, vint64m2_t value,
@@ -825,7 +825,7 @@ void test_vsuxei16_v_i64m2(int64_t *base, vuint16mf2_t bindex, vint64m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i64m4(int64_t *base, vuint16m1_t bindex, vint64m4_t value,
@@ -837,7 +837,7 @@ void test_vsuxei16_v_i64m4(int64_t *base, vuint16m1_t bindex, vint64m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i64m8(int64_t *base, vuint16m2_t bindex, vint64m8_t value,
@@ -849,7 +849,7 @@ void test_vsuxei16_v_i64m8(int64_t *base, vuint16m2_t bindex, vint64m8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i64m1(int64_t *base, vuint32mf2_t bindex, vint64m1_t value,
@@ -861,7 +861,7 @@ void test_vsuxei32_v_i64m1(int64_t *base, vuint32mf2_t bindex, vint64m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i64m2(int64_t *base, vuint32m1_t bindex, vint64m2_t value,
@@ -873,7 +873,7 @@ void test_vsuxei32_v_i64m2(int64_t *base, vuint32m1_t bindex, vint64m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i64m4(int64_t *base, vuint32m2_t bindex, vint64m4_t value,
@@ -885,7 +885,7 @@ void test_vsuxei32_v_i64m4(int64_t *base, vuint32m2_t bindex, vint64m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i64m8(int64_t *base, vuint32m4_t bindex, vint64m8_t value,
@@ -897,7 +897,7 @@ void test_vsuxei32_v_i64m8(int64_t *base, vuint32m4_t bindex, vint64m8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i64m1(int64_t *base, vuint64m1_t bindex, vint64m1_t value,
@@ -909,7 +909,7 @@ void test_vsuxei64_v_i64m1(int64_t *base, vuint64m1_t bindex, vint64m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i64m2(int64_t *base, vuint64m2_t bindex, vint64m2_t value,
@@ -921,7 +921,7 @@ void test_vsuxei64_v_i64m2(int64_t *base, vuint64m2_t bindex, vint64m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i64m4(int64_t *base, vuint64m4_t bindex, vint64m4_t value,
@@ -933,7 +933,7 @@ void test_vsuxei64_v_i64m4(int64_t *base, vuint64m4_t bindex, vint64m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i64m8(int64_t *base, vuint64m8_t bindex, vint64m8_t value,
@@ -945,7 +945,7 @@ void test_vsuxei64_v_i64m8(int64_t *base, vuint64m8_t bindex, vint64m8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u8mf8(uint8_t *base, vuint8mf8_t bindex, vuint8mf8_t value,
@@ -957,7 +957,7 @@ void test_vsuxei8_v_u8mf8(uint8_t *base, vuint8mf8_t bindex, vuint8mf8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u8mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u8mf4(uint8_t *base, vuint8mf4_t bindex, vuint8mf4_t value,
@@ -969,7 +969,7 @@ void test_vsuxei8_v_u8mf4(uint8_t *base, vuint8mf4_t bindex, vuint8mf4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u8mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u8mf2(uint8_t *base, vuint8mf2_t bindex, vuint8mf2_t value,
@@ -981,7 +981,7 @@ void test_vsuxei8_v_u8mf2(uint8_t *base, vuint8mf2_t bindex, vuint8mf2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u8m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u8m1(uint8_t *base, vuint8m1_t bindex, vuint8m1_t value,
@@ -993,7 +993,7 @@ void test_vsuxei8_v_u8m1(uint8_t *base, vuint8m1_t bindex, vuint8m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u8m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u8m2(uint8_t *base, vuint8m2_t bindex, vuint8m2_t value,
@@ -1005,7 +1005,7 @@ void test_vsuxei8_v_u8m2(uint8_t *base, vuint8m2_t bindex, vuint8m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u8m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u8m4(uint8_t *base, vuint8m4_t bindex, vuint8m4_t value,
@@ -1017,7 +1017,7 @@ void test_vsuxei8_v_u8m4(uint8_t *base, vuint8m4_t bindex, vuint8m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u8m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u8m8(uint8_t *base, vuint8m8_t bindex, vuint8m8_t value,
@@ -1029,7 +1029,7 @@ void test_vsuxei8_v_u8m8(uint8_t *base, vuint8m8_t bindex, vuint8m8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u8mf8(uint8_t *base, vuint16mf4_t bindex,
@@ -1041,7 +1041,7 @@ void test_vsuxei16_v_u8mf8(uint8_t *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u8mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u8mf4(uint8_t *base, vuint16mf2_t bindex,
@@ -1053,7 +1053,7 @@ void test_vsuxei16_v_u8mf4(uint8_t *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u8mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u8mf2(uint8_t *base, vuint16m1_t bindex, vuint8mf2_t value,
@@ -1065,7 +1065,7 @@ void test_vsuxei16_v_u8mf2(uint8_t *base, vuint16m1_t bindex, vuint8mf2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u8m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u8m1(uint8_t *base, vuint16m2_t bindex, vuint8m1_t value,
@@ -1077,7 +1077,7 @@ void test_vsuxei16_v_u8m1(uint8_t *base, vuint16m2_t bindex, vuint8m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u8m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u8m2(uint8_t *base, vuint16m4_t bindex, vuint8m2_t value,
@@ -1089,7 +1089,7 @@ void test_vsuxei16_v_u8m2(uint8_t *base, vuint16m4_t bindex, vuint8m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u8m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u8m4(uint8_t *base, vuint16m8_t bindex, vuint8m4_t value,
@@ -1101,7 +1101,7 @@ void test_vsuxei16_v_u8m4(uint8_t *base, vuint16m8_t bindex, vuint8m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u8mf8(uint8_t *base, vuint32mf2_t bindex,
@@ -1113,7 +1113,7 @@ void test_vsuxei32_v_u8mf8(uint8_t *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u8mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u8mf4(uint8_t *base, vuint32m1_t bindex, vuint8mf4_t value,
@@ -1125,7 +1125,7 @@ void test_vsuxei32_v_u8mf4(uint8_t *base, vuint32m1_t bindex, vuint8mf4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u8mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u8mf2(uint8_t *base, vuint32m2_t bindex, vuint8mf2_t value,
@@ -1137,7 +1137,7 @@ void test_vsuxei32_v_u8mf2(uint8_t *base, vuint32m2_t bindex, vuint8mf2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u8m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u8m1(uint8_t *base, vuint32m4_t bindex, vuint8m1_t value,
@@ -1149,7 +1149,7 @@ void test_vsuxei32_v_u8m1(uint8_t *base, vuint32m4_t bindex, vuint8m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u8m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u8m2(uint8_t *base, vuint32m8_t bindex, vuint8m2_t value,
@@ -1161,7 +1161,7 @@ void test_vsuxei32_v_u8m2(uint8_t *base, vuint32m8_t bindex, vuint8m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u8mf8(uint8_t *base, vuint64m1_t bindex, vuint8mf8_t value,
@@ -1173,7 +1173,7 @@ void test_vsuxei64_v_u8mf8(uint8_t *base, vuint64m1_t bindex, vuint8mf8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u8mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u8mf4(uint8_t *base, vuint64m2_t bindex, vuint8mf4_t value,
@@ -1185,7 +1185,7 @@ void test_vsuxei64_v_u8mf4(uint8_t *base, vuint64m2_t bindex, vuint8mf4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u8mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u8mf2(uint8_t *base, vuint64m4_t bindex, vuint8mf2_t value,
@@ -1197,7 +1197,7 @@ void test_vsuxei64_v_u8mf2(uint8_t *base, vuint64m4_t bindex, vuint8mf2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u8m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u8m1(uint8_t *base, vuint64m8_t bindex, vuint8m1_t value,
@@ -1209,7 +1209,7 @@ void test_vsuxei64_v_u8m1(uint8_t *base, vuint64m8_t bindex, vuint8m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u16mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u16mf4(uint16_t *base, vuint8mf8_t bindex,
@@ -1221,7 +1221,7 @@ void test_vsuxei8_v_u16mf4(uint16_t *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u16mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u16mf2(uint16_t *base, vuint8mf4_t bindex,
@@ -1233,7 +1233,7 @@ void test_vsuxei8_v_u16mf2(uint16_t *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u16m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u16m1(uint16_t *base, vuint8mf2_t bindex, vuint16m1_t value,
@@ -1245,7 +1245,7 @@ void test_vsuxei8_v_u16m1(uint16_t *base, vuint8mf2_t bindex, vuint16m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u16m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u16m2(uint16_t *base, vuint8m1_t bindex, vuint16m2_t value,
@@ -1257,7 +1257,7 @@ void test_vsuxei8_v_u16m2(uint16_t *base, vuint8m1_t bindex, vuint16m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u16m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u16m4(uint16_t *base, vuint8m2_t bindex, vuint16m4_t value,
@@ -1269,7 +1269,7 @@ void test_vsuxei8_v_u16m4(uint16_t *base, vuint8m2_t bindex, vuint16m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u16m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u16m8(uint16_t *base, vuint8m4_t bindex, vuint16m8_t value,
@@ -1281,7 +1281,7 @@ void test_vsuxei8_v_u16m8(uint16_t *base, vuint8m4_t bindex, vuint16m8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u16mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u16mf4(uint16_t *base, vuint16mf4_t bindex,
@@ -1293,7 +1293,7 @@ void test_vsuxei16_v_u16mf4(uint16_t *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u16mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u16mf2(uint16_t *base, vuint16mf2_t bindex,
@@ -1305,7 +1305,7 @@ void test_vsuxei16_v_u16mf2(uint16_t *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u16m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u16m1(uint16_t *base, vuint16m1_t bindex,
@@ -1317,7 +1317,7 @@ void test_vsuxei16_v_u16m1(uint16_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u16m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u16m2(uint16_t *base, vuint16m2_t bindex,
@@ -1329,7 +1329,7 @@ void test_vsuxei16_v_u16m2(uint16_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u16m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u16m4(uint16_t *base, vuint16m4_t bindex,
@@ -1341,7 +1341,7 @@ void test_vsuxei16_v_u16m4(uint16_t *base, vuint16m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u16m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u16m8(uint16_t *base, vuint16m8_t bindex,
@@ -1353,7 +1353,7 @@ void test_vsuxei16_v_u16m8(uint16_t *base, vuint16m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u16mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u16mf4(uint16_t *base, vuint32mf2_t bindex,
@@ -1365,7 +1365,7 @@ void test_vsuxei32_v_u16mf4(uint16_t *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u16mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u16mf2(uint16_t *base, vuint32m1_t bindex,
@@ -1377,7 +1377,7 @@ void test_vsuxei32_v_u16mf2(uint16_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u16m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u16m1(uint16_t *base, vuint32m2_t bindex,
@@ -1389,7 +1389,7 @@ void test_vsuxei32_v_u16m1(uint16_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u16m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u16m2(uint16_t *base, vuint32m4_t bindex,
@@ -1401,7 +1401,7 @@ void test_vsuxei32_v_u16m2(uint16_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u16m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u16m4(uint16_t *base, vuint32m8_t bindex,
@@ -1413,7 +1413,7 @@ void test_vsuxei32_v_u16m4(uint16_t *base, vuint32m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u16mf4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u16mf4(uint16_t *base, vuint64m1_t bindex,
@@ -1425,7 +1425,7 @@ void test_vsuxei64_v_u16mf4(uint16_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u16mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u16mf2(uint16_t *base, vuint64m2_t bindex,
@@ -1437,7 +1437,7 @@ void test_vsuxei64_v_u16mf2(uint16_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u16m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u16m1(uint16_t *base, vuint64m4_t bindex,
@@ -1449,7 +1449,7 @@ void test_vsuxei64_v_u16m1(uint16_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u16m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u16m2(uint16_t *base, vuint64m8_t bindex,
@@ -1461,7 +1461,7 @@ void test_vsuxei64_v_u16m2(uint16_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u32mf2(uint32_t *base, vuint8mf8_t bindex,
@@ -1473,7 +1473,7 @@ void test_vsuxei8_v_u32mf2(uint32_t *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u32m1(uint32_t *base, vuint8mf4_t bindex, vuint32m1_t value,
@@ -1485,7 +1485,7 @@ void test_vsuxei8_v_u32m1(uint32_t *base, vuint8mf4_t bindex, vuint32m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u32m2(uint32_t *base, vuint8mf2_t bindex, vuint32m2_t value,
@@ -1497,7 +1497,7 @@ void test_vsuxei8_v_u32m2(uint32_t *base, vuint8mf2_t bindex, vuint32m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u32m4(uint32_t *base, vuint8m1_t bindex, vuint32m4_t value,
@@ -1509,7 +1509,7 @@ void test_vsuxei8_v_u32m4(uint32_t *base, vuint8m1_t bindex, vuint32m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u32m8(uint32_t *base, vuint8m2_t bindex, vuint32m8_t value,
@@ -1521,7 +1521,7 @@ void test_vsuxei8_v_u32m8(uint32_t *base, vuint8m2_t bindex, vuint32m8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u32mf2(uint32_t *base, vuint16mf4_t bindex,
@@ -1533,7 +1533,7 @@ void test_vsuxei16_v_u32mf2(uint32_t *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u32m1(uint32_t *base, vuint16mf2_t bindex,
@@ -1545,7 +1545,7 @@ void test_vsuxei16_v_u32m1(uint32_t *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u32m2(uint32_t *base, vuint16m1_t bindex,
@@ -1557,7 +1557,7 @@ void test_vsuxei16_v_u32m2(uint32_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u32m4(uint32_t *base, vuint16m2_t bindex,
@@ -1569,7 +1569,7 @@ void test_vsuxei16_v_u32m4(uint32_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u32m8(uint32_t *base, vuint16m4_t bindex,
@@ -1581,7 +1581,7 @@ void test_vsuxei16_v_u32m8(uint32_t *base, vuint16m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u32mf2(uint32_t *base, vuint32mf2_t bindex,
@@ -1593,7 +1593,7 @@ void test_vsuxei32_v_u32mf2(uint32_t *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u32m1(uint32_t *base, vuint32m1_t bindex,
@@ -1605,7 +1605,7 @@ void test_vsuxei32_v_u32m1(uint32_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u32m2(uint32_t *base, vuint32m2_t bindex,
@@ -1617,7 +1617,7 @@ void test_vsuxei32_v_u32m2(uint32_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u32m4(uint32_t *base, vuint32m4_t bindex,
@@ -1629,7 +1629,7 @@ void test_vsuxei32_v_u32m4(uint32_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u32m8(uint32_t *base, vuint32m8_t bindex,
@@ -1641,7 +1641,7 @@ void test_vsuxei32_v_u32m8(uint32_t *base, vuint32m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u32mf2(uint32_t *base, vuint64m1_t bindex,
@@ -1653,7 +1653,7 @@ void test_vsuxei64_v_u32mf2(uint32_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u32m1(uint32_t *base, vuint64m2_t bindex,
@@ -1665,7 +1665,7 @@ void test_vsuxei64_v_u32m1(uint32_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u32m2(uint32_t *base, vuint64m4_t bindex,
@@ -1677,7 +1677,7 @@ void test_vsuxei64_v_u32m2(uint32_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u32m4(uint32_t *base, vuint64m8_t bindex,
@@ -1689,7 +1689,7 @@ void test_vsuxei64_v_u32m4(uint32_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u64m1(uint64_t *base, vuint8mf8_t bindex, vuint64m1_t value,
@@ -1701,7 +1701,7 @@ void test_vsuxei8_v_u64m1(uint64_t *base, vuint8mf8_t bindex, vuint64m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u64m2(uint64_t *base, vuint8mf4_t bindex, vuint64m2_t value,
@@ -1713,7 +1713,7 @@ void test_vsuxei8_v_u64m2(uint64_t *base, vuint8mf4_t bindex, vuint64m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u64m4(uint64_t *base, vuint8mf2_t bindex, vuint64m4_t value,
@@ -1725,7 +1725,7 @@ void test_vsuxei8_v_u64m4(uint64_t *base, vuint8mf2_t bindex, vuint64m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u64m8(uint64_t *base, vuint8m1_t bindex, vuint64m8_t value,
@@ -1737,7 +1737,7 @@ void test_vsuxei8_v_u64m8(uint64_t *base, vuint8m1_t bindex, vuint64m8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u64m1(uint64_t *base, vuint16mf4_t bindex,
@@ -1749,7 +1749,7 @@ void test_vsuxei16_v_u64m1(uint64_t *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u64m2(uint64_t *base, vuint16mf2_t bindex,
@@ -1761,7 +1761,7 @@ void test_vsuxei16_v_u64m2(uint64_t *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u64m4(uint64_t *base, vuint16m1_t bindex,
@@ -1773,7 +1773,7 @@ void test_vsuxei16_v_u64m4(uint64_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u64m8(uint64_t *base, vuint16m2_t bindex,
@@ -1785,7 +1785,7 @@ void test_vsuxei16_v_u64m8(uint64_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u64m1(uint64_t *base, vuint32mf2_t bindex,
@@ -1797,7 +1797,7 @@ void test_vsuxei32_v_u64m1(uint64_t *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u64m2(uint64_t *base, vuint32m1_t bindex,
@@ -1809,7 +1809,7 @@ void test_vsuxei32_v_u64m2(uint64_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u64m4(uint64_t *base, vuint32m2_t bindex,
@@ -1821,7 +1821,7 @@ void test_vsuxei32_v_u64m4(uint64_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u64m8(uint64_t *base, vuint32m4_t bindex,
@@ -1833,7 +1833,7 @@ void test_vsuxei32_v_u64m8(uint64_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u64m1(uint64_t *base, vuint64m1_t bindex,
@@ -1845,7 +1845,7 @@ void test_vsuxei64_v_u64m1(uint64_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u64m2(uint64_t *base, vuint64m2_t bindex,
@@ -1857,7 +1857,7 @@ void test_vsuxei64_v_u64m2(uint64_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u64m4(uint64_t *base, vuint64m4_t bindex,
@@ -1869,7 +1869,7 @@ void test_vsuxei64_v_u64m4(uint64_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u64m8(uint64_t *base, vuint64m8_t bindex,
@@ -1881,7 +1881,7 @@ void test_vsuxei64_v_u64m8(uint64_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_f32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1f32.nxv1i8.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1f32.nxv1i8.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_f32mf2(float *base, vuint8mf8_t bindex, vfloat32mf2_t value,
@@ -1893,7 +1893,7 @@ void test_vsuxei8_v_f32mf2(float *base, vuint8mf8_t bindex, vfloat32mf2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_f32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2f32.nxv2i8.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2f32.nxv2i8.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_f32m1(float *base, vuint8mf4_t bindex, vfloat32m1_t value,
@@ -1905,7 +1905,7 @@ void test_vsuxei8_v_f32m1(float *base, vuint8mf4_t bindex, vfloat32m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_f32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4f32.nxv4i8.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4f32.nxv4i8.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_f32m2(float *base, vuint8mf2_t bindex, vfloat32m2_t value,
@@ -1917,7 +1917,7 @@ void test_vsuxei8_v_f32m2(float *base, vuint8mf2_t bindex, vfloat32m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_f32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8f32.nxv8i8.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8f32.nxv8i8.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_f32m4(float *base, vuint8m1_t bindex, vfloat32m4_t value,
@@ -1929,7 +1929,7 @@ void test_vsuxei8_v_f32m4(float *base, vuint8m1_t bindex, vfloat32m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_f32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16f32.nxv16i8.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16f32.nxv16i8.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_f32m8(float *base, vuint8m2_t bindex, vfloat32m8_t value,
@@ -1941,7 +1941,7 @@ void test_vsuxei8_v_f32m8(float *base, vuint8m2_t bindex, vfloat32m8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_f32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1f32.nxv1i16.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1f32.nxv1i16.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_f32mf2(float *base, vuint16mf4_t bindex,
@@ -1953,7 +1953,7 @@ void test_vsuxei16_v_f32mf2(float *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_f32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2f32.nxv2i16.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2f32.nxv2i16.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_f32m1(float *base, vuint16mf2_t bindex, vfloat32m1_t value,
@@ -1965,7 +1965,7 @@ void test_vsuxei16_v_f32m1(float *base, vuint16mf2_t bindex, vfloat32m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_f32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4f32.nxv4i16.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4f32.nxv4i16.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_f32m2(float *base, vuint16m1_t bindex, vfloat32m2_t value,
@@ -1977,7 +1977,7 @@ void test_vsuxei16_v_f32m2(float *base, vuint16m1_t bindex, vfloat32m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_f32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8f32.nxv8i16.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8f32.nxv8i16.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_f32m4(float *base, vuint16m2_t bindex, vfloat32m4_t value,
@@ -1989,7 +1989,7 @@ void test_vsuxei16_v_f32m4(float *base, vuint16m2_t bindex, vfloat32m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_f32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16f32.nxv16i16.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16f32.nxv16i16.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_f32m8(float *base, vuint16m4_t bindex, vfloat32m8_t value,
@@ -2001,7 +2001,7 @@ void test_vsuxei16_v_f32m8(float *base, vuint16m4_t bindex, vfloat32m8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_f32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1f32.nxv1i32.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1f32.nxv1i32.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_f32mf2(float *base, vuint32mf2_t bindex,
@@ -2013,7 +2013,7 @@ void test_vsuxei32_v_f32mf2(float *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_f32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2f32.nxv2i32.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2f32.nxv2i32.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_f32m1(float *base, vuint32m1_t bindex, vfloat32m1_t value,
@@ -2025,7 +2025,7 @@ void test_vsuxei32_v_f32m1(float *base, vuint32m1_t bindex, vfloat32m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_f32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4f32.nxv4i32.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4f32.nxv4i32.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_f32m2(float *base, vuint32m2_t bindex, vfloat32m2_t value,
@@ -2037,7 +2037,7 @@ void test_vsuxei32_v_f32m2(float *base, vuint32m2_t bindex, vfloat32m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_f32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8f32.nxv8i32.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8f32.nxv8i32.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_f32m4(float *base, vuint32m4_t bindex, vfloat32m4_t value,
@@ -2049,7 +2049,7 @@ void test_vsuxei32_v_f32m4(float *base, vuint32m4_t bindex, vfloat32m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_f32m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16f32.nxv16i32.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv16f32.nxv16i32.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_f32m8(float *base, vuint32m8_t bindex, vfloat32m8_t value,
@@ -2061,7 +2061,7 @@ void test_vsuxei32_v_f32m8(float *base, vuint32m8_t bindex, vfloat32m8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_f32mf2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1f32.nxv1i64.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1f32.nxv1i64.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_f32mf2(float *base, vuint64m1_t bindex,
@@ -2073,7 +2073,7 @@ void test_vsuxei64_v_f32mf2(float *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_f32m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2f32.nxv2i64.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2f32.nxv2i64.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_f32m1(float *base, vuint64m2_t bindex, vfloat32m1_t value,
@@ -2085,7 +2085,7 @@ void test_vsuxei64_v_f32m1(float *base, vuint64m2_t bindex, vfloat32m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_f32m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4f32.nxv4i64.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4f32.nxv4i64.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_f32m2(float *base, vuint64m4_t bindex, vfloat32m2_t value,
@@ -2097,7 +2097,7 @@ void test_vsuxei64_v_f32m2(float *base, vuint64m4_t bindex, vfloat32m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_f32m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8f32.nxv8i64.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8f32.nxv8i64.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_f32m4(float *base, vuint64m8_t bindex, vfloat32m4_t value,
@@ -2109,7 +2109,7 @@ void test_vsuxei64_v_f32m4(float *base, vuint64m8_t bindex, vfloat32m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_f64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1f64.nxv1i8.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1f64.nxv1i8.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_f64m1(double *base, vuint8mf8_t bindex, vfloat64m1_t value,
@@ -2121,7 +2121,7 @@ void test_vsuxei8_v_f64m1(double *base, vuint8mf8_t bindex, vfloat64m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_f64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2f64.nxv2i8.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2f64.nxv2i8.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_f64m2(double *base, vuint8mf4_t bindex, vfloat64m2_t value,
@@ -2133,7 +2133,7 @@ void test_vsuxei8_v_f64m2(double *base, vuint8mf4_t bindex, vfloat64m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_f64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4f64.nxv4i8.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4f64.nxv4i8.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_f64m4(double *base, vuint8mf2_t bindex, vfloat64m4_t value,
@@ -2145,7 +2145,7 @@ void test_vsuxei8_v_f64m4(double *base, vuint8mf2_t bindex, vfloat64m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_f64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8f64.nxv8i8.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8f64.nxv8i8.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_f64m8(double *base, vuint8m1_t bindex, vfloat64m8_t value,
@@ -2157,7 +2157,7 @@ void test_vsuxei8_v_f64m8(double *base, vuint8m1_t bindex, vfloat64m8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_f64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1f64.nxv1i16.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1f64.nxv1i16.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_f64m1(double *base, vuint16mf4_t bindex,
@@ -2169,7 +2169,7 @@ void test_vsuxei16_v_f64m1(double *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_f64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2f64.nxv2i16.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2f64.nxv2i16.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_f64m2(double *base, vuint16mf2_t bindex,
@@ -2181,7 +2181,7 @@ void test_vsuxei16_v_f64m2(double *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_f64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4f64.nxv4i16.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4f64.nxv4i16.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_f64m4(double *base, vuint16m1_t bindex, vfloat64m4_t value,
@@ -2193,7 +2193,7 @@ void test_vsuxei16_v_f64m4(double *base, vuint16m1_t bindex, vfloat64m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_f64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8f64.nxv8i16.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8f64.nxv8i16.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_f64m8(double *base, vuint16m2_t bindex, vfloat64m8_t value,
@@ -2205,7 +2205,7 @@ void test_vsuxei16_v_f64m8(double *base, vuint16m2_t bindex, vfloat64m8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_f64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1f64.nxv1i32.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1f64.nxv1i32.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_f64m1(double *base, vuint32mf2_t bindex,
@@ -2217,7 +2217,7 @@ void test_vsuxei32_v_f64m1(double *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_f64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2f64.nxv2i32.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2f64.nxv2i32.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_f64m2(double *base, vuint32m1_t bindex, vfloat64m2_t value,
@@ -2229,7 +2229,7 @@ void test_vsuxei32_v_f64m2(double *base, vuint32m1_t bindex, vfloat64m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_f64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4f64.nxv4i32.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4f64.nxv4i32.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_f64m4(double *base, vuint32m2_t bindex, vfloat64m4_t value,
@@ -2241,7 +2241,7 @@ void test_vsuxei32_v_f64m4(double *base, vuint32m2_t bindex, vfloat64m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_f64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8f64.nxv8i32.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8f64.nxv8i32.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_f64m8(double *base, vuint32m4_t bindex, vfloat64m8_t value,
@@ -2253,7 +2253,7 @@ void test_vsuxei32_v_f64m8(double *base, vuint32m4_t bindex, vfloat64m8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_f64m1(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1f64.nxv1i64.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv1f64.nxv1i64.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_f64m1(double *base, vuint64m1_t bindex, vfloat64m1_t value,
@@ -2265,7 +2265,7 @@ void test_vsuxei64_v_f64m1(double *base, vuint64m1_t bindex, vfloat64m1_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_f64m2(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2f64.nxv2i64.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv2f64.nxv2i64.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_f64m2(double *base, vuint64m2_t bindex, vfloat64m2_t value,
@@ -2277,7 +2277,7 @@ void test_vsuxei64_v_f64m2(double *base, vuint64m2_t bindex, vfloat64m2_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_f64m4(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4f64.nxv4i64.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv4f64.nxv4i64.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_f64m4(double *base, vuint64m4_t bindex, vfloat64m4_t value,
@@ -2289,7 +2289,7 @@ void test_vsuxei64_v_f64m4(double *base, vuint64m4_t bindex, vfloat64m4_t value,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_f64m8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8f64.nxv8i64.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.nxv8f64.nxv8i64.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_f64m8(double *base, vuint64m8_t bindex, vfloat64m8_t value,
@@ -2301,7 +2301,7 @@ void test_vsuxei64_v_f64m8(double *base, vuint64m8_t bindex, vfloat64m8_t value,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i8mf8_m(vbool64_t mask, int8_t *base, vuint8mf8_t bindex,
@@ -2313,7 +2313,7 @@ void test_vsuxei8_v_i8mf8_m(vbool64_t mask, int8_t *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i8mf4_m(vbool32_t mask, int8_t *base, vuint8mf4_t bindex,
@@ -2325,7 +2325,7 @@ void test_vsuxei8_v_i8mf4_m(vbool32_t mask, int8_t *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i8mf2_m(vbool16_t mask, int8_t *base, vuint8mf2_t bindex,
@@ -2337,7 +2337,7 @@ void test_vsuxei8_v_i8mf2_m(vbool16_t mask, int8_t *base, vuint8mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i8m1_m(vbool8_t mask, int8_t *base, vuint8m1_t bindex,
@@ -2349,7 +2349,7 @@ void test_vsuxei8_v_i8m1_m(vbool8_t mask, int8_t *base, vuint8m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i8m2_m(vbool4_t mask, int8_t *base, vuint8m2_t bindex,
@@ -2361,7 +2361,7 @@ void test_vsuxei8_v_i8m2_m(vbool4_t mask, int8_t *base, vuint8m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i8m4_m(vbool2_t mask, int8_t *base, vuint8m4_t bindex,
@@ -2373,7 +2373,7 @@ void test_vsuxei8_v_i8m4_m(vbool2_t mask, int8_t *base, vuint8m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i8m8_m(vbool1_t mask, int8_t *base, vuint8m8_t bindex,
@@ -2385,7 +2385,7 @@ void test_vsuxei8_v_i8m8_m(vbool1_t mask, int8_t *base, vuint8m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i8mf8_m(vbool64_t mask, int8_t *base, vuint16mf4_t bindex,
@@ -2397,7 +2397,7 @@ void test_vsuxei16_v_i8mf8_m(vbool64_t mask, int8_t *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i8mf4_m(vbool32_t mask, int8_t *base, vuint16mf2_t bindex,
@@ -2409,7 +2409,7 @@ void test_vsuxei16_v_i8mf4_m(vbool32_t mask, int8_t *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i8mf2_m(vbool16_t mask, int8_t *base, vuint16m1_t bindex,
@@ -2421,7 +2421,7 @@ void test_vsuxei16_v_i8mf2_m(vbool16_t mask, int8_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i8m1_m(vbool8_t mask, int8_t *base, vuint16m2_t bindex,
@@ -2433,7 +2433,7 @@ void test_vsuxei16_v_i8m1_m(vbool8_t mask, int8_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i8m2_m(vbool4_t mask, int8_t *base, vuint16m4_t bindex,
@@ -2445,7 +2445,7 @@ void test_vsuxei16_v_i8m2_m(vbool4_t mask, int8_t *base, vuint16m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i8m4_m(vbool2_t mask, int8_t *base, vuint16m8_t bindex,
@@ -2457,7 +2457,7 @@ void test_vsuxei16_v_i8m4_m(vbool2_t mask, int8_t *base, vuint16m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i8mf8_m(vbool64_t mask, int8_t *base, vuint32mf2_t bindex,
@@ -2469,7 +2469,7 @@ void test_vsuxei32_v_i8mf8_m(vbool64_t mask, int8_t *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i8mf4_m(vbool32_t mask, int8_t *base, vuint32m1_t bindex,
@@ -2481,7 +2481,7 @@ void test_vsuxei32_v_i8mf4_m(vbool32_t mask, int8_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i8mf2_m(vbool16_t mask, int8_t *base, vuint32m2_t bindex,
@@ -2493,7 +2493,7 @@ void test_vsuxei32_v_i8mf2_m(vbool16_t mask, int8_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i8m1_m(vbool8_t mask, int8_t *base, vuint32m4_t bindex,
@@ -2505,7 +2505,7 @@ void test_vsuxei32_v_i8m1_m(vbool8_t mask, int8_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i8m2_m(vbool4_t mask, int8_t *base, vuint32m8_t bindex,
@@ -2517,7 +2517,7 @@ void test_vsuxei32_v_i8m2_m(vbool4_t mask, int8_t *base, vuint32m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i8mf8_m(vbool64_t mask, int8_t *base, vuint64m1_t bindex,
@@ -2529,7 +2529,7 @@ void test_vsuxei64_v_i8mf8_m(vbool64_t mask, int8_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i8mf4_m(vbool32_t mask, int8_t *base, vuint64m2_t bindex,
@@ -2541,7 +2541,7 @@ void test_vsuxei64_v_i8mf4_m(vbool32_t mask, int8_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i8mf2_m(vbool16_t mask, int8_t *base, vuint64m4_t bindex,
@@ -2553,7 +2553,7 @@ void test_vsuxei64_v_i8mf2_m(vbool16_t mask, int8_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i8m1_m(vbool8_t mask, int8_t *base, vuint64m8_t bindex,
@@ -2565,7 +2565,7 @@ void test_vsuxei64_v_i8m1_m(vbool8_t mask, int8_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i16mf4_m(vbool64_t mask, int16_t *base, vuint8mf8_t bindex,
@@ -2577,7 +2577,7 @@ void test_vsuxei8_v_i16mf4_m(vbool64_t mask, int16_t *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i16mf2_m(vbool32_t mask, int16_t *base, vuint8mf4_t bindex,
@@ -2589,7 +2589,7 @@ void test_vsuxei8_v_i16mf2_m(vbool32_t mask, int16_t *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i16m1_m(vbool16_t mask, int16_t *base, vuint8mf2_t bindex,
@@ -2601,7 +2601,7 @@ void test_vsuxei8_v_i16m1_m(vbool16_t mask, int16_t *base, vuint8mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i16m2_m(vbool8_t mask, int16_t *base, vuint8m1_t bindex,
@@ -2613,7 +2613,7 @@ void test_vsuxei8_v_i16m2_m(vbool8_t mask, int16_t *base, vuint8m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i16m4_m(vbool4_t mask, int16_t *base, vuint8m2_t bindex,
@@ -2625,7 +2625,7 @@ void test_vsuxei8_v_i16m4_m(vbool4_t mask, int16_t *base, vuint8m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i16m8_m(vbool2_t mask, int16_t *base, vuint8m4_t bindex,
@@ -2637,7 +2637,7 @@ void test_vsuxei8_v_i16m8_m(vbool2_t mask, int16_t *base, vuint8m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i16mf4_m(vbool64_t mask, int16_t *base,
@@ -2650,7 +2650,7 @@ void test_vsuxei16_v_i16mf4_m(vbool64_t mask, int16_t *base,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i16mf2_m(vbool32_t mask, int16_t *base,
@@ -2663,7 +2663,7 @@ void test_vsuxei16_v_i16mf2_m(vbool32_t mask, int16_t *base,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i16m1_m(vbool16_t mask, int16_t *base, vuint16m1_t bindex,
@@ -2675,7 +2675,7 @@ void test_vsuxei16_v_i16m1_m(vbool16_t mask, int16_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i16m2_m(vbool8_t mask, int16_t *base, vuint16m2_t bindex,
@@ -2687,7 +2687,7 @@ void test_vsuxei16_v_i16m2_m(vbool8_t mask, int16_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i16m4_m(vbool4_t mask, int16_t *base, vuint16m4_t bindex,
@@ -2699,7 +2699,7 @@ void test_vsuxei16_v_i16m4_m(vbool4_t mask, int16_t *base, vuint16m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i16m8_m(vbool2_t mask, int16_t *base, vuint16m8_t bindex,
@@ -2711,7 +2711,7 @@ void test_vsuxei16_v_i16m8_m(vbool2_t mask, int16_t *base, vuint16m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i16mf4_m(vbool64_t mask, int16_t *base,
@@ -2724,7 +2724,7 @@ void test_vsuxei32_v_i16mf4_m(vbool64_t mask, int16_t *base,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i16mf2_m(vbool32_t mask, int16_t *base, vuint32m1_t bindex,
@@ -2736,7 +2736,7 @@ void test_vsuxei32_v_i16mf2_m(vbool32_t mask, int16_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i16m1_m(vbool16_t mask, int16_t *base, vuint32m2_t bindex,
@@ -2748,7 +2748,7 @@ void test_vsuxei32_v_i16m1_m(vbool16_t mask, int16_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i16m2_m(vbool8_t mask, int16_t *base, vuint32m4_t bindex,
@@ -2760,7 +2760,7 @@ void test_vsuxei32_v_i16m2_m(vbool8_t mask, int16_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i16m4_m(vbool4_t mask, int16_t *base, vuint32m8_t bindex,
@@ -2772,7 +2772,7 @@ void test_vsuxei32_v_i16m4_m(vbool4_t mask, int16_t *base, vuint32m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i16mf4_m(vbool64_t mask, int16_t *base, vuint64m1_t bindex,
@@ -2784,7 +2784,7 @@ void test_vsuxei64_v_i16mf4_m(vbool64_t mask, int16_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i16mf2_m(vbool32_t mask, int16_t *base, vuint64m2_t bindex,
@@ -2796,7 +2796,7 @@ void test_vsuxei64_v_i16mf2_m(vbool32_t mask, int16_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i16m1_m(vbool16_t mask, int16_t *base, vuint64m4_t bindex,
@@ -2808,7 +2808,7 @@ void test_vsuxei64_v_i16m1_m(vbool16_t mask, int16_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i16m2_m(vbool8_t mask, int16_t *base, vuint64m8_t bindex,
@@ -2820,7 +2820,7 @@ void test_vsuxei64_v_i16m2_m(vbool8_t mask, int16_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i32mf2_m(vbool64_t mask, int32_t *base, vuint8mf8_t bindex,
@@ -2832,7 +2832,7 @@ void test_vsuxei8_v_i32mf2_m(vbool64_t mask, int32_t *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i32m1_m(vbool32_t mask, int32_t *base, vuint8mf4_t bindex,
@@ -2844,7 +2844,7 @@ void test_vsuxei8_v_i32m1_m(vbool32_t mask, int32_t *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i32m2_m(vbool16_t mask, int32_t *base, vuint8mf2_t bindex,
