From: Simon Pilgrim <llvm-dev@redking.me.uk>
Date: Fri, 21 Oct 2016 10:50:52 +0000 (+0000)
Subject: [X86][AVX] Add 32-bit target tests for vector lzcnt/tzcnt to demonstrate missed foldi... 
X-Git-Tag: llvmorg-4.0.0-rc1~6606
X-Git-Url: http://review.tizen.org/git/?a=commitdiff_plain;h=f447fbf913dcdbe651764e452c8abc1d32656e46;p=platform%2Fupstream%2Fllvm.git

[X86][AVX] Add 32-bit target tests for vector lzcnt/tzcnt to demonstrate missed folding opportunities

llvm-svn: 284816
---

diff --git a/llvm/test/CodeGen/X86/vector-lzcnt-256.ll b/llvm/test/CodeGen/X86/vector-lzcnt-256.ll
index a145621..e288f32 100644
--- a/llvm/test/CodeGen/X86/vector-lzcnt-256.ll
+++ b/llvm/test/CodeGen/X86/vector-lzcnt-256.ll
@@ -3,6 +3,9 @@
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX2
 ; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=knl -mattr=+avx512cd -mattr=+avx512vl | FileCheck %s --check-prefix=ALL --check-prefix=AVX512 --check-prefix=AVX512VLCD
 ; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=knl -mattr=+avx512cd | FileCheck %s --check-prefix=ALL --check-prefix=AVX512 --check-prefix=AVX512CD
+;
+; Just one 32-bit run to make sure we do reasonable things for i64 lzcnt.
+; RUN: llc < %s -mtriple=i686-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefix=ALL --check-prefix=X32-AVX --check-prefix=X32-AVX2
 
 define <4 x i64> @testv4i64(<4 x i64> %in) nounwind {
 ; AVX1-LABEL: testv4i64:
@@ -101,6 +104,36 @@ define <4 x i64> @testv4i64(<4 x i64> %in) nounwind {
 ; AVX512CD-NEXT:    vplzcntq %zmm0, %zmm0
 ; AVX512CD-NEXT:    ## kill: %YMM0<def> %YMM0<kill> %ZMM0<kill>
 ; AVX512CD-NEXT:    retq
+;
+; X32-AVX-LABEL: testv4i64:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm1 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15]
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm2
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm3 = [4,3,2,2,1,1,1,1,0,0,0,0,0,0,0,0,4,3,2,2,1,1,1,1,0,0,0,0,0,0,0,0]
+; X32-AVX-NEXT:    vpshufb %ymm2, %ymm3, %ymm2
+; X32-AVX-NEXT:    vpsrlw $4, %ymm0, %ymm4
+; X32-AVX-NEXT:    vpand %ymm1, %ymm4, %ymm1
+; X32-AVX-NEXT:    vpxor %ymm4, %ymm4, %ymm4
+; X32-AVX-NEXT:    vpcmpeqb %ymm4, %ymm1, %ymm5
+; X32-AVX-NEXT:    vpand %ymm5, %ymm2, %ymm2
+; X32-AVX-NEXT:    vpshufb %ymm1, %ymm3, %ymm1
+; X32-AVX-NEXT:    vpaddb %ymm1, %ymm2, %ymm1
+; X32-AVX-NEXT:    vpcmpeqb %ymm4, %ymm0, %ymm2
+; X32-AVX-NEXT:    vpsrlw $8, %ymm2, %ymm2
+; X32-AVX-NEXT:    vpand %ymm2, %ymm1, %ymm2
+; X32-AVX-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpaddw %ymm2, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpcmpeqw %ymm4, %ymm0, %ymm2
+; X32-AVX-NEXT:    vpsrld $16, %ymm2, %ymm2
+; X32-AVX-NEXT:    vpand %ymm2, %ymm1, %ymm2
+; X32-AVX-NEXT:    vpsrld $16, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpaddd %ymm2, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpcmpeqd %ymm4, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpsrlq $32, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpand %ymm0, %ymm1, %ymm0
+; X32-AVX-NEXT:    vpsrlq $32, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpaddq %ymm0, %ymm1, %ymm0
+; X32-AVX-NEXT:    retl
 
   %out = call <4 x i64> @llvm.ctlz.v4i64(<4 x i64> %in, i1 0)
   ret <4 x i64> %out
@@ -203,6 +236,36 @@ define <4 x i64> @testv4i64u(<4 x i64> %in) nounwind {
 ; AVX512CD-NEXT:    vplzcntq %zmm0, %zmm0
 ; AVX512CD-NEXT:    ## kill: %YMM0<def> %YMM0<kill> %ZMM0<kill>
 ; AVX512CD-NEXT:    retq
+;
+; X32-AVX-LABEL: testv4i64u:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm1 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15]
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm2
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm3 = [4,3,2,2,1,1,1,1,0,0,0,0,0,0,0,0,4,3,2,2,1,1,1,1,0,0,0,0,0,0,0,0]
+; X32-AVX-NEXT:    vpshufb %ymm2, %ymm3, %ymm2
+; X32-AVX-NEXT:    vpsrlw $4, %ymm0, %ymm4
+; X32-AVX-NEXT:    vpand %ymm1, %ymm4, %ymm1
+; X32-AVX-NEXT:    vpxor %ymm4, %ymm4, %ymm4
+; X32-AVX-NEXT:    vpcmpeqb %ymm4, %ymm1, %ymm5
+; X32-AVX-NEXT:    vpand %ymm5, %ymm2, %ymm2
+; X32-AVX-NEXT:    vpshufb %ymm1, %ymm3, %ymm1
+; X32-AVX-NEXT:    vpaddb %ymm1, %ymm2, %ymm1
+; X32-AVX-NEXT:    vpcmpeqb %ymm4, %ymm0, %ymm2
+; X32-AVX-NEXT:    vpsrlw $8, %ymm2, %ymm2
+; X32-AVX-NEXT:    vpand %ymm2, %ymm1, %ymm2
+; X32-AVX-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpaddw %ymm2, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpcmpeqw %ymm4, %ymm0, %ymm2
+; X32-AVX-NEXT:    vpsrld $16, %ymm2, %ymm2
+; X32-AVX-NEXT:    vpand %ymm2, %ymm1, %ymm2
+; X32-AVX-NEXT:    vpsrld $16, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpaddd %ymm2, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpcmpeqd %ymm4, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpsrlq $32, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpand %ymm0, %ymm1, %ymm0
+; X32-AVX-NEXT:    vpsrlq $32, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpaddq %ymm0, %ymm1, %ymm0
+; X32-AVX-NEXT:    retl
 
