[ARM] Fix for MVE VREV64

The VREV64 instruction is apparently unpredictable if Qd == Qm, due to the
cross-beat nature of the instruction. This adds an earlyclobber to Qd, which
seems to be the same way we deal with this on other instructions like the
write-back on loads and stores.

Differential Revision: https://reviews.llvm.org/D65502

llvm-svn: 367544

diff --git a/llvm/lib/Target/ARM/ARMInstrMVE.td b/llvm/lib/Target/ARM/ARMInstrMVE.td
index 198d31f5784b32156d7a13339183bb82ab32d2fc..3e8f726bde4c0ecd17d762c3a6538daf3f399559 100644 (file)
--- a/llvm/lib/Target/ARM/ARMInstrMVE.td
+++ b/llvm/lib/Target/ARM/ARMInstrMVE.td
@@ -985,9 +985,9 @@ def MVE_VBIC : MVE_bit_arith<(outs MQPR:$Qd), (ins MQPR:$Qn, MQPR:$Qm),
   let Inst{0} = 0b0;
 }
 
-class MVE_VREV<string iname, string suffix, bits<2> size, bits<2> bit_8_7>
+class MVE_VREV<string iname, string suffix, bits<2> size, bits<2> bit_8_7, string cstr="">
   : MVE_bit_arith<(outs MQPR:$Qd), (ins MQPR:$Qm), iname,
-                  suffix, "$Qd, $Qm", ""> {
+                  suffix, "$Qd, $Qm", cstr> {
 
   let Inst{28} = 0b1;
   let Inst{25-23} = 0b111;
@@ -1001,9 +1001,9 @@ class MVE_VREV<string iname, string suffix, bits<2> size, bits<2> bit_8_7>
   let Inst{0} = 0b0;
 }
 
-def MVE_VREV64_8  : MVE_VREV<"vrev64", "8", 0b00, 0b00>;
-def MVE_VREV64_16 : MVE_VREV<"vrev64", "16", 0b01, 0b00>;
-def MVE_VREV64_32 : MVE_VREV<"vrev64", "32", 0b10, 0b00>;
+def MVE_VREV64_8  : MVE_VREV<"vrev64", "8", 0b00, 0b00, "@earlyclobber $Qd">;
+def MVE_VREV64_16 : MVE_VREV<"vrev64", "16", 0b01, 0b00, "@earlyclobber $Qd">;
+def MVE_VREV64_32 : MVE_VREV<"vrev64", "32", 0b10, 0b00, "@earlyclobber $Qd">;
 
 def MVE_VREV32_8  : MVE_VREV<"vrev32", "8", 0b00, 0b01>;
 def MVE_VREV32_16 : MVE_VREV<"vrev32", "16", 0b01, 0b01>;

diff --git a/llvm/test/CodeGen/Thumb2/mve-shuffle.ll b/llvm/test/CodeGen/Thumb2/mve-shuffle.ll
index b5c96f9f86dc3d665c01b4aa4390b07b689bc482..90f09af135a29443cad71de55c9cccbbb785a791 100644 (file)
--- a/llvm/test/CodeGen/Thumb2/mve-shuffle.ll
+++ b/llvm/test/CodeGen/Thumb2/mve-shuffle.ll
@@ -42,7 +42,8 @@ entry:
 define arm_aapcs_vfpcc <4 x i32> @shuffle5_i32(<4 x i32> %src) {
 ; CHECK-LABEL: shuffle5_i32:
 ; CHECK:       @ %bb.0: @ %entry
-; CHECK-NEXT:    vrev64.32 q0, q0
+; CHECK-NEXT:    vrev64.32 q1, q0
+; CHECK-NEXT:    vmov q0, q1
 ; CHECK-NEXT:    bx lr
 entry:
   %out = shufflevector <4 x i32> %src, <4 x i32> undef, <4 x i32> <i32 1, i32 0, i32 3, i32 2>
@@ -122,7 +123,8 @@ entry:
 define arm_aapcs_vfpcc <8 x i16> @shuffle5_i16(<8 x i16> %src) {
 ; CHECK-LABEL: shuffle5_i16:
 ; CHECK:       @ %bb.0: @ %entry
-; CHECK-NEXT:    vrev64.16 q0, q0
+; CHECK-NEXT:    vrev64.16 q1, q0
+; CHECK-NEXT:    vmov q0, q1
 ; CHECK-NEXT:    bx lr
 entry:
   %out = shufflevector <8 x i16> %src, <8 x i16> undef, <8 x i32> <i32 3, i32 2, i32 1, i32 0, i32 7, i32 6, i32 5, i32 4>
@@ -235,7 +237,8 @@ entry:
 define arm_aapcs_vfpcc <16 x i8> @shuffle5_i8(<16 x i8> %src) {
 ; CHECK-LABEL: shuffle5_i8:
 ; CHECK:       @ %bb.0: @ %entry
-; CHECK-NEXT:    vrev64.8 q0, q0
+; CHECK-NEXT:    vrev64.8 q1, q0
+; CHECK-NEXT:    vmov q0, q1
 ; CHECK-NEXT:    bx lr
 entry:
   %out = shufflevector <16 x i8> %src, <16 x i8> undef, <16 x i32> <i32 7, i32 6, i32 5, i32 4, i32 3, i32 2, i32 1, i32 0, i32 15, i32 14, i32 13, i32 12, i32 11, i32 10, i32 9, i32 8>
@@ -334,7 +337,8 @@ entry:
 define arm_aapcs_vfpcc <4 x float> @shuffle5_f32(<4 x float> %src) {
 ; CHECK-LABEL: shuffle5_f32:
 ; CHECK:       @ %bb.0: @ %entry
-; CHECK-NEXT:    vrev64.32 q0, q0
+; CHECK-NEXT:    vrev64.32 q1, q0
+; CHECK-NEXT:    vmov q0, q1
 ; CHECK-NEXT:    bx lr
 entry:
   %out = shufflevector <4 x float> %src, <4 x float> undef, <4 x i32> <i32 1, i32 0, i32 3, i32 2>
@@ -405,7 +409,8 @@ entry:
 define arm_aapcs_vfpcc <8 x half> @shuffle5_f16(<8 x half> %src) {
 ; CHECK-LABEL: shuffle5_f16:
 ; CHECK:       @ %bb.0: @ %entry
-; CHECK-NEXT:    vrev64.16 q0, q0
+; CHECK-NEXT:    vrev64.16 q1, q0
+; CHECK-NEXT:    vmov q0, q1
 ; CHECK-NEXT:    bx lr
 entry:
   %out = shufflevector <8 x half> %src, <8 x half> undef, <8 x i32> <i32 3, i32 2, i32 1, i32 0, i32 7, i32 6, i32 5, i32 4>