@@ -2856,7 +2856,7 @@ void test_vsuxei8_v_i32m2_m(vbool16_t mask, int32_t *base, vuint8mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i32m4_m(vbool8_t mask, int32_t *base, vuint8m1_t bindex,
@@ -2868,7 +2868,7 @@ void test_vsuxei8_v_i32m4_m(vbool8_t mask, int32_t *base, vuint8m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i32m8_m(vbool4_t mask, int32_t *base, vuint8m2_t bindex,
@@ -2880,7 +2880,7 @@ void test_vsuxei8_v_i32m8_m(vbool4_t mask, int32_t *base, vuint8m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i32mf2_m(vbool64_t mask, int32_t *base,
@@ -2893,7 +2893,7 @@ void test_vsuxei16_v_i32mf2_m(vbool64_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i32m1_m(vbool32_t mask, int32_t *base, vuint16mf2_t bindex,
@@ -2905,7 +2905,7 @@ void test_vsuxei16_v_i32m1_m(vbool32_t mask, int32_t *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i32m2_m(vbool16_t mask, int32_t *base, vuint16m1_t bindex,
@@ -2917,7 +2917,7 @@ void test_vsuxei16_v_i32m2_m(vbool16_t mask, int32_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i32m4_m(vbool8_t mask, int32_t *base, vuint16m2_t bindex,
@@ -2929,7 +2929,7 @@ void test_vsuxei16_v_i32m4_m(vbool8_t mask, int32_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i32m8_m(vbool4_t mask, int32_t *base, vuint16m4_t bindex,
@@ -2941,7 +2941,7 @@ void test_vsuxei16_v_i32m8_m(vbool4_t mask, int32_t *base, vuint16m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i32mf2_m(vbool64_t mask, int32_t *base,
@@ -2954,7 +2954,7 @@ void test_vsuxei32_v_i32mf2_m(vbool64_t mask, int32_t *base,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i32m1_m(vbool32_t mask, int32_t *base, vuint32m1_t bindex,
@@ -2966,7 +2966,7 @@ void test_vsuxei32_v_i32m1_m(vbool32_t mask, int32_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i32m2_m(vbool16_t mask, int32_t *base, vuint32m2_t bindex,
@@ -2978,7 +2978,7 @@ void test_vsuxei32_v_i32m2_m(vbool16_t mask, int32_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i32m4_m(vbool8_t mask, int32_t *base, vuint32m4_t bindex,
@@ -2990,7 +2990,7 @@ void test_vsuxei32_v_i32m4_m(vbool8_t mask, int32_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i32m8_m(vbool4_t mask, int32_t *base, vuint32m8_t bindex,
@@ -3002,7 +3002,7 @@ void test_vsuxei32_v_i32m8_m(vbool4_t mask, int32_t *base, vuint32m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i32mf2_m(vbool64_t mask, int32_t *base, vuint64m1_t bindex,
@@ -3014,7 +3014,7 @@ void test_vsuxei64_v_i32mf2_m(vbool64_t mask, int32_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i32m1_m(vbool32_t mask, int32_t *base, vuint64m2_t bindex,
@@ -3026,7 +3026,7 @@ void test_vsuxei64_v_i32m1_m(vbool32_t mask, int32_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i32m2_m(vbool16_t mask, int32_t *base, vuint64m4_t bindex,
@@ -3038,7 +3038,7 @@ void test_vsuxei64_v_i32m2_m(vbool16_t mask, int32_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i32m4_m(vbool8_t mask, int32_t *base, vuint64m8_t bindex,
@@ -3050,7 +3050,7 @@ void test_vsuxei64_v_i32m4_m(vbool8_t mask, int32_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i64m1_m(vbool64_t mask, int64_t *base, vuint8mf8_t bindex,
@@ -3062,7 +3062,7 @@ void test_vsuxei8_v_i64m1_m(vbool64_t mask, int64_t *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i64m2_m(vbool32_t mask, int64_t *base, vuint8mf4_t bindex,
@@ -3074,7 +3074,7 @@ void test_vsuxei8_v_i64m2_m(vbool32_t mask, int64_t *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i64m4_m(vbool16_t mask, int64_t *base, vuint8mf2_t bindex,
@@ -3086,7 +3086,7 @@ void test_vsuxei8_v_i64m4_m(vbool16_t mask, int64_t *base, vuint8mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_i64m8_m(vbool8_t mask, int64_t *base, vuint8m1_t bindex,
@@ -3098,7 +3098,7 @@ void test_vsuxei8_v_i64m8_m(vbool8_t mask, int64_t *base, vuint8m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i64m1_m(vbool64_t mask, int64_t *base, vuint16mf4_t bindex,
@@ -3110,7 +3110,7 @@ void test_vsuxei16_v_i64m1_m(vbool64_t mask, int64_t *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i64m2_m(vbool32_t mask, int64_t *base, vuint16mf2_t bindex,
@@ -3122,7 +3122,7 @@ void test_vsuxei16_v_i64m2_m(vbool32_t mask, int64_t *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i64m4_m(vbool16_t mask, int64_t *base, vuint16m1_t bindex,
@@ -3134,7 +3134,7 @@ void test_vsuxei16_v_i64m4_m(vbool16_t mask, int64_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_i64m8_m(vbool8_t mask, int64_t *base, vuint16m2_t bindex,
@@ -3146,7 +3146,7 @@ void test_vsuxei16_v_i64m8_m(vbool8_t mask, int64_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i64m1_m(vbool64_t mask, int64_t *base, vuint32mf2_t bindex,
@@ -3158,7 +3158,7 @@ void test_vsuxei32_v_i64m1_m(vbool64_t mask, int64_t *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i64m2_m(vbool32_t mask, int64_t *base, vuint32m1_t bindex,
@@ -3170,7 +3170,7 @@ void test_vsuxei32_v_i64m2_m(vbool32_t mask, int64_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i64m4_m(vbool16_t mask, int64_t *base, vuint32m2_t bindex,
@@ -3182,7 +3182,7 @@ void test_vsuxei32_v_i64m4_m(vbool16_t mask, int64_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_i64m8_m(vbool8_t mask, int64_t *base, vuint32m4_t bindex,
@@ -3194,7 +3194,7 @@ void test_vsuxei32_v_i64m8_m(vbool8_t mask, int64_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i64m1_m(vbool64_t mask, int64_t *base, vuint64m1_t bindex,
@@ -3206,7 +3206,7 @@ void test_vsuxei64_v_i64m1_m(vbool64_t mask, int64_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i64m2_m(vbool32_t mask, int64_t *base, vuint64m2_t bindex,
@@ -3218,7 +3218,7 @@ void test_vsuxei64_v_i64m2_m(vbool32_t mask, int64_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i64m4_m(vbool16_t mask, int64_t *base, vuint64m4_t bindex,
@@ -3230,7 +3230,7 @@ void test_vsuxei64_v_i64m4_m(vbool16_t mask, int64_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_i64m8_m(vbool8_t mask, int64_t *base, vuint64m8_t bindex,
@@ -3242,7 +3242,7 @@ void test_vsuxei64_v_i64m8_m(vbool8_t mask, int64_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u8mf8_m(vbool64_t mask, uint8_t *base, vuint8mf8_t bindex,
@@ -3254,7 +3254,7 @@ void test_vsuxei8_v_u8mf8_m(vbool64_t mask, uint8_t *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u8mf4_m(vbool32_t mask, uint8_t *base, vuint8mf4_t bindex,
@@ -3266,7 +3266,7 @@ void test_vsuxei8_v_u8mf4_m(vbool32_t mask, uint8_t *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u8mf2_m(vbool16_t mask, uint8_t *base, vuint8mf2_t bindex,
@@ -3278,7 +3278,7 @@ void test_vsuxei8_v_u8mf2_m(vbool16_t mask, uint8_t *base, vuint8mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u8m1_m(vbool8_t mask, uint8_t *base, vuint8m1_t bindex,
@@ -3290,7 +3290,7 @@ void test_vsuxei8_v_u8m1_m(vbool8_t mask, uint8_t *base, vuint8m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u8m2_m(vbool4_t mask, uint8_t *base, vuint8m2_t bindex,
@@ -3302,7 +3302,7 @@ void test_vsuxei8_v_u8m2_m(vbool4_t mask, uint8_t *base, vuint8m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u8m4_m(vbool2_t mask, uint8_t *base, vuint8m4_t bindex,
@@ -3314,7 +3314,7 @@ void test_vsuxei8_v_u8m4_m(vbool2_t mask, uint8_t *base, vuint8m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 64 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[VALUE:%.*]], <vscale x 64 x i8>* [[TMP0]], <vscale x 64 x i8> [[BINDEX:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u8m8_m(vbool1_t mask, uint8_t *base, vuint8m8_t bindex,
@@ -3326,7 +3326,7 @@ void test_vsuxei8_v_u8m8_m(vbool1_t mask, uint8_t *base, vuint8m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i8.nxv1i16.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u8mf8_m(vbool64_t mask, uint8_t *base, vuint16mf4_t bindex,
@@ -3338,7 +3338,7 @@ void test_vsuxei16_v_u8mf8_m(vbool64_t mask, uint8_t *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i8.nxv2i16.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u8mf4_m(vbool32_t mask, uint8_t *base, vuint16mf2_t bindex,
@@ -3350,7 +3350,7 @@ void test_vsuxei16_v_u8mf4_m(vbool32_t mask, uint8_t *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i8.nxv4i16.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u8mf2_m(vbool16_t mask, uint8_t *base, vuint16m1_t bindex,
@@ -3362,7 +3362,7 @@ void test_vsuxei16_v_u8mf2_m(vbool16_t mask, uint8_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i8.nxv8i16.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u8m1_m(vbool8_t mask, uint8_t *base, vuint16m2_t bindex,
@@ -3374,7 +3374,7 @@ void test_vsuxei16_v_u8m1_m(vbool8_t mask, uint8_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i8.nxv16i16.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u8m2_m(vbool4_t mask, uint8_t *base, vuint16m4_t bindex,
@@ -3386,7 +3386,7 @@ void test_vsuxei16_v_u8m2_m(vbool4_t mask, uint8_t *base, vuint16m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 32 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv32i8.nxv32i16.i64(<vscale x 32 x i8> [[VALUE:%.*]], <vscale x 32 x i8>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u8m4_m(vbool2_t mask, uint8_t *base, vuint16m8_t bindex,
@@ -3398,7 +3398,7 @@ void test_vsuxei16_v_u8m4_m(vbool2_t mask, uint8_t *base, vuint16m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i8.nxv1i32.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u8mf8_m(vbool64_t mask, uint8_t *base, vuint32mf2_t bindex,
@@ -3410,7 +3410,7 @@ void test_vsuxei32_v_u8mf8_m(vbool64_t mask, uint8_t *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i8.nxv2i32.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u8mf4_m(vbool32_t mask, uint8_t *base, vuint32m1_t bindex,
@@ -3422,7 +3422,7 @@ void test_vsuxei32_v_u8mf4_m(vbool32_t mask, uint8_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i8.nxv4i32.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u8mf2_m(vbool16_t mask, uint8_t *base, vuint32m2_t bindex,
@@ -3434,7 +3434,7 @@ void test_vsuxei32_v_u8mf2_m(vbool16_t mask, uint8_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i8.nxv8i32.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u8m1_m(vbool8_t mask, uint8_t *base, vuint32m4_t bindex,
@@ -3446,7 +3446,7 @@ void test_vsuxei32_v_u8m1_m(vbool8_t mask, uint8_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i8.nxv16i32.i64(<vscale x 16 x i8> [[VALUE:%.*]], <vscale x 16 x i8>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u8m2_m(vbool4_t mask, uint8_t *base, vuint32m8_t bindex,
@@ -3458,7 +3458,7 @@ void test_vsuxei32_v_u8m2_m(vbool4_t mask, uint8_t *base, vuint32m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i8.nxv1i64.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u8mf8_m(vbool64_t mask, uint8_t *base, vuint64m1_t bindex,
@@ -3470,7 +3470,7 @@ void test_vsuxei64_v_u8mf8_m(vbool64_t mask, uint8_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 2 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i8.nxv2i64.i64(<vscale x 2 x i8> [[VALUE:%.*]], <vscale x 2 x i8>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u8mf4_m(vbool32_t mask, uint8_t *base, vuint64m2_t bindex,
@@ -3482,7 +3482,7 @@ void test_vsuxei64_v_u8mf4_m(vbool32_t mask, uint8_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 4 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i8.nxv4i64.i64(<vscale x 4 x i8> [[VALUE:%.*]], <vscale x 4 x i8>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u8mf2_m(vbool16_t mask, uint8_t *base, vuint64m4_t bindex,
@@ -3494,7 +3494,7 @@ void test_vsuxei64_v_u8mf2_m(vbool16_t mask, uint8_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 8 x i8>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i8.nxv8i64.i64(<vscale x 8 x i8> [[VALUE:%.*]], <vscale x 8 x i8>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u8m1_m(vbool8_t mask, uint8_t *base, vuint64m8_t bindex,
@@ -3506,7 +3506,7 @@ void test_vsuxei64_v_u8m1_m(vbool8_t mask, uint8_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u16mf4_m(vbool64_t mask, uint16_t *base, vuint8mf8_t bindex,
@@ -3518,7 +3518,7 @@ void test_vsuxei8_v_u16mf4_m(vbool64_t mask, uint16_t *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u16mf2_m(vbool32_t mask, uint16_t *base, vuint8mf4_t bindex,
@@ -3530,7 +3530,7 @@ void test_vsuxei8_v_u16mf2_m(vbool32_t mask, uint16_t *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u16m1_m(vbool16_t mask, uint16_t *base, vuint8mf2_t bindex,
@@ -3542,7 +3542,7 @@ void test_vsuxei8_v_u16m1_m(vbool16_t mask, uint16_t *base, vuint8mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u16m2_m(vbool8_t mask, uint16_t *base, vuint8m1_t bindex,
@@ -3554,7 +3554,7 @@ void test_vsuxei8_v_u16m2_m(vbool8_t mask, uint16_t *base, vuint8m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u16m4_m(vbool4_t mask, uint16_t *base, vuint8m2_t bindex,
@@ -3566,7 +3566,7 @@ void test_vsuxei8_v_u16m4_m(vbool4_t mask, uint16_t *base, vuint8m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i8> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u16m8_m(vbool2_t mask, uint16_t *base, vuint8m4_t bindex,
@@ -3578,7 +3578,7 @@ void test_vsuxei8_v_u16m8_m(vbool2_t mask, uint16_t *base, vuint8m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u16mf4_m(vbool64_t mask, uint16_t *base,
@@ -3591,7 +3591,7 @@ void test_vsuxei16_v_u16mf4_m(vbool64_t mask, uint16_t *base,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u16mf2_m(vbool32_t mask, uint16_t *base,
@@ -3604,7 +3604,7 @@ void test_vsuxei16_v_u16mf2_m(vbool32_t mask, uint16_t *base,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u16m1_m(vbool16_t mask, uint16_t *base, vuint16m1_t bindex,
@@ -3616,7 +3616,7 @@ void test_vsuxei16_v_u16m1_m(vbool16_t mask, uint16_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u16m2_m(vbool8_t mask, uint16_t *base, vuint16m2_t bindex,
@@ -3628,7 +3628,7 @@ void test_vsuxei16_v_u16m2_m(vbool8_t mask, uint16_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u16m4_m(vbool4_t mask, uint16_t *base, vuint16m4_t bindex,
@@ -3640,7 +3640,7 @@ void test_vsuxei16_v_u16m4_m(vbool4_t mask, uint16_t *base, vuint16m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 32 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[VALUE:%.*]], <vscale x 32 x i16>* [[TMP0]], <vscale x 32 x i16> [[BINDEX:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u16m8_m(vbool2_t mask, uint16_t *base, vuint16m8_t bindex,
@@ -3652,7 +3652,7 @@ void test_vsuxei16_v_u16m8_m(vbool2_t mask, uint16_t *base, vuint16m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i16.nxv1i32.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u16mf4_m(vbool64_t mask, uint16_t *base,
@@ -3665,7 +3665,7 @@ void test_vsuxei32_v_u16mf4_m(vbool64_t mask, uint16_t *base,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i16.nxv2i32.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u16mf2_m(vbool32_t mask, uint16_t *base,
@@ -3678,7 +3678,7 @@ void test_vsuxei32_v_u16mf2_m(vbool32_t mask, uint16_t *base,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i16.nxv4i32.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u16m1_m(vbool16_t mask, uint16_t *base, vuint32m2_t bindex,
@@ -3690,7 +3690,7 @@ void test_vsuxei32_v_u16m1_m(vbool16_t mask, uint16_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i16.nxv8i32.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u16m2_m(vbool8_t mask, uint16_t *base, vuint32m4_t bindex,
@@ -3702,7 +3702,7 @@ void test_vsuxei32_v_u16m2_m(vbool8_t mask, uint16_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 16 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i16.nxv16i32.i64(<vscale x 16 x i16> [[VALUE:%.*]], <vscale x 16 x i16>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u16m4_m(vbool4_t mask, uint16_t *base, vuint32m8_t bindex,
@@ -3714,7 +3714,7 @@ void test_vsuxei32_v_u16m4_m(vbool4_t mask, uint16_t *base, vuint32m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 1 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i16.nxv1i64.i64(<vscale x 1 x i16> [[VALUE:%.*]], <vscale x 1 x i16>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u16mf4_m(vbool64_t mask, uint16_t *base,
@@ -3727,7 +3727,7 @@ void test_vsuxei64_v_u16mf4_m(vbool64_t mask, uint16_t *base,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 2 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i16.nxv2i64.i64(<vscale x 2 x i16> [[VALUE:%.*]], <vscale x 2 x i16>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u16mf2_m(vbool32_t mask, uint16_t *base,
@@ -3740,7 +3740,7 @@ void test_vsuxei64_v_u16mf2_m(vbool32_t mask, uint16_t *base,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 4 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i16.nxv4i64.i64(<vscale x 4 x i16> [[VALUE:%.*]], <vscale x 4 x i16>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u16m1_m(vbool16_t mask, uint16_t *base, vuint64m4_t bindex,
@@ -3752,7 +3752,7 @@ void test_vsuxei64_v_u16m1_m(vbool16_t mask, uint16_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i16.nxv8i64.i64(<vscale x 8 x i16> [[VALUE:%.*]], <vscale x 8 x i16>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u16m2_m(vbool8_t mask, uint16_t *base, vuint64m8_t bindex,
@@ -3764,7 +3764,7 @@ void test_vsuxei64_v_u16m2_m(vbool8_t mask, uint16_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u32mf2_m(vbool64_t mask, uint32_t *base, vuint8mf8_t bindex,
@@ -3776,7 +3776,7 @@ void test_vsuxei8_v_u32mf2_m(vbool64_t mask, uint32_t *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u32m1_m(vbool32_t mask, uint32_t *base, vuint8mf4_t bindex,
@@ -3788,7 +3788,7 @@ void test_vsuxei8_v_u32m1_m(vbool32_t mask, uint32_t *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u32m2_m(vbool16_t mask, uint32_t *base, vuint8mf2_t bindex,
@@ -3800,7 +3800,7 @@ void test_vsuxei8_v_u32m2_m(vbool16_t mask, uint32_t *base, vuint8mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u32m4_m(vbool8_t mask, uint32_t *base, vuint8m1_t bindex,
@@ -3812,7 +3812,7 @@ void test_vsuxei8_v_u32m4_m(vbool8_t mask, uint32_t *base, vuint8m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u32m8_m(vbool4_t mask, uint32_t *base, vuint8m2_t bindex,
@@ -3824,7 +3824,7 @@ void test_vsuxei8_v_u32m8_m(vbool4_t mask, uint32_t *base, vuint8m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u32mf2_m(vbool64_t mask, uint32_t *base,
@@ -3837,7 +3837,7 @@ void test_vsuxei16_v_u32mf2_m(vbool64_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u32m1_m(vbool32_t mask, uint32_t *base,
@@ -3850,7 +3850,7 @@ void test_vsuxei16_v_u32m1_m(vbool32_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u32m2_m(vbool16_t mask, uint32_t *base, vuint16m1_t bindex,
@@ -3862,7 +3862,7 @@ void test_vsuxei16_v_u32m2_m(vbool16_t mask, uint32_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u32m4_m(vbool8_t mask, uint32_t *base, vuint16m2_t bindex,
@@ -3874,7 +3874,7 @@ void test_vsuxei16_v_u32m4_m(vbool8_t mask, uint32_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u32m8_m(vbool4_t mask, uint32_t *base, vuint16m4_t bindex,
@@ -3886,7 +3886,7 @@ void test_vsuxei16_v_u32m8_m(vbool4_t mask, uint32_t *base, vuint16m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u32mf2_m(vbool64_t mask, uint32_t *base,
@@ -3899,7 +3899,7 @@ void test_vsuxei32_v_u32mf2_m(vbool64_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u32m1_m(vbool32_t mask, uint32_t *base, vuint32m1_t bindex,
@@ -3911,7 +3911,7 @@ void test_vsuxei32_v_u32m1_m(vbool32_t mask, uint32_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u32m2_m(vbool16_t mask, uint32_t *base, vuint32m2_t bindex,
@@ -3923,7 +3923,7 @@ void test_vsuxei32_v_u32m2_m(vbool16_t mask, uint32_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u32m4_m(vbool8_t mask, uint32_t *base, vuint32m4_t bindex,
@@ -3935,7 +3935,7 @@ void test_vsuxei32_v_u32m4_m(vbool8_t mask, uint32_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 16 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[VALUE:%.*]], <vscale x 16 x i32>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u32m8_m(vbool4_t mask, uint32_t *base, vuint32m8_t bindex,
@@ -3947,7 +3947,7 @@ void test_vsuxei32_v_u32m8_m(vbool4_t mask, uint32_t *base, vuint32m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 1 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i32.nxv1i64.i64(<vscale x 1 x i32> [[VALUE:%.*]], <vscale x 1 x i32>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u32mf2_m(vbool64_t mask, uint32_t *base,
@@ -3960,7 +3960,7 @@ void test_vsuxei64_v_u32mf2_m(vbool64_t mask, uint32_t *base,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 2 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i32.nxv2i64.i64(<vscale x 2 x i32> [[VALUE:%.*]], <vscale x 2 x i32>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u32m1_m(vbool32_t mask, uint32_t *base, vuint64m2_t bindex,
@@ -3972,7 +3972,7 @@ void test_vsuxei64_v_u32m1_m(vbool32_t mask, uint32_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i32.nxv4i64.i64(<vscale x 4 x i32> [[VALUE:%.*]], <vscale x 4 x i32>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u32m2_m(vbool16_t mask, uint32_t *base, vuint64m4_t bindex,
@@ -3984,7 +3984,7 @@ void test_vsuxei64_v_u32m2_m(vbool16_t mask, uint32_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 8 x i32>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i32.nxv8i64.i64(<vscale x 8 x i32> [[VALUE:%.*]], <vscale x 8 x i32>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u32m4_m(vbool8_t mask, uint32_t *base, vuint64m8_t bindex,
@@ -3996,7 +3996,7 @@ void test_vsuxei64_v_u32m4_m(vbool8_t mask, uint32_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u64m1_m(vbool64_t mask, uint64_t *base, vuint8mf8_t bindex,
@@ -4008,7 +4008,7 @@ void test_vsuxei8_v_u64m1_m(vbool64_t mask, uint64_t *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u64m2_m(vbool32_t mask, uint64_t *base, vuint8mf4_t bindex,
@@ -4020,7 +4020,7 @@ void test_vsuxei8_v_u64m2_m(vbool32_t mask, uint64_t *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u64m4_m(vbool16_t mask, uint64_t *base, vuint8mf2_t bindex,
@@ -4032,7 +4032,7 @@ void test_vsuxei8_v_u64m4_m(vbool16_t mask, uint64_t *base, vuint8mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_u64m8_m(vbool8_t mask, uint64_t *base, vuint8m1_t bindex,
@@ -4044,7 +4044,7 @@ void test_vsuxei8_v_u64m8_m(vbool8_t mask, uint64_t *base, vuint8m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u64m1_m(vbool64_t mask, uint64_t *base,
@@ -4057,7 +4057,7 @@ void test_vsuxei16_v_u64m1_m(vbool64_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u64m2_m(vbool32_t mask, uint64_t *base,
@@ -4070,7 +4070,7 @@ void test_vsuxei16_v_u64m2_m(vbool32_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u64m4_m(vbool16_t mask, uint64_t *base, vuint16m1_t bindex,
@@ -4082,7 +4082,7 @@ void test_vsuxei16_v_u64m4_m(vbool16_t mask, uint64_t *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_u64m8_m(vbool8_t mask, uint64_t *base, vuint16m2_t bindex,
@@ -4094,7 +4094,7 @@ void test_vsuxei16_v_u64m8_m(vbool8_t mask, uint64_t *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u64m1_m(vbool64_t mask, uint64_t *base,
@@ -4107,7 +4107,7 @@ void test_vsuxei32_v_u64m1_m(vbool64_t mask, uint64_t *base,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u64m2_m(vbool32_t mask, uint64_t *base, vuint32m1_t bindex,
@@ -4119,7 +4119,7 @@ void test_vsuxei32_v_u64m2_m(vbool32_t mask, uint64_t *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u64m4_m(vbool16_t mask, uint64_t *base, vuint32m2_t bindex,
@@ -4131,7 +4131,7 @@ void test_vsuxei32_v_u64m4_m(vbool16_t mask, uint64_t *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_u64m8_m(vbool8_t mask, uint64_t *base, vuint32m4_t bindex,
@@ -4143,7 +4143,7 @@ void test_vsuxei32_v_u64m8_m(vbool8_t mask, uint64_t *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 1 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[VALUE:%.*]], <vscale x 1 x i64>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u64m1_m(vbool64_t mask, uint64_t *base, vuint64m1_t bindex,
@@ -4155,7 +4155,7 @@ void test_vsuxei64_v_u64m1_m(vbool64_t mask, uint64_t *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[VALUE:%.*]], <vscale x 2 x i64>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u64m2_m(vbool32_t mask, uint64_t *base, vuint64m2_t bindex,
@@ -4167,7 +4167,7 @@ void test_vsuxei64_v_u64m2_m(vbool32_t mask, uint64_t *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 4 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[VALUE:%.*]], <vscale x 4 x i64>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u64m4_m(vbool16_t mask, uint64_t *base, vuint64m4_t bindex,
@@ -4179,7 +4179,7 @@ void test_vsuxei64_v_u64m4_m(vbool16_t mask, uint64_t *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 8 x i64>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[VALUE:%.*]], <vscale x 8 x i64>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_u64m8_m(vbool8_t mask, uint64_t *base, vuint64m8_t bindex,
@@ -4191,7 +4191,7 @@ void test_vsuxei64_v_u64m8_m(vbool8_t mask, uint64_t *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1f32.nxv1i8.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1f32.nxv1i8.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_f32mf2_m(vbool64_t mask, float *base, vuint8mf8_t bindex,
@@ -4203,7 +4203,7 @@ void test_vsuxei8_v_f32mf2_m(vbool64_t mask, float *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2f32.nxv2i8.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2f32.nxv2i8.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_f32m1_m(vbool32_t mask, float *base, vuint8mf4_t bindex,
@@ -4215,7 +4215,7 @@ void test_vsuxei8_v_f32m1_m(vbool32_t mask, float *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4f32.nxv4i8.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4f32.nxv4i8.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_f32m2_m(vbool16_t mask, float *base, vuint8mf2_t bindex,
@@ -4227,7 +4227,7 @@ void test_vsuxei8_v_f32m2_m(vbool16_t mask, float *base, vuint8mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8f32.nxv8i8.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8f32.nxv8i8.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_f32m4_m(vbool8_t mask, float *base, vuint8m1_t bindex,
@@ -4239,7 +4239,7 @@ void test_vsuxei8_v_f32m4_m(vbool8_t mask, float *base, vuint8m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16f32.nxv16i8.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16f32.nxv16i8.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i8> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_f32m8_m(vbool4_t mask, float *base, vuint8m2_t bindex,
@@ -4251,7 +4251,7 @@ void test_vsuxei8_v_f32m8_m(vbool4_t mask, float *base, vuint8m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1f32.nxv1i16.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1f32.nxv1i16.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_f32mf2_m(vbool64_t mask, float *base, vuint16mf4_t bindex,
@@ -4263,7 +4263,7 @@ void test_vsuxei16_v_f32mf2_m(vbool64_t mask, float *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2f32.nxv2i16.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2f32.nxv2i16.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_f32m1_m(vbool32_t mask, float *base, vuint16mf2_t bindex,
@@ -4275,7 +4275,7 @@ void test_vsuxei16_v_f32m1_m(vbool32_t mask, float *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4f32.nxv4i16.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4f32.nxv4i16.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_f32m2_m(vbool16_t mask, float *base, vuint16m1_t bindex,
@@ -4287,7 +4287,7 @@ void test_vsuxei16_v_f32m2_m(vbool16_t mask, float *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8f32.nxv8i16.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8f32.nxv8i16.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_f32m4_m(vbool8_t mask, float *base, vuint16m2_t bindex,
@@ -4299,7 +4299,7 @@ void test_vsuxei16_v_f32m4_m(vbool8_t mask, float *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16f32.nxv16i16.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16f32.nxv16i16.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i16> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_f32m8_m(vbool4_t mask, float *base, vuint16m4_t bindex,
@@ -4311,7 +4311,7 @@ void test_vsuxei16_v_f32m8_m(vbool4_t mask, float *base, vuint16m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1f32.nxv1i32.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1f32.nxv1i32.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_f32mf2_m(vbool64_t mask, float *base, vuint32mf2_t bindex,
@@ -4323,7 +4323,7 @@ void test_vsuxei32_v_f32mf2_m(vbool64_t mask, float *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2f32.nxv2i32.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2f32.nxv2i32.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_f32m1_m(vbool32_t mask, float *base, vuint32m1_t bindex,
@@ -4335,7 +4335,7 @@ void test_vsuxei32_v_f32m1_m(vbool32_t mask, float *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4f32.nxv4i32.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4f32.nxv4i32.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_f32m2_m(vbool16_t mask, float *base, vuint32m2_t bindex,
@@ -4347,7 +4347,7 @@ void test_vsuxei32_v_f32m2_m(vbool16_t mask, float *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8f32.nxv8i32.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8f32.nxv8i32.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_f32m4_m(vbool8_t mask, float *base, vuint32m4_t bindex,
@@ -4359,7 +4359,7 @@ void test_vsuxei32_v_f32m4_m(vbool8_t mask, float *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_f32m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 16 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16f32.nxv16i32.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv16f32.nxv16i32.i64(<vscale x 16 x float> [[VALUE:%.*]], <vscale x 16 x float>* [[TMP0]], <vscale x 16 x i32> [[BINDEX:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_f32m8_m(vbool4_t mask, float *base, vuint32m8_t bindex,
@@ -4371,7 +4371,7 @@ void test_vsuxei32_v_f32m8_m(vbool4_t mask, float *base, vuint32m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_f32mf2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 1 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1f32.nxv1i64.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1f32.nxv1i64.i64(<vscale x 1 x float> [[VALUE:%.*]], <vscale x 1 x float>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_f32mf2_m(vbool64_t mask, float *base, vuint64m1_t bindex,
@@ -4383,7 +4383,7 @@ void test_vsuxei64_v_f32mf2_m(vbool64_t mask, float *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_f32m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 2 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2f32.nxv2i64.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2f32.nxv2i64.i64(<vscale x 2 x float> [[VALUE:%.*]], <vscale x 2 x float>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_f32m1_m(vbool32_t mask, float *base, vuint64m2_t bindex,
@@ -4395,7 +4395,7 @@ void test_vsuxei64_v_f32m1_m(vbool32_t mask, float *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_f32m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4f32.nxv4i64.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4f32.nxv4i64.i64(<vscale x 4 x float> [[VALUE:%.*]], <vscale x 4 x float>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_f32m2_m(vbool16_t mask, float *base, vuint64m4_t bindex,
@@ -4407,7 +4407,7 @@ void test_vsuxei64_v_f32m2_m(vbool16_t mask, float *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_f32m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 8 x float>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8f32.nxv8i64.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8f32.nxv8i64.i64(<vscale x 8 x float> [[VALUE:%.*]], <vscale x 8 x float>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_f32m4_m(vbool8_t mask, float *base, vuint64m8_t bindex,
@@ -4419,7 +4419,7 @@ void test_vsuxei64_v_f32m4_m(vbool8_t mask, float *base, vuint64m8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1f64.nxv1i8.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1f64.nxv1i8.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i8> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_f64m1_m(vbool64_t mask, double *base, vuint8mf8_t bindex,
@@ -4431,7 +4431,7 @@ void test_vsuxei8_v_f64m1_m(vbool64_t mask, double *base, vuint8mf8_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2f64.nxv2i8.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2f64.nxv2i8.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i8> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_f64m2_m(vbool32_t mask, double *base, vuint8mf4_t bindex,
@@ -4443,7 +4443,7 @@ void test_vsuxei8_v_f64m2_m(vbool32_t mask, double *base, vuint8mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4f64.nxv4i8.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4f64.nxv4i8.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i8> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_f64m4_m(vbool16_t mask, double *base, vuint8mf2_t bindex,
@@ -4455,7 +4455,7 @@ void test_vsuxei8_v_f64m4_m(vbool16_t mask, double *base, vuint8mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei8_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8f64.nxv8i8.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8f64.nxv8i8.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i8> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei8_v_f64m8_m(vbool8_t mask, double *base, vuint8m1_t bindex,
@@ -4467,7 +4467,7 @@ void test_vsuxei8_v_f64m8_m(vbool8_t mask, double *base, vuint8m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1f64.nxv1i16.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1f64.nxv1i16.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i16> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_f64m1_m(vbool64_t mask, double *base, vuint16mf4_t bindex,
@@ -4479,7 +4479,7 @@ void test_vsuxei16_v_f64m1_m(vbool64_t mask, double *base, vuint16mf4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2f64.nxv2i16.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2f64.nxv2i16.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i16> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_f64m2_m(vbool32_t mask, double *base, vuint16mf2_t bindex,
@@ -4491,7 +4491,7 @@ void test_vsuxei16_v_f64m2_m(vbool32_t mask, double *base, vuint16mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4f64.nxv4i16.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4f64.nxv4i16.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i16> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_f64m4_m(vbool16_t mask, double *base, vuint16m1_t bindex,
@@ -4503,7 +4503,7 @@ void test_vsuxei16_v_f64m4_m(vbool16_t mask, double *base, vuint16m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei16_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8f64.nxv8i16.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8f64.nxv8i16.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i16> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei16_v_f64m8_m(vbool8_t mask, double *base, vuint16m2_t bindex,
@@ -4515,7 +4515,7 @@ void test_vsuxei16_v_f64m8_m(vbool8_t mask, double *base, vuint16m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1f64.nxv1i32.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1f64.nxv1i32.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i32> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_f64m1_m(vbool64_t mask, double *base, vuint32mf2_t bindex,
@@ -4527,7 +4527,7 @@ void test_vsuxei32_v_f64m1_m(vbool64_t mask, double *base, vuint32mf2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2f64.nxv2i32.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2f64.nxv2i32.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i32> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_f64m2_m(vbool32_t mask, double *base, vuint32m1_t bindex,
@@ -4539,7 +4539,7 @@ void test_vsuxei32_v_f64m2_m(vbool32_t mask, double *base, vuint32m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4f64.nxv4i32.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4f64.nxv4i32.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i32> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_f64m4_m(vbool16_t mask, double *base, vuint32m2_t bindex,
@@ -4551,7 +4551,7 @@ void test_vsuxei32_v_f64m4_m(vbool16_t mask, double *base, vuint32m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei32_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8f64.nxv8i32.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8f64.nxv8i32.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i32> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei32_v_f64m8_m(vbool8_t mask, double *base, vuint32m4_t bindex,
@@ -4563,7 +4563,7 @@ void test_vsuxei32_v_f64m8_m(vbool8_t mask, double *base, vuint32m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_f64m1_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 1 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1f64.nxv1i64.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv1f64.nxv1i64.i64(<vscale x 1 x double> [[VALUE:%.*]], <vscale x 1 x double>* [[TMP0]], <vscale x 1 x i64> [[BINDEX:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_f64m1_m(vbool64_t mask, double *base, vuint64m1_t bindex,
@@ -4575,7 +4575,7 @@ void test_vsuxei64_v_f64m1_m(vbool64_t mask, double *base, vuint64m1_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_f64m2_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2f64.nxv2i64.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv2f64.nxv2i64.i64(<vscale x 2 x double> [[VALUE:%.*]], <vscale x 2 x double>* [[TMP0]], <vscale x 2 x i64> [[BINDEX:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_f64m2_m(vbool32_t mask, double *base, vuint64m2_t bindex,
@@ -4587,7 +4587,7 @@ void test_vsuxei64_v_f64m2_m(vbool32_t mask, double *base, vuint64m2_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_f64m4_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 4 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4f64.nxv4i64.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv4f64.nxv4i64.i64(<vscale x 4 x double> [[VALUE:%.*]], <vscale x 4 x double>* [[TMP0]], <vscale x 4 x i64> [[BINDEX:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_f64m4_m(vbool16_t mask, double *base, vuint64m4_t bindex,
@@ -4599,7 +4599,7 @@ void test_vsuxei64_v_f64m4_m(vbool16_t mask, double *base, vuint64m4_t bindex,
 // CHECK-RV64-LABEL: @test_vsuxei64_v_f64m8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
-// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8f64.nxv8i64.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) [[ATTR8]]
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsuxei.mask.nxv8f64.nxv8i64.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i64> [[BINDEX:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsuxei64_v_f64m8_m(vbool8_t mask, double *base, vuint64m8_t bindex,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwadd.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwadd.c
index 7b83551..62c5076 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwadd.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwadd.c