   %out = call <4 x i64> @llvm.ctlz.v4i64(<4 x i64> %in, i1 -1)
   ret <4 x i64> %out
@@ -290,6 +353,31 @@ define <8 x i32> @testv8i32(<8 x i32> %in) nounwind {
 ; AVX512CD-NEXT:    vplzcntd %zmm0, %zmm0
 ; AVX512CD-NEXT:    ## kill: %YMM0<def> %YMM0<kill> %ZMM0<kill>
 ; AVX512CD-NEXT:    retq
+;
+; X32-AVX-LABEL: testv8i32:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm1 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15]
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm2
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm3 = [4,3,2,2,1,1,1,1,0,0,0,0,0,0,0,0,4,3,2,2,1,1,1,1,0,0,0,0,0,0,0,0]
+; X32-AVX-NEXT:    vpshufb %ymm2, %ymm3, %ymm2
+; X32-AVX-NEXT:    vpsrlw $4, %ymm0, %ymm4
+; X32-AVX-NEXT:    vpand %ymm1, %ymm4, %ymm1
+; X32-AVX-NEXT:    vpxor %ymm4, %ymm4, %ymm4
+; X32-AVX-NEXT:    vpcmpeqb %ymm4, %ymm1, %ymm5
+; X32-AVX-NEXT:    vpand %ymm5, %ymm2, %ymm2
+; X32-AVX-NEXT:    vpshufb %ymm1, %ymm3, %ymm1
+; X32-AVX-NEXT:    vpaddb %ymm1, %ymm2, %ymm1
+; X32-AVX-NEXT:    vpcmpeqb %ymm4, %ymm0, %ymm2
+; X32-AVX-NEXT:    vpsrlw $8, %ymm2, %ymm2
+; X32-AVX-NEXT:    vpand %ymm2, %ymm1, %ymm2
+; X32-AVX-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpaddw %ymm2, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpcmpeqw %ymm4, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpsrld $16, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpand %ymm0, %ymm1, %ymm0
+; X32-AVX-NEXT:    vpsrld $16, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpaddd %ymm0, %ymm1, %ymm0
+; X32-AVX-NEXT:    retl
 
   %out = call <8 x i32> @llvm.ctlz.v8i32(<8 x i32> %in, i1 0)
   ret <8 x i32> %out
@@ -377,6 +465,31 @@ define <8 x i32> @testv8i32u(<8 x i32> %in) nounwind {
 ; AVX512CD-NEXT:    vplzcntd %zmm0, %zmm0
 ; AVX512CD-NEXT:    ## kill: %YMM0<def> %YMM0<kill> %ZMM0<kill>
 ; AVX512CD-NEXT:    retq
+;
+; X32-AVX-LABEL: testv8i32u:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm1 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15]
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm2
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm3 = [4,3,2,2,1,1,1,1,0,0,0,0,0,0,0,0,4,3,2,2,1,1,1,1,0,0,0,0,0,0,0,0]
+; X32-AVX-NEXT:    vpshufb %ymm2, %ymm3, %ymm2
+; X32-AVX-NEXT:    vpsrlw $4, %ymm0, %ymm4
+; X32-AVX-NEXT:    vpand %ymm1, %ymm4, %ymm1
+; X32-AVX-NEXT:    vpxor %ymm4, %ymm4, %ymm4
+; X32-AVX-NEXT:    vpcmpeqb %ymm4, %ymm1, %ymm5
+; X32-AVX-NEXT:    vpand %ymm5, %ymm2, %ymm2
+; X32-AVX-NEXT:    vpshufb %ymm1, %ymm3, %ymm1
+; X32-AVX-NEXT:    vpaddb %ymm1, %ymm2, %ymm1
+; X32-AVX-NEXT:    vpcmpeqb %ymm4, %ymm0, %ymm2
+; X32-AVX-NEXT:    vpsrlw $8, %ymm2, %ymm2
+; X32-AVX-NEXT:    vpand %ymm2, %ymm1, %ymm2
+; X32-AVX-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpaddw %ymm2, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpcmpeqw %ymm4, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpsrld $16, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpand %ymm0, %ymm1, %ymm0
+; X32-AVX-NEXT:    vpsrld $16, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpaddd %ymm0, %ymm1, %ymm0
+; X32-AVX-NEXT:    retl
 