@@ -1215,7 +1215,7 @@ vuint64m8_t test_vwaddu_wx_u64m8(vuint64m8_t op1, uint32_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vwadd_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwadd.mask.nxv1i16.nxv1i8.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwadd.mask.nxv1i16.nxv1i8.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vwadd_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -1226,7 +1226,7 @@ vint16mf4_t test_vwadd_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwadd.mask.nxv1i16.nxv1i8.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwadd.mask.nxv1i16.nxv1i8.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vwadd_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -1237,7 +1237,7 @@ vint16mf4_t test_vwadd_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwadd.w.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwadd.w.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vwadd_wv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -1248,7 +1248,7 @@ vint16mf4_t test_vwadd_wv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwadd.w.mask.nxv1i16.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwadd.w.mask.nxv1i16.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vwadd_wx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -1259,7 +1259,7 @@ vint16mf4_t test_vwadd_wx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwadd.mask.nxv2i16.nxv2i8.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwadd.mask.nxv2i16.nxv2i8.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vwadd_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -1270,7 +1270,7 @@ vint16mf2_t test_vwadd_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwadd.mask.nxv2i16.nxv2i8.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwadd.mask.nxv2i16.nxv2i8.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vwadd_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -1281,7 +1281,7 @@ vint16mf2_t test_vwadd_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwadd.w.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwadd.w.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vwadd_wv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -1292,7 +1292,7 @@ vint16mf2_t test_vwadd_wv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwadd.w.mask.nxv2i16.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwadd.w.mask.nxv2i16.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vwadd_wx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -1303,7 +1303,7 @@ vint16mf2_t test_vwadd_wx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwadd.mask.nxv4i16.nxv4i8.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwadd.mask.nxv4i16.nxv4i8.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwadd_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -1314,7 +1314,7 @@ vint16m1_t test_vwadd_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwadd.mask.nxv4i16.nxv4i8.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwadd.mask.nxv4i16.nxv4i8.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwadd_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -1325,7 +1325,7 @@ vint16m1_t test_vwadd_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwadd.w.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwadd.w.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwadd_wv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -1336,7 +1336,7 @@ vint16m1_t test_vwadd_wv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwadd.w.mask.nxv4i16.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwadd.w.mask.nxv4i16.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwadd_wx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -1347,7 +1347,7 @@ vint16m1_t test_vwadd_wx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwadd.mask.nxv8i16.nxv8i8.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwadd.mask.nxv8i16.nxv8i8.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vwadd_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -1358,7 +1358,7 @@ vint16m2_t test_vwadd_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwadd.mask.nxv8i16.nxv8i8.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwadd.mask.nxv8i16.nxv8i8.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vwadd_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -1369,7 +1369,7 @@ vint16m2_t test_vwadd_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwadd.w.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwadd.w.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vwadd_wv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -1380,7 +1380,7 @@ vint16m2_t test_vwadd_wv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwadd.w.mask.nxv8i16.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwadd.w.mask.nxv8i16.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vwadd_wx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -1391,7 +1391,7 @@ vint16m2_t test_vwadd_wx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwadd.mask.nxv16i16.nxv16i8.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwadd.mask.nxv16i16.nxv16i8.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vwadd_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -1402,7 +1402,7 @@ vint16m4_t test_vwadd_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwadd.mask.nxv16i16.nxv16i8.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwadd.mask.nxv16i16.nxv16i8.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vwadd_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -1413,7 +1413,7 @@ vint16m4_t test_vwadd_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwadd.w.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwadd.w.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vwadd_wv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -1424,7 +1424,7 @@ vint16m4_t test_vwadd_wv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwadd.w.mask.nxv16i16.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwadd.w.mask.nxv16i16.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vwadd_wx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -1435,7 +1435,7 @@ vint16m4_t test_vwadd_wx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwadd.mask.nxv32i16.nxv32i8.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwadd.mask.nxv32i16.nxv32i8.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vwadd_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -1446,7 +1446,7 @@ vint16m8_t test_vwadd_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwadd.mask.nxv32i16.nxv32i8.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwadd.mask.nxv32i16.nxv32i8.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vwadd_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -1457,7 +1457,7 @@ vint16m8_t test_vwadd_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwadd.w.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwadd.w.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vwadd_wv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -1468,7 +1468,7 @@ vint16m8_t test_vwadd_wv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwadd.w.mask.nxv32i16.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwadd.w.mask.nxv32i16.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vwadd_wx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -1479,7 +1479,7 @@ vint16m8_t test_vwadd_wx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwadd.mask.nxv1i32.nxv1i16.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwadd.mask.nxv1i32.nxv1i16.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vwadd_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -1491,7 +1491,7 @@ vint32mf2_t test_vwadd_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwadd.mask.nxv1i32.nxv1i16.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwadd.mask.nxv1i32.nxv1i16.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vwadd_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -1502,7 +1502,7 @@ vint32mf2_t test_vwadd_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwadd.w.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwadd.w.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vwadd_wv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -1514,7 +1514,7 @@ vint32mf2_t test_vwadd_wv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwadd.w.mask.nxv1i32.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwadd.w.mask.nxv1i32.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vwadd_wx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -1525,7 +1525,7 @@ vint32mf2_t test_vwadd_wx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwadd.mask.nxv2i32.nxv2i16.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwadd.mask.nxv2i32.nxv2i16.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwadd_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -1536,7 +1536,7 @@ vint32m1_t test_vwadd_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwadd.mask.nxv2i32.nxv2i16.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwadd.mask.nxv2i32.nxv2i16.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwadd_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -1547,7 +1547,7 @@ vint32m1_t test_vwadd_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwadd.w.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwadd.w.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwadd_wv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -1558,7 +1558,7 @@ vint32m1_t test_vwadd_wv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwadd.w.mask.nxv2i32.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwadd.w.mask.nxv2i32.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwadd_wx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -1569,7 +1569,7 @@ vint32m1_t test_vwadd_wx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwadd.mask.nxv4i32.nxv4i16.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwadd.mask.nxv4i32.nxv4i16.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vwadd_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -1580,7 +1580,7 @@ vint32m2_t test_vwadd_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwadd.mask.nxv4i32.nxv4i16.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwadd.mask.nxv4i32.nxv4i16.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vwadd_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -1591,7 +1591,7 @@ vint32m2_t test_vwadd_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwadd.w.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwadd.w.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vwadd_wv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -1602,7 +1602,7 @@ vint32m2_t test_vwadd_wv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwadd.w.mask.nxv4i32.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwadd.w.mask.nxv4i32.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vwadd_wx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -1613,7 +1613,7 @@ vint32m2_t test_vwadd_wx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwadd.mask.nxv8i32.nxv8i16.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwadd.mask.nxv8i32.nxv8i16.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vwadd_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -1624,7 +1624,7 @@ vint32m4_t test_vwadd_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwadd.mask.nxv8i32.nxv8i16.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwadd.mask.nxv8i32.nxv8i16.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vwadd_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -1635,7 +1635,7 @@ vint32m4_t test_vwadd_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwadd.w.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwadd.w.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vwadd_wv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -1646,7 +1646,7 @@ vint32m4_t test_vwadd_wv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwadd.w.mask.nxv8i32.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwadd.w.mask.nxv8i32.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vwadd_wx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -1657,7 +1657,7 @@ vint32m4_t test_vwadd_wx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwadd.mask.nxv16i32.nxv16i16.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwadd.mask.nxv16i32.nxv16i16.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vwadd_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -1668,7 +1668,7 @@ vint32m8_t test_vwadd_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwadd.mask.nxv16i32.nxv16i16.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwadd.mask.nxv16i32.nxv16i16.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vwadd_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -1679,7 +1679,7 @@ vint32m8_t test_vwadd_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwadd.w.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwadd.w.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vwadd_wv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -1690,7 +1690,7 @@ vint32m8_t test_vwadd_wv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwadd.w.mask.nxv16i32.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwadd.w.mask.nxv16i32.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vwadd_wx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -1701,7 +1701,7 @@ vint32m8_t test_vwadd_wx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwadd.mask.nxv1i64.nxv1i32.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwadd.mask.nxv1i64.nxv1i32.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwadd_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -1712,7 +1712,7 @@ vint64m1_t test_vwadd_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwadd.mask.nxv1i64.nxv1i32.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwadd.mask.nxv1i64.nxv1i32.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwadd_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -1723,7 +1723,7 @@ vint64m1_t test_vwadd_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwadd.w.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwadd.w.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwadd_wv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -1734,7 +1734,7 @@ vint64m1_t test_vwadd_wv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwadd.w.mask.nxv1i64.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwadd.w.mask.nxv1i64.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwadd_wx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -1745,7 +1745,7 @@ vint64m1_t test_vwadd_wx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwadd.mask.nxv2i64.nxv2i32.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwadd.mask.nxv2i64.nxv2i32.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vwadd_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -1756,7 +1756,7 @@ vint64m2_t test_vwadd_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwadd.mask.nxv2i64.nxv2i32.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwadd.mask.nxv2i64.nxv2i32.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vwadd_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -1767,7 +1767,7 @@ vint64m2_t test_vwadd_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwadd.w.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwadd.w.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vwadd_wv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -1778,7 +1778,7 @@ vint64m2_t test_vwadd_wv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwadd.w.mask.nxv2i64.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwadd.w.mask.nxv2i64.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vwadd_wx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -1789,7 +1789,7 @@ vint64m2_t test_vwadd_wx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwadd.mask.nxv4i64.nxv4i32.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwadd.mask.nxv4i64.nxv4i32.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vwadd_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -1800,7 +1800,7 @@ vint64m4_t test_vwadd_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwadd.mask.nxv4i64.nxv4i32.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwadd.mask.nxv4i64.nxv4i32.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vwadd_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -1811,7 +1811,7 @@ vint64m4_t test_vwadd_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwadd.w.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwadd.w.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vwadd_wv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -1822,7 +1822,7 @@ vint64m4_t test_vwadd_wv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwadd.w.mask.nxv4i64.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwadd.w.mask.nxv4i64.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vwadd_wx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -1833,7 +1833,7 @@ vint64m4_t test_vwadd_wx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwadd.mask.nxv8i64.nxv8i32.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwadd.mask.nxv8i64.nxv8i32.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vwadd_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -1844,7 +1844,7 @@ vint64m8_t test_vwadd_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwadd.mask.nxv8i64.nxv8i32.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwadd.mask.nxv8i64.nxv8i32.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vwadd_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -1855,7 +1855,7 @@ vint64m8_t test_vwadd_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwadd.w.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwadd.w.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vwadd_wv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -1866,7 +1866,7 @@ vint64m8_t test_vwadd_wv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwadd_wx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwadd.w.mask.nxv8i64.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwadd.w.mask.nxv8i64.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vwadd_wx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -1877,7 +1877,7 @@ vint64m8_t test_vwadd_wx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwaddu.mask.nxv1i16.nxv1i8.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwaddu.mask.nxv1i16.nxv1i8.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vwaddu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -1889,7 +1889,7 @@ vuint16mf4_t test_vwaddu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwaddu.mask.nxv1i16.nxv1i8.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwaddu.mask.nxv1i16.nxv1i8.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vwaddu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -1900,7 +1900,7 @@ vuint16mf4_t test_vwaddu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwaddu.w.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwaddu.w.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vwaddu_wv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -1912,7 +1912,7 @@ vuint16mf4_t test_vwaddu_wv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwaddu.w.mask.nxv1i16.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwaddu.w.mask.nxv1i16.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vwaddu_wx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -1923,7 +1923,7 @@ vuint16mf4_t test_vwaddu_wx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwaddu.mask.nxv2i16.nxv2i8.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwaddu.mask.nxv2i16.nxv2i8.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vwaddu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -1935,7 +1935,7 @@ vuint16mf2_t test_vwaddu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwaddu.mask.nxv2i16.nxv2i8.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwaddu.mask.nxv2i16.nxv2i8.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vwaddu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -1946,7 +1946,7 @@ vuint16mf2_t test_vwaddu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwaddu.w.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwaddu.w.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vwaddu_wv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -1958,7 +1958,7 @@ vuint16mf2_t test_vwaddu_wv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwaddu.w.mask.nxv2i16.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwaddu.w.mask.nxv2i16.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vwaddu_wx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -1969,7 +1969,7 @@ vuint16mf2_t test_vwaddu_wx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwaddu.mask.nxv4i16.nxv4i8.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwaddu.mask.nxv4i16.nxv4i8.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vwaddu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -1981,7 +1981,7 @@ vuint16m1_t test_vwaddu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwaddu.mask.nxv4i16.nxv4i8.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwaddu.mask.nxv4i16.nxv4i8.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vwaddu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -1992,7 +1992,7 @@ vuint16m1_t test_vwaddu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwaddu.w.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwaddu.w.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vwaddu_wv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -2004,7 +2004,7 @@ vuint16m1_t test_vwaddu_wv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwaddu.w.mask.nxv4i16.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwaddu.w.mask.nxv4i16.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vwaddu_wx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -2015,7 +2015,7 @@ vuint16m1_t test_vwaddu_wx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwaddu.mask.nxv8i16.nxv8i8.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwaddu.mask.nxv8i16.nxv8i8.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vwaddu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -2026,7 +2026,7 @@ vuint16m2_t test_vwaddu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwaddu.mask.nxv8i16.nxv8i8.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwaddu.mask.nxv8i16.nxv8i8.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vwaddu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -2037,7 +2037,7 @@ vuint16m2_t test_vwaddu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwaddu.w.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwaddu.w.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vwaddu_wv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -2048,7 +2048,7 @@ vuint16m2_t test_vwaddu_wv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwaddu.w.mask.nxv8i16.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwaddu.w.mask.nxv8i16.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vwaddu_wx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -2059,7 +2059,7 @@ vuint16m2_t test_vwaddu_wx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwaddu.mask.nxv16i16.nxv16i8.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwaddu.mask.nxv16i16.nxv16i8.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vwaddu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -2070,7 +2070,7 @@ vuint16m4_t test_vwaddu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwaddu.mask.nxv16i16.nxv16i8.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwaddu.mask.nxv16i16.nxv16i8.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vwaddu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -2081,7 +2081,7 @@ vuint16m4_t test_vwaddu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwaddu.w.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwaddu.w.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vwaddu_wv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -2092,7 +2092,7 @@ vuint16m4_t test_vwaddu_wv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwaddu.w.mask.nxv16i16.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwaddu.w.mask.nxv16i16.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vwaddu_wx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -2103,7 +2103,7 @@ vuint16m4_t test_vwaddu_wx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwaddu.mask.nxv32i16.nxv32i8.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwaddu.mask.nxv32i16.nxv32i8.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vwaddu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -2114,7 +2114,7 @@ vuint16m8_t test_vwaddu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwaddu.mask.nxv32i16.nxv32i8.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwaddu.mask.nxv32i16.nxv32i8.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vwaddu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -2125,7 +2125,7 @@ vuint16m8_t test_vwaddu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwaddu.w.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwaddu.w.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vwaddu_wv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -2136,7 +2136,7 @@ vuint16m8_t test_vwaddu_wv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwaddu.w.mask.nxv32i16.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwaddu.w.mask.nxv32i16.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vwaddu_wx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -2147,7 +2147,7 @@ vuint16m8_t test_vwaddu_wx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwaddu.mask.nxv1i32.nxv1i16.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwaddu.mask.nxv1i32.nxv1i16.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vwaddu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -2159,7 +2159,7 @@ vuint32mf2_t test_vwaddu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwaddu.mask.nxv1i32.nxv1i16.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwaddu.mask.nxv1i32.nxv1i16.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vwaddu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -2171,7 +2171,7 @@ vuint32mf2_t test_vwaddu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwaddu.w.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwaddu.w.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vwaddu_wv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -2183,7 +2183,7 @@ vuint32mf2_t test_vwaddu_wv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwaddu.w.mask.nxv1i32.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwaddu.w.mask.nxv1i32.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vwaddu_wx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -2195,7 +2195,7 @@ vuint32mf2_t test_vwaddu_wx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwaddu.mask.nxv2i32.nxv2i16.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwaddu.mask.nxv2i32.nxv2i16.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vwaddu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -2207,7 +2207,7 @@ vuint32m1_t test_vwaddu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwaddu.mask.nxv2i32.nxv2i16.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwaddu.mask.nxv2i32.nxv2i16.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vwaddu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -2218,7 +2218,7 @@ vuint32m1_t test_vwaddu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwaddu.w.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwaddu.w.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vwaddu_wv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -2230,7 +2230,7 @@ vuint32m1_t test_vwaddu_wv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwaddu.w.mask.nxv2i32.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwaddu.w.mask.nxv2i32.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vwaddu_wx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -2241,7 +2241,7 @@ vuint32m1_t test_vwaddu_wx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwaddu.mask.nxv4i32.nxv4i16.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwaddu.mask.nxv4i32.nxv4i16.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vwaddu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -2253,7 +2253,7 @@ vuint32m2_t test_vwaddu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwaddu.mask.nxv4i32.nxv4i16.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwaddu.mask.nxv4i32.nxv4i16.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vwaddu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -2264,7 +2264,7 @@ vuint32m2_t test_vwaddu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwaddu.w.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwaddu.w.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vwaddu_wv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -2276,7 +2276,7 @@ vuint32m2_t test_vwaddu_wv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwaddu.w.mask.nxv4i32.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwaddu.w.mask.nxv4i32.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vwaddu_wx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -2287,7 +2287,7 @@ vuint32m2_t test_vwaddu_wx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwaddu.mask.nxv8i32.nxv8i16.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwaddu.mask.nxv8i32.nxv8i16.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vwaddu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -2299,7 +2299,7 @@ vuint32m4_t test_vwaddu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwaddu.mask.nxv8i32.nxv8i16.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwaddu.mask.nxv8i32.nxv8i16.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vwaddu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -2310,7 +2310,7 @@ vuint32m4_t test_vwaddu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwaddu.w.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwaddu.w.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vwaddu_wv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -2322,7 +2322,7 @@ vuint32m4_t test_vwaddu_wv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwaddu.w.mask.nxv8i32.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwaddu.w.mask.nxv8i32.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vwaddu_wx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -2333,7 +2333,7 @@ vuint32m4_t test_vwaddu_wx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwaddu.mask.nxv16i32.nxv16i16.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwaddu.mask.nxv16i32.nxv16i16.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vwaddu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -2345,7 +2345,7 @@ vuint32m8_t test_vwaddu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwaddu.mask.nxv16i32.nxv16i16.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwaddu.mask.nxv16i32.nxv16i16.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vwaddu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -2356,7 +2356,7 @@ vuint32m8_t test_vwaddu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwaddu.w.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwaddu.w.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vwaddu_wv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -2368,7 +2368,7 @@ vuint32m8_t test_vwaddu_wv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwaddu.w.mask.nxv16i32.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwaddu.w.mask.nxv16i32.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vwaddu_wx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -2379,7 +2379,7 @@ vuint32m8_t test_vwaddu_wx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwaddu.mask.nxv1i64.nxv1i32.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwaddu.mask.nxv1i64.nxv1i32.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vwaddu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -2391,7 +2391,7 @@ vuint64m1_t test_vwaddu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwaddu.mask.nxv1i64.nxv1i32.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwaddu.mask.nxv1i64.nxv1i32.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vwaddu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -2402,7 +2402,7 @@ vuint64m1_t test_vwaddu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwaddu.w.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwaddu.w.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vwaddu_wv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -2414,7 +2414,7 @@ vuint64m1_t test_vwaddu_wv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwaddu.w.mask.nxv1i64.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwaddu.w.mask.nxv1i64.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vwaddu_wx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -2425,7 +2425,7 @@ vuint64m1_t test_vwaddu_wx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwaddu.mask.nxv2i64.nxv2i32.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwaddu.mask.nxv2i64.nxv2i32.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vwaddu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -2437,7 +2437,7 @@ vuint64m2_t test_vwaddu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwaddu.mask.nxv2i64.nxv2i32.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwaddu.mask.nxv2i64.nxv2i32.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vwaddu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -2448,7 +2448,7 @@ vuint64m2_t test_vwaddu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwaddu.w.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwaddu.w.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vwaddu_wv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -2460,7 +2460,7 @@ vuint64m2_t test_vwaddu_wv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwaddu.w.mask.nxv2i64.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwaddu.w.mask.nxv2i64.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vwaddu_wx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -2471,7 +2471,7 @@ vuint64m2_t test_vwaddu_wx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwaddu.mask.nxv4i64.nxv4i32.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwaddu.mask.nxv4i64.nxv4i32.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vwaddu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -2483,7 +2483,7 @@ vuint64m4_t test_vwaddu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwaddu.mask.nxv4i64.nxv4i32.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwaddu.mask.nxv4i64.nxv4i32.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vwaddu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -2494,7 +2494,7 @@ vuint64m4_t test_vwaddu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwaddu.w.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwaddu.w.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vwaddu_wv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -2506,7 +2506,7 @@ vuint64m4_t test_vwaddu_wv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwaddu.w.mask.nxv4i64.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwaddu.w.mask.nxv4i64.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vwaddu_wx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -2517,7 +2517,7 @@ vuint64m4_t test_vwaddu_wx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwaddu.mask.nxv8i64.nxv8i32.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwaddu.mask.nxv8i64.nxv8i32.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vwaddu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
@@ -2529,7 +2529,7 @@ vuint64m8_t test_vwaddu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwaddu.mask.nxv8i64.nxv8i32.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwaddu.mask.nxv8i64.nxv8i32.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vwaddu_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
@@ -2540,7 +2540,7 @@ vuint64m8_t test_vwaddu_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwaddu.w.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwaddu.w.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vwaddu_wv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
@@ -2552,7 +2552,7 @@ vuint64m8_t test_vwaddu_wv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwaddu_wx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwaddu.w.mask.nxv8i64.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwaddu.w.mask.nxv8i64.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vwaddu_wx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwcvt.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwcvt.c
index 189f0fb..1166db5 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwcvt.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwcvt.c
@@ -307,7 +307,7 @@ vuint64m8_t test_vwcvtu_x_x_v_u64m8 (vuint32m4_t src, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vwcvt_x_x_v_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwadd.mask.nxv1i16.nxv1i8.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[SRC:%.*]], i8 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwadd.mask.nxv1i16.nxv1i8.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[SRC:%.*]], i8 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vwcvt_x_x_v_i16mf4_m (vbool64_t mask, vint16mf4_t maskedoff, vint8mf8_t src, size_t vl) {
@@ -317,7 +317,7 @@ vint16mf4_t test_vwcvt_x_x_v_i16mf4_m (vbool64_t mask, vint16mf4_t maskedoff, vi
 //
 // CHECK-RV64-LABEL: @test_vwcvt_x_x_v_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwadd.mask.nxv2i16.nxv2i8.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[SRC:%.*]], i8 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwadd.mask.nxv2i16.nxv2i8.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[SRC:%.*]], i8 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vwcvt_x_x_v_i16mf2_m (vbool32_t mask, vint16mf2_t maskedoff, vint8mf4_t src, size_t vl) {
@@ -327,7 +327,7 @@ vint16mf2_t test_vwcvt_x_x_v_i16mf2_m (vbool32_t mask, vint16mf2_t maskedoff, vi
 //
 // CHECK-RV64-LABEL: @test_vwcvt_x_x_v_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwadd.mask.nxv4i16.nxv4i8.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[SRC:%.*]], i8 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwadd.mask.nxv4i16.nxv4i8.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[SRC:%.*]], i8 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwcvt_x_x_v_i16m1_m (vbool16_t mask, vint16m1_t maskedoff, vint8mf2_t src, size_t vl) {
@@ -337,7 +337,7 @@ vint16m1_t test_vwcvt_x_x_v_i16m1_m (vbool16_t mask, vint16m1_t maskedoff, vint8
 //
 // CHECK-RV64-LABEL: @test_vwcvt_x_x_v_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwadd.mask.nxv8i16.nxv8i8.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[SRC:%.*]], i8 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwadd.mask.nxv8i16.nxv8i8.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[SRC:%.*]], i8 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vwcvt_x_x_v_i16m2_m (vbool8_t mask, vint16m2_t maskedoff, vint8m1_t src, size_t vl) {
@@ -347,7 +347,7 @@ vint16m2_t test_vwcvt_x_x_v_i16m2_m (vbool8_t mask, vint16m2_t maskedoff, vint8m
 //
 // CHECK-RV64-LABEL: @test_vwcvt_x_x_v_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwadd.mask.nxv16i16.nxv16i8.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[SRC:%.*]], i8 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwadd.mask.nxv16i16.nxv16i8.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[SRC:%.*]], i8 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vwcvt_x_x_v_i16m4_m (vbool4_t mask, vint16m4_t maskedoff, vint8m2_t src, size_t vl) {
@@ -357,7 +357,7 @@ vint16m4_t test_vwcvt_x_x_v_i16m4_m (vbool4_t mask, vint16m4_t maskedoff, vint8m
 //
 // CHECK-RV64-LABEL: @test_vwcvt_x_x_v_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwadd.mask.nxv32i16.nxv32i8.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[SRC:%.*]], i8 0, <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwadd.mask.nxv32i16.nxv32i8.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[SRC:%.*]], i8 0, <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vwcvt_x_x_v_i16m8_m (vbool2_t mask, vint16m8_t maskedoff, vint8m4_t src, size_t vl) {
@@ -367,7 +367,7 @@ vint16m8_t test_vwcvt_x_x_v_i16m8_m (vbool2_t mask, vint16m8_t maskedoff, vint8m
 //
 // CHECK-RV64-LABEL: @test_vwcvtu_x_x_v_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwaddu.mask.nxv1i16.nxv1i8.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[SRC:%.*]], i8 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwaddu.mask.nxv1i16.nxv1i8.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[SRC:%.*]], i8 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vwcvtu_x_x_v_u16mf4_m (vbool64_t mask, vuint16mf4_t maskedoff, vuint8mf8_t src, size_t vl) {
@@ -377,7 +377,7 @@ vuint16mf4_t test_vwcvtu_x_x_v_u16mf4_m (vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwcvtu_x_x_v_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwaddu.mask.nxv2i16.nxv2i8.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[SRC:%.*]], i8 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwaddu.mask.nxv2i16.nxv2i8.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[SRC:%.*]], i8 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vwcvtu_x_x_v_u16mf2_m (vbool32_t mask, vuint16mf2_t maskedoff, vuint8mf4_t src, size_t vl) {
@@ -387,7 +387,7 @@ vuint16mf2_t test_vwcvtu_x_x_v_u16mf2_m (vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwcvtu_x_x_v_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwaddu.mask.nxv4i16.nxv4i8.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[SRC:%.*]], i8 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwaddu.mask.nxv4i16.nxv4i8.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[SRC:%.*]], i8 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vwcvtu_x_x_v_u16m1_m (vbool16_t mask, vuint16m1_t maskedoff, vuint8mf2_t src, size_t vl) {
@@ -397,7 +397,7 @@ vuint16m1_t test_vwcvtu_x_x_v_u16m1_m (vbool16_t mask, vuint16m1_t maskedoff, vu
 //
 // CHECK-RV64-LABEL: @test_vwcvtu_x_x_v_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwaddu.mask.nxv8i16.nxv8i8.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[SRC:%.*]], i8 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwaddu.mask.nxv8i16.nxv8i8.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[SRC:%.*]], i8 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vwcvtu_x_x_v_u16m2_m (vbool8_t mask, vuint16m2_t maskedoff, vuint8m1_t src, size_t vl) {
@@ -407,7 +407,7 @@ vuint16m2_t test_vwcvtu_x_x_v_u16m2_m (vbool8_t mask, vuint16m2_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vwcvtu_x_x_v_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwaddu.mask.nxv16i16.nxv16i8.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[SRC:%.*]], i8 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwaddu.mask.nxv16i16.nxv16i8.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[SRC:%.*]], i8 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vwcvtu_x_x_v_u16m4_m (vbool4_t mask, vuint16m4_t maskedoff, vuint8m2_t src, size_t vl) {
@@ -417,7 +417,7 @@ vuint16m4_t test_vwcvtu_x_x_v_u16m4_m (vbool4_t mask, vuint16m4_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vwcvtu_x_x_v_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwaddu.mask.nxv32i16.nxv32i8.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[SRC:%.*]], i8 0, <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwaddu.mask.nxv32i16.nxv32i8.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[SRC:%.*]], i8 0, <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vwcvtu_x_x_v_u16m8_m (vbool2_t mask, vuint16m8_t maskedoff, vuint8m4_t src, size_t vl) {
@@ -427,7 +427,7 @@ vuint16m8_t test_vwcvtu_x_x_v_u16m8_m (vbool2_t mask, vuint16m8_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vwcvt_x_x_v_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwadd.mask.nxv1i32.nxv1i16.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i16 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwadd.mask.nxv1i32.nxv1i16.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i16 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vwcvt_x_x_v_i32mf2_m (vbool64_t mask, vint32mf2_t maskedoff, vint16mf4_t src, size_t vl) {
@@ -437,7 +437,7 @@ vint32mf2_t test_vwcvt_x_x_v_i32mf2_m (vbool64_t mask, vint32mf2_t maskedoff, vi
 //
 // CHECK-RV64-LABEL: @test_vwcvt_x_x_v_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwadd.mask.nxv2i32.nxv2i16.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i16 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwadd.mask.nxv2i32.nxv2i16.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i16 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwcvt_x_x_v_i32m1_m (vbool32_t mask, vint32m1_t maskedoff, vint16mf2_t src, size_t vl) {
@@ -447,7 +447,7 @@ vint32m1_t test_vwcvt_x_x_v_i32m1_m (vbool32_t mask, vint32m1_t maskedoff, vint1
 //
 // CHECK-RV64-LABEL: @test_vwcvt_x_x_v_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwadd.mask.nxv4i32.nxv4i16.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i16 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwadd.mask.nxv4i32.nxv4i16.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i16 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vwcvt_x_x_v_i32m2_m (vbool16_t mask, vint32m2_t maskedoff, vint16m1_t src, size_t vl) {
@@ -457,7 +457,7 @@ vint32m2_t test_vwcvt_x_x_v_i32m2_m (vbool16_t mask, vint32m2_t maskedoff, vint1
 //
 // CHECK-RV64-LABEL: @test_vwcvt_x_x_v_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwadd.mask.nxv8i32.nxv8i16.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i16 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwadd.mask.nxv8i32.nxv8i16.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i16 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vwcvt_x_x_v_i32m4_m (vbool8_t mask, vint32m4_t maskedoff, vint16m2_t src, size_t vl) {
@@ -467,7 +467,7 @@ vint32m4_t test_vwcvt_x_x_v_i32m4_m (vbool8_t mask, vint32m4_t maskedoff, vint16
 //
 // CHECK-RV64-LABEL: @test_vwcvt_x_x_v_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwadd.mask.nxv16i32.nxv16i16.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i16 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwadd.mask.nxv16i32.nxv16i16.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i16 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vwcvt_x_x_v_i32m8_m (vbool4_t mask, vint32m8_t maskedoff, vint16m4_t src, size_t vl) {
@@ -477,7 +477,7 @@ vint32m8_t test_vwcvt_x_x_v_i32m8_m (vbool4_t mask, vint32m8_t maskedoff, vint16
 //
 // CHECK-RV64-LABEL: @test_vwcvtu_x_x_v_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwaddu.mask.nxv1i32.nxv1i16.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i16 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwaddu.mask.nxv1i32.nxv1i16.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[SRC:%.*]], i16 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vwcvtu_x_x_v_u32mf2_m (vbool64_t mask, vuint32mf2_t maskedoff, vuint16mf4_t src, size_t vl) {
@@ -487,7 +487,7 @@ vuint32mf2_t test_vwcvtu_x_x_v_u32mf2_m (vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwcvtu_x_x_v_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwaddu.mask.nxv2i32.nxv2i16.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i16 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwaddu.mask.nxv2i32.nxv2i16.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[SRC:%.*]], i16 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vwcvtu_x_x_v_u32m1_m (vbool32_t mask, vuint32m1_t maskedoff, vuint16mf2_t src, size_t vl) {
@@ -497,7 +497,7 @@ vuint32m1_t test_vwcvtu_x_x_v_u32m1_m (vbool32_t mask, vuint32m1_t maskedoff, vu
 //
 // CHECK-RV64-LABEL: @test_vwcvtu_x_x_v_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwaddu.mask.nxv4i32.nxv4i16.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i16 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwaddu.mask.nxv4i32.nxv4i16.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[SRC:%.*]], i16 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vwcvtu_x_x_v_u32m2_m (vbool16_t mask, vuint32m2_t maskedoff, vuint16m1_t src, size_t vl) {
@@ -507,7 +507,7 @@ vuint32m2_t test_vwcvtu_x_x_v_u32m2_m (vbool16_t mask, vuint32m2_t maskedoff, vu
 //
 // CHECK-RV64-LABEL: @test_vwcvtu_x_x_v_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwaddu.mask.nxv8i32.nxv8i16.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i16 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwaddu.mask.nxv8i32.nxv8i16.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[SRC:%.*]], i16 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vwcvtu_x_x_v_u32m4_m (vbool8_t mask, vuint32m4_t maskedoff, vuint16m2_t src, size_t vl) {
@@ -517,7 +517,7 @@ vuint32m4_t test_vwcvtu_x_x_v_u32m4_m (vbool8_t mask, vuint32m4_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vwcvtu_x_x_v_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwaddu.mask.nxv16i32.nxv16i16.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i16 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwaddu.mask.nxv16i32.nxv16i16.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[SRC:%.*]], i16 0, <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vwcvtu_x_x_v_u32m8_m (vbool4_t mask, vuint32m8_t maskedoff, vuint16m4_t src, size_t vl) {
@@ -527,7 +527,7 @@ vuint32m8_t test_vwcvtu_x_x_v_u32m8_m (vbool4_t mask, vuint32m8_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vwcvt_x_x_v_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwadd.mask.nxv1i64.nxv1i32.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i32 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwadd.mask.nxv1i64.nxv1i32.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i32 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwcvt_x_x_v_i64m1_m (vbool64_t mask, vint64m1_t maskedoff, vint32mf2_t src, size_t vl) {
@@ -537,7 +537,7 @@ vint64m1_t test_vwcvt_x_x_v_i64m1_m (vbool64_t mask, vint64m1_t maskedoff, vint3
 //
 // CHECK-RV64-LABEL: @test_vwcvt_x_x_v_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwadd.mask.nxv2i64.nxv2i32.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i32 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwadd.mask.nxv2i64.nxv2i32.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i32 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vwcvt_x_x_v_i64m2_m (vbool32_t mask, vint64m2_t maskedoff, vint32m1_t src, size_t vl) {
@@ -547,7 +547,7 @@ vint64m2_t test_vwcvt_x_x_v_i64m2_m (vbool32_t mask, vint64m2_t maskedoff, vint3
 //
 // CHECK-RV64-LABEL: @test_vwcvt_x_x_v_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwadd.mask.nxv4i64.nxv4i32.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i32 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwadd.mask.nxv4i64.nxv4i32.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i32 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vwcvt_x_x_v_i64m4_m (vbool16_t mask, vint64m4_t maskedoff, vint32m2_t src, size_t vl) {
@@ -557,7 +557,7 @@ vint64m4_t test_vwcvt_x_x_v_i64m4_m (vbool16_t mask, vint64m4_t maskedoff, vint3
 //
 // CHECK-RV64-LABEL: @test_vwcvt_x_x_v_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwadd.mask.nxv8i64.nxv8i32.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i32 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwadd.mask.nxv8i64.nxv8i32.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i32 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vwcvt_x_x_v_i64m8_m (vbool8_t mask, vint64m8_t maskedoff, vint32m4_t src, size_t vl) {
@@ -567,7 +567,7 @@ vint64m8_t test_vwcvt_x_x_v_i64m8_m (vbool8_t mask, vint64m8_t maskedoff, vint32
 //
 // CHECK-RV64-LABEL: @test_vwcvtu_x_x_v_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwaddu.mask.nxv1i64.nxv1i32.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i32 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwaddu.mask.nxv1i64.nxv1i32.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[SRC:%.*]], i32 0, <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vwcvtu_x_x_v_u64m1_m (vbool64_t mask, vuint64m1_t maskedoff, vuint32mf2_t src, size_t vl) {
@@ -577,7 +577,7 @@ vuint64m1_t test_vwcvtu_x_x_v_u64m1_m (vbool64_t mask, vuint64m1_t maskedoff, vu
 //
 // CHECK-RV64-LABEL: @test_vwcvtu_x_x_v_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwaddu.mask.nxv2i64.nxv2i32.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i32 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwaddu.mask.nxv2i64.nxv2i32.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[SRC:%.*]], i32 0, <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vwcvtu_x_x_v_u64m2_m (vbool32_t mask, vuint64m2_t maskedoff, vuint32m1_t src, size_t vl) {
@@ -587,7 +587,7 @@ vuint64m2_t test_vwcvtu_x_x_v_u64m2_m (vbool32_t mask, vuint64m2_t maskedoff, vu
 //
 // CHECK-RV64-LABEL: @test_vwcvtu_x_x_v_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwaddu.mask.nxv4i64.nxv4i32.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i32 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwaddu.mask.nxv4i64.nxv4i32.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[SRC:%.*]], i32 0, <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vwcvtu_x_x_v_u64m4_m (vbool16_t mask, vuint64m4_t maskedoff, vuint32m2_t src, size_t vl) {
@@ -597,7 +597,7 @@ vuint64m4_t test_vwcvtu_x_x_v_u64m4_m (vbool16_t mask, vuint64m4_t maskedoff, vu
 //
 // CHECK-RV64-LABEL: @test_vwcvtu_x_x_v_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwaddu.mask.nxv8i64.nxv8i32.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i32 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwaddu.mask.nxv8i64.nxv8i32.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[SRC:%.*]], i32 0, <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vwcvtu_x_x_v_u64m8_m (vbool8_t mask, vuint64m8_t maskedoff, vuint32m4_t src, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwmacc.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwmacc.c
index 20c64de..ebf63d8 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwmacc.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwmacc.c
@@ -1162,7 +1162,7 @@ vint64m8_t test_vwmaccus_vx_i64m8(vint64m8_t acc, uint32_t op1, vint32m4_t op2,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmacc.mask.nxv1i16.nxv1i8.nxv1i8.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmacc.mask.nxv1i16.nxv1i8.nxv1i8.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vwmacc_vv_i16mf4_m(vbool64_t mask, vint16mf4_t acc,
@@ -1173,7 +1173,7 @@ vint16mf4_t test_vwmacc_vv_i16mf4_m(vbool64_t mask, vint16mf4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmacc.mask.nxv1i16.i8.nxv1i8.i64(<vscale x 1 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmacc.mask.nxv1i16.i8.nxv1i8.i64(<vscale x 1 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vwmacc_vx_i16mf4_m(vbool64_t mask, vint16mf4_t acc, int8_t op1,
@@ -1184,7 +1184,7 @@ vint16mf4_t test_vwmacc_vx_i16mf4_m(vbool64_t mask, vint16mf4_t acc, int8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmacc.mask.nxv2i16.nxv2i8.nxv2i8.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmacc.mask.nxv2i16.nxv2i8.nxv2i8.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vwmacc_vv_i16mf2_m(vbool32_t mask, vint16mf2_t acc,
@@ -1195,7 +1195,7 @@ vint16mf2_t test_vwmacc_vv_i16mf2_m(vbool32_t mask, vint16mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmacc.mask.nxv2i16.i8.nxv2i8.i64(<vscale x 2 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmacc.mask.nxv2i16.i8.nxv2i8.i64(<vscale x 2 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vwmacc_vx_i16mf2_m(vbool32_t mask, vint16mf2_t acc, int8_t op1,
@@ -1206,7 +1206,7 @@ vint16mf2_t test_vwmacc_vx_i16mf2_m(vbool32_t mask, vint16mf2_t acc, int8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmacc.mask.nxv4i16.nxv4i8.nxv4i8.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmacc.mask.nxv4i16.nxv4i8.nxv4i8.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwmacc_vv_i16m1_m(vbool16_t mask, vint16m1_t acc,
@@ -1217,7 +1217,7 @@ vint16m1_t test_vwmacc_vv_i16m1_m(vbool16_t mask, vint16m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmacc.mask.nxv4i16.i8.nxv4i8.i64(<vscale x 4 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmacc.mask.nxv4i16.i8.nxv4i8.i64(<vscale x 4 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwmacc_vx_i16m1_m(vbool16_t mask, vint16m1_t acc, int8_t op1,
@@ -1228,7 +1228,7 @@ vint16m1_t test_vwmacc_vx_i16m1_m(vbool16_t mask, vint16m1_t acc, int8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmacc.mask.nxv8i16.nxv8i8.nxv8i8.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmacc.mask.nxv8i16.nxv8i8.nxv8i8.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vwmacc_vv_i16m2_m(vbool8_t mask, vint16m2_t acc, vint8m1_t op1,
@@ -1239,7 +1239,7 @@ vint16m2_t test_vwmacc_vv_i16m2_m(vbool8_t mask, vint16m2_t acc, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmacc.mask.nxv8i16.i8.nxv8i8.i64(<vscale x 8 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmacc.mask.nxv8i16.i8.nxv8i8.i64(<vscale x 8 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vwmacc_vx_i16m2_m(vbool8_t mask, vint16m2_t acc, int8_t op1,
@@ -1250,7 +1250,7 @@ vint16m2_t test_vwmacc_vx_i16m2_m(vbool8_t mask, vint16m2_t acc, int8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmacc.mask.nxv16i16.nxv16i8.nxv16i8.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmacc.mask.nxv16i16.nxv16i8.nxv16i8.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vwmacc_vv_i16m4_m(vbool4_t mask, vint16m4_t acc, vint8m2_t op1,
@@ -1261,7 +1261,7 @@ vint16m4_t test_vwmacc_vv_i16m4_m(vbool4_t mask, vint16m4_t acc, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmacc.mask.nxv16i16.i8.nxv16i8.i64(<vscale x 16 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmacc.mask.nxv16i16.i8.nxv16i8.i64(<vscale x 16 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vwmacc_vx_i16m4_m(vbool4_t mask, vint16m4_t acc, int8_t op1,
@@ -1272,7 +1272,7 @@ vint16m4_t test_vwmacc_vx_i16m4_m(vbool4_t mask, vint16m4_t acc, int8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmacc.mask.nxv32i16.nxv32i8.nxv32i8.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmacc.mask.nxv32i16.nxv32i8.nxv32i8.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vwmacc_vv_i16m8_m(vbool2_t mask, vint16m8_t acc, vint8m4_t op1,
@@ -1283,7 +1283,7 @@ vint16m8_t test_vwmacc_vv_i16m8_m(vbool2_t mask, vint16m8_t acc, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmacc.mask.nxv32i16.i8.nxv32i8.i64(<vscale x 32 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmacc.mask.nxv32i16.i8.nxv32i8.i64(<vscale x 32 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vwmacc_vx_i16m8_m(vbool2_t mask, vint16m8_t acc, int8_t op1,
@@ -1294,7 +1294,7 @@ vint16m8_t test_vwmacc_vx_i16m8_m(vbool2_t mask, vint16m8_t acc, int8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmacc.mask.nxv1i32.nxv1i16.nxv1i16.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmacc.mask.nxv1i32.nxv1i16.nxv1i16.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vwmacc_vv_i32mf2_m(vbool64_t mask, vint32mf2_t acc,
@@ -1306,7 +1306,7 @@ vint32mf2_t test_vwmacc_vv_i32mf2_m(vbool64_t mask, vint32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmacc.mask.nxv1i32.i16.nxv1i16.i64(<vscale x 1 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmacc.mask.nxv1i32.i16.nxv1i16.i64(<vscale x 1 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vwmacc_vx_i32mf2_m(vbool64_t mask, vint32mf2_t acc,
@@ -1317,7 +1317,7 @@ vint32mf2_t test_vwmacc_vx_i32mf2_m(vbool64_t mask, vint32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmacc.mask.nxv2i32.nxv2i16.nxv2i16.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmacc.mask.nxv2i32.nxv2i16.nxv2i16.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwmacc_vv_i32m1_m(vbool32_t mask, vint32m1_t acc,
@@ -1328,7 +1328,7 @@ vint32m1_t test_vwmacc_vv_i32m1_m(vbool32_t mask, vint32m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmacc.mask.nxv2i32.i16.nxv2i16.i64(<vscale x 2 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmacc.mask.nxv2i32.i16.nxv2i16.i64(<vscale x 2 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwmacc_vx_i32m1_m(vbool32_t mask, vint32m1_t acc, int16_t op1,
@@ -1339,7 +1339,7 @@ vint32m1_t test_vwmacc_vx_i32m1_m(vbool32_t mask, vint32m1_t acc, int16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmacc.mask.nxv4i32.nxv4i16.nxv4i16.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmacc.mask.nxv4i32.nxv4i16.nxv4i16.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vwmacc_vv_i32m2_m(vbool16_t mask, vint32m2_t acc,
@@ -1350,7 +1350,7 @@ vint32m2_t test_vwmacc_vv_i32m2_m(vbool16_t mask, vint32m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmacc.mask.nxv4i32.i16.nxv4i16.i64(<vscale x 4 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmacc.mask.nxv4i32.i16.nxv4i16.i64(<vscale x 4 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vwmacc_vx_i32m2_m(vbool16_t mask, vint32m2_t acc, int16_t op1,
@@ -1361,7 +1361,7 @@ vint32m2_t test_vwmacc_vx_i32m2_m(vbool16_t mask, vint32m2_t acc, int16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmacc.mask.nxv8i32.nxv8i16.nxv8i16.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmacc.mask.nxv8i32.nxv8i16.nxv8i16.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vwmacc_vv_i32m4_m(vbool8_t mask, vint32m4_t acc, vint16m2_t op1,
@@ -1372,7 +1372,7 @@ vint32m4_t test_vwmacc_vv_i32m4_m(vbool8_t mask, vint32m4_t acc, vint16m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmacc.mask.nxv8i32.i16.nxv8i16.i64(<vscale x 8 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmacc.mask.nxv8i32.i16.nxv8i16.i64(<vscale x 8 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vwmacc_vx_i32m4_m(vbool8_t mask, vint32m4_t acc, int16_t op1,
@@ -1383,7 +1383,7 @@ vint32m4_t test_vwmacc_vx_i32m4_m(vbool8_t mask, vint32m4_t acc, int16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmacc.mask.nxv16i32.nxv16i16.nxv16i16.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmacc.mask.nxv16i32.nxv16i16.nxv16i16.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vwmacc_vv_i32m8_m(vbool4_t mask, vint32m8_t acc, vint16m4_t op1,
@@ -1394,7 +1394,7 @@ vint32m8_t test_vwmacc_vv_i32m8_m(vbool4_t mask, vint32m8_t acc, vint16m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmacc.mask.nxv16i32.i16.nxv16i16.i64(<vscale x 16 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmacc.mask.nxv16i32.i16.nxv16i16.i64(<vscale x 16 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vwmacc_vx_i32m8_m(vbool4_t mask, vint32m8_t acc, int16_t op1,
@@ -1405,7 +1405,7 @@ vint32m8_t test_vwmacc_vx_i32m8_m(vbool4_t mask, vint32m8_t acc, int16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmacc.mask.nxv1i64.nxv1i32.nxv1i32.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmacc.mask.nxv1i64.nxv1i32.nxv1i32.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwmacc_vv_i64m1_m(vbool64_t mask, vint64m1_t acc,
@@ -1416,7 +1416,7 @@ vint64m1_t test_vwmacc_vv_i64m1_m(vbool64_t mask, vint64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmacc.mask.nxv1i64.i32.nxv1i32.i64(<vscale x 1 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmacc.mask.nxv1i64.i32.nxv1i32.i64(<vscale x 1 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwmacc_vx_i64m1_m(vbool64_t mask, vint64m1_t acc, int32_t op1,
@@ -1427,7 +1427,7 @@ vint64m1_t test_vwmacc_vx_i64m1_m(vbool64_t mask, vint64m1_t acc, int32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmacc.mask.nxv2i64.nxv2i32.nxv2i32.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmacc.mask.nxv2i64.nxv2i32.nxv2i32.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vwmacc_vv_i64m2_m(vbool32_t mask, vint64m2_t acc,
@@ -1438,7 +1438,7 @@ vint64m2_t test_vwmacc_vv_i64m2_m(vbool32_t mask, vint64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmacc.mask.nxv2i64.i32.nxv2i32.i64(<vscale x 2 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmacc.mask.nxv2i64.i32.nxv2i32.i64(<vscale x 2 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vwmacc_vx_i64m2_m(vbool32_t mask, vint64m2_t acc, int32_t op1,
@@ -1449,7 +1449,7 @@ vint64m2_t test_vwmacc_vx_i64m2_m(vbool32_t mask, vint64m2_t acc, int32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmacc.mask.nxv4i64.nxv4i32.nxv4i32.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmacc.mask.nxv4i64.nxv4i32.nxv4i32.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vwmacc_vv_i64m4_m(vbool16_t mask, vint64m4_t acc,
@@ -1460,7 +1460,7 @@ vint64m4_t test_vwmacc_vv_i64m4_m(vbool16_t mask, vint64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmacc.mask.nxv4i64.i32.nxv4i32.i64(<vscale x 4 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmacc.mask.nxv4i64.i32.nxv4i32.i64(<vscale x 4 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vwmacc_vx_i64m4_m(vbool16_t mask, vint64m4_t acc, int32_t op1,
@@ -1471,7 +1471,7 @@ vint64m4_t test_vwmacc_vx_i64m4_m(vbool16_t mask, vint64m4_t acc, int32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmacc.mask.nxv8i64.nxv8i32.nxv8i32.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmacc.mask.nxv8i64.nxv8i32.nxv8i32.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vwmacc_vv_i64m8_m(vbool8_t mask, vint64m8_t acc, vint32m4_t op1,
@@ -1482,7 +1482,7 @@ vint64m8_t test_vwmacc_vv_i64m8_m(vbool8_t mask, vint64m8_t acc, vint32m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmacc_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmacc.mask.nxv8i64.i32.nxv8i32.i64(<vscale x 8 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmacc.mask.nxv8i64.i32.nxv8i32.i64(<vscale x 8 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vwmacc_vx_i64m8_m(vbool8_t mask, vint64m8_t acc, int32_t op1,
@@ -1493,7 +1493,7 @@ vint64m8_t test_vwmacc_vx_i64m8_m(vbool8_t mask, vint64m8_t acc, int32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmaccu.mask.nxv1i16.nxv1i8.nxv1i8.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmaccu.mask.nxv1i16.nxv1i8.nxv1i8.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vwmaccu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t acc,
@@ -1505,7 +1505,7 @@ vuint16mf4_t test_vwmaccu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmaccu.mask.nxv1i16.i8.nxv1i8.i64(<vscale x 1 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmaccu.mask.nxv1i16.i8.nxv1i8.i64(<vscale x 1 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vwmaccu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t acc,
@@ -1516,7 +1516,7 @@ vuint16mf4_t test_vwmaccu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmaccu.mask.nxv2i16.nxv2i8.nxv2i8.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmaccu.mask.nxv2i16.nxv2i8.nxv2i8.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vwmaccu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t acc,
@@ -1528,7 +1528,7 @@ vuint16mf2_t test_vwmaccu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmaccu.mask.nxv2i16.i8.nxv2i8.i64(<vscale x 2 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmaccu.mask.nxv2i16.i8.nxv2i8.i64(<vscale x 2 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vwmaccu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t acc,
@@ -1539,7 +1539,7 @@ vuint16mf2_t test_vwmaccu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmaccu.mask.nxv4i16.nxv4i8.nxv4i8.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmaccu.mask.nxv4i16.nxv4i8.nxv4i8.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vwmaccu_vv_u16m1_m(vbool16_t mask, vuint16m1_t acc,
@@ -1551,7 +1551,7 @@ vuint16m1_t test_vwmaccu_vv_u16m1_m(vbool16_t mask, vuint16m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmaccu.mask.nxv4i16.i8.nxv4i8.i64(<vscale x 4 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmaccu.mask.nxv4i16.i8.nxv4i8.i64(<vscale x 4 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vwmaccu_vx_u16m1_m(vbool16_t mask, vuint16m1_t acc,
@@ -1562,7 +1562,7 @@ vuint16m1_t test_vwmaccu_vx_u16m1_m(vbool16_t mask, vuint16m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmaccu.mask.nxv8i16.nxv8i8.nxv8i8.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmaccu.mask.nxv8i16.nxv8i8.nxv8i8.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vwmaccu_vv_u16m2_m(vbool8_t mask, vuint16m2_t acc,
@@ -1573,7 +1573,7 @@ vuint16m2_t test_vwmaccu_vv_u16m2_m(vbool8_t mask, vuint16m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmaccu.mask.nxv8i16.i8.nxv8i8.i64(<vscale x 8 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmaccu.mask.nxv8i16.i8.nxv8i8.i64(<vscale x 8 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vwmaccu_vx_u16m2_m(vbool8_t mask, vuint16m2_t acc, uint8_t op1,
@@ -1584,7 +1584,7 @@ vuint16m2_t test_vwmaccu_vx_u16m2_m(vbool8_t mask, vuint16m2_t acc, uint8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmaccu.mask.nxv16i16.nxv16i8.nxv16i8.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmaccu.mask.nxv16i16.nxv16i8.nxv16i8.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vwmaccu_vv_u16m4_m(vbool4_t mask, vuint16m4_t acc,
@@ -1595,7 +1595,7 @@ vuint16m4_t test_vwmaccu_vv_u16m4_m(vbool4_t mask, vuint16m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmaccu.mask.nxv16i16.i8.nxv16i8.i64(<vscale x 16 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmaccu.mask.nxv16i16.i8.nxv16i8.i64(<vscale x 16 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vwmaccu_vx_u16m4_m(vbool4_t mask, vuint16m4_t acc, uint8_t op1,
@@ -1606,7 +1606,7 @@ vuint16m4_t test_vwmaccu_vx_u16m4_m(vbool4_t mask, vuint16m4_t acc, uint8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmaccu.mask.nxv32i16.nxv32i8.nxv32i8.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmaccu.mask.nxv32i16.nxv32i8.nxv32i8.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vwmaccu_vv_u16m8_m(vbool2_t mask, vuint16m8_t acc,
@@ -1617,7 +1617,7 @@ vuint16m8_t test_vwmaccu_vv_u16m8_m(vbool2_t mask, vuint16m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmaccu.mask.nxv32i16.i8.nxv32i8.i64(<vscale x 32 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmaccu.mask.nxv32i16.i8.nxv32i8.i64(<vscale x 32 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vwmaccu_vx_u16m8_m(vbool2_t mask, vuint16m8_t acc, uint8_t op1,
@@ -1628,7 +1628,7 @@ vuint16m8_t test_vwmaccu_vx_u16m8_m(vbool2_t mask, vuint16m8_t acc, uint8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmaccu.mask.nxv1i32.nxv1i16.nxv1i16.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmaccu.mask.nxv1i32.nxv1i16.nxv1i16.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vwmaccu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t acc,
@@ -1640,7 +1640,7 @@ vuint32mf2_t test_vwmaccu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmaccu.mask.nxv1i32.i16.nxv1i16.i64(<vscale x 1 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmaccu.mask.nxv1i32.i16.nxv1i16.i64(<vscale x 1 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vwmaccu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t acc,
@@ -1652,7 +1652,7 @@ vuint32mf2_t test_vwmaccu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmaccu.mask.nxv2i32.nxv2i16.nxv2i16.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmaccu.mask.nxv2i32.nxv2i16.nxv2i16.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vwmaccu_vv_u32m1_m(vbool32_t mask, vuint32m1_t acc,
@@ -1664,7 +1664,7 @@ vuint32m1_t test_vwmaccu_vv_u32m1_m(vbool32_t mask, vuint32m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmaccu.mask.nxv2i32.i16.nxv2i16.i64(<vscale x 2 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmaccu.mask.nxv2i32.i16.nxv2i16.i64(<vscale x 2 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vwmaccu_vx_u32m1_m(vbool32_t mask, vuint32m1_t acc,
@@ -1675,7 +1675,7 @@ vuint32m1_t test_vwmaccu_vx_u32m1_m(vbool32_t mask, vuint32m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmaccu.mask.nxv4i32.nxv4i16.nxv4i16.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmaccu.mask.nxv4i32.nxv4i16.nxv4i16.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vwmaccu_vv_u32m2_m(vbool16_t mask, vuint32m2_t acc,
@@ -1687,7 +1687,7 @@ vuint32m2_t test_vwmaccu_vv_u32m2_m(vbool16_t mask, vuint32m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmaccu.mask.nxv4i32.i16.nxv4i16.i64(<vscale x 4 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmaccu.mask.nxv4i32.i16.nxv4i16.i64(<vscale x 4 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vwmaccu_vx_u32m2_m(vbool16_t mask, vuint32m2_t acc,
@@ -1698,7 +1698,7 @@ vuint32m2_t test_vwmaccu_vx_u32m2_m(vbool16_t mask, vuint32m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmaccu.mask.nxv8i32.nxv8i16.nxv8i16.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmaccu.mask.nxv8i32.nxv8i16.nxv8i16.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vwmaccu_vv_u32m4_m(vbool8_t mask, vuint32m4_t acc,
@@ -1710,7 +1710,7 @@ vuint32m4_t test_vwmaccu_vv_u32m4_m(vbool8_t mask, vuint32m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmaccu.mask.nxv8i32.i16.nxv8i16.i64(<vscale x 8 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmaccu.mask.nxv8i32.i16.nxv8i16.i64(<vscale x 8 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vwmaccu_vx_u32m4_m(vbool8_t mask, vuint32m4_t acc,
@@ -1721,7 +1721,7 @@ vuint32m4_t test_vwmaccu_vx_u32m4_m(vbool8_t mask, vuint32m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmaccu.mask.nxv16i32.nxv16i16.nxv16i16.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmaccu.mask.nxv16i32.nxv16i16.nxv16i16.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vwmaccu_vv_u32m8_m(vbool4_t mask, vuint32m8_t acc,
@@ -1733,7 +1733,7 @@ vuint32m8_t test_vwmaccu_vv_u32m8_m(vbool4_t mask, vuint32m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmaccu.mask.nxv16i32.i16.nxv16i16.i64(<vscale x 16 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmaccu.mask.nxv16i32.i16.nxv16i16.i64(<vscale x 16 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vwmaccu_vx_u32m8_m(vbool4_t mask, vuint32m8_t acc,
@@ -1744,7 +1744,7 @@ vuint32m8_t test_vwmaccu_vx_u32m8_m(vbool4_t mask, vuint32m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmaccu.mask.nxv1i64.nxv1i32.nxv1i32.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmaccu.mask.nxv1i64.nxv1i32.nxv1i32.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vwmaccu_vv_u64m1_m(vbool64_t mask, vuint64m1_t acc,
@@ -1756,7 +1756,7 @@ vuint64m1_t test_vwmaccu_vv_u64m1_m(vbool64_t mask, vuint64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmaccu.mask.nxv1i64.i32.nxv1i32.i64(<vscale x 1 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmaccu.mask.nxv1i64.i32.nxv1i32.i64(<vscale x 1 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vwmaccu_vx_u64m1_m(vbool64_t mask, vuint64m1_t acc,
@@ -1767,7 +1767,7 @@ vuint64m1_t test_vwmaccu_vx_u64m1_m(vbool64_t mask, vuint64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmaccu.mask.nxv2i64.nxv2i32.nxv2i32.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmaccu.mask.nxv2i64.nxv2i32.nxv2i32.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vwmaccu_vv_u64m2_m(vbool32_t mask, vuint64m2_t acc,
@@ -1779,7 +1779,7 @@ vuint64m2_t test_vwmaccu_vv_u64m2_m(vbool32_t mask, vuint64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmaccu.mask.nxv2i64.i32.nxv2i32.i64(<vscale x 2 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmaccu.mask.nxv2i64.i32.nxv2i32.i64(<vscale x 2 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vwmaccu_vx_u64m2_m(vbool32_t mask, vuint64m2_t acc,
@@ -1790,7 +1790,7 @@ vuint64m2_t test_vwmaccu_vx_u64m2_m(vbool32_t mask, vuint64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmaccu.mask.nxv4i64.nxv4i32.nxv4i32.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmaccu.mask.nxv4i64.nxv4i32.nxv4i32.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vwmaccu_vv_u64m4_m(vbool16_t mask, vuint64m4_t acc,
@@ -1802,7 +1802,7 @@ vuint64m4_t test_vwmaccu_vv_u64m4_m(vbool16_t mask, vuint64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmaccu.mask.nxv4i64.i32.nxv4i32.i64(<vscale x 4 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmaccu.mask.nxv4i64.i32.nxv4i32.i64(<vscale x 4 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vwmaccu_vx_u64m4_m(vbool16_t mask, vuint64m4_t acc,
@@ -1813,7 +1813,7 @@ vuint64m4_t test_vwmaccu_vx_u64m4_m(vbool16_t mask, vuint64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmaccu.mask.nxv8i64.nxv8i32.nxv8i32.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmaccu.mask.nxv8i64.nxv8i32.nxv8i32.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vwmaccu_vv_u64m8_m(vbool8_t mask, vuint64m8_t acc,
@@ -1825,7 +1825,7 @@ vuint64m8_t test_vwmaccu_vv_u64m8_m(vbool8_t mask, vuint64m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccu_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmaccu.mask.nxv8i64.i32.nxv8i32.i64(<vscale x 8 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmaccu.mask.nxv8i64.i32.nxv8i32.i64(<vscale x 8 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vwmaccu_vx_u64m8_m(vbool8_t mask, vuint64m8_t acc,
@@ -1836,7 +1836,7 @@ vuint64m8_t test_vwmaccu_vx_u64m8_m(vbool8_t mask, vuint64m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmaccsu.mask.nxv1i16.nxv1i8.nxv1i8.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmaccsu.mask.nxv1i16.nxv1i8.nxv1i8.i64(<vscale x 1 x i16> [[ACC:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vwmaccsu_vv_i16mf4_m(vbool64_t mask, vint16mf4_t acc,
@@ -1848,7 +1848,7 @@ vint16mf4_t test_vwmaccsu_vv_i16mf4_m(vbool64_t mask, vint16mf4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmaccsu.mask.nxv1i16.i8.nxv1i8.i64(<vscale x 1 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmaccsu.mask.nxv1i16.i8.nxv1i8.i64(<vscale x 1 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vwmaccsu_vx_i16mf4_m(vbool64_t mask, vint16mf4_t acc,
@@ -1859,7 +1859,7 @@ vint16mf4_t test_vwmaccsu_vx_i16mf4_m(vbool64_t mask, vint16mf4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmaccsu.mask.nxv2i16.nxv2i8.nxv2i8.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmaccsu.mask.nxv2i16.nxv2i8.nxv2i8.i64(<vscale x 2 x i16> [[ACC:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vwmaccsu_vv_i16mf2_m(vbool32_t mask, vint16mf2_t acc,
@@ -1871,7 +1871,7 @@ vint16mf2_t test_vwmaccsu_vv_i16mf2_m(vbool32_t mask, vint16mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmaccsu.mask.nxv2i16.i8.nxv2i8.i64(<vscale x 2 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmaccsu.mask.nxv2i16.i8.nxv2i8.i64(<vscale x 2 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vwmaccsu_vx_i16mf2_m(vbool32_t mask, vint16mf2_t acc,
@@ -1882,7 +1882,7 @@ vint16mf2_t test_vwmaccsu_vx_i16mf2_m(vbool32_t mask, vint16mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmaccsu.mask.nxv4i16.nxv4i8.nxv4i8.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmaccsu.mask.nxv4i16.nxv4i8.nxv4i8.i64(<vscale x 4 x i16> [[ACC:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwmaccsu_vv_i16m1_m(vbool16_t mask, vint16m1_t acc,
@@ -1894,7 +1894,7 @@ vint16m1_t test_vwmaccsu_vv_i16m1_m(vbool16_t mask, vint16m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmaccsu.mask.nxv4i16.i8.nxv4i8.i64(<vscale x 4 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmaccsu.mask.nxv4i16.i8.nxv4i8.i64(<vscale x 4 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwmaccsu_vx_i16m1_m(vbool16_t mask, vint16m1_t acc, int8_t op1,
@@ -1905,7 +1905,7 @@ vint16m1_t test_vwmaccsu_vx_i16m1_m(vbool16_t mask, vint16m1_t acc, int8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmaccsu.mask.nxv8i16.nxv8i8.nxv8i8.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmaccsu.mask.nxv8i16.nxv8i8.nxv8i8.i64(<vscale x 8 x i16> [[ACC:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vwmaccsu_vv_i16m2_m(vbool8_t mask, vint16m2_t acc,
@@ -1916,7 +1916,7 @@ vint16m2_t test_vwmaccsu_vv_i16m2_m(vbool8_t mask, vint16m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmaccsu.mask.nxv8i16.i8.nxv8i8.i64(<vscale x 8 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmaccsu.mask.nxv8i16.i8.nxv8i8.i64(<vscale x 8 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vwmaccsu_vx_i16m2_m(vbool8_t mask, vint16m2_t acc, int8_t op1,
@@ -1927,7 +1927,7 @@ vint16m2_t test_vwmaccsu_vx_i16m2_m(vbool8_t mask, vint16m2_t acc, int8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmaccsu.mask.nxv16i16.nxv16i8.nxv16i8.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmaccsu.mask.nxv16i16.nxv16i8.nxv16i8.i64(<vscale x 16 x i16> [[ACC:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vwmaccsu_vv_i16m4_m(vbool4_t mask, vint16m4_t acc,