   %out = call <8 x i32> @llvm.ctlz.v8i32(<8 x i32> %in, i1 -1)
   ret <8 x i32> %out
@@ -445,6 +558,26 @@ define <16 x i16> @testv16i16(<16 x i16> %in) nounwind {
 ; AVX512-NEXT:    vpmovdw %zmm0, %ymm0
 ; AVX512-NEXT:    vpsubw {{.*}}(%rip), %ymm0, %ymm0
 ; AVX512-NEXT:    retq
+;
+; X32-AVX-LABEL: testv16i16:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm1 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15]
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm2
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm3 = [4,3,2,2,1,1,1,1,0,0,0,0,0,0,0,0,4,3,2,2,1,1,1,1,0,0,0,0,0,0,0,0]
+; X32-AVX-NEXT:    vpshufb %ymm2, %ymm3, %ymm2
+; X32-AVX-NEXT:    vpsrlw $4, %ymm0, %ymm4
+; X32-AVX-NEXT:    vpand %ymm1, %ymm4, %ymm1
+; X32-AVX-NEXT:    vpxor %ymm4, %ymm4, %ymm4
+; X32-AVX-NEXT:    vpcmpeqb %ymm4, %ymm1, %ymm5
+; X32-AVX-NEXT:    vpand %ymm5, %ymm2, %ymm2
+; X32-AVX-NEXT:    vpshufb %ymm1, %ymm3, %ymm1
+; X32-AVX-NEXT:    vpaddb %ymm1, %ymm2, %ymm1
+; X32-AVX-NEXT:    vpcmpeqb %ymm4, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpsrlw $8, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpand %ymm0, %ymm1, %ymm0
+; X32-AVX-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpaddw %ymm0, %ymm1, %ymm0
+; X32-AVX-NEXT:    retl
   %out = call <16 x i16> @llvm.ctlz.v16i16(<16 x i16> %in, i1 0)
   ret <16 x i16> %out
 }
@@ -512,6 +645,26 @@ define <16 x i16> @testv16i16u(<16 x i16> %in) nounwind {
 ; AVX512-NEXT:    vpmovdw %zmm0, %ymm0
 ; AVX512-NEXT:    vpsubw {{.*}}(%rip), %ymm0, %ymm0
 ; AVX512-NEXT:    retq
+;
+; X32-AVX-LABEL: testv16i16u:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm1 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15]
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm2
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm3 = [4,3,2,2,1,1,1,1,0,0,0,0,0,0,0,0,4,3,2,2,1,1,1,1,0,0,0,0,0,0,0,0]
+; X32-AVX-NEXT:    vpshufb %ymm2, %ymm3, %ymm2
+; X32-AVX-NEXT:    vpsrlw $4, %ymm0, %ymm4
+; X32-AVX-NEXT:    vpand %ymm1, %ymm4, %ymm1
+; X32-AVX-NEXT:    vpxor %ymm4, %ymm4, %ymm4
+; X32-AVX-NEXT:    vpcmpeqb %ymm4, %ymm1, %ymm5
+; X32-AVX-NEXT:    vpand %ymm5, %ymm2, %ymm2
+; X32-AVX-NEXT:    vpshufb %ymm1, %ymm3, %ymm1
+; X32-AVX-NEXT:    vpaddb %ymm1, %ymm2, %ymm1
+; X32-AVX-NEXT:    vpcmpeqb %ymm4, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpsrlw $8, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpand %ymm0, %ymm1, %ymm0
+; X32-AVX-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpaddw %ymm0, %ymm1, %ymm0
+; X32-AVX-NEXT:    retl
   %out = call <16 x i16> @llvm.ctlz.v16i16(<16 x i16> %in, i1 -1)
   ret <16 x i16> %out
 }
@@ -586,6 +739,21 @@ define <32 x i8> @testv32i8(<32 x i8> %in) nounwind {
 ; AVX512CD-NEXT:    vpsubb %xmm2, %xmm0, %xmm0
 ; AVX512CD-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
 ; AVX512CD-NEXT:    retq
+;
+; X32-AVX-LABEL: testv32i8:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm1 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15]
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm2
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm3 = [4,3,2,2,1,1,1,1,0,0,0,0,0,0,0,0,4,3,2,2,1,1,1,1,0,0,0,0,0,0,0,0]
+; X32-AVX-NEXT:    vpshufb %ymm2, %ymm3, %ymm2
+; X32-AVX-NEXT:    vpsrlw $4, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpxor %ymm1, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpcmpeqb %ymm1, %ymm0, %ymm1
+; X32-AVX-NEXT:    vpand %ymm1, %ymm2, %ymm1
+; X32-AVX-NEXT:    vpshufb %ymm0, %ymm3, %ymm0
+; X32-AVX-NEXT:    vpaddb %ymm0, %ymm1, %ymm0
+; X32-AVX-NEXT:    retl
   %out = call <32 x i8> @llvm.ctlz.v32i8(<32 x i8> %in, i1 0)
   ret <32 x i8> %out
 }
@@ -660,6 +828,21 @@ define <32 x i8> @testv32i8u(<32 x i8> %in) nounwind {
 ; AVX512CD-NEXT:    vpsubb %xmm2, %xmm0, %xmm0
 ; AVX512CD-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
 ; AVX512CD-NEXT:    retq
+;
+; X32-AVX-LABEL: testv32i8u:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm1 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15]
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm2
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm3 = [4,3,2,2,1,1,1,1,0,0,0,0,0,0,0,0,4,3,2,2,1,1,1,1,0,0,0,0,0,0,0,0]
+; X32-AVX-NEXT:    vpshufb %ymm2, %ymm3, %ymm2
+; X32-AVX-NEXT:    vpsrlw $4, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpxor %ymm1, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpcmpeqb %ymm1, %ymm0, %ymm1
+; X32-AVX-NEXT:    vpand %ymm1, %ymm2, %ymm1
+; X32-AVX-NEXT:    vpshufb %ymm0, %ymm3, %ymm0
+; X32-AVX-NEXT:    vpaddb %ymm0, %ymm1, %ymm0
+; X32-AVX-NEXT:    retl
   %out = call <32 x i8> @llvm.ctlz.v32i8(<32 x i8> %in, i1 -1)
   ret <32 x i8> %out
 }
@@ -674,6 +857,37 @@ define <4 x i64> @foldv4i64() nounwind {
 ; AVX512:       ## BB#0:
 ; AVX512-NEXT:    vmovaps {{.*#+}} ymm0 = [55,0,64,56]
 ; AVX512-NEXT:    retq
+;
+; X32-AVX-LABEL: foldv4i64:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm0 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15]
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm1 = [256,0,4294967295,4294967295,0,0,255,0]
+; X32-AVX-NEXT:    vpand %ymm0, %ymm1, %ymm2
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm3 = [4,3,2,2,1,1,1,1,0,0,0,0,0,0,0,0,4,3,2,2,1,1,1,1,0,0,0,0,0,0,0,0]
+; X32-AVX-NEXT:    vpshufb %ymm2, %ymm3, %ymm2
+; X32-AVX-NEXT:    vpsrlw $4, %ymm1, %ymm4
+; X32-AVX-NEXT:    vpand %ymm0, %ymm4, %ymm0
+; X32-AVX-NEXT:    vpxor %ymm4, %ymm4, %ymm4
+; X32-AVX-NEXT:    vpcmpeqb %ymm4, %ymm0, %ymm5
+; X32-AVX-NEXT:    vpand %ymm5, %ymm2, %ymm2
+; X32-AVX-NEXT:    vpshufb %ymm0, %ymm3, %ymm0
+; X32-AVX-NEXT:    vpaddb %ymm0, %ymm2, %ymm0
+; X32-AVX-NEXT:    vpcmpeqb %ymm4, %ymm1, %ymm2
+; X32-AVX-NEXT:    vpsrlw $8, %ymm2, %ymm2
+; X32-AVX-NEXT:    vpand %ymm2, %ymm0, %ymm2
+; X32-AVX-NEXT:    vpsrlw $8, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpaddw %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpcmpeqw %ymm4, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpsrld $16, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm1
+; X32-AVX-NEXT:    vpsrld $16, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpaddd %ymm1, %ymm0, %ymm0
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm1 = [0,4294967295,0,0,4294967295,4294967295,0,4294967295]
+; X32-AVX-NEXT:    vpsrlq $32, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm1
+; X32-AVX-NEXT:    vpsrlq $32, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpaddq %ymm1, %ymm0, %ymm0