@@ -1938,7 +1938,7 @@ vint16m4_t test_vwmaccsu_vv_i16m4_m(vbool4_t mask, vint16m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmaccsu.mask.nxv16i16.i8.nxv16i8.i64(<vscale x 16 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmaccsu.mask.nxv16i16.i8.nxv16i8.i64(<vscale x 16 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vwmaccsu_vx_i16m4_m(vbool4_t mask, vint16m4_t acc, int8_t op1,
@@ -1949,7 +1949,7 @@ vint16m4_t test_vwmaccsu_vx_i16m4_m(vbool4_t mask, vint16m4_t acc, int8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmaccsu.mask.nxv32i16.nxv32i8.nxv32i8.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmaccsu.mask.nxv32i16.nxv32i8.nxv32i8.i64(<vscale x 32 x i16> [[ACC:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vwmaccsu_vv_i16m8_m(vbool2_t mask, vint16m8_t acc,
@@ -1960,7 +1960,7 @@ vint16m8_t test_vwmaccsu_vv_i16m8_m(vbool2_t mask, vint16m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmaccsu.mask.nxv32i16.i8.nxv32i8.i64(<vscale x 32 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmaccsu.mask.nxv32i16.i8.nxv32i8.i64(<vscale x 32 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vwmaccsu_vx_i16m8_m(vbool2_t mask, vint16m8_t acc, int8_t op1,
@@ -1971,7 +1971,7 @@ vint16m8_t test_vwmaccsu_vx_i16m8_m(vbool2_t mask, vint16m8_t acc, int8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmaccsu.mask.nxv1i32.nxv1i16.nxv1i16.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmaccsu.mask.nxv1i32.nxv1i16.nxv1i16.i64(<vscale x 1 x i32> [[ACC:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vwmaccsu_vv_i32mf2_m(vbool64_t mask, vint32mf2_t acc,
@@ -1983,7 +1983,7 @@ vint32mf2_t test_vwmaccsu_vv_i32mf2_m(vbool64_t mask, vint32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmaccsu.mask.nxv1i32.i16.nxv1i16.i64(<vscale x 1 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmaccsu.mask.nxv1i32.i16.nxv1i16.i64(<vscale x 1 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vwmaccsu_vx_i32mf2_m(vbool64_t mask, vint32mf2_t acc,
@@ -1995,7 +1995,7 @@ vint32mf2_t test_vwmaccsu_vx_i32mf2_m(vbool64_t mask, vint32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmaccsu.mask.nxv2i32.nxv2i16.nxv2i16.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmaccsu.mask.nxv2i32.nxv2i16.nxv2i16.i64(<vscale x 2 x i32> [[ACC:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwmaccsu_vv_i32m1_m(vbool32_t mask, vint32m1_t acc,
@@ -2007,7 +2007,7 @@ vint32m1_t test_vwmaccsu_vv_i32m1_m(vbool32_t mask, vint32m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmaccsu.mask.nxv2i32.i16.nxv2i16.i64(<vscale x 2 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmaccsu.mask.nxv2i32.i16.nxv2i16.i64(<vscale x 2 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwmaccsu_vx_i32m1_m(vbool32_t mask, vint32m1_t acc, int16_t op1,
@@ -2018,7 +2018,7 @@ vint32m1_t test_vwmaccsu_vx_i32m1_m(vbool32_t mask, vint32m1_t acc, int16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmaccsu.mask.nxv4i32.nxv4i16.nxv4i16.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmaccsu.mask.nxv4i32.nxv4i16.nxv4i16.i64(<vscale x 4 x i32> [[ACC:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vwmaccsu_vv_i32m2_m(vbool16_t mask, vint32m2_t acc,
@@ -2030,7 +2030,7 @@ vint32m2_t test_vwmaccsu_vv_i32m2_m(vbool16_t mask, vint32m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmaccsu.mask.nxv4i32.i16.nxv4i16.i64(<vscale x 4 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmaccsu.mask.nxv4i32.i16.nxv4i16.i64(<vscale x 4 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vwmaccsu_vx_i32m2_m(vbool16_t mask, vint32m2_t acc, int16_t op1,
@@ -2041,7 +2041,7 @@ vint32m2_t test_vwmaccsu_vx_i32m2_m(vbool16_t mask, vint32m2_t acc, int16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmaccsu.mask.nxv8i32.nxv8i16.nxv8i16.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmaccsu.mask.nxv8i32.nxv8i16.nxv8i16.i64(<vscale x 8 x i32> [[ACC:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vwmaccsu_vv_i32m4_m(vbool8_t mask, vint32m4_t acc,
@@ -2053,7 +2053,7 @@ vint32m4_t test_vwmaccsu_vv_i32m4_m(vbool8_t mask, vint32m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmaccsu.mask.nxv8i32.i16.nxv8i16.i64(<vscale x 8 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmaccsu.mask.nxv8i32.i16.nxv8i16.i64(<vscale x 8 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vwmaccsu_vx_i32m4_m(vbool8_t mask, vint32m4_t acc, int16_t op1,
@@ -2064,7 +2064,7 @@ vint32m4_t test_vwmaccsu_vx_i32m4_m(vbool8_t mask, vint32m4_t acc, int16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmaccsu.mask.nxv16i32.nxv16i16.nxv16i16.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmaccsu.mask.nxv16i32.nxv16i16.nxv16i16.i64(<vscale x 16 x i32> [[ACC:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vwmaccsu_vv_i32m8_m(vbool4_t mask, vint32m8_t acc,
@@ -2076,7 +2076,7 @@ vint32m8_t test_vwmaccsu_vv_i32m8_m(vbool4_t mask, vint32m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmaccsu.mask.nxv16i32.i16.nxv16i16.i64(<vscale x 16 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmaccsu.mask.nxv16i32.i16.nxv16i16.i64(<vscale x 16 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vwmaccsu_vx_i32m8_m(vbool4_t mask, vint32m8_t acc, int16_t op1,
@@ -2087,7 +2087,7 @@ vint32m8_t test_vwmaccsu_vx_i32m8_m(vbool4_t mask, vint32m8_t acc, int16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmaccsu.mask.nxv1i64.nxv1i32.nxv1i32.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmaccsu.mask.nxv1i64.nxv1i32.nxv1i32.i64(<vscale x 1 x i64> [[ACC:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwmaccsu_vv_i64m1_m(vbool64_t mask, vint64m1_t acc,
@@ -2099,7 +2099,7 @@ vint64m1_t test_vwmaccsu_vv_i64m1_m(vbool64_t mask, vint64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmaccsu.mask.nxv1i64.i32.nxv1i32.i64(<vscale x 1 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmaccsu.mask.nxv1i64.i32.nxv1i32.i64(<vscale x 1 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwmaccsu_vx_i64m1_m(vbool64_t mask, vint64m1_t acc, int32_t op1,
@@ -2110,7 +2110,7 @@ vint64m1_t test_vwmaccsu_vx_i64m1_m(vbool64_t mask, vint64m1_t acc, int32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmaccsu.mask.nxv2i64.nxv2i32.nxv2i32.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmaccsu.mask.nxv2i64.nxv2i32.nxv2i32.i64(<vscale x 2 x i64> [[ACC:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vwmaccsu_vv_i64m2_m(vbool32_t mask, vint64m2_t acc,
@@ -2122,7 +2122,7 @@ vint64m2_t test_vwmaccsu_vv_i64m2_m(vbool32_t mask, vint64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmaccsu.mask.nxv2i64.i32.nxv2i32.i64(<vscale x 2 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmaccsu.mask.nxv2i64.i32.nxv2i32.i64(<vscale x 2 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vwmaccsu_vx_i64m2_m(vbool32_t mask, vint64m2_t acc, int32_t op1,
@@ -2133,7 +2133,7 @@ vint64m2_t test_vwmaccsu_vx_i64m2_m(vbool32_t mask, vint64m2_t acc, int32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmaccsu.mask.nxv4i64.nxv4i32.nxv4i32.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmaccsu.mask.nxv4i64.nxv4i32.nxv4i32.i64(<vscale x 4 x i64> [[ACC:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vwmaccsu_vv_i64m4_m(vbool16_t mask, vint64m4_t acc,
@@ -2145,7 +2145,7 @@ vint64m4_t test_vwmaccsu_vv_i64m4_m(vbool16_t mask, vint64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmaccsu.mask.nxv4i64.i32.nxv4i32.i64(<vscale x 4 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmaccsu.mask.nxv4i64.i32.nxv4i32.i64(<vscale x 4 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vwmaccsu_vx_i64m4_m(vbool16_t mask, vint64m4_t acc, int32_t op1,
@@ -2156,7 +2156,7 @@ vint64m4_t test_vwmaccsu_vx_i64m4_m(vbool16_t mask, vint64m4_t acc, int32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmaccsu.mask.nxv8i64.nxv8i32.nxv8i32.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmaccsu.mask.nxv8i64.nxv8i32.nxv8i32.i64(<vscale x 8 x i64> [[ACC:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vwmaccsu_vv_i64m8_m(vbool8_t mask, vint64m8_t acc,
@@ -2168,7 +2168,7 @@ vint64m8_t test_vwmaccsu_vv_i64m8_m(vbool8_t mask, vint64m8_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccsu_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmaccsu.mask.nxv8i64.i32.nxv8i32.i64(<vscale x 8 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmaccsu.mask.nxv8i64.i32.nxv8i32.i64(<vscale x 8 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vwmaccsu_vx_i64m8_m(vbool8_t mask, vint64m8_t acc, int32_t op1,
@@ -2179,7 +2179,7 @@ vint64m8_t test_vwmaccsu_vx_i64m8_m(vbool8_t mask, vint64m8_t acc, int32_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccus_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmaccus.mask.nxv1i16.i8.nxv1i8.i64(<vscale x 1 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmaccus.mask.nxv1i16.i8.nxv1i8.i64(<vscale x 1 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vwmaccus_vx_i16mf4_m(vbool64_t mask, vint16mf4_t acc,
@@ -2190,7 +2190,7 @@ vint16mf4_t test_vwmaccus_vx_i16mf4_m(vbool64_t mask, vint16mf4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccus_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmaccus.mask.nxv2i16.i8.nxv2i8.i64(<vscale x 2 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmaccus.mask.nxv2i16.i8.nxv2i8.i64(<vscale x 2 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vwmaccus_vx_i16mf2_m(vbool32_t mask, vint16mf2_t acc,
@@ -2201,7 +2201,7 @@ vint16mf2_t test_vwmaccus_vx_i16mf2_m(vbool32_t mask, vint16mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccus_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmaccus.mask.nxv4i16.i8.nxv4i8.i64(<vscale x 4 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmaccus.mask.nxv4i16.i8.nxv4i8.i64(<vscale x 4 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwmaccus_vx_i16m1_m(vbool16_t mask, vint16m1_t acc, uint8_t op1,
@@ -2212,7 +2212,7 @@ vint16m1_t test_vwmaccus_vx_i16m1_m(vbool16_t mask, vint16m1_t acc, uint8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccus_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmaccus.mask.nxv8i16.i8.nxv8i8.i64(<vscale x 8 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmaccus.mask.nxv8i16.i8.nxv8i8.i64(<vscale x 8 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vwmaccus_vx_i16m2_m(vbool8_t mask, vint16m2_t acc, uint8_t op1,
@@ -2223,7 +2223,7 @@ vint16m2_t test_vwmaccus_vx_i16m2_m(vbool8_t mask, vint16m2_t acc, uint8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccus_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmaccus.mask.nxv16i16.i8.nxv16i8.i64(<vscale x 16 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmaccus.mask.nxv16i16.i8.nxv16i8.i64(<vscale x 16 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vwmaccus_vx_i16m4_m(vbool4_t mask, vint16m4_t acc, uint8_t op1,
@@ -2234,7 +2234,7 @@ vint16m4_t test_vwmaccus_vx_i16m4_m(vbool4_t mask, vint16m4_t acc, uint8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccus_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmaccus.mask.nxv32i16.i8.nxv32i8.i64(<vscale x 32 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmaccus.mask.nxv32i16.i8.nxv32i8.i64(<vscale x 32 x i16> [[ACC:%.*]], i8 [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vwmaccus_vx_i16m8_m(vbool2_t mask, vint16m8_t acc, uint8_t op1,
@@ -2245,7 +2245,7 @@ vint16m8_t test_vwmaccus_vx_i16m8_m(vbool2_t mask, vint16m8_t acc, uint8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccus_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmaccus.mask.nxv1i32.i16.nxv1i16.i64(<vscale x 1 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmaccus.mask.nxv1i32.i16.nxv1i16.i64(<vscale x 1 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vwmaccus_vx_i32mf2_m(vbool64_t mask, vint32mf2_t acc,
@@ -2257,7 +2257,7 @@ vint32mf2_t test_vwmaccus_vx_i32mf2_m(vbool64_t mask, vint32mf2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccus_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmaccus.mask.nxv2i32.i16.nxv2i16.i64(<vscale x 2 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmaccus.mask.nxv2i32.i16.nxv2i16.i64(<vscale x 2 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwmaccus_vx_i32m1_m(vbool32_t mask, vint32m1_t acc,
@@ -2268,7 +2268,7 @@ vint32m1_t test_vwmaccus_vx_i32m1_m(vbool32_t mask, vint32m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccus_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmaccus.mask.nxv4i32.i16.nxv4i16.i64(<vscale x 4 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmaccus.mask.nxv4i32.i16.nxv4i16.i64(<vscale x 4 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vwmaccus_vx_i32m2_m(vbool16_t mask, vint32m2_t acc,
@@ -2279,7 +2279,7 @@ vint32m2_t test_vwmaccus_vx_i32m2_m(vbool16_t mask, vint32m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccus_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmaccus.mask.nxv8i32.i16.nxv8i16.i64(<vscale x 8 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmaccus.mask.nxv8i32.i16.nxv8i16.i64(<vscale x 8 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vwmaccus_vx_i32m4_m(vbool8_t mask, vint32m4_t acc, uint16_t op1,
@@ -2290,7 +2290,7 @@ vint32m4_t test_vwmaccus_vx_i32m4_m(vbool8_t mask, vint32m4_t acc, uint16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccus_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmaccus.mask.nxv16i32.i16.nxv16i16.i64(<vscale x 16 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmaccus.mask.nxv16i32.i16.nxv16i16.i64(<vscale x 16 x i32> [[ACC:%.*]], i16 [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vwmaccus_vx_i32m8_m(vbool4_t mask, vint32m8_t acc, uint16_t op1,
@@ -2301,7 +2301,7 @@ vint32m8_t test_vwmaccus_vx_i32m8_m(vbool4_t mask, vint32m8_t acc, uint16_t op1,
 //
 // CHECK-RV64-LABEL: @test_vwmaccus_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmaccus.mask.nxv1i64.i32.nxv1i32.i64(<vscale x 1 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmaccus.mask.nxv1i64.i32.nxv1i32.i64(<vscale x 1 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwmaccus_vx_i64m1_m(vbool64_t mask, vint64m1_t acc,
@@ -2312,7 +2312,7 @@ vint64m1_t test_vwmaccus_vx_i64m1_m(vbool64_t mask, vint64m1_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccus_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmaccus.mask.nxv2i64.i32.nxv2i32.i64(<vscale x 2 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmaccus.mask.nxv2i64.i32.nxv2i32.i64(<vscale x 2 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vwmaccus_vx_i64m2_m(vbool32_t mask, vint64m2_t acc,
@@ -2323,7 +2323,7 @@ vint64m2_t test_vwmaccus_vx_i64m2_m(vbool32_t mask, vint64m2_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccus_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmaccus.mask.nxv4i64.i32.nxv4i32.i64(<vscale x 4 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmaccus.mask.nxv4i64.i32.nxv4i32.i64(<vscale x 4 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vwmaccus_vx_i64m4_m(vbool16_t mask, vint64m4_t acc,
@@ -2334,7 +2334,7 @@ vint64m4_t test_vwmaccus_vx_i64m4_m(vbool16_t mask, vint64m4_t acc,
 //
 // CHECK-RV64-LABEL: @test_vwmaccus_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmaccus.mask.nxv8i64.i32.nxv8i32.i64(<vscale x 8 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmaccus.mask.nxv8i64.i32.nxv8i32.i64(<vscale x 8 x i64> [[ACC:%.*]], i32 [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vwmaccus_vx_i64m8_m(vbool8_t mask, vint64m8_t acc, uint32_t op1,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwmul.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwmul.c
index 5554be7..e985433 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwmul.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwmul.c
@@ -907,7 +907,7 @@ vint64m8_t test_vwmulsu_vx_i64m8(vint32m4_t op1, uint32_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vwmul_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmul.mask.nxv1i16.nxv1i8.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmul.mask.nxv1i16.nxv1i8.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vwmul_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint8mf8_t op1, vint8mf8_t op2, size_t vl) {
@@ -917,7 +917,7 @@ vint16mf4_t test_vwmul_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint8m
 //
 // CHECK-RV64-LABEL: @test_vwmul_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmul.mask.nxv1i16.nxv1i8.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmul.mask.nxv1i16.nxv1i8.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vwmul_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint8mf8_t op1, int8_t op2, size_t vl) {
@@ -927,7 +927,7 @@ vint16mf4_t test_vwmul_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint8m
 //
 // CHECK-RV64-LABEL: @test_vwmul_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmul.mask.nxv2i16.nxv2i8.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmul.mask.nxv2i16.nxv2i8.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vwmul_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint8mf4_t op1, vint8mf4_t op2, size_t vl) {
@@ -937,7 +937,7 @@ vint16mf2_t test_vwmul_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint8m
 //
 // CHECK-RV64-LABEL: @test_vwmul_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmul.mask.nxv2i16.nxv2i8.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmul.mask.nxv2i16.nxv2i8.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vwmul_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint8mf4_t op1, int8_t op2, size_t vl) {
@@ -947,7 +947,7 @@ vint16mf2_t test_vwmul_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint8m
 //
 // CHECK-RV64-LABEL: @test_vwmul_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmul.mask.nxv4i16.nxv4i8.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmul.mask.nxv4i16.nxv4i8.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwmul_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint8mf2_t op1, vint8mf2_t op2, size_t vl) {
@@ -957,7 +957,7 @@ vint16m1_t test_vwmul_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint8mf2_
 //
 // CHECK-RV64-LABEL: @test_vwmul_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmul.mask.nxv4i16.nxv4i8.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmul.mask.nxv4i16.nxv4i8.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwmul_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint8mf2_t op1, int8_t op2, size_t vl) {
@@ -967,7 +967,7 @@ vint16m1_t test_vwmul_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint8mf2_
 //
 // CHECK-RV64-LABEL: @test_vwmul_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmul.mask.nxv8i16.nxv8i8.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmul.mask.nxv8i16.nxv8i8.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vwmul_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint8m1_t op1, vint8m1_t op2, size_t vl) {
@@ -977,7 +977,7 @@ vint16m2_t test_vwmul_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint8m1_t
 //
 // CHECK-RV64-LABEL: @test_vwmul_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmul.mask.nxv8i16.nxv8i8.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmul.mask.nxv8i16.nxv8i8.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vwmul_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint8m1_t op1, int8_t op2, size_t vl) {
@@ -987,7 +987,7 @@ vint16m2_t test_vwmul_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint8m1_t
 //
 // CHECK-RV64-LABEL: @test_vwmul_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmul.mask.nxv16i16.nxv16i8.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmul.mask.nxv16i16.nxv16i8.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vwmul_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint8m2_t op1, vint8m2_t op2, size_t vl) {
@@ -997,7 +997,7 @@ vint16m4_t test_vwmul_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint8m2_t
 //
 // CHECK-RV64-LABEL: @test_vwmul_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmul.mask.nxv16i16.nxv16i8.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmul.mask.nxv16i16.nxv16i8.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vwmul_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint8m2_t op1, int8_t op2, size_t vl) {
@@ -1007,7 +1007,7 @@ vint16m4_t test_vwmul_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint8m2_t
 //
 // CHECK-RV64-LABEL: @test_vwmul_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmul.mask.nxv32i16.nxv32i8.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmul.mask.nxv32i16.nxv32i8.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vwmul_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint8m4_t op1, vint8m4_t op2, size_t vl) {
@@ -1017,7 +1017,7 @@ vint16m8_t test_vwmul_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint8m4_t
 //
 // CHECK-RV64-LABEL: @test_vwmul_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmul.mask.nxv32i16.nxv32i8.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmul.mask.nxv32i16.nxv32i8.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vwmul_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint8m4_t op1, int8_t op2, size_t vl) {
@@ -1027,7 +1027,7 @@ vint16m8_t test_vwmul_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint8m4_t
 //
 // CHECK-RV64-LABEL: @test_vwmul_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmul.mask.nxv1i32.nxv1i16.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmul.mask.nxv1i32.nxv1i16.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vwmul_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint16mf4_t op1, vint16mf4_t op2, size_t vl) {
@@ -1037,7 +1037,7 @@ vint32mf2_t test_vwmul_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint16
 //
 // CHECK-RV64-LABEL: @test_vwmul_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmul.mask.nxv1i32.nxv1i16.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmul.mask.nxv1i32.nxv1i16.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vwmul_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint16mf4_t op1, int16_t op2, size_t vl) {
@@ -1047,7 +1047,7 @@ vint32mf2_t test_vwmul_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint16
 //
 // CHECK-RV64-LABEL: @test_vwmul_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmul.mask.nxv2i32.nxv2i16.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmul.mask.nxv2i32.nxv2i16.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwmul_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint16mf2_t op1, vint16mf2_t op2, size_t vl) {
@@ -1057,7 +1057,7 @@ vint32m1_t test_vwmul_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint16mf2
 //
 // CHECK-RV64-LABEL: @test_vwmul_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmul.mask.nxv2i32.nxv2i16.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmul.mask.nxv2i32.nxv2i16.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwmul_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint16mf2_t op1, int16_t op2, size_t vl) {
@@ -1067,7 +1067,7 @@ vint32m1_t test_vwmul_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint16mf2
 //
 // CHECK-RV64-LABEL: @test_vwmul_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmul.mask.nxv4i32.nxv4i16.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmul.mask.nxv4i32.nxv4i16.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vwmul_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint16m1_t op1, vint16m1_t op2, size_t vl) {
@@ -1077,7 +1077,7 @@ vint32m2_t test_vwmul_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint16m1_
 //
 // CHECK-RV64-LABEL: @test_vwmul_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmul.mask.nxv4i32.nxv4i16.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmul.mask.nxv4i32.nxv4i16.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vwmul_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint16m1_t op1, int16_t op2, size_t vl) {
@@ -1087,7 +1087,7 @@ vint32m2_t test_vwmul_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint16m1_
 //
 // CHECK-RV64-LABEL: @test_vwmul_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmul.mask.nxv8i32.nxv8i16.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmul.mask.nxv8i32.nxv8i16.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vwmul_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint16m2_t op1, vint16m2_t op2, size_t vl) {
@@ -1097,7 +1097,7 @@ vint32m4_t test_vwmul_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vwmul_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmul.mask.nxv8i32.nxv8i16.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmul.mask.nxv8i32.nxv8i16.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vwmul_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint16m2_t op1, int16_t op2, size_t vl) {
@@ -1107,7 +1107,7 @@ vint32m4_t test_vwmul_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vwmul_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmul.mask.nxv16i32.nxv16i16.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmul.mask.nxv16i32.nxv16i16.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vwmul_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint16m4_t op1, vint16m4_t op2, size_t vl) {
@@ -1117,7 +1117,7 @@ vint32m8_t test_vwmul_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vwmul_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmul.mask.nxv16i32.nxv16i16.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmul.mask.nxv16i32.nxv16i16.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vwmul_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint16m4_t op1, int16_t op2, size_t vl) {
@@ -1127,7 +1127,7 @@ vint32m8_t test_vwmul_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vwmul_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmul.mask.nxv1i64.nxv1i32.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmul.mask.nxv1i64.nxv1i32.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwmul_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint32mf2_t op1, vint32mf2_t op2, size_t vl) {
@@ -1137,7 +1137,7 @@ vint64m1_t test_vwmul_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint32mf2
 //
 // CHECK-RV64-LABEL: @test_vwmul_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmul.mask.nxv1i64.nxv1i32.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmul.mask.nxv1i64.nxv1i32.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwmul_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint32mf2_t op1, int32_t op2, size_t vl) {
@@ -1147,7 +1147,7 @@ vint64m1_t test_vwmul_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint32mf2
 //
 // CHECK-RV64-LABEL: @test_vwmul_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmul.mask.nxv2i64.nxv2i32.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmul.mask.nxv2i64.nxv2i32.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vwmul_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint32m1_t op1, vint32m1_t op2, size_t vl) {
@@ -1157,7 +1157,7 @@ vint64m2_t test_vwmul_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint32m1_
 //
 // CHECK-RV64-LABEL: @test_vwmul_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmul.mask.nxv2i64.nxv2i32.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmul.mask.nxv2i64.nxv2i32.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vwmul_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint32m1_t op1, int32_t op2, size_t vl) {
@@ -1167,7 +1167,7 @@ vint64m2_t test_vwmul_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint32m1_
 //
 // CHECK-RV64-LABEL: @test_vwmul_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmul.mask.nxv4i64.nxv4i32.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmul.mask.nxv4i64.nxv4i32.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vwmul_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint32m2_t op1, vint32m2_t op2, size_t vl) {
@@ -1177,7 +1177,7 @@ vint64m4_t test_vwmul_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint32m2_
 //
 // CHECK-RV64-LABEL: @test_vwmul_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmul.mask.nxv4i64.nxv4i32.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmul.mask.nxv4i64.nxv4i32.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vwmul_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint32m2_t op1, int32_t op2, size_t vl) {
@@ -1187,7 +1187,7 @@ vint64m4_t test_vwmul_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint32m2_
 //
 // CHECK-RV64-LABEL: @test_vwmul_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmul.mask.nxv8i64.nxv8i32.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmul.mask.nxv8i64.nxv8i32.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vwmul_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint32m4_t op1, vint32m4_t op2, size_t vl) {
@@ -1197,7 +1197,7 @@ vint64m8_t test_vwmul_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vwmul_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmul.mask.nxv8i64.nxv8i32.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmul.mask.nxv8i64.nxv8i32.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vwmul_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint32m4_t op1, int32_t op2, size_t vl) {
@@ -1207,7 +1207,7 @@ vint64m8_t test_vwmul_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmulu.mask.nxv1i16.nxv1i8.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmulu.mask.nxv1i16.nxv1i8.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vwmulu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint8mf8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -1217,7 +1217,7 @@ vuint16mf4_t test_vwmulu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmulu.mask.nxv1i16.nxv1i8.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmulu.mask.nxv1i16.nxv1i8.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vwmulu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint8mf8_t op1, uint8_t op2, size_t vl) {
@@ -1227,7 +1227,7 @@ vuint16mf4_t test_vwmulu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmulu.mask.nxv2i16.nxv2i8.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmulu.mask.nxv2i16.nxv2i8.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vwmulu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint8mf4_t op1, vuint8mf4_t op2, size_t vl) {
@@ -1237,7 +1237,7 @@ vuint16mf2_t test_vwmulu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmulu.mask.nxv2i16.nxv2i8.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmulu.mask.nxv2i16.nxv2i8.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vwmulu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint8mf4_t op1, uint8_t op2, size_t vl) {
@@ -1247,7 +1247,7 @@ vuint16mf2_t test_vwmulu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmulu.mask.nxv4i16.nxv4i8.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmulu.mask.nxv4i16.nxv4i8.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vwmulu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint8mf2_t op1, vuint8mf2_t op2, size_t vl) {
@@ -1257,7 +1257,7 @@ vuint16m1_t test_vwmulu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint8
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmulu.mask.nxv4i16.nxv4i8.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmulu.mask.nxv4i16.nxv4i8.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vwmulu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint8mf2_t op1, uint8_t op2, size_t vl) {
@@ -1267,7 +1267,7 @@ vuint16m1_t test_vwmulu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint8
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmulu.mask.nxv8i16.nxv8i8.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmulu.mask.nxv8i16.nxv8i8.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vwmulu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint8m1_t op1, vuint8m1_t op2, size_t vl) {
@@ -1277,7 +1277,7 @@ vuint16m2_t test_vwmulu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmulu.mask.nxv8i16.nxv8i8.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmulu.mask.nxv8i16.nxv8i8.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vwmulu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint8m1_t op1, uint8_t op2, size_t vl) {
@@ -1287,7 +1287,7 @@ vuint16m2_t test_vwmulu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmulu.mask.nxv16i16.nxv16i8.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmulu.mask.nxv16i16.nxv16i8.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vwmulu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint8m2_t op1, vuint8m2_t op2, size_t vl) {
@@ -1297,7 +1297,7 @@ vuint16m4_t test_vwmulu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmulu.mask.nxv16i16.nxv16i8.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmulu.mask.nxv16i16.nxv16i8.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vwmulu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint8m2_t op1, uint8_t op2, size_t vl) {
@@ -1307,7 +1307,7 @@ vuint16m4_t test_vwmulu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmulu.mask.nxv32i16.nxv32i8.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmulu.mask.nxv32i16.nxv32i8.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vwmulu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint8m4_t op1, vuint8m4_t op2, size_t vl) {
@@ -1317,7 +1317,7 @@ vuint16m8_t test_vwmulu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmulu.mask.nxv32i16.nxv32i8.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmulu.mask.nxv32i16.nxv32i8.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vwmulu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint8m4_t op1, uint8_t op2, size_t vl) {
@@ -1327,7 +1327,7 @@ vuint16m8_t test_vwmulu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint8m
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmulu.mask.nxv1i32.nxv1i16.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmulu.mask.nxv1i32.nxv1i16.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vwmulu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint16mf4_t op1, vuint16mf4_t op2, size_t vl) {
@@ -1337,7 +1337,7 @@ vuint32mf2_t test_vwmulu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmulu.mask.nxv1i32.nxv1i16.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmulu.mask.nxv1i32.nxv1i16.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vwmulu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint16mf4_t op1, uint16_t op2, size_t vl) {
@@ -1347,7 +1347,7 @@ vuint32mf2_t test_vwmulu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vui
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmulu.mask.nxv2i32.nxv2i16.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmulu.mask.nxv2i32.nxv2i16.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vwmulu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint16mf2_t op1, vuint16mf2_t op2, size_t vl) {
@@ -1357,7 +1357,7 @@ vuint32m1_t test_vwmulu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint1
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmulu.mask.nxv2i32.nxv2i16.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmulu.mask.nxv2i32.nxv2i16.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vwmulu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint16mf2_t op1, uint16_t op2, size_t vl) {
@@ -1367,7 +1367,7 @@ vuint32m1_t test_vwmulu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint1
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmulu.mask.nxv4i32.nxv4i16.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmulu.mask.nxv4i32.nxv4i16.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vwmulu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint16m1_t op1, vuint16m1_t op2, size_t vl) {
@@ -1377,7 +1377,7 @@ vuint32m2_t test_vwmulu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint1
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmulu.mask.nxv4i32.nxv4i16.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmulu.mask.nxv4i32.nxv4i16.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vwmulu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint16m1_t op1, uint16_t op2, size_t vl) {
@@ -1387,7 +1387,7 @@ vuint32m2_t test_vwmulu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint1
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmulu.mask.nxv8i32.nxv8i16.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmulu.mask.nxv8i32.nxv8i16.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vwmulu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint16m2_t op1, vuint16m2_t op2, size_t vl) {
@@ -1397,7 +1397,7 @@ vuint32m4_t test_vwmulu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmulu.mask.nxv8i32.nxv8i16.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmulu.mask.nxv8i32.nxv8i16.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vwmulu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint16m2_t op1, uint16_t op2, size_t vl) {
@@ -1407,7 +1407,7 @@ vuint32m4_t test_vwmulu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmulu.mask.nxv16i32.nxv16i16.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmulu.mask.nxv16i32.nxv16i16.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vwmulu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint16m4_t op1, vuint16m4_t op2, size_t vl) {
@@ -1417,7 +1417,7 @@ vuint32m8_t test_vwmulu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmulu.mask.nxv16i32.nxv16i16.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmulu.mask.nxv16i32.nxv16i16.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vwmulu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint16m4_t op1, uint16_t op2, size_t vl) {
@@ -1427,7 +1427,7 @@ vuint32m8_t test_vwmulu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint16
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmulu.mask.nxv1i64.nxv1i32.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmulu.mask.nxv1i64.nxv1i32.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vwmulu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint32mf2_t op1, vuint32mf2_t op2, size_t vl) {
@@ -1437,7 +1437,7 @@ vuint64m1_t test_vwmulu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint3
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmulu.mask.nxv1i64.nxv1i32.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmulu.mask.nxv1i64.nxv1i32.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vwmulu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint32mf2_t op1, uint32_t op2, size_t vl) {
@@ -1447,7 +1447,7 @@ vuint64m1_t test_vwmulu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint3
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmulu.mask.nxv2i64.nxv2i32.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmulu.mask.nxv2i64.nxv2i32.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vwmulu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint32m1_t op1, vuint32m1_t op2, size_t vl) {
@@ -1457,7 +1457,7 @@ vuint64m2_t test_vwmulu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint3
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmulu.mask.nxv2i64.nxv2i32.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmulu.mask.nxv2i64.nxv2i32.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vwmulu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint32m1_t op1, uint32_t op2, size_t vl) {
@@ -1467,7 +1467,7 @@ vuint64m2_t test_vwmulu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint3
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmulu.mask.nxv4i64.nxv4i32.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmulu.mask.nxv4i64.nxv4i32.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vwmulu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint32m2_t op1, vuint32m2_t op2, size_t vl) {
@@ -1477,7 +1477,7 @@ vuint64m4_t test_vwmulu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint3
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmulu.mask.nxv4i64.nxv4i32.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmulu.mask.nxv4i64.nxv4i32.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vwmulu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint32m2_t op1, uint32_t op2, size_t vl) {
@@ -1487,7 +1487,7 @@ vuint64m4_t test_vwmulu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint3