+; X32-AVX-NEXT:    retl
   %out = call <4 x i64> @llvm.ctlz.v4i64(<4 x i64> <i64 256, i64 -1, i64 0, i64 255>, i1 0)
   ret <4 x i64> %out
 }
@@ -688,6 +902,37 @@ define <4 x i64> @foldv4i64u() nounwind {
 ; AVX512:       ## BB#0:
 ; AVX512-NEXT:    vmovaps {{.*#+}} ymm0 = [55,0,64,56]
 ; AVX512-NEXT:    retq
+;
+; X32-AVX-LABEL: foldv4i64u:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm0 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15]
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm1 = [256,0,4294967295,4294967295,0,0,255,0]
+; X32-AVX-NEXT:    vpand %ymm0, %ymm1, %ymm2
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm3 = [4,3,2,2,1,1,1,1,0,0,0,0,0,0,0,0,4,3,2,2,1,1,1,1,0,0,0,0,0,0,0,0]
+; X32-AVX-NEXT:    vpshufb %ymm2, %ymm3, %ymm2
+; X32-AVX-NEXT:    vpsrlw $4, %ymm1, %ymm4
+; X32-AVX-NEXT:    vpand %ymm0, %ymm4, %ymm0
+; X32-AVX-NEXT:    vpxor %ymm4, %ymm4, %ymm4
+; X32-AVX-NEXT:    vpcmpeqb %ymm4, %ymm0, %ymm5
+; X32-AVX-NEXT:    vpand %ymm5, %ymm2, %ymm2
+; X32-AVX-NEXT:    vpshufb %ymm0, %ymm3, %ymm0
+; X32-AVX-NEXT:    vpaddb %ymm0, %ymm2, %ymm0
+; X32-AVX-NEXT:    vpcmpeqb %ymm4, %ymm1, %ymm2
+; X32-AVX-NEXT:    vpsrlw $8, %ymm2, %ymm2
+; X32-AVX-NEXT:    vpand %ymm2, %ymm0, %ymm2
+; X32-AVX-NEXT:    vpsrlw $8, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpaddw %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpcmpeqw %ymm4, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpsrld $16, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm1
+; X32-AVX-NEXT:    vpsrld $16, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpaddd %ymm1, %ymm0, %ymm0
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm1 = [0,4294967295,0,0,4294967295,4294967295,0,4294967295]
+; X32-AVX-NEXT:    vpsrlq $32, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm1
+; X32-AVX-NEXT:    vpsrlq $32, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpaddq %ymm1, %ymm0, %ymm0
+; X32-AVX-NEXT:    retl
   %out = call <4 x i64> @llvm.ctlz.v4i64(<4 x i64> <i64 256, i64 -1, i64 0, i64 255>, i1 -1)
   ret <4 x i64> %out
 }
@@ -702,6 +947,11 @@ define <8 x i32> @foldv8i32() nounwind {
 ; AVX512:       ## BB#0:
 ; AVX512-NEXT:    vmovaps {{.*#+}} ymm0 = [23,0,32,24,0,29,27,25]
 ; AVX512-NEXT:    retq
+;
+; X32-AVX-LABEL: foldv8i32:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [23,0,32,24,0,29,27,25]
+; X32-AVX-NEXT:    retl
   %out = call <8 x i32> @llvm.ctlz.v8i32(<8 x i32> <i32 256, i32 -1, i32 0, i32 255, i32 -65536, i32 7, i32 24, i32 88>, i1 0)
   ret <8 x i32> %out
 }
@@ -716,6 +966,11 @@ define <8 x i32> @foldv8i32u() nounwind {
 ; AVX512:       ## BB#0:
 ; AVX512-NEXT:    vmovaps {{.*#+}} ymm0 = [23,0,32,24,0,29,27,25]
 ; AVX512-NEXT:    retq
+;
+; X32-AVX-LABEL: foldv8i32u:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [23,0,32,24,0,29,27,25]
+; X32-AVX-NEXT:    retl
   %out = call <8 x i32> @llvm.ctlz.v8i32(<8 x i32> <i32 256, i32 -1, i32 0, i32 255, i32 -65536, i32 7, i32 24, i32 88>, i1 -1)
   ret <8 x i32> %out
 }
@@ -730,6 +985,11 @@ define <16 x i16> @foldv16i16() nounwind {
 ; AVX512:       ## BB#0:
 ; AVX512-NEXT:    vmovaps {{.*#+}} ymm0 = [7,0,16,8,16,13,11,9,0,8,15,14,13,12,11,10]
 ; AVX512-NEXT:    retq
+;
+; X32-AVX-LABEL: foldv16i16:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [7,0,16,8,16,13,11,9,0,8,15,14,13,12,11,10]
+; X32-AVX-NEXT:    retl
   %out = call <16 x i16> @llvm.ctlz.v16i16(<16 x i16> <i16 256, i16 -1, i16 0, i16 255, i16 -65536, i16 7, i16 24, i16 88, i16 -2, i16 254, i16 1, i16 2, i16 4, i16 8, i16 16, i16 32>, i1 0)
   ret <16 x i16> %out
 }
@@ -744,6 +1004,11 @@ define <16 x i16> @foldv16i16u() nounwind {
 ; AVX512:       ## BB#0:
 ; AVX512-NEXT:    vmovaps {{.*#+}} ymm0 = [7,0,16,8,16,13,11,9,0,8,15,14,13,12,11,10]
 ; AVX512-NEXT:    retq
+;
+; X32-AVX-LABEL: foldv16i16u:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [7,0,16,8,16,13,11,9,0,8,15,14,13,12,11,10]
+; X32-AVX-NEXT:    retl
   %out = call <16 x i16> @llvm.ctlz.v16i16(<16 x i16> <i16 256, i16 -1, i16 0, i16 255, i16 -65536, i16 7, i16 24, i16 88, i16 -2, i16 254, i16 1, i16 2, i16 4, i16 8, i16 16, i16 32>, i1 -1)
   ret <16 x i16> %out
 }
@@ -758,6 +1023,11 @@ define <32 x i8> @foldv32i8() nounwind {
 ; AVX512:       ## BB#0:
 ; AVX512-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,8,0,8,5,3,1,0,0,7,6,5,4,3,2,1,0,8,8,0,0,0,0,0,0,0,0,6,5,5,1]
 ; AVX512-NEXT:    retq
+;
+; X32-AVX-LABEL: foldv32i8:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,8,0,8,5,3,1,0,0,7,6,5,4,3,2,1,0,8,8,0,0,0,0,0,0,0,0,6,5,5,1]
+; X32-AVX-NEXT:    retl
   %out = call <32 x i8> @llvm.ctlz.v32i8(<32 x i8> <i8 256, i8 -1, i8 0, i8 255, i8 -65536, i8 7, i8 24, i8 88, i8 -2, i8 254, i8 1, i8 2, i8 4, i8 8, i8 16, i8 32, i8 64, i8 128, i8 256, i8 -256, i8 -128, i8 -64, i8 -32, i8 -16, i8 -8, i8 -4, i8 -2, i8 -1, i8 3, i8 5, i8 7, i8 127>, i1 0)
   ret <32 x i8> %out
 }
@@ -772,6 +1042,11 @@ define <32 x i8> @foldv32i8u() nounwind {
 ; AVX512:       ## BB#0:
 ; AVX512-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,8,0,8,5,3,1,0,0,7,6,5,4,3,2,1,0,8,8,0,0,0,0,0,0,0,0,6,5,5,1]
 ; AVX512-NEXT:    retq
+;
+; X32-AVX-LABEL: foldv32i8u:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,8,0,8,5,3,1,0,0,7,6,5,4,3,2,1,0,8,8,0,0,0,0,0,0,0,0,6,5,5,1]
+; X32-AVX-NEXT:    retl
   %out = call <32 x i8> @llvm.ctlz.v32i8(<32 x i8> <i8 256, i8 -1, i8 0, i8 255, i8 -65536, i8 7, i8 24, i8 88, i8 -2, i8 254, i8 1, i8 2, i8 4, i8 8, i8 16, i8 32, i8 64, i8 128, i8 256, i8 -256, i8 -128, i8 -64, i8 -32, i8 -16, i8 -8, i8 -4, i8 -2, i8 -1, i8 3, i8 5, i8 7, i8 127>, i1 -1)
   ret <32 x i8> %out
 }
diff --git a/llvm/test/CodeGen/X86/vector-tzcnt-256.ll b/llvm/test/CodeGen/X86/vector-tzcnt-256.ll
index ffac692..db9081d 100644
--- a/llvm/test/CodeGen/X86/vector-tzcnt-256.ll
+++ b/llvm/test/CodeGen/X86/vector-tzcnt-256.ll
@@ -3,6 +3,9 @@
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX2
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512cd,+avx512vl | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX512CDVL
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512cd,-avx512vl | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX512CD
+;
+; Just one 32-bit run to make sure we do reasonable things for i64 tzcnt.
+; RUN: llc < %s -mtriple=i686-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefix=ALL --check-prefix=X32-AVX --check-prefix=X32-AVX2
 