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmulu.mask.nxv8i64.nxv8i32.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmulu.mask.nxv8i64.nxv8i32.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vwmulu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint32m4_t op1, vuint32m4_t op2, size_t vl) {
@@ -1497,7 +1497,7 @@ vuint64m8_t test_vwmulu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vwmulu_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmulu.mask.nxv8i64.nxv8i32.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmulu.mask.nxv8i64.nxv8i32.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vwmulu_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint32m4_t op1, uint32_t op2, size_t vl) {
@@ -1507,7 +1507,7 @@ vuint64m8_t test_vwmulu_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint32
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmulsu.mask.nxv1i16.nxv1i8.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmulsu.mask.nxv1i16.nxv1i8.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vwmulsu_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint8mf8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -1517,7 +1517,7 @@ vint16mf4_t test_vwmulsu_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmulsu.mask.nxv1i16.nxv1i8.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwmulsu.mask.nxv1i16.nxv1i8.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vwmulsu_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint8mf8_t op1, uint8_t op2, size_t vl) {
@@ -1527,7 +1527,7 @@ vint16mf4_t test_vwmulsu_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmulsu.mask.nxv2i16.nxv2i8.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmulsu.mask.nxv2i16.nxv2i8.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vwmulsu_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint8mf4_t op1, vuint8mf4_t op2, size_t vl) {
@@ -1537,7 +1537,7 @@ vint16mf2_t test_vwmulsu_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmulsu.mask.nxv2i16.nxv2i8.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwmulsu.mask.nxv2i16.nxv2i8.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vwmulsu_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint8mf4_t op1, uint8_t op2, size_t vl) {
@@ -1547,7 +1547,7 @@ vint16mf2_t test_vwmulsu_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmulsu.mask.nxv4i16.nxv4i8.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmulsu.mask.nxv4i16.nxv4i8.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwmulsu_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint8mf2_t op1, vuint8mf2_t op2, size_t vl) {
@@ -1557,7 +1557,7 @@ vint16m1_t test_vwmulsu_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint8mf
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmulsu.mask.nxv4i16.nxv4i8.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwmulsu.mask.nxv4i16.nxv4i8.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwmulsu_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint8mf2_t op1, uint8_t op2, size_t vl) {
@@ -1567,7 +1567,7 @@ vint16m1_t test_vwmulsu_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint8mf
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmulsu.mask.nxv8i16.nxv8i8.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmulsu.mask.nxv8i16.nxv8i8.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vwmulsu_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint8m1_t op1, vuint8m1_t op2, size_t vl) {
@@ -1577,7 +1577,7 @@ vint16m2_t test_vwmulsu_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint8m1_
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmulsu.mask.nxv8i16.nxv8i8.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwmulsu.mask.nxv8i16.nxv8i8.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vwmulsu_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint8m1_t op1, uint8_t op2, size_t vl) {
@@ -1587,7 +1587,7 @@ vint16m2_t test_vwmulsu_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint8m1_
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmulsu.mask.nxv16i16.nxv16i8.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmulsu.mask.nxv16i16.nxv16i8.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vwmulsu_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint8m2_t op1, vuint8m2_t op2, size_t vl) {
@@ -1597,7 +1597,7 @@ vint16m4_t test_vwmulsu_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint8m2_
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmulsu.mask.nxv16i16.nxv16i8.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwmulsu.mask.nxv16i16.nxv16i8.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vwmulsu_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint8m2_t op1, uint8_t op2, size_t vl) {
@@ -1607,7 +1607,7 @@ vint16m4_t test_vwmulsu_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint8m2_
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmulsu.mask.nxv32i16.nxv32i8.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmulsu.mask.nxv32i16.nxv32i8.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vwmulsu_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint8m4_t op1, vuint8m4_t op2, size_t vl) {
@@ -1617,7 +1617,7 @@ vint16m8_t test_vwmulsu_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint8m4_
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmulsu.mask.nxv32i16.nxv32i8.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwmulsu.mask.nxv32i16.nxv32i8.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vwmulsu_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint8m4_t op1, uint8_t op2, size_t vl) {
@@ -1627,7 +1627,7 @@ vint16m8_t test_vwmulsu_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint8m4_
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmulsu.mask.nxv1i32.nxv1i16.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmulsu.mask.nxv1i32.nxv1i16.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vwmulsu_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint16mf4_t op1, vuint16mf4_t op2, size_t vl) {
@@ -1637,7 +1637,7 @@ vint32mf2_t test_vwmulsu_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmulsu.mask.nxv1i32.nxv1i16.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwmulsu.mask.nxv1i32.nxv1i16.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vwmulsu_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint16mf4_t op1, uint16_t op2, size_t vl) {
@@ -1647,7 +1647,7 @@ vint32mf2_t test_vwmulsu_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmulsu.mask.nxv2i32.nxv2i16.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmulsu.mask.nxv2i32.nxv2i16.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwmulsu_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint16mf2_t op1, vuint16mf2_t op2, size_t vl) {
@@ -1657,7 +1657,7 @@ vint32m1_t test_vwmulsu_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmulsu.mask.nxv2i32.nxv2i16.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwmulsu.mask.nxv2i32.nxv2i16.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwmulsu_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint16mf2_t op1, uint16_t op2, size_t vl) {
@@ -1667,7 +1667,7 @@ vint32m1_t test_vwmulsu_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmulsu.mask.nxv4i32.nxv4i16.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmulsu.mask.nxv4i32.nxv4i16.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vwmulsu_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint16m1_t op1, vuint16m1_t op2, size_t vl) {
@@ -1677,7 +1677,7 @@ vint32m2_t test_vwmulsu_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmulsu.mask.nxv4i32.nxv4i16.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwmulsu.mask.nxv4i32.nxv4i16.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vwmulsu_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint16m1_t op1, uint16_t op2, size_t vl) {
@@ -1687,7 +1687,7 @@ vint32m2_t test_vwmulsu_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmulsu.mask.nxv8i32.nxv8i16.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmulsu.mask.nxv8i32.nxv8i16.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vwmulsu_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint16m2_t op1, vuint16m2_t op2, size_t vl) {
@@ -1697,7 +1697,7 @@ vint32m4_t test_vwmulsu_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint16m2
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmulsu.mask.nxv8i32.nxv8i16.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwmulsu.mask.nxv8i32.nxv8i16.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vwmulsu_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint16m2_t op1, uint16_t op2, size_t vl) {
@@ -1707,7 +1707,7 @@ vint32m4_t test_vwmulsu_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint16m2
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmulsu.mask.nxv16i32.nxv16i16.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmulsu.mask.nxv16i32.nxv16i16.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vwmulsu_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint16m4_t op1, vuint16m4_t op2, size_t vl) {
@@ -1717,7 +1717,7 @@ vint32m8_t test_vwmulsu_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint16m4
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmulsu.mask.nxv16i32.nxv16i16.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwmulsu.mask.nxv16i32.nxv16i16.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vwmulsu_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint16m4_t op1, uint16_t op2, size_t vl) {
@@ -1727,7 +1727,7 @@ vint32m8_t test_vwmulsu_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint16m4
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmulsu.mask.nxv1i64.nxv1i32.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmulsu.mask.nxv1i64.nxv1i32.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwmulsu_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint32mf2_t op1, vuint32mf2_t op2, size_t vl) {
@@ -1737,7 +1737,7 @@ vint64m1_t test_vwmulsu_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmulsu.mask.nxv1i64.nxv1i32.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwmulsu.mask.nxv1i64.nxv1i32.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwmulsu_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint32mf2_t op1, uint32_t op2, size_t vl) {
@@ -1747,7 +1747,7 @@ vint64m1_t test_vwmulsu_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmulsu.mask.nxv2i64.nxv2i32.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmulsu.mask.nxv2i64.nxv2i32.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vwmulsu_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint32m1_t op1, vuint32m1_t op2, size_t vl) {
@@ -1757,7 +1757,7 @@ vint64m2_t test_vwmulsu_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmulsu.mask.nxv2i64.nxv2i32.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwmulsu.mask.nxv2i64.nxv2i32.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vwmulsu_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint32m1_t op1, uint32_t op2, size_t vl) {
@@ -1767,7 +1767,7 @@ vint64m2_t test_vwmulsu_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmulsu.mask.nxv4i64.nxv4i32.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmulsu.mask.nxv4i64.nxv4i32.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vwmulsu_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint32m2_t op1, vuint32m2_t op2, size_t vl) {
@@ -1777,7 +1777,7 @@ vint64m4_t test_vwmulsu_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmulsu.mask.nxv4i64.nxv4i32.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwmulsu.mask.nxv4i64.nxv4i32.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vwmulsu_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint32m2_t op1, uint32_t op2, size_t vl) {
@@ -1787,7 +1787,7 @@ vint64m4_t test_vwmulsu_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmulsu.mask.nxv8i64.nxv8i32.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmulsu.mask.nxv8i64.nxv8i32.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vwmulsu_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint32m4_t op1, vuint32m4_t op2, size_t vl) {
@@ -1797,7 +1797,7 @@ vint64m8_t test_vwmulsu_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint32m4
 //
 // CHECK-RV64-LABEL: @test_vwmulsu_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmulsu.mask.nxv8i64.nxv8i32.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwmulsu.mask.nxv8i64.nxv8i32.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vwmulsu_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint32m4_t op1, uint32_t op2, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwredsum.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwredsum.c
index 5205bfa..718f688 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwredsum.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwredsum.c
@@ -403,7 +403,7 @@ vuint64m1_t test_vwredsumu_vs_u32m8_u64m1(vuint64m1_t dst, vuint32m8_t vector,
 //
 // CHECK-RV64-LABEL: @test_vwredsum_vs_i8mf8_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsum.mask.nxv4i16.nxv1i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsum.mask.nxv4i16.nxv1i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwredsum_vs_i8mf8_i16m1_m(vbool64_t mask, vint16m1_t dst,
@@ -415,7 +415,7 @@ vint16m1_t test_vwredsum_vs_i8mf8_i16m1_m(vbool64_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsum_vs_i8mf4_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsum.mask.nxv4i16.nxv2i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsum.mask.nxv4i16.nxv2i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwredsum_vs_i8mf4_i16m1_m(vbool32_t mask, vint16m1_t dst,
@@ -427,7 +427,7 @@ vint16m1_t test_vwredsum_vs_i8mf4_i16m1_m(vbool32_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsum_vs_i8mf2_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsum.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsum.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwredsum_vs_i8mf2_i16m1_m(vbool16_t mask, vint16m1_t dst,
@@ -439,7 +439,7 @@ vint16m1_t test_vwredsum_vs_i8mf2_i16m1_m(vbool16_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsum_vs_i8m1_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsum.mask.nxv4i16.nxv8i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsum.mask.nxv4i16.nxv8i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwredsum_vs_i8m1_i16m1_m(vbool8_t mask, vint16m1_t dst,
@@ -451,7 +451,7 @@ vint16m1_t test_vwredsum_vs_i8m1_i16m1_m(vbool8_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsum_vs_i8m2_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsum.mask.nxv4i16.nxv16i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsum.mask.nxv4i16.nxv16i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwredsum_vs_i8m2_i16m1_m(vbool4_t mask, vint16m1_t dst,
@@ -463,7 +463,7 @@ vint16m1_t test_vwredsum_vs_i8m2_i16m1_m(vbool4_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsum_vs_i8m4_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsum.mask.nxv4i16.nxv32i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsum.mask.nxv4i16.nxv32i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwredsum_vs_i8m4_i16m1_m(vbool2_t mask, vint16m1_t dst,
@@ -475,7 +475,7 @@ vint16m1_t test_vwredsum_vs_i8m4_i16m1_m(vbool2_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsum_vs_i8m8_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsum.mask.nxv4i16.nxv64i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsum.mask.nxv4i16.nxv64i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwredsum_vs_i8m8_i16m1_m(vbool1_t mask, vint16m1_t dst,
@@ -487,7 +487,7 @@ vint16m1_t test_vwredsum_vs_i8m8_i16m1_m(vbool1_t mask, vint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsum_vs_i16mf4_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsum.mask.nxv2i32.nxv1i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsum.mask.nxv2i32.nxv1i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwredsum_vs_i16mf4_i32m1_m(vbool64_t mask, vint32m1_t dst,
@@ -499,7 +499,7 @@ vint32m1_t test_vwredsum_vs_i16mf4_i32m1_m(vbool64_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsum_vs_i16mf2_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsum.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsum.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwredsum_vs_i16mf2_i32m1_m(vbool32_t mask, vint32m1_t dst,
@@ -511,7 +511,7 @@ vint32m1_t test_vwredsum_vs_i16mf2_i32m1_m(vbool32_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsum_vs_i16m1_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsum.mask.nxv2i32.nxv4i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsum.mask.nxv2i32.nxv4i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwredsum_vs_i16m1_i32m1_m(vbool16_t mask, vint32m1_t dst,
@@ -523,7 +523,7 @@ vint32m1_t test_vwredsum_vs_i16m1_i32m1_m(vbool16_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsum_vs_i16m2_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsum.mask.nxv2i32.nxv8i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsum.mask.nxv2i32.nxv8i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwredsum_vs_i16m2_i32m1_m(vbool8_t mask, vint32m1_t dst,
@@ -535,7 +535,7 @@ vint32m1_t test_vwredsum_vs_i16m2_i32m1_m(vbool8_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsum_vs_i16m4_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsum.mask.nxv2i32.nxv16i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsum.mask.nxv2i32.nxv16i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwredsum_vs_i16m4_i32m1_m(vbool4_t mask, vint32m1_t dst,
@@ -547,7 +547,7 @@ vint32m1_t test_vwredsum_vs_i16m4_i32m1_m(vbool4_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsum_vs_i16m8_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsum.mask.nxv2i32.nxv32i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsum.mask.nxv2i32.nxv32i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwredsum_vs_i16m8_i32m1_m(vbool2_t mask, vint32m1_t dst,
@@ -559,7 +559,7 @@ vint32m1_t test_vwredsum_vs_i16m8_i32m1_m(vbool2_t mask, vint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsum_vs_i32mf2_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwredsum.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwredsum.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwredsum_vs_i32mf2_i64m1_m(vbool64_t mask, vint64m1_t dst,
@@ -571,7 +571,7 @@ vint64m1_t test_vwredsum_vs_i32mf2_i64m1_m(vbool64_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsum_vs_i32m1_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwredsum.mask.nxv1i64.nxv2i32.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwredsum.mask.nxv1i64.nxv2i32.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwredsum_vs_i32m1_i64m1_m(vbool32_t mask, vint64m1_t dst,
@@ -583,7 +583,7 @@ vint64m1_t test_vwredsum_vs_i32m1_i64m1_m(vbool32_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsum_vs_i32m2_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwredsum.mask.nxv1i64.nxv4i32.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwredsum.mask.nxv1i64.nxv4i32.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwredsum_vs_i32m2_i64m1_m(vbool16_t mask, vint64m1_t dst,
@@ -595,7 +595,7 @@ vint64m1_t test_vwredsum_vs_i32m2_i64m1_m(vbool16_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsum_vs_i32m4_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwredsum.mask.nxv1i64.nxv8i32.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwredsum.mask.nxv1i64.nxv8i32.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwredsum_vs_i32m4_i64m1_m(vbool8_t mask, vint64m1_t dst,
@@ -607,7 +607,7 @@ vint64m1_t test_vwredsum_vs_i32m4_i64m1_m(vbool8_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsum_vs_i32m8_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwredsum.mask.nxv1i64.nxv16i32.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwredsum.mask.nxv1i64.nxv16i32.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwredsum_vs_i32m8_i64m1_m(vbool4_t mask, vint64m1_t dst,
@@ -619,7 +619,7 @@ vint64m1_t test_vwredsum_vs_i32m8_i64m1_m(vbool4_t mask, vint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsumu_vs_u8mf8_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsumu.mask.nxv4i16.nxv1i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsumu.mask.nxv4i16.nxv1i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 1 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vwredsumu_vs_u8mf8_u16m1_m(vbool64_t mask, vuint16m1_t dst,
@@ -631,7 +631,7 @@ vuint16m1_t test_vwredsumu_vs_u8mf8_u16m1_m(vbool64_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsumu_vs_u8mf4_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsumu.mask.nxv4i16.nxv2i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsumu.mask.nxv4i16.nxv2i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 2 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vwredsumu_vs_u8mf4_u16m1_m(vbool32_t mask, vuint16m1_t dst,
@@ -643,7 +643,7 @@ vuint16m1_t test_vwredsumu_vs_u8mf4_u16m1_m(vbool32_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsumu_vs_u8mf2_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsumu.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsumu.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 4 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vwredsumu_vs_u8mf2_u16m1_m(vbool16_t mask, vuint16m1_t dst,
@@ -655,7 +655,7 @@ vuint16m1_t test_vwredsumu_vs_u8mf2_u16m1_m(vbool16_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsumu_vs_u8m1_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsumu.mask.nxv4i16.nxv8i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsumu.mask.nxv4i16.nxv8i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 8 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vwredsumu_vs_u8m1_u16m1_m(vbool8_t mask, vuint16m1_t dst,
@@ -667,7 +667,7 @@ vuint16m1_t test_vwredsumu_vs_u8m1_u16m1_m(vbool8_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsumu_vs_u8m2_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsumu.mask.nxv4i16.nxv16i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsumu.mask.nxv4i16.nxv16i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 16 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vwredsumu_vs_u8m2_u16m1_m(vbool4_t mask, vuint16m1_t dst,
@@ -679,7 +679,7 @@ vuint16m1_t test_vwredsumu_vs_u8m2_u16m1_m(vbool4_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsumu_vs_u8m4_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsumu.mask.nxv4i16.nxv32i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsumu.mask.nxv4i16.nxv32i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 32 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vwredsumu_vs_u8m4_u16m1_m(vbool2_t mask, vuint16m1_t dst,
@@ -691,7 +691,7 @@ vuint16m1_t test_vwredsumu_vs_u8m4_u16m1_m(vbool2_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsumu_vs_u8m8_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsumu.mask.nxv4i16.nxv64i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwredsumu.mask.nxv4i16.nxv64i8.i64(<vscale x 4 x i16> [[DST:%.*]], <vscale x 64 x i8> [[VECTOR:%.*]], <vscale x 4 x i16> [[SCALAR:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vwredsumu_vs_u8m8_u16m1_m(vbool1_t mask, vuint16m1_t dst,
@@ -703,7 +703,7 @@ vuint16m1_t test_vwredsumu_vs_u8m8_u16m1_m(vbool1_t mask, vuint16m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsumu_vs_u16mf4_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsumu.mask.nxv2i32.nxv1i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsumu.mask.nxv2i32.nxv1i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 1 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vwredsumu_vs_u16mf4_u32m1_m(vbool64_t mask, vuint32m1_t dst,
@@ -715,7 +715,7 @@ vuint32m1_t test_vwredsumu_vs_u16mf4_u32m1_m(vbool64_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsumu_vs_u16mf2_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsumu.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsumu.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 2 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vwredsumu_vs_u16mf2_u32m1_m(vbool32_t mask, vuint32m1_t dst,
@@ -727,7 +727,7 @@ vuint32m1_t test_vwredsumu_vs_u16mf2_u32m1_m(vbool32_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsumu_vs_u16m1_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsumu.mask.nxv2i32.nxv4i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsumu.mask.nxv2i32.nxv4i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 4 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vwredsumu_vs_u16m1_u32m1_m(vbool16_t mask, vuint32m1_t dst,
@@ -739,7 +739,7 @@ vuint32m1_t test_vwredsumu_vs_u16m1_u32m1_m(vbool16_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsumu_vs_u16m2_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsumu.mask.nxv2i32.nxv8i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsumu.mask.nxv2i32.nxv8i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 8 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vwredsumu_vs_u16m2_u32m1_m(vbool8_t mask, vuint32m1_t dst,
@@ -751,7 +751,7 @@ vuint32m1_t test_vwredsumu_vs_u16m2_u32m1_m(vbool8_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsumu_vs_u16m4_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsumu.mask.nxv2i32.nxv16i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsumu.mask.nxv2i32.nxv16i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 16 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vwredsumu_vs_u16m4_u32m1_m(vbool4_t mask, vuint32m1_t dst,
@@ -763,7 +763,7 @@ vuint32m1_t test_vwredsumu_vs_u16m4_u32m1_m(vbool4_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsumu_vs_u16m8_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsumu.mask.nxv2i32.nxv32i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwredsumu.mask.nxv2i32.nxv32i16.i64(<vscale x 2 x i32> [[DST:%.*]], <vscale x 32 x i16> [[VECTOR:%.*]], <vscale x 2 x i32> [[SCALAR:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vwredsumu_vs_u16m8_u32m1_m(vbool2_t mask, vuint32m1_t dst,
@@ -775,7 +775,7 @@ vuint32m1_t test_vwredsumu_vs_u16m8_u32m1_m(vbool2_t mask, vuint32m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsumu_vs_u32mf2_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwredsumu.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwredsumu.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 1 x i32> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vwredsumu_vs_u32mf2_u64m1_m(vbool64_t mask, vuint64m1_t dst,
@@ -787,7 +787,7 @@ vuint64m1_t test_vwredsumu_vs_u32mf2_u64m1_m(vbool64_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsumu_vs_u32m1_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwredsumu.mask.nxv1i64.nxv2i32.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwredsumu.mask.nxv1i64.nxv2i32.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 2 x i32> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vwredsumu_vs_u32m1_u64m1_m(vbool32_t mask, vuint64m1_t dst,
@@ -799,7 +799,7 @@ vuint64m1_t test_vwredsumu_vs_u32m1_u64m1_m(vbool32_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsumu_vs_u32m2_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwredsumu.mask.nxv1i64.nxv4i32.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwredsumu.mask.nxv1i64.nxv4i32.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 4 x i32> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vwredsumu_vs_u32m2_u64m1_m(vbool16_t mask, vuint64m1_t dst,
@@ -811,7 +811,7 @@ vuint64m1_t test_vwredsumu_vs_u32m2_u64m1_m(vbool16_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsumu_vs_u32m4_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwredsumu.mask.nxv1i64.nxv8i32.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwredsumu.mask.nxv1i64.nxv8i32.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 8 x i32> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vwredsumu_vs_u32m4_u64m1_m(vbool8_t mask, vuint64m1_t dst,
@@ -823,7 +823,7 @@ vuint64m1_t test_vwredsumu_vs_u32m4_u64m1_m(vbool8_t mask, vuint64m1_t dst,
 //
 // CHECK-RV64-LABEL: @test_vwredsumu_vs_u32m8_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwredsumu.mask.nxv1i64.nxv16i32.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR5]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwredsumu.mask.nxv1i64.nxv16i32.i64(<vscale x 1 x i64> [[DST:%.*]], <vscale x 16 x i32> [[VECTOR:%.*]], <vscale x 1 x i64> [[SCALAR:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vwredsumu_vs_u32m8_u64m1_m(vbool4_t mask, vuint64m1_t dst,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwsub.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwsub.c
index 649f7e8..7d94594 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwsub.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vwsub.c
@@ -1215,7 +1215,7 @@ vuint64m8_t test_vwsubu_wx_u64m8(vuint64m8_t op1, uint32_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vwsub_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwsub.mask.nxv1i16.nxv1i8.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwsub.mask.nxv1i16.nxv1i8.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vwsub_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -1226,7 +1226,7 @@ vint16mf4_t test_vwsub_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwsub.mask.nxv1i16.nxv1i8.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwsub.mask.nxv1i16.nxv1i8.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vwsub_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -1237,7 +1237,7 @@ vint16mf4_t test_vwsub_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwsub.w.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwsub.w.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vwsub_wv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -1248,7 +1248,7 @@ vint16mf4_t test_vwsub_wv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwsub.w.mask.nxv1i16.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwsub.w.mask.nxv1i16.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vwsub_wx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
@@ -1259,7 +1259,7 @@ vint16mf4_t test_vwsub_wx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwsub.mask.nxv2i16.nxv2i8.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwsub.mask.nxv2i16.nxv2i8.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vwsub_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -1270,7 +1270,7 @@ vint16mf2_t test_vwsub_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwsub.mask.nxv2i16.nxv2i8.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwsub.mask.nxv2i16.nxv2i8.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vwsub_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -1281,7 +1281,7 @@ vint16mf2_t test_vwsub_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwsub.w.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwsub.w.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vwsub_wv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -1292,7 +1292,7 @@ vint16mf2_t test_vwsub_wv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwsub.w.mask.nxv2i16.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwsub.w.mask.nxv2i16.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vwsub_wx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
@@ -1303,7 +1303,7 @@ vint16mf2_t test_vwsub_wx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwsub.mask.nxv4i16.nxv4i8.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwsub.mask.nxv4i16.nxv4i8.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwsub_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -1314,7 +1314,7 @@ vint16m1_t test_vwsub_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwsub.mask.nxv4i16.nxv4i8.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwsub.mask.nxv4i16.nxv4i8.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwsub_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -1325,7 +1325,7 @@ vint16m1_t test_vwsub_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwsub.w.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwsub.w.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwsub_wv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -1336,7 +1336,7 @@ vint16m1_t test_vwsub_wv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwsub.w.mask.nxv4i16.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwsub.w.mask.nxv4i16.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vwsub_wx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
@@ -1347,7 +1347,7 @@ vint16m1_t test_vwsub_wx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwsub.mask.nxv8i16.nxv8i8.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwsub.mask.nxv8i16.nxv8i8.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vwsub_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -1358,7 +1358,7 @@ vint16m2_t test_vwsub_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwsub.mask.nxv8i16.nxv8i8.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwsub.mask.nxv8i16.nxv8i8.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vwsub_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -1369,7 +1369,7 @@ vint16m2_t test_vwsub_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwsub.w.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwsub.w.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vwsub_wv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -1380,7 +1380,7 @@ vint16m2_t test_vwsub_wv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwsub.w.mask.nxv8i16.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwsub.w.mask.nxv8i16.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vwsub_wx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
@@ -1391,7 +1391,7 @@ vint16m2_t test_vwsub_wx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwsub.mask.nxv16i16.nxv16i8.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwsub.mask.nxv16i16.nxv16i8.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vwsub_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -1402,7 +1402,7 @@ vint16m4_t test_vwsub_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwsub.mask.nxv16i16.nxv16i8.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwsub.mask.nxv16i16.nxv16i8.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vwsub_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -1413,7 +1413,7 @@ vint16m4_t test_vwsub_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwsub.w.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwsub.w.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vwsub_wv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -1424,7 +1424,7 @@ vint16m4_t test_vwsub_wv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwsub.w.mask.nxv16i16.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwsub.w.mask.nxv16i16.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vwsub_wx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
@@ -1435,7 +1435,7 @@ vint16m4_t test_vwsub_wx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwsub.mask.nxv32i16.nxv32i8.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwsub.mask.nxv32i16.nxv32i8.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vwsub_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -1446,7 +1446,7 @@ vint16m8_t test_vwsub_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwsub.mask.nxv32i16.nxv32i8.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwsub.mask.nxv32i16.nxv32i8.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vwsub_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -1457,7 +1457,7 @@ vint16m8_t test_vwsub_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwsub.w.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwsub.w.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vwsub_wv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -1468,7 +1468,7 @@ vint16m8_t test_vwsub_wv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwsub.w.mask.nxv32i16.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwsub.w.mask.nxv32i16.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vwsub_wx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
@@ -1479,7 +1479,7 @@ vint16m8_t test_vwsub_wx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwsub.mask.nxv1i32.nxv1i16.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwsub.mask.nxv1i32.nxv1i16.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vwsub_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -1491,7 +1491,7 @@ vint32mf2_t test_vwsub_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwsub.mask.nxv1i32.nxv1i16.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwsub.mask.nxv1i32.nxv1i16.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vwsub_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -1502,7 +1502,7 @@ vint32mf2_t test_vwsub_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwsub.w.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwsub.w.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vwsub_wv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -1514,7 +1514,7 @@ vint32mf2_t test_vwsub_wv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwsub.w.mask.nxv1i32.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwsub.w.mask.nxv1i32.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vwsub_wx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
@@ -1525,7 +1525,7 @@ vint32mf2_t test_vwsub_wx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwsub.mask.nxv2i32.nxv2i16.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwsub.mask.nxv2i32.nxv2i16.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwsub_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -1536,7 +1536,7 @@ vint32m1_t test_vwsub_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwsub.mask.nxv2i32.nxv2i16.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwsub.mask.nxv2i32.nxv2i16.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwsub_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -1547,7 +1547,7 @@ vint32m1_t test_vwsub_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwsub.w.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwsub.w.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwsub_wv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -1558,7 +1558,7 @@ vint32m1_t test_vwsub_wv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwsub.w.mask.nxv2i32.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwsub.w.mask.nxv2i32.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vwsub_wx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
@@ -1569,7 +1569,7 @@ vint32m1_t test_vwsub_wx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwsub.mask.nxv4i32.nxv4i16.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwsub.mask.nxv4i32.nxv4i16.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vwsub_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -1580,7 +1580,7 @@ vint32m2_t test_vwsub_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwsub.mask.nxv4i32.nxv4i16.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwsub.mask.nxv4i32.nxv4i16.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vwsub_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -1591,7 +1591,7 @@ vint32m2_t test_vwsub_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwsub.w.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwsub.w.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vwsub_wv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -1602,7 +1602,7 @@ vint32m2_t test_vwsub_wv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwsub.w.mask.nxv4i32.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwsub.w.mask.nxv4i32.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vwsub_wx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