 define <4 x i64> @testv4i64(<4 x i64> %in) nounwind {
 ; AVX1-LABEL: testv4i64:
@@ -88,6 +91,23 @@ define <4 x i64> @testv4i64(<4 x i64> %in) nounwind {
 ; AVX512CD-NEXT:    vpaddb %ymm3, %ymm0, %ymm0
 ; AVX512CD-NEXT:    vpsadbw %ymm1, %ymm0, %ymm0
 ; AVX512CD-NEXT:    retq
+;
+; X32-AVX-LABEL: testv4i64:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vpxor %ymm1, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpsubq %ymm0, %ymm1, %ymm2
+; X32-AVX-NEXT:    vpand %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpsubq {{\.LCPI.*}}, %ymm0, %ymm0
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm2 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15]
+; X32-AVX-NEXT:    vpand %ymm2, %ymm0, %ymm3
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm4 = [0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4,0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4]
+; X32-AVX-NEXT:    vpshufb %ymm3, %ymm4, %ymm3
+; X32-AVX-NEXT:    vpsrlw $4, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpand %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpshufb %ymm0, %ymm4, %ymm0
+; X32-AVX-NEXT:    vpaddb %ymm3, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpsadbw %ymm1, %ymm0, %ymm0
+; X32-AVX-NEXT:    retl
   %out = call <4 x i64> @llvm.cttz.v4i64(<4 x i64> %in, i1 0)
   ret <4 x i64> %out
 }
@@ -161,6 +181,23 @@ define <4 x i64> @testv4i64u(<4 x i64> %in) nounwind {
 ; AVX512CD-NEXT:    vpbroadcastq {{.*}}(%rip), %ymm1
 ; AVX512CD-NEXT:    vpsubq %ymm0, %ymm1, %ymm0
 ; AVX512CD-NEXT:    retq
+;
+; X32-AVX-LABEL: testv4i64u:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vpxor %ymm1, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpsubq %ymm0, %ymm1, %ymm2
+; X32-AVX-NEXT:    vpand %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpsubq {{\.LCPI.*}}, %ymm0, %ymm0
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm2 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15]
+; X32-AVX-NEXT:    vpand %ymm2, %ymm0, %ymm3
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm4 = [0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4,0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4]
+; X32-AVX-NEXT:    vpshufb %ymm3, %ymm4, %ymm3
+; X32-AVX-NEXT:    vpsrlw $4, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpand %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpshufb %ymm0, %ymm4, %ymm0
+; X32-AVX-NEXT:    vpaddb %ymm3, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpsadbw %ymm1, %ymm0, %ymm0
+; X32-AVX-NEXT:    retl
   %out = call <4 x i64> @llvm.cttz.v4i64(<4 x i64> %in, i1 -1)
   ret <4 x i64> %out
 }
@@ -269,6 +306,28 @@ define <8 x i32> @testv8i32(<8 x i32> %in) nounwind {
 ; AVX512CD-NEXT:    vpsadbw %ymm1, %ymm0, %ymm0
 ; AVX512CD-NEXT:    vpackuswb %ymm2, %ymm0, %ymm0
 ; AVX512CD-NEXT:    retq
+;
+; X32-AVX-LABEL: testv8i32:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vpxor %ymm1, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpsubd %ymm0, %ymm1, %ymm2
+; X32-AVX-NEXT:    vpand %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpbroadcastd {{\.LCPI.*}}, %ymm2
+; X32-AVX-NEXT:    vpsubd %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm2 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15]
+; X32-AVX-NEXT:    vpand %ymm2, %ymm0, %ymm3
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm4 = [0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4,0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4]
+; X32-AVX-NEXT:    vpshufb %ymm3, %ymm4, %ymm3
+; X32-AVX-NEXT:    vpsrlw $4, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpand %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpshufb %ymm0, %ymm4, %ymm0
+; X32-AVX-NEXT:    vpaddb %ymm3, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpunpckhdq {{.*#+}} ymm2 = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; X32-AVX-NEXT:    vpsadbw %ymm1, %ymm2, %ymm2
+; X32-AVX-NEXT:    vpunpckldq {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; X32-AVX-NEXT:    vpsadbw %ymm1, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpackuswb %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    retl
   %out = call <8 x i32> @llvm.cttz.v8i32(<8 x i32> %in, i1 0)
   ret <8 x i32> %out
 }
@@ -354,6 +413,28 @@ define <8 x i32> @testv8i32u(<8 x i32> %in) nounwind {
 ; AVX512CD-NEXT:    vpbroadcastd {{.*}}(%rip), %ymm1
 ; AVX512CD-NEXT:    vpsubd %ymm0, %ymm1, %ymm0
 ; AVX512CD-NEXT:    retq
+;
+; X32-AVX-LABEL: testv8i32u:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vpxor %ymm1, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpsubd %ymm0, %ymm1, %ymm2
+; X32-AVX-NEXT:    vpand %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpbroadcastd {{\.LCPI.*}}, %ymm2
+; X32-AVX-NEXT:    vpsubd %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm2 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15]
+; X32-AVX-NEXT:    vpand %ymm2, %ymm0, %ymm3
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm4 = [0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4,0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4]
+; X32-AVX-NEXT:    vpshufb %ymm3, %ymm4, %ymm3
+; X32-AVX-NEXT:    vpsrlw $4, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpand %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpshufb %ymm0, %ymm4, %ymm0
+; X32-AVX-NEXT:    vpaddb %ymm3, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpunpckhdq {{.*#+}} ymm2 = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
+; X32-AVX-NEXT:    vpsadbw %ymm1, %ymm2, %ymm2
+; X32-AVX-NEXT:    vpunpckldq {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
+; X32-AVX-NEXT:    vpsadbw %ymm1, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpackuswb %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    retl
   %out = call <8 x i32> @llvm.cttz.v8i32(<8 x i32> %in, i1 -1)
   ret <8 x i32> %out
 }
@@ -450,6 +531,25 @@ define <16 x i16> @testv16i16(<16 x i16> %in) nounwind {
 ; AVX512CD-NEXT:    vpaddb %ymm0, %ymm1, %ymm0
 ; AVX512CD-NEXT:    vpsrlw $8, %ymm0, %ymm0
 ; AVX512CD-NEXT:    retq
+;
+; X32-AVX-LABEL: testv16i16:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vpxor %ymm1, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpsubw %ymm0, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpsubw {{\.LCPI.*}}, %ymm0, %ymm0
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm1 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15]
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm2
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm3 = [0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4,0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4]
+; X32-AVX-NEXT:    vpshufb %ymm2, %ymm3, %ymm2
+; X32-AVX-NEXT:    vpsrlw $4, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpshufb %ymm0, %ymm3, %ymm0
+; X32-AVX-NEXT:    vpaddb %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpsllw $8, %ymm0, %ymm1
+; X32-AVX-NEXT:    vpaddb %ymm0, %ymm1, %ymm0
+; X32-AVX-NEXT:    vpsrlw $8, %ymm0, %ymm0
+; X32-AVX-NEXT:    retl
   %out = call <16 x i16> @llvm.cttz.v16i16(<16 x i16> %in, i1 0)
   ret <16 x i16> %out
 }
@@ -546,6 +646,25 @@ define <16 x i16> @testv16i16u(<16 x i16> %in) nounwind {
 ; AVX512CD-NEXT:    vpaddb %ymm0, %ymm1, %ymm0
 ; AVX512CD-NEXT:    vpsrlw $8, %ymm0, %ymm0
 ; AVX512CD-NEXT:    retq
+;
+; X32-AVX-LABEL: testv16i16u:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vpxor %ymm1, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpsubw %ymm0, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpsubw {{\.LCPI.*}}, %ymm0, %ymm0
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm1 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15]
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm2
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm3 = [0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4,0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4]
+; X32-AVX-NEXT:    vpshufb %ymm2, %ymm3, %ymm2
+; X32-AVX-NEXT:    vpsrlw $4, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpshufb %ymm0, %ymm3, %ymm0
+; X32-AVX-NEXT:    vpaddb %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpsllw $8, %ymm0, %ymm1
+; X32-AVX-NEXT:    vpaddb %ymm0, %ymm1, %ymm0
+; X32-AVX-NEXT:    vpsrlw $8, %ymm0, %ymm0
+; X32-AVX-NEXT:    retl
   %out = call <16 x i16> @llvm.cttz.v16i16(<16 x i16> %in, i1 -1)
   ret <16 x i16> %out
 }
@@ -627,6 +746,22 @@ define <32 x i8> @testv32i8(<32 x i8> %in) nounwind {
 ; AVX512CD-NEXT:    vpshufb %ymm0, %ymm3, %ymm0
 ; AVX512CD-NEXT:    vpaddb %ymm2, %ymm0, %ymm0
 ; AVX512CD-NEXT:    retq
+;
+; X32-AVX-LABEL: testv32i8:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vpxor %ymm1, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpsubb %ymm0, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpsubb {{\.LCPI.*}}, %ymm0, %ymm0
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm1 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15]
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm2
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm3 = [0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4,0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4]
+; X32-AVX-NEXT:    vpshufb %ymm2, %ymm3, %ymm2
+; X32-AVX-NEXT:    vpsrlw $4, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpshufb %ymm0, %ymm3, %ymm0
+; X32-AVX-NEXT:    vpaddb %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    retl
   %out = call <32 x i8> @llvm.cttz.v32i8(<32 x i8> %in, i1 0)
   ret <32 x i8> %out
 }
@@ -708,78 +843,160 @@ define <32 x i8> @testv32i8u(<32 x i8> %in) nounwind {
 ; AVX512CD-NEXT:    vpshufb %ymm0, %ymm3, %ymm0
 ; AVX512CD-NEXT:    vpaddb %ymm2, %ymm0, %ymm0
 ; AVX512CD-NEXT:    retq
+;
+; X32-AVX-LABEL: testv32i8u:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vpxor %ymm1, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpsubb %ymm0, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpsubb {{\.LCPI.*}}, %ymm0, %ymm0
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm1 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15]
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm2
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm3 = [0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4,0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4]
+; X32-AVX-NEXT:    vpshufb %ymm2, %ymm3, %ymm2
+; X32-AVX-NEXT:    vpsrlw $4, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpand %ymm1, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpshufb %ymm0, %ymm3, %ymm0
+; X32-AVX-NEXT:    vpaddb %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    retl
   %out = call <32 x i8> @llvm.cttz.v32i8(<32 x i8> %in, i1 -1)
   ret <32 x i8> %out
 }
 