@@ -1613,7 +1613,7 @@ vint32m2_t test_vwsub_wx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwsub.mask.nxv8i32.nxv8i16.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwsub.mask.nxv8i32.nxv8i16.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vwsub_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -1624,7 +1624,7 @@ vint32m4_t test_vwsub_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwsub.mask.nxv8i32.nxv8i16.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwsub.mask.nxv8i32.nxv8i16.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vwsub_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -1635,7 +1635,7 @@ vint32m4_t test_vwsub_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwsub.w.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwsub.w.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vwsub_wv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -1646,7 +1646,7 @@ vint32m4_t test_vwsub_wv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwsub.w.mask.nxv8i32.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwsub.w.mask.nxv8i32.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vwsub_wx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
@@ -1657,7 +1657,7 @@ vint32m4_t test_vwsub_wx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwsub.mask.nxv16i32.nxv16i16.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwsub.mask.nxv16i32.nxv16i16.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vwsub_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -1668,7 +1668,7 @@ vint32m8_t test_vwsub_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwsub.mask.nxv16i32.nxv16i16.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwsub.mask.nxv16i32.nxv16i16.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vwsub_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -1679,7 +1679,7 @@ vint32m8_t test_vwsub_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwsub.w.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwsub.w.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vwsub_wv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -1690,7 +1690,7 @@ vint32m8_t test_vwsub_wv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwsub.w.mask.nxv16i32.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwsub.w.mask.nxv16i32.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vwsub_wx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
@@ -1701,7 +1701,7 @@ vint32m8_t test_vwsub_wx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwsub.mask.nxv1i64.nxv1i32.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwsub.mask.nxv1i64.nxv1i32.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwsub_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -1712,7 +1712,7 @@ vint64m1_t test_vwsub_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwsub.mask.nxv1i64.nxv1i32.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwsub.mask.nxv1i64.nxv1i32.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwsub_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -1723,7 +1723,7 @@ vint64m1_t test_vwsub_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwsub.w.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwsub.w.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwsub_wv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -1734,7 +1734,7 @@ vint64m1_t test_vwsub_wv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwsub.w.mask.nxv1i64.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwsub.w.mask.nxv1i64.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vwsub_wx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
@@ -1745,7 +1745,7 @@ vint64m1_t test_vwsub_wx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwsub.mask.nxv2i64.nxv2i32.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwsub.mask.nxv2i64.nxv2i32.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vwsub_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -1756,7 +1756,7 @@ vint64m2_t test_vwsub_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwsub.mask.nxv2i64.nxv2i32.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwsub.mask.nxv2i64.nxv2i32.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vwsub_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -1767,7 +1767,7 @@ vint64m2_t test_vwsub_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwsub.w.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwsub.w.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vwsub_wv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -1778,7 +1778,7 @@ vint64m2_t test_vwsub_wv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwsub.w.mask.nxv2i64.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwsub.w.mask.nxv2i64.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vwsub_wx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
@@ -1789,7 +1789,7 @@ vint64m2_t test_vwsub_wx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwsub.mask.nxv4i64.nxv4i32.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwsub.mask.nxv4i64.nxv4i32.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vwsub_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -1800,7 +1800,7 @@ vint64m4_t test_vwsub_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwsub.mask.nxv4i64.nxv4i32.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwsub.mask.nxv4i64.nxv4i32.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vwsub_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -1811,7 +1811,7 @@ vint64m4_t test_vwsub_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwsub.w.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwsub.w.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vwsub_wv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -1822,7 +1822,7 @@ vint64m4_t test_vwsub_wv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwsub.w.mask.nxv4i64.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwsub.w.mask.nxv4i64.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vwsub_wx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
@@ -1833,7 +1833,7 @@ vint64m4_t test_vwsub_wx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwsub.mask.nxv8i64.nxv8i32.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwsub.mask.nxv8i64.nxv8i32.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vwsub_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -1844,7 +1844,7 @@ vint64m8_t test_vwsub_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwsub.mask.nxv8i64.nxv8i32.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwsub.mask.nxv8i64.nxv8i32.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vwsub_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -1855,7 +1855,7 @@ vint64m8_t test_vwsub_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwsub.w.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwsub.w.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vwsub_wv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -1866,7 +1866,7 @@ vint64m8_t test_vwsub_wv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsub_wx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwsub.w.mask.nxv8i64.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwsub.w.mask.nxv8i64.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vwsub_wx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
@@ -1877,7 +1877,7 @@ vint64m8_t test_vwsub_wx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwsubu.mask.nxv1i16.nxv1i8.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwsubu.mask.nxv1i16.nxv1i8.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vwsubu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -1889,7 +1889,7 @@ vuint16mf4_t test_vwsubu_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwsubu.mask.nxv1i16.nxv1i8.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwsubu.mask.nxv1i16.nxv1i8.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vwsubu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -1900,7 +1900,7 @@ vuint16mf4_t test_vwsubu_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwsubu.w.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwsubu.w.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vwsubu_wv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -1912,7 +1912,7 @@ vuint16mf4_t test_vwsubu_wv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwsubu.w.mask.nxv1i16.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vwsubu.w.mask.nxv1i16.i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vwsubu_wx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -1923,7 +1923,7 @@ vuint16mf4_t test_vwsubu_wx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwsubu.mask.nxv2i16.nxv2i8.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwsubu.mask.nxv2i16.nxv2i8.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vwsubu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -1935,7 +1935,7 @@ vuint16mf2_t test_vwsubu_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwsubu.mask.nxv2i16.nxv2i8.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwsubu.mask.nxv2i16.nxv2i8.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vwsubu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -1946,7 +1946,7 @@ vuint16mf2_t test_vwsubu_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwsubu.w.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwsubu.w.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vwsubu_wv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -1958,7 +1958,7 @@ vuint16mf2_t test_vwsubu_wv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwsubu.w.mask.nxv2i16.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vwsubu.w.mask.nxv2i16.i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vwsubu_wx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -1969,7 +1969,7 @@ vuint16mf2_t test_vwsubu_wx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwsubu.mask.nxv4i16.nxv4i8.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwsubu.mask.nxv4i16.nxv4i8.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vwsubu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -1981,7 +1981,7 @@ vuint16m1_t test_vwsubu_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwsubu.mask.nxv4i16.nxv4i8.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwsubu.mask.nxv4i16.nxv4i8.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vwsubu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -1992,7 +1992,7 @@ vuint16m1_t test_vwsubu_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwsubu.w.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwsubu.w.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vwsubu_wv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -2004,7 +2004,7 @@ vuint16m1_t test_vwsubu_wv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwsubu.w.mask.nxv4i16.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vwsubu.w.mask.nxv4i16.i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vwsubu_wx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -2015,7 +2015,7 @@ vuint16m1_t test_vwsubu_wx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwsubu.mask.nxv8i16.nxv8i8.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwsubu.mask.nxv8i16.nxv8i8.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vwsubu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -2026,7 +2026,7 @@ vuint16m2_t test_vwsubu_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwsubu.mask.nxv8i16.nxv8i8.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwsubu.mask.nxv8i16.nxv8i8.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vwsubu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -2037,7 +2037,7 @@ vuint16m2_t test_vwsubu_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwsubu.w.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwsubu.w.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vwsubu_wv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -2048,7 +2048,7 @@ vuint16m2_t test_vwsubu_wv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwsubu.w.mask.nxv8i16.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vwsubu.w.mask.nxv8i16.i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vwsubu_wx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -2059,7 +2059,7 @@ vuint16m2_t test_vwsubu_wx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwsubu.mask.nxv16i16.nxv16i8.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwsubu.mask.nxv16i16.nxv16i8.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vwsubu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -2070,7 +2070,7 @@ vuint16m4_t test_vwsubu_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwsubu.mask.nxv16i16.nxv16i8.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwsubu.mask.nxv16i16.nxv16i8.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vwsubu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -2081,7 +2081,7 @@ vuint16m4_t test_vwsubu_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwsubu.w.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwsubu.w.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vwsubu_wv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -2092,7 +2092,7 @@ vuint16m4_t test_vwsubu_wv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwsubu.w.mask.nxv16i16.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vwsubu.w.mask.nxv16i16.i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vwsubu_wx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -2103,7 +2103,7 @@ vuint16m4_t test_vwsubu_wx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwsubu.mask.nxv32i16.nxv32i8.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwsubu.mask.nxv32i16.nxv32i8.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vwsubu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -2114,7 +2114,7 @@ vuint16m8_t test_vwsubu_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwsubu.mask.nxv32i16.nxv32i8.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwsubu.mask.nxv32i16.nxv32i8.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vwsubu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -2125,7 +2125,7 @@ vuint16m8_t test_vwsubu_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwsubu.w.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwsubu.w.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vwsubu_wv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -2136,7 +2136,7 @@ vuint16m8_t test_vwsubu_wv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwsubu.w.mask.nxv32i16.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vwsubu.w.mask.nxv32i16.i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vwsubu_wx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -2147,7 +2147,7 @@ vuint16m8_t test_vwsubu_wx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwsubu.mask.nxv1i32.nxv1i16.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwsubu.mask.nxv1i32.nxv1i16.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vwsubu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -2159,7 +2159,7 @@ vuint32mf2_t test_vwsubu_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwsubu.mask.nxv1i32.nxv1i16.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwsubu.mask.nxv1i32.nxv1i16.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vwsubu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -2171,7 +2171,7 @@ vuint32mf2_t test_vwsubu_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwsubu.w.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwsubu.w.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vwsubu_wv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -2183,7 +2183,7 @@ vuint32mf2_t test_vwsubu_wv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwsubu.w.mask.nxv1i32.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vwsubu.w.mask.nxv1i32.i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vwsubu_wx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -2195,7 +2195,7 @@ vuint32mf2_t test_vwsubu_wx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwsubu.mask.nxv2i32.nxv2i16.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwsubu.mask.nxv2i32.nxv2i16.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vwsubu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -2207,7 +2207,7 @@ vuint32m1_t test_vwsubu_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwsubu.mask.nxv2i32.nxv2i16.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwsubu.mask.nxv2i32.nxv2i16.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vwsubu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -2218,7 +2218,7 @@ vuint32m1_t test_vwsubu_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwsubu.w.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwsubu.w.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vwsubu_wv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -2230,7 +2230,7 @@ vuint32m1_t test_vwsubu_wv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwsubu.w.mask.nxv2i32.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vwsubu.w.mask.nxv2i32.i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vwsubu_wx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -2241,7 +2241,7 @@ vuint32m1_t test_vwsubu_wx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwsubu.mask.nxv4i32.nxv4i16.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwsubu.mask.nxv4i32.nxv4i16.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vwsubu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -2253,7 +2253,7 @@ vuint32m2_t test_vwsubu_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwsubu.mask.nxv4i32.nxv4i16.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwsubu.mask.nxv4i32.nxv4i16.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vwsubu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -2264,7 +2264,7 @@ vuint32m2_t test_vwsubu_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwsubu.w.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwsubu.w.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vwsubu_wv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -2276,7 +2276,7 @@ vuint32m2_t test_vwsubu_wv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwsubu.w.mask.nxv4i32.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vwsubu.w.mask.nxv4i32.i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vwsubu_wx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -2287,7 +2287,7 @@ vuint32m2_t test_vwsubu_wx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwsubu.mask.nxv8i32.nxv8i16.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwsubu.mask.nxv8i32.nxv8i16.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vwsubu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -2299,7 +2299,7 @@ vuint32m4_t test_vwsubu_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwsubu.mask.nxv8i32.nxv8i16.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwsubu.mask.nxv8i32.nxv8i16.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vwsubu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -2310,7 +2310,7 @@ vuint32m4_t test_vwsubu_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwsubu.w.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwsubu.w.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vwsubu_wv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -2322,7 +2322,7 @@ vuint32m4_t test_vwsubu_wv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwsubu.w.mask.nxv8i32.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vwsubu.w.mask.nxv8i32.i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vwsubu_wx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -2333,7 +2333,7 @@ vuint32m4_t test_vwsubu_wx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwsubu.mask.nxv16i32.nxv16i16.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwsubu.mask.nxv16i32.nxv16i16.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vwsubu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -2345,7 +2345,7 @@ vuint32m8_t test_vwsubu_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwsubu.mask.nxv16i32.nxv16i16.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwsubu.mask.nxv16i32.nxv16i16.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vwsubu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -2356,7 +2356,7 @@ vuint32m8_t test_vwsubu_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwsubu.w.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwsubu.w.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vwsubu_wv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -2368,7 +2368,7 @@ vuint32m8_t test_vwsubu_wv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwsubu.w.mask.nxv16i32.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vwsubu.w.mask.nxv16i32.i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vwsubu_wx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -2379,7 +2379,7 @@ vuint32m8_t test_vwsubu_wx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwsubu.mask.nxv1i64.nxv1i32.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwsubu.mask.nxv1i64.nxv1i32.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vwsubu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -2391,7 +2391,7 @@ vuint64m1_t test_vwsubu_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwsubu.mask.nxv1i64.nxv1i32.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwsubu.mask.nxv1i64.nxv1i32.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vwsubu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -2402,7 +2402,7 @@ vuint64m1_t test_vwsubu_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwsubu.w.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwsubu.w.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vwsubu_wv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -2414,7 +2414,7 @@ vuint64m1_t test_vwsubu_wv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwsubu.w.mask.nxv1i64.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vwsubu.w.mask.nxv1i64.i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vwsubu_wx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -2425,7 +2425,7 @@ vuint64m1_t test_vwsubu_wx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwsubu.mask.nxv2i64.nxv2i32.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwsubu.mask.nxv2i64.nxv2i32.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vwsubu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -2437,7 +2437,7 @@ vuint64m2_t test_vwsubu_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwsubu.mask.nxv2i64.nxv2i32.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwsubu.mask.nxv2i64.nxv2i32.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vwsubu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -2448,7 +2448,7 @@ vuint64m2_t test_vwsubu_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwsubu.w.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwsubu.w.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vwsubu_wv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -2460,7 +2460,7 @@ vuint64m2_t test_vwsubu_wv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwsubu.w.mask.nxv2i64.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vwsubu.w.mask.nxv2i64.i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vwsubu_wx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -2471,7 +2471,7 @@ vuint64m2_t test_vwsubu_wx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwsubu.mask.nxv4i64.nxv4i32.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwsubu.mask.nxv4i64.nxv4i32.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vwsubu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -2483,7 +2483,7 @@ vuint64m4_t test_vwsubu_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwsubu.mask.nxv4i64.nxv4i32.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwsubu.mask.nxv4i64.nxv4i32.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vwsubu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -2494,7 +2494,7 @@ vuint64m4_t test_vwsubu_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwsubu.w.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwsubu.w.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vwsubu_wv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -2506,7 +2506,7 @@ vuint64m4_t test_vwsubu_wv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwsubu.w.mask.nxv4i64.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vwsubu.w.mask.nxv4i64.i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vwsubu_wx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -2517,7 +2517,7 @@ vuint64m4_t test_vwsubu_wx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwsubu.mask.nxv8i64.nxv8i32.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwsubu.mask.nxv8i64.nxv8i32.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vwsubu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
@@ -2529,7 +2529,7 @@ vuint64m8_t test_vwsubu_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwsubu.mask.nxv8i64.nxv8i32.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwsubu.mask.nxv8i64.nxv8i32.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vwsubu_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
@@ -2540,7 +2540,7 @@ vuint64m8_t test_vwsubu_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwsubu.w.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwsubu.w.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vwsubu_wv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
@@ -2552,7 +2552,7 @@ vuint64m8_t test_vwsubu_wv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vwsubu_wx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwsubu.w.mask.nxv8i64.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vwsubu.w.mask.nxv8i64.i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vwsubu_wx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vxor.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vxor.c
index 4a48cc5..047093a 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vxor.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vxor.c
@@ -887,7 +887,7 @@ vuint64m8_t test_vxor_vx_u64m8(vuint64m8_t op1, uint64_t op2, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vxor.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vxor.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vxor_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, vint8mf8_t op2, size_t vl) {
@@ -897,7 +897,7 @@ vint8mf8_t test_vxor_vv_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vxor.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vxor.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vint8mf8_t test_vxor_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t op1, int8_t op2, size_t vl) {
@@ -907,7 +907,7 @@ vint8mf8_t test_vxor_vx_i8mf8_m(vbool64_t mask, vint8mf8_t maskedoff, vint8mf8_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vxor.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vxor.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vxor_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, vint8mf4_t op2, size_t vl) {
@@ -917,7 +917,7 @@ vint8mf4_t test_vxor_vv_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vxor.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vxor.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vint8mf4_t test_vxor_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t op1, int8_t op2, size_t vl) {
@@ -927,7 +927,7 @@ vint8mf4_t test_vxor_vx_i8mf4_m(vbool32_t mask, vint8mf4_t maskedoff, vint8mf4_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vxor.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vxor.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vxor_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, vint8mf2_t op2, size_t vl) {
@@ -937,7 +937,7 @@ vint8mf2_t test_vxor_vv_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vxor.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vxor.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vint8mf2_t test_vxor_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t op1, int8_t op2, size_t vl) {
@@ -947,7 +947,7 @@ vint8mf2_t test_vxor_vx_i8mf2_m(vbool16_t mask, vint8mf2_t maskedoff, vint8mf2_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vxor.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vxor.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vxor_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, vint8m1_t op2, size_t vl) {
@@ -957,7 +957,7 @@ vint8m1_t test_vxor_vv_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vxor.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vxor.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vint8m1_t test_vxor_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1, int8_t op2, size_t vl) {
@@ -967,7 +967,7 @@ vint8m1_t test_vxor_vx_i8m1_m(vbool8_t mask, vint8m1_t maskedoff, vint8m1_t op1,
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vxor.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vxor.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vxor_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, vint8m2_t op2, size_t vl) {
@@ -977,7 +977,7 @@ vint8m2_t test_vxor_vv_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vxor.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vxor.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vint8m2_t test_vxor_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1, int8_t op2, size_t vl) {
@@ -987,7 +987,7 @@ vint8m2_t test_vxor_vx_i8m2_m(vbool4_t mask, vint8m2_t maskedoff, vint8m2_t op1,
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vxor.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vxor.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vxor_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, vint8m4_t op2, size_t vl) {
@@ -997,7 +997,7 @@ vint8m4_t test_vxor_vv_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vxor.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vxor.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vint8m4_t test_vxor_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1, int8_t op2, size_t vl) {
@@ -1007,7 +1007,7 @@ vint8m4_t test_vxor_vx_i8m4_m(vbool2_t mask, vint8m4_t maskedoff, vint8m4_t op1,
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vxor.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vxor.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vxor_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, vint8m8_t op2, size_t vl) {
@@ -1017,7 +1017,7 @@ vint8m8_t test_vxor_vv_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vxor.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vxor.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vint8m8_t test_vxor_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1, int8_t op2, size_t vl) {
@@ -1027,7 +1027,7 @@ vint8m8_t test_vxor_vx_i8m8_m(vbool1_t mask, vint8m8_t maskedoff, vint8m8_t op1,
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vxor.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vxor.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vxor_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, vint16mf4_t op2, size_t vl) {
@@ -1037,7 +1037,7 @@ vint16mf4_t test_vxor_vv_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vxor.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vxor.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vint16mf4_t test_vxor_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16mf4_t op1, int16_t op2, size_t vl) {
@@ -1047,7 +1047,7 @@ vint16mf4_t test_vxor_vx_i16mf4_m(vbool64_t mask, vint16mf4_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vxor.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vxor.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vxor_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, vint16mf2_t op2, size_t vl) {
@@ -1057,7 +1057,7 @@ vint16mf2_t test_vxor_vv_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vxor.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vxor.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vint16mf2_t test_vxor_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16mf2_t op1, int16_t op2, size_t vl) {
@@ -1067,7 +1067,7 @@ vint16mf2_t test_vxor_vx_i16mf2_m(vbool32_t mask, vint16mf2_t maskedoff, vint16m
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vxor.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vxor.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vxor_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, vint16m1_t op2, size_t vl) {
@@ -1077,7 +1077,7 @@ vint16m1_t test_vxor_vv_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vxor.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vxor.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vint16m1_t test_vxor_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t op1, int16_t op2, size_t vl) {
@@ -1087,7 +1087,7 @@ vint16m1_t test_vxor_vx_i16m1_m(vbool16_t mask, vint16m1_t maskedoff, vint16m1_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vxor.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vxor.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vxor_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, vint16m2_t op2, size_t vl) {
@@ -1097,7 +1097,7 @@ vint16m2_t test_vxor_vv_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vxor.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vxor.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vint16m2_t test_vxor_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t op1, int16_t op2, size_t vl) {
@@ -1107,7 +1107,7 @@ vint16m2_t test_vxor_vx_i16m2_m(vbool8_t mask, vint16m2_t maskedoff, vint16m2_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vxor.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vxor.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vxor_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, vint16m4_t op2, size_t vl) {
@@ -1117,7 +1117,7 @@ vint16m4_t test_vxor_vv_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vxor.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vxor.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vint16m4_t test_vxor_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t op1, int16_t op2, size_t vl) {
@@ -1127,7 +1127,7 @@ vint16m4_t test_vxor_vx_i16m4_m(vbool4_t mask, vint16m4_t maskedoff, vint16m4_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vxor.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vxor.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vxor_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, vint16m8_t op2, size_t vl) {
@@ -1137,7 +1137,7 @@ vint16m8_t test_vxor_vv_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vxor.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vxor.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vint16m8_t test_vxor_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t op1, int16_t op2, size_t vl) {
@@ -1147,7 +1147,7 @@ vint16m8_t test_vxor_vx_i16m8_m(vbool2_t mask, vint16m8_t maskedoff, vint16m8_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vxor.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vxor.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vxor_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, vint32mf2_t op2, size_t vl) {
@@ -1157,7 +1157,7 @@ vint32mf2_t test_vxor_vv_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vxor.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vxor.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vint32mf2_t test_vxor_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32mf2_t op1, int32_t op2, size_t vl) {
@@ -1167,7 +1167,7 @@ vint32mf2_t test_vxor_vx_i32mf2_m(vbool64_t mask, vint32mf2_t maskedoff, vint32m
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vxor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vxor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vxor_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, vint32m1_t op2, size_t vl) {
@@ -1177,7 +1177,7 @@ vint32m1_t test_vxor_vv_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vxor.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vxor.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vint32m1_t test_vxor_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t op1, int32_t op2, size_t vl) {
@@ -1187,7 +1187,7 @@ vint32m1_t test_vxor_vx_i32m1_m(vbool32_t mask, vint32m1_t maskedoff, vint32m1_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vxor.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vxor.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vxor_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, vint32m2_t op2, size_t vl) {
@@ -1197,7 +1197,7 @@ vint32m2_t test_vxor_vv_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vxor.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vxor.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vint32m2_t test_vxor_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t op1, int32_t op2, size_t vl) {
@@ -1207,7 +1207,7 @@ vint32m2_t test_vxor_vx_i32m2_m(vbool16_t mask, vint32m2_t maskedoff, vint32m2_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vxor.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vxor.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vxor_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, vint32m4_t op2, size_t vl) {
@@ -1217,7 +1217,7 @@ vint32m4_t test_vxor_vv_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vxor.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vxor.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vint32m4_t test_vxor_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t op1, int32_t op2, size_t vl) {
@@ -1227,7 +1227,7 @@ vint32m4_t test_vxor_vx_i32m4_m(vbool8_t mask, vint32m4_t maskedoff, vint32m4_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vxor.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vxor.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vxor_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, vint32m8_t op2, size_t vl) {
@@ -1237,7 +1237,7 @@ vint32m8_t test_vxor_vv_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vxor.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vxor.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vint32m8_t test_vxor_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t op1, int32_t op2, size_t vl) {
@@ -1247,7 +1247,7 @@ vint32m8_t test_vxor_vx_i32m8_m(vbool4_t mask, vint32m8_t maskedoff, vint32m8_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vxor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vxor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vxor_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, vint64m1_t op2, size_t vl) {
@@ -1257,7 +1257,7 @@ vint64m1_t test_vxor_vv_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vxor.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vxor.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vint64m1_t test_vxor_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t op1, int64_t op2, size_t vl) {
@@ -1267,7 +1267,7 @@ vint64m1_t test_vxor_vx_i64m1_m(vbool64_t mask, vint64m1_t maskedoff, vint64m1_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vxor.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vxor.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vxor_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, vint64m2_t op2, size_t vl) {
@@ -1277,7 +1277,7 @@ vint64m2_t test_vxor_vv_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vxor.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vxor.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vint64m2_t test_vxor_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t op1, int64_t op2, size_t vl) {
@@ -1287,7 +1287,7 @@ vint64m2_t test_vxor_vx_i64m2_m(vbool32_t mask, vint64m2_t maskedoff, vint64m2_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vxor.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vxor.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vxor_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, vint64m4_t op2, size_t vl) {
@@ -1297,7 +1297,7 @@ vint64m4_t test_vxor_vv_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vxor.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vxor.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vint64m4_t test_vxor_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t op1, int64_t op2, size_t vl) {
@@ -1307,7 +1307,7 @@ vint64m4_t test_vxor_vx_i64m4_m(vbool16_t mask, vint64m4_t maskedoff, vint64m4_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vxor.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vxor.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vxor_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, vint64m8_t op2, size_t vl) {
@@ -1317,7 +1317,7 @@ vint64m8_t test_vxor_vv_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_i64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vxor.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vxor.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vint64m8_t test_vxor_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t op1, int64_t op2, size_t vl) {
@@ -1327,7 +1327,7 @@ vint64m8_t test_vxor_vx_i64m8_m(vbool8_t mask, vint64m8_t maskedoff, vint64m8_t
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vxor.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vxor.mask.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i8> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vxor_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, vuint8mf8_t op2, size_t vl) {
@@ -1337,7 +1337,7 @@ vuint8mf8_t test_vxor_vv_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u8mf8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vxor.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vxor.mask.nxv1i8.i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
 //
 vuint8mf8_t test_vxor_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf8_t op1, uint8_t op2, size_t vl) {
@@ -1347,7 +1347,7 @@ vuint8mf8_t test_vxor_vx_u8mf8_m(vbool64_t mask, vuint8mf8_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vxor.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vxor.mask.nxv2i8.nxv2i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i8> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vxor_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, vuint8mf4_t op2, size_t vl) {
@@ -1357,7 +1357,7 @@ vuint8mf4_t test_vxor_vv_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u8mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vxor.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i8> @llvm.riscv.vxor.mask.nxv2i8.i8.i64(<vscale x 2 x i8> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i8> [[TMP0]]
 //
 vuint8mf4_t test_vxor_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf4_t op1, uint8_t op2, size_t vl) {
@@ -1367,7 +1367,7 @@ vuint8mf4_t test_vxor_vx_u8mf4_m(vbool32_t mask, vuint8mf4_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vxor.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vxor.mask.nxv4i8.nxv4i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i8> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vxor_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, vuint8mf2_t op2, size_t vl) {
@@ -1377,7 +1377,7 @@ vuint8mf2_t test_vxor_vv_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u8mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vxor.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i8> @llvm.riscv.vxor.mask.nxv4i8.i8.i64(<vscale x 4 x i8> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i8> [[TMP0]]
 //
 vuint8mf2_t test_vxor_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf2_t op1, uint8_t op2, size_t vl) {
@@ -1387,7 +1387,7 @@ vuint8mf2_t test_vxor_vx_u8mf2_m(vbool16_t mask, vuint8mf2_t maskedoff, vuint8mf
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vxor.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vxor.mask.nxv8i8.nxv8i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i8> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vxor_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, vuint8m1_t op2, size_t vl) {
@@ -1397,7 +1397,7 @@ vuint8m1_t test_vxor_vv_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t o
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u8m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vxor.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i8> @llvm.riscv.vxor.mask.nxv8i8.i8.i64(<vscale x 8 x i8> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i8> [[TMP0]]
 //
 vuint8m1_t test_vxor_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t op1, uint8_t op2, size_t vl) {
@@ -1407,7 +1407,7 @@ vuint8m1_t test_vxor_vx_u8m1_m(vbool8_t mask, vuint8m1_t maskedoff, vuint8m1_t o
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vxor.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vxor.mask.nxv16i8.nxv16i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vxor_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, vuint8m2_t op2, size_t vl) {
@@ -1417,7 +1417,7 @@ vuint8m2_t test_vxor_vv_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t o
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u8m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vxor.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i8> @llvm.riscv.vxor.mask.nxv16i8.i8.i64(<vscale x 16 x i8> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i8> [[TMP0]]
 //
 vuint8m2_t test_vxor_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t op1, uint8_t op2, size_t vl) {
@@ -1427,7 +1427,7 @@ vuint8m2_t test_vxor_vx_u8m2_m(vbool4_t mask, vuint8m2_t maskedoff, vuint8m2_t o
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vxor.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vxor.mask.nxv32i8.nxv32i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i8> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vxor_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, vuint8m4_t op2, size_t vl) {
@@ -1437,7 +1437,7 @@ vuint8m4_t test_vxor_vv_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t o
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u8m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vxor.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i8> @llvm.riscv.vxor.mask.nxv32i8.i8.i64(<vscale x 32 x i8> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i8> [[TMP0]]
 //
 vuint8m4_t test_vxor_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t op1, uint8_t op2, size_t vl) {
@@ -1447,7 +1447,7 @@ vuint8m4_t test_vxor_vx_u8m4_m(vbool2_t mask, vuint8m4_t maskedoff, vuint8m4_t o
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vxor.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vxor.mask.nxv64i8.nxv64i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], <vscale x 64 x i8> [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vxor_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, vuint8m8_t op2, size_t vl) {
@@ -1457,7 +1457,7 @@ vuint8m8_t test_vxor_vv_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t o
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u8m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vxor.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 64 x i8> @llvm.riscv.vxor.mask.nxv64i8.i8.i64(<vscale x 64 x i8> [[MASKEDOFF:%.*]], <vscale x 64 x i8> [[OP1:%.*]], i8 [[OP2:%.*]], <vscale x 64 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 64 x i8> [[TMP0]]
 //
 vuint8m8_t test_vxor_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t op1, uint8_t op2, size_t vl) {
@@ -1467,7 +1467,7 @@ vuint8m8_t test_vxor_vx_u8m8_m(vbool1_t mask, vuint8m8_t maskedoff, vuint8m8_t o
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vxor.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vxor.mask.nxv1i16.nxv1i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i16> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vxor_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, vuint16mf4_t op2, size_t vl) {
@@ -1477,7 +1477,7 @@ vuint16mf4_t test_vxor_vv_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vxor.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vxor.mask.nxv1i16.i16.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vxor_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint16mf4_t op1, uint16_t op2, size_t vl) {
@@ -1487,7 +1487,7 @@ vuint16mf4_t test_vxor_vx_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vxor.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vxor.mask.nxv2i16.nxv2i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i16> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vxor_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, vuint16mf2_t op2, size_t vl) {
@@ -1497,7 +1497,7 @@ vuint16mf2_t test_vxor_vv_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vxor.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vxor.mask.nxv2i16.i16.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vxor_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint16mf2_t op1, uint16_t op2, size_t vl) {
@@ -1507,7 +1507,7 @@ vuint16mf2_t test_vxor_vx_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vxor.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vxor.mask.nxv4i16.nxv4i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i16> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vxor_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, vuint16m1_t op2, size_t vl) {
@@ -1517,7 +1517,7 @@ vuint16m1_t test_vxor_vv_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vxor.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vxor.mask.nxv4i16.i16.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vxor_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m1_t op1, uint16_t op2, size_t vl) {
@@ -1527,7 +1527,7 @@ vuint16m1_t test_vxor_vx_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff, vuint16m
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vxor.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vxor.mask.nxv8i16.nxv8i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vxor_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, vuint16m2_t op2, size_t vl) {
@@ -1537,7 +1537,7 @@ vuint16m2_t test_vxor_vv_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vxor.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vxor.mask.nxv8i16.i16.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vxor_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2_t op1, uint16_t op2, size_t vl) {
@@ -1547,7 +1547,7 @@ vuint16m2_t test_vxor_vx_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff, vuint16m2
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vxor.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vxor.mask.nxv16i16.nxv16i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i16> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vxor_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, vuint16m4_t op2, size_t vl) {
@@ -1557,7 +1557,7 @@ vuint16m4_t test_vxor_vv_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vxor.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vxor.mask.nxv16i16.i16.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vxor_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4_t op1, uint16_t op2, size_t vl) {
@@ -1567,7 +1567,7 @@ vuint16m4_t test_vxor_vx_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff, vuint16m4
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vxor.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vxor.mask.nxv32i16.nxv32i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], <vscale x 32 x i16> [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vxor_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, vuint16m8_t op2, size_t vl) {
@@ -1577,7 +1577,7 @@ vuint16m8_t test_vxor_vv_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vxor.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vxor.mask.nxv32i16.i16.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i16> [[OP1:%.*]], i16 [[OP2:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vxor_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8_t op1, uint16_t op2, size_t vl) {
@@ -1587,7 +1587,7 @@ vuint16m8_t test_vxor_vx_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff, vuint16m8
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vxor.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vxor.mask.nxv1i32.nxv1i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i32> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vxor_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, vuint32mf2_t op2, size_t vl) {
@@ -1597,7 +1597,7 @@ vuint32mf2_t test_vxor_vv_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vxor.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vxor.mask.nxv1i32.i32.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vxor_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint32mf2_t op1, uint32_t op2, size_t vl) {
@@ -1607,7 +1607,7 @@ vuint32mf2_t test_vxor_vx_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff, vuint
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vxor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vxor.mask.nxv2i32.nxv2i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i32> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vxor_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, vuint32m1_t op2, size_t vl) {
@@ -1617,7 +1617,7 @@ vuint32m1_t test_vxor_vv_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vxor.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vxor.mask.nxv2i32.i32.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vxor_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m1_t op1, uint32_t op2, size_t vl) {
@@ -1627,7 +1627,7 @@ vuint32m1_t test_vxor_vx_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vxor.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vxor.mask.nxv4i32.nxv4i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vxor_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, vuint32m2_t op2, size_t vl) {
@@ -1637,7 +1637,7 @@ vuint32m2_t test_vxor_vv_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vxor.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vxor.mask.nxv4i32.i32.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vxor_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m2_t op1, uint32_t op2, size_t vl) {
@@ -1647,7 +1647,7 @@ vuint32m2_t test_vxor_vx_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff, vuint32m
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vxor.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vxor.mask.nxv8i32.nxv8i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i32> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vxor_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, vuint32m4_t op2, size_t vl) {
@@ -1657,7 +1657,7 @@ vuint32m4_t test_vxor_vv_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vxor.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vxor.mask.nxv8i32.i32.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vxor_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4_t op1, uint32_t op2, size_t vl) {
@@ -1667,7 +1667,7 @@ vuint32m4_t test_vxor_vx_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff, vuint32m4
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vxor.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vxor.mask.nxv16i32.nxv16i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], <vscale x 16 x i32> [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vxor_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, vuint32m8_t op2, size_t vl) {
@@ -1677,7 +1677,7 @@ vuint32m8_t test_vxor_vv_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vxor.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vxor.mask.nxv16i32.i32.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i32> [[OP1:%.*]], i32 [[OP2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vxor_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8_t op1, uint32_t op2, size_t vl) {
@@ -1687,7 +1687,7 @@ vuint32m8_t test_vxor_vx_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff, vuint32m8
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vxor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vxor.mask.nxv1i64.nxv1i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], <vscale x 1 x i64> [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vxor_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, vuint64m1_t op2, size_t vl) {
@@ -1697,7 +1697,7 @@ vuint64m1_t test_vxor_vv_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vxor.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vxor.mask.nxv1i64.i64.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vxor_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m1_t op1, uint64_t op2, size_t vl) {
@@ -1707,7 +1707,7 @@ vuint64m1_t test_vxor_vx_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vxor.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vxor.mask.nxv2i64.nxv2i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vxor_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, vuint64m2_t op2, size_t vl) {
@@ -1717,7 +1717,7 @@ vuint64m2_t test_vxor_vv_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vxor.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vxor.mask.nxv2i64.i64.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vxor_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m2_t op1, uint64_t op2, size_t vl) {
@@ -1727,7 +1727,7 @@ vuint64m2_t test_vxor_vx_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vxor.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vxor.mask.nxv4i64.nxv4i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], <vscale x 4 x i64> [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vxor_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, vuint64m4_t op2, size_t vl) {
@@ -1737,7 +1737,7 @@ vuint64m4_t test_vxor_vv_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vxor.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vxor.mask.nxv4i64.i64.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vxor_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m4_t op1, uint64_t op2, size_t vl) {
@@ -1747,7 +1747,7 @@ vuint64m4_t test_vxor_vx_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff, vuint64m
 //
 // CHECK-RV64-LABEL: @test_vxor_vv_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vxor.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vxor.mask.nxv8i64.nxv8i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], <vscale x 8 x i64> [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vxor_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, vuint64m8_t op2, size_t vl) {
@@ -1757,7 +1757,7 @@ vuint64m8_t test_vxor_vv_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8
 //
 // CHECK-RV64-LABEL: @test_vxor_vx_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vxor.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR8]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vxor.mask.nxv8i64.i64.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i64> [[OP1:%.*]], i64 [[OP2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vxor_vx_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff, vuint64m8_t op1, uint64_t op2, size_t vl) {
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vzext.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vzext.c
index 299025b..05a60c5 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vzext.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vzext.c
@@ -287,7 +287,7 @@ vuint64m8_t test_vzext_vf2_u64m8(vuint32m4_t op1, size_t vl) {
 //
 // CHECK-RV64-LABEL: @test_vzext_vf2_u16mf4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vzext.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7:[0-9]+]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i16> @llvm.riscv.vzext.mask.nxv1i16.nxv1i8.i64(<vscale x 1 x i16> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i16> [[TMP0]]
 //
 vuint16mf4_t test_vzext_vf2_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
@@ -298,7 +298,7 @@ vuint16mf4_t test_vzext_vf2_u16mf4_m(vbool64_t mask, vuint16mf4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf2_u16mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vzext.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i16> @llvm.riscv.vzext.mask.nxv2i16.nxv2i8.i64(<vscale x 2 x i16> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i16> [[TMP0]]
 //
 vuint16mf2_t test_vzext_vf2_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
@@ -309,7 +309,7 @@ vuint16mf2_t test_vzext_vf2_u16mf2_m(vbool32_t mask, vuint16mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf2_u16m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vzext.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i16> @llvm.riscv.vzext.mask.nxv4i16.nxv4i8.i64(<vscale x 4 x i16> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i16> [[TMP0]]
 //
 vuint16m1_t test_vzext_vf2_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
@@ -320,7 +320,7 @@ vuint16m1_t test_vzext_vf2_u16m1_m(vbool16_t mask, vuint16m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf2_u16m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vzext.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i16> @llvm.riscv.vzext.mask.nxv8i16.nxv8i8.i64(<vscale x 8 x i16> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i16> [[TMP0]]
 //
 vuint16m2_t test_vzext_vf2_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
@@ -331,7 +331,7 @@ vuint16m2_t test_vzext_vf2_u16m2_m(vbool8_t mask, vuint16m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf2_u16m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vzext.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i16> @llvm.riscv.vzext.mask.nxv16i16.nxv16i8.i64(<vscale x 16 x i16> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i16> [[TMP0]]
 //
 vuint16m4_t test_vzext_vf2_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
@@ -342,7 +342,7 @@ vuint16m4_t test_vzext_vf2_u16m4_m(vbool4_t mask, vuint16m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf2_u16m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vzext.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 32 x i16> @llvm.riscv.vzext.mask.nxv32i16.nxv32i8.i64(<vscale x 32 x i16> [[MASKEDOFF:%.*]], <vscale x 32 x i8> [[OP1:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 32 x i16> [[TMP0]]
 //
 vuint16m8_t test_vzext_vf2_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
@@ -353,7 +353,7 @@ vuint16m8_t test_vzext_vf2_u16m8_m(vbool2_t mask, vuint16m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf4_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vzext.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vzext.mask.nxv1i32.nxv1i8.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vzext_vf4_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -364,7 +364,7 @@ vuint32mf2_t test_vzext_vf4_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf4_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vzext.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vzext.mask.nxv2i32.nxv2i8.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vzext_vf4_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -375,7 +375,7 @@ vuint32m1_t test_vzext_vf4_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf4_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vzext.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vzext.mask.nxv4i32.nxv4i8.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vzext_vf4_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -386,7 +386,7 @@ vuint32m2_t test_vzext_vf4_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf4_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vzext.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vzext.mask.nxv8i32.nxv8i8.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vzext_vf4_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -397,7 +397,7 @@ vuint32m4_t test_vzext_vf4_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf4_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vzext.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vzext.mask.nxv16i32.nxv16i8.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vzext_vf4_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -408,7 +408,7 @@ vuint32m8_t test_vzext_vf4_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf8_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vzext.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vzext.mask.nxv1i64.nxv1i8.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i8> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vzext_vf8_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -419,7 +419,7 @@ vuint64m1_t test_vzext_vf8_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf8_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vzext.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vzext.mask.nxv2i64.nxv2i8.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i8> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vzext_vf8_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -430,7 +430,7 @@ vuint64m2_t test_vzext_vf8_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf8_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vzext.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vzext.mask.nxv4i64.nxv4i8.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i8> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vzext_vf8_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -441,7 +441,7 @@ vuint64m4_t test_vzext_vf8_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf8_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vzext.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vzext.mask.nxv8i64.nxv8i8.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i8> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vzext_vf8_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
@@ -452,7 +452,7 @@ vuint64m8_t test_vzext_vf8_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf2_u32mf2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vzext.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.vzext.mask.nxv1i32.nxv1i16.i64(<vscale x 1 x i32> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i32> [[TMP0]]
 //
 vuint32mf2_t test_vzext_vf2_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
@@ -463,7 +463,7 @@ vuint32mf2_t test_vzext_vf2_u32mf2_m(vbool64_t mask, vuint32mf2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf2_u32m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vzext.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.vzext.mask.nxv2i32.nxv2i16.i64(<vscale x 2 x i32> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
 vuint32m1_t test_vzext_vf2_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
@@ -474,7 +474,7 @@ vuint32m1_t test_vzext_vf2_u32m1_m(vbool32_t mask, vuint32m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf2_u32m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vzext.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.vzext.mask.nxv4i32.nxv4i16.i64(<vscale x 4 x i32> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
 vuint32m2_t test_vzext_vf2_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
@@ -485,7 +485,7 @@ vuint32m2_t test_vzext_vf2_u32m2_m(vbool16_t mask, vuint32m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf2_u32m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vzext.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.vzext.mask.nxv8i32.nxv8i16.i64(<vscale x 8 x i32> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
 vuint32m4_t test_vzext_vf2_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
@@ -496,7 +496,7 @@ vuint32m4_t test_vzext_vf2_u32m4_m(vbool8_t mask, vuint32m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf2_u32m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vzext.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.vzext.mask.nxv16i32.nxv16i16.i64(<vscale x 16 x i32> [[MASKEDOFF:%.*]], <vscale x 16 x i16> [[OP1:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
 vuint32m8_t test_vzext_vf2_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
@@ -507,7 +507,7 @@ vuint32m8_t test_vzext_vf2_u32m8_m(vbool4_t mask, vuint32m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf4_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vzext.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vzext.mask.nxv1i64.nxv1i16.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i16> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vzext_vf4_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -518,7 +518,7 @@ vuint64m1_t test_vzext_vf4_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf4_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vzext.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vzext.mask.nxv2i64.nxv2i16.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i16> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vzext_vf4_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -529,7 +529,7 @@ vuint64m2_t test_vzext_vf4_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf4_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vzext.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vzext.mask.nxv4i64.nxv4i16.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i16> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vzext_vf4_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -540,7 +540,7 @@ vuint64m4_t test_vzext_vf4_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf4_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vzext.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vzext.mask.nxv8i64.nxv8i16.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vzext_vf4_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
@@ -551,7 +551,7 @@ vuint64m8_t test_vzext_vf4_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf2_u64m1_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vzext.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.vzext.mask.nxv1i64.nxv1i32.i64(<vscale x 1 x i64> [[MASKEDOFF:%.*]], <vscale x 1 x i32> [[OP1:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i64> [[TMP0]]
 //
 vuint64m1_t test_vzext_vf2_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
@@ -562,7 +562,7 @@ vuint64m1_t test_vzext_vf2_u64m1_m(vbool64_t mask, vuint64m1_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf2_u64m2_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vzext.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.vzext.mask.nxv2i64.nxv2i32.i64(<vscale x 2 x i64> [[MASKEDOFF:%.*]], <vscale x 2 x i32> [[OP1:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i64> [[TMP0]]
 //
 vuint64m2_t test_vzext_vf2_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
@@ -573,7 +573,7 @@ vuint64m2_t test_vzext_vf2_u64m2_m(vbool32_t mask, vuint64m2_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf2_u64m4_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vzext.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.vzext.mask.nxv4i64.nxv4i32.i64(<vscale x 4 x i64> [[MASKEDOFF:%.*]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i64> [[TMP0]]
 //
 vuint64m4_t test_vzext_vf2_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
@@ -584,7 +584,7 @@ vuint64m4_t test_vzext_vf2_u64m4_m(vbool16_t mask, vuint64m4_t maskedoff,
 //
 // CHECK-RV64-LABEL: @test_vzext_vf2_u64m8_m(
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vzext.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]]) #[[ATTR7]]
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.vzext.mask.nxv8i64.nxv8i32.i64(<vscale x 8 x i64> [[MASKEDOFF:%.*]], <vscale x 8 x i32> [[OP1:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i64> [[TMP0]]
 //
 vuint64m8_t test_vzext_vf2_u64m8_m(vbool8_t mask, vuint64m8_t maskedoff,
diff --git a/clang/utils/TableGen/RISCVVEmitter.cpp b/clang/utils/TableGen/RISCVVEmitter.cpp
index c71469a..423e547 100644
--- a/clang/utils/TableGen/RISCVVEmitter.cpp
+++ b/clang/utils/TableGen/RISCVVEmitter.cpp
@@ -161,11 +161,6 @@ private:
   // The types we use to obtain the specific LLVM intrinsic. They are index of
   // InputTypes. -1 means the return type.
   std::vector<int64_t> IntrinsicTypes;
-  // C/C++ intrinsic operand order is different to builtin operand order. Record
-  // the mapping of InputTypes index.
-  SmallVector<unsigned> CTypeOrder;
-  // Operands are reordered in the header.
-  bool IsOperandReordered = false;
   uint8_t RISCVExtensions = 0;
 
 public:
@@ -174,7 +169,6 @@ public:
                bool HasMaskedOffOperand, bool HasVL, bool HasNoMaskedOverloaded,
                bool HasAutoDef, StringRef ManualCodegen, const RVVTypes &Types,
                const std::vector<int64_t> &IntrinsicTypes,
-               const std::vector<int64_t> &PermuteOperands,
                StringRef RequiredExtension);
   ~RVVIntrinsic() = default;
 
@@ -187,8 +181,6 @@ public:
   bool hasManualCodegen() const { return !ManualCodegen.empty(); }
   bool hasAutoDef() const { return HasAutoDef; }
   bool isMask() const { return IsMask; }
-  bool isOperandReordered() const { return IsOperandReordered; }
-  size_t getNumOperand() const { return InputTypes.size(); }
   StringRef getIRName() const { return IRName; }
   StringRef getManualCodegen() const { return ManualCodegen; }
   uint8_t getRISCVExtensions() const { return RISCVExtensions; }
@@ -754,7 +746,6 @@ RVVIntrinsic::RVVIntrinsic(StringRef NewName, StringRef Suffix,
                            bool HasNoMaskedOverloaded, bool HasAutoDef,
                            StringRef ManualCodegen, const RVVTypes &OutInTypes,
                            const std::vector<int64_t> &NewIntrinsicTypes,
-                           const std::vector<int64_t> &PermuteOperands,
                            StringRef RequiredExtension)
     : IRName(IRName), HasSideEffects(HasSideEffects), IsMask(IsMask),
       HasMaskedOffOperand(HasMaskedOffOperand), HasVL(HasVL),
@@ -787,43 +778,6 @@ RVVIntrinsic::RVVIntrinsic(StringRef NewName, StringRef Suffix,
   // Init OutputType and InputTypes
   OutputType = OutInTypes[0];
   InputTypes.assign(OutInTypes.begin() + 1, OutInTypes.end());
-  CTypeOrder.resize(InputTypes.size());
-  std::iota(CTypeOrder.begin(), CTypeOrder.end(), 0);
-  // Update default order if we need permutate.
-  if (!PermuteOperands.empty()) {
-    IsOperandReordered = true;
-    // PermuteOperands is nonmasked version index. Update index when there is
-    // maskedoff operand which is always in first operand.
-
-    unsigned Skew = HasMaskedOffOperand ? 1 : 0;
-    for (unsigned i = 0; i < PermuteOperands.size(); ++i) {
-      if (i != PermuteOperands[i])
-        CTypeOrder[i] = PermuteOperands[i] + Skew;
-    }
-    // Verify the result of CTypeOrder has legal value.
-    if (*std::max_element(CTypeOrder.begin(), CTypeOrder.end()) >=
-        CTypeOrder.size())
-      PrintFatalError(
-          "The index of PermuteOperand is bigger than the operand number");
-    SmallSet<unsigned, 8> Seen;
-    for (auto Idx : CTypeOrder) {
-      if (!Seen.insert(Idx).second)
-        PrintFatalError(
-            "The different element in PermuteOperand could not be equal");
-    }
-  }
-
-  if (IsMask) {
-    if (HasVL)
-      // Builtin type order: op0, op1, ..., mask, vl
-      // C type order: mask, op0, op1, ..., vl
-      std::rotate(CTypeOrder.begin(), CTypeOrder.end() - 2,
-                  CTypeOrder.end() - 1);
-    else
-      // Builtin type order: op0, op1, ..., mask
-      // C type order: mask, op0, op1, ...,
-      std::rotate(CTypeOrder.begin(), CTypeOrder.end() - 1, CTypeOrder.end());
-  }
 
   // IntrinsicTypes is nonmasked version index. Need to update it
   // if there is maskedoff operand (It is always in first operand).
@@ -853,6 +807,15 @@ void RVVIntrinsic::emitCodeGenSwitchBody(raw_ostream &OS) const {
     OS << "break;\n";
     return;
   }
+
+  if (isMask()) {
+    if (hasVL()) {
+      OS << "  std::rotate(Ops.begin(), Ops.begin() + 1, Ops.end() - 1);\n";
+    } else {
+      OS << "  std::rotate(Ops.begin(), Ops.begin() + 1, Ops.end());\n";
+    }
+  }
+
   OS << "  IntrinsicTypes = {";
   ListSeparator LS;
   for (const auto &Idx : IntrinsicTypes) {
@@ -865,55 +828,39 @@ void RVVIntrinsic::emitCodeGenSwitchBody(raw_ostream &OS) const {
   // VL could be i64 or i32, need to encode it in IntrinsicTypes. VL is
   // always last operand.
   if (hasVL())
-    OS << ", Ops[" << getNumOperand() - 1 << "]->getType()";
+    OS << ", Ops.back()->getType()";
   OS << "};\n";
   OS << "  break;\n";
 }
 
 void RVVIntrinsic::emitIntrinsicMacro(raw_ostream &OS) const {
   OS << "#define " << getName() << "(";
-  if (getNumOperand() > 0) {
+  if (!InputTypes.empty()) {
     ListSeparator LS;
-    for (const auto &I : CTypeOrder)
-      OS << LS << "op" << I;
+    for (unsigned i = 0, e = InputTypes.size(); i != e; ++i)
+      OS << LS << "op" << i;
   }
   OS << ") \\\n";
   OS << "__builtin_rvv_" << getName() << "(";
-  if (getNumOperand() > 0) {
+  if (!InputTypes.empty()) {
     ListSeparator LS;
-    for (unsigned i = 0; i < InputTypes.size(); ++i)
+    for (unsigned i = 0, e = InputTypes.size(); i != e; ++i)
       OS << LS << "(" << InputTypes[i]->getTypeStr() << ")(op" << i << ")";
   }
   OS << ")\n";
 }
 
 void RVVIntrinsic::emitMangledFuncDef(raw_ostream &OS) const {
-  bool UseAliasAttr = !isMask() && !isOperandReordered();
-  if (UseAliasAttr) {
-    OS << "__attribute__((clang_builtin_alias(";
-    OS << "__builtin_rvv_" << getName() << ")))\n";
-  }
+  OS << "__attribute__((clang_builtin_alias(";
+  OS << "__builtin_rvv_" << getName() << ")))\n";
   OS << OutputType->getTypeStr() << " " << getMangledName() << "(";
   // Emit function arguments
-  if (getNumOperand() > 0) {
+  if (!InputTypes.empty()) {
     ListSeparator LS;
-    for (unsigned i = 0; i < CTypeOrder.size(); ++i)
-      OS << LS << InputTypes[CTypeOrder[i]]->getTypeStr() << " op" << i;
-  }
-  if (UseAliasAttr) {
-    OS << ");\n\n";
-  } else {
-    OS << "){\n";
-    OS << "  return " << getName() << "(";
-    // Emit parameter variables
-    if (getNumOperand() > 0) {
-      ListSeparator LS;
-      for (unsigned i = 0; i < CTypeOrder.size(); ++i)
-        OS << LS << "op" << i;
-    }
-    OS << ");\n";
-    OS << "}\n\n";
+    for (unsigned i = 0; i < InputTypes.size(); ++i)
+      OS << LS << InputTypes[i]->getTypeStr() << " op" << i;
   }
+  OS << ");\n\n";
 }
 
 //===----------------------------------------------------------------------===//
@@ -1123,8 +1070,6 @@ void RVVEmitter::createRVVIntrinsics(
     StringRef ManualCodegenMask = R->getValueAsString("ManualCodegenMask");
     std::vector<int64_t> IntrinsicTypes =
         R->getValueAsListOfInts("IntrinsicTypes");
-    std::vector<int64_t> PermuteOperands =
-        R->getValueAsListOfInts("PermuteOperands");
     StringRef RequiredExtension = R->getValueAsString("RequiredExtension");
     StringRef IRName = R->getValueAsString("IRName");
     StringRef IRNameMask = R->getValueAsString("IRNameMask");
@@ -1144,11 +1089,11 @@ void RVVEmitter::createRVVIntrinsics(
     // Compute Builtin types
     SmallVector<std::string> ProtoMaskSeq = ProtoSeq;
     if (HasMask) {
-      // If HasMask, append 'm' to last operand.
-      ProtoMaskSeq.push_back("m");
       // If HasMaskedOffOperand, insert result type as first input operand.
       if (HasMaskedOffOperand)
         ProtoMaskSeq.insert(ProtoMaskSeq.begin() + 1, ProtoSeq[0]);
+      // If HasMask, insert 'm' as first input operand.
+      ProtoMaskSeq.insert(ProtoMaskSeq.begin() + 1, "m");
     }
     // If HasVL, append 'z' to last operand
     if (HasVL) {
@@ -1170,7 +1115,7 @@ void RVVEmitter::createRVVIntrinsics(
             Name, SuffixStr, MangledName, IRName, HasSideEffects,
             /*IsMask=*/false, /*HasMaskedOffOperand=*/false, HasVL,
             HasNoMaskedOverloaded, HasAutoDef, ManualCodegen, Types.getValue(),
-            IntrinsicTypes, PermuteOperands, RequiredExtension));
+            IntrinsicTypes, RequiredExtension));
         if (HasMask) {
           // Create a mask intrinsic
           Optional<RVVTypes> MaskTypes =
@@ -1179,8 +1124,7 @@ void RVVEmitter::createRVVIntrinsics(
               Name, SuffixStr, MangledName, IRNameMask, HasSideEffects,
               /*IsMask=*/true, HasMaskedOffOperand, HasVL,
               HasNoMaskedOverloaded, HasAutoDef, ManualCodegenMask,
-              MaskTypes.getValue(), IntrinsicTypes, PermuteOperands,
-              RequiredExtension));
+              MaskTypes.getValue(), IntrinsicTypes, RequiredExtension));
         }
       } // end for Log2LMULList
     }   // end for TypeRange
-- 
2.7.4