 define <4 x i64> @foldv4i64() nounwind {
-; ALL-LABEL: foldv4i64:
-; ALL:       # BB#0:
-; ALL-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,64,0]
-; ALL-NEXT:    retq
+; AVX-LABEL: foldv4i64:
+; AVX:       # BB#0:
+; AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,64,0]
+; AVX-NEXT:    retq
+;
+; X32-AVX-LABEL: foldv4i64:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm0 = [256,0,4294967295,4294967295,0,0,255,0]
+; X32-AVX-NEXT:    vpxor %ymm1, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpsubq %ymm0, %ymm1, %ymm2
+; X32-AVX-NEXT:    vpand %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpsubq {{\.LCPI.*}}, %ymm0, %ymm0
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm2 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15]
+; X32-AVX-NEXT:    vpand %ymm2, %ymm0, %ymm3
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm4 = [0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4,0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4]
+; X32-AVX-NEXT:    vpshufb %ymm3, %ymm4, %ymm3
+; X32-AVX-NEXT:    vpsrlw $4, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpand %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpshufb %ymm0, %ymm4, %ymm0
+; X32-AVX-NEXT:    vpaddb %ymm3, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpsadbw %ymm1, %ymm0, %ymm0
+; X32-AVX-NEXT:    retl
   %out = call <4 x i64> @llvm.cttz.v4i64(<4 x i64> <i64 256, i64 -1, i64 0, i64 255>, i1 0)
   ret <4 x i64> %out
 }
 
 define <4 x i64> @foldv4i64u() nounwind {
-; ALL-LABEL: foldv4i64u:
-; ALL:       # BB#0:
-; ALL-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,64,0]
-; ALL-NEXT:    retq
+; AVX-LABEL: foldv4i64u:
+; AVX:       # BB#0:
+; AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,64,0]
+; AVX-NEXT:    retq
+;
+; X32-AVX-LABEL: foldv4i64u:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm0 = [256,0,4294967295,4294967295,0,0,255,0]
+; X32-AVX-NEXT:    vpxor %ymm1, %ymm1, %ymm1
+; X32-AVX-NEXT:    vpsubq %ymm0, %ymm1, %ymm2
+; X32-AVX-NEXT:    vpand %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpsubq {{\.LCPI.*}}, %ymm0, %ymm0
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm2 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15]
+; X32-AVX-NEXT:    vpand %ymm2, %ymm0, %ymm3
+; X32-AVX-NEXT:    vmovdqa {{.*#+}} ymm4 = [0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4,0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4]
+; X32-AVX-NEXT:    vpshufb %ymm3, %ymm4, %ymm3
+; X32-AVX-NEXT:    vpsrlw $4, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpand %ymm2, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpshufb %ymm0, %ymm4, %ymm0
+; X32-AVX-NEXT:    vpaddb %ymm3, %ymm0, %ymm0
+; X32-AVX-NEXT:    vpsadbw %ymm1, %ymm0, %ymm0
+; X32-AVX-NEXT:    retl
   %out = call <4 x i64> @llvm.cttz.v4i64(<4 x i64> <i64 256, i64 -1, i64 0, i64 255>, i1 -1)
   ret <4 x i64> %out
 }
 
 define <8 x i32> @foldv8i32() nounwind {
-; ALL-LABEL: foldv8i32:
-; ALL:       # BB#0:
-; ALL-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,32,0,16,0,3,3]
-; ALL-NEXT:    retq
+; AVX-LABEL: foldv8i32:
+; AVX:       # BB#0:
+; AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,32,0,16,0,3,3]
+; AVX-NEXT:    retq
+;
+; X32-AVX-LABEL: foldv8i32:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,32,0,16,0,3,3]
+; X32-AVX-NEXT:    retl
   %out = call <8 x i32> @llvm.cttz.v8i32(<8 x i32> <i32 256, i32 -1, i32 0, i32 255, i32 -65536, i32 7, i32 24, i32 88>, i1 0)
   ret <8 x i32> %out
 }
 
 define <8 x i32> @foldv8i32u() nounwind {
-; ALL-LABEL: foldv8i32u:
-; ALL:       # BB#0:
-; ALL-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,32,0,16,0,3,3]
-; ALL-NEXT:    retq
+; AVX-LABEL: foldv8i32u:
+; AVX:       # BB#0:
+; AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,32,0,16,0,3,3]
+; AVX-NEXT:    retq
+;
+; X32-AVX-LABEL: foldv8i32u:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,32,0,16,0,3,3]
+; X32-AVX-NEXT:    retl
   %out = call <8 x i32> @llvm.cttz.v8i32(<8 x i32> <i32 256, i32 -1, i32 0, i32 255, i32 -65536, i32 7, i32 24, i32 88>, i1 -1)
   ret <8 x i32> %out
 }
 
 define <16 x i16> @foldv16i16() nounwind {
-; ALL-LABEL: foldv16i16:
-; ALL:       # BB#0:
-; ALL-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,16,0,16,0,3,3,1,1,0,1,2,3,4,5]
-; ALL-NEXT:    retq
+; AVX-LABEL: foldv16i16:
+; AVX:       # BB#0:
+; AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,16,0,16,0,3,3,1,1,0,1,2,3,4,5]
+; AVX-NEXT:    retq
+;
+; X32-AVX-LABEL: foldv16i16:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,16,0,16,0,3,3,1,1,0,1,2,3,4,5]
+; X32-AVX-NEXT:    retl
   %out = call <16 x i16> @llvm.cttz.v16i16(<16 x i16> <i16 256, i16 -1, i16 0, i16 255, i16 -65536, i16 7, i16 24, i16 88, i16 -2, i16 254, i16 1, i16 2, i16 4, i16 8, i16 16, i16 32>, i1 0)
   ret <16 x i16> %out
 }
 
 define <16 x i16> @foldv16i16u() nounwind {
-; ALL-LABEL: foldv16i16u:
-; ALL:       # BB#0:
-; ALL-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,16,0,16,0,3,3,1,1,0,1,2,3,4,5]
-; ALL-NEXT:    retq
+; AVX-LABEL: foldv16i16u:
+; AVX:       # BB#0:
+; AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,16,0,16,0,3,3,1,1,0,1,2,3,4,5]
+; AVX-NEXT:    retq
+;
+; X32-AVX-LABEL: foldv16i16u:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,16,0,16,0,3,3,1,1,0,1,2,3,4,5]
+; X32-AVX-NEXT:    retl
   %out = call <16 x i16> @llvm.cttz.v16i16(<16 x i16> <i16 256, i16 -1, i16 0, i16 255, i16 -65536, i16 7, i16 24, i16 88, i16 -2, i16 254, i16 1, i16 2, i16 4, i16 8, i16 16, i16 32>, i1 -1)
   ret <16 x i16> %out
 }
 
 define <32 x i8> @foldv32i8() nounwind {
-; ALL-LABEL: foldv32i8:
-; ALL:       # BB#0:
-; ALL-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,8,0,8,0,3,3,1,1,0,1,2,3,4,5,6,7,8,8,7,6,5,4,3,2,1,0,0,0,0,0]
-; ALL-NEXT:    retq
+; AVX-LABEL: foldv32i8:
+; AVX:       # BB#0:
+; AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,8,0,8,0,3,3,1,1,0,1,2,3,4,5,6,7,8,8,7,6,5,4,3,2,1,0,0,0,0,0]
+; AVX-NEXT:    retq
+;
+; X32-AVX-LABEL: foldv32i8:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,8,0,8,0,3,3,1,1,0,1,2,3,4,5,6,7,8,8,7,6,5,4,3,2,1,0,0,0,0,0]
+; X32-AVX-NEXT:    retl
   %out = call <32 x i8> @llvm.cttz.v32i8(<32 x i8> <i8 256, i8 -1, i8 0, i8 255, i8 -65536, i8 7, i8 24, i8 88, i8 -2, i8 254, i8 1, i8 2, i8 4, i8 8, i8 16, i8 32, i8 64, i8 128, i8 256, i8 -256, i8 -128, i8 -64, i8 -32, i8 -16, i8 -8, i8 -4, i8 -2, i8 -1, i8 3, i8 5, i8 7, i8 127>, i1 0)
   ret <32 x i8> %out
 }
 
 define <32 x i8> @foldv32i8u() nounwind {
-; ALL-LABEL: foldv32i8u:
-; ALL:       # BB#0:
-; ALL-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,8,0,8,0,3,3,1,1,0,1,2,3,4,5,6,7,8,8,7,6,5,4,3,2,1,0,0,0,0,0]
-; ALL-NEXT:    retq
+; AVX-LABEL: foldv32i8u:
+; AVX:       # BB#0:
+; AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,8,0,8,0,3,3,1,1,0,1,2,3,4,5,6,7,8,8,7,6,5,4,3,2,1,0,0,0,0,0]
+; AVX-NEXT:    retq
+;
+; X32-AVX-LABEL: foldv32i8u:
+; X32-AVX:       # BB#0:
+; X32-AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [8,0,8,0,8,0,3,3,1,1,0,1,2,3,4,5,6,7,8,8,7,6,5,4,3,2,1,0,0,0,0,0]
+; X32-AVX-NEXT:    retl
   %out = call <32 x i8> @llvm.cttz.v32i8(<32 x i8> <i8 256, i8 -1, i8 0, i8 255, i8 -65536, i8 7, i8 24, i8 88, i8 -2, i8 254, i8 1, i8 2, i8 4, i8 8, i8 16, i8 32, i8 64, i8 128, i8 256, i8 -256, i8 -128, i8 -64, i8 -32, i8 -16, i8 -8, i8 -4, i8 -2, i8 -1, i8 3, i8 5, i8 7, i8 127>, i1 -1)
   ret <32 x i8> %out
 }