[X86] Add use condition for combineSetCCMOVMSK

Reviewed by RKSimon, LuoYuanke Differential Revision: https://reviews.llvm.org/D123652
llvm · Apr 26, 2022 · c430f0f · c430f0f
1 parent d6f9000
commit c430f0f
Show file tree

Hide file tree

Showing 3 changed files with 15 additions and 12 deletions.
diff --git a/llvm/lib/Target/X86/X86ISelLowering.cpp b/llvm/lib/Target/X86/X86ISelLowering.cpp
@@ -45012,6 +45012,12 @@ static SDValue combineSetCCMOVMSK(SDValue EFLAGS, X86::CondCode &CC,
   if (!IsAnyOf && !IsAllOf)
     return SDValue();
 
+  // TODO: Check more combining cases for me.
+  // Here we check the cmp use number to decide do combining or not.
+  // Currently we only get 2 tests about combining "MOVMSK(CONCAT(..))"
+  // and "MOVMSK(PCMPEQ(..))" are fit to use this constraint.
+  bool IsOneUse = CmpOp.getNode()->hasOneUse();
+
   // See if we can peek through to a vector with a wider element type, if the
   // signbits extend down to all the sub-elements as well.
   // Calling MOVMSK with the wider type, avoiding the bitcast, helps expose
@@ -45040,7 +45046,7 @@ static SDValue combineSetCCMOVMSK(SDValue EFLAGS, X86::CondCode &CC,
   // MOVMSK(CONCAT(X,Y)) != 0 ->  MOVMSK(OR(X,Y)).
   // MOVMSK(CONCAT(X,Y)) == -1 ->  MOVMSK(AND(X,Y)).
   // MOVMSK(CONCAT(X,Y)) != -1 ->  MOVMSK(AND(X,Y)).
-  if (VecVT.is256BitVector() && NumElts <= CmpBits) {
+  if (VecVT.is256BitVector() && NumElts <= CmpBits && IsOneUse) {
     SmallVector<SDValue> Ops;
     if (collectConcatOps(peekThroughBitcasts(Vec).getNode(), Ops) &&
         Ops.size() == 2) {
@@ -45061,7 +45067,7 @@ static SDValue combineSetCCMOVMSK(SDValue EFLAGS, X86::CondCode &CC,
   // MOVMSK(PCMPEQ(X,0)) != -1 -> !PTESTZ(X,X).
   // MOVMSK(PCMPEQ(X,Y)) == -1 -> PTESTZ(SUB(X,Y),SUB(X,Y)).
   // MOVMSK(PCMPEQ(X,Y)) != -1 -> !PTESTZ(SUB(X,Y),SUB(X,Y)).
-  if (IsAllOf && Subtarget.hasSSE41()) {
+  if (IsAllOf && Subtarget.hasSSE41() && IsOneUse) {
     MVT TestVT = VecVT.is128BitVector() ? MVT::v2i64 : MVT::v4i64;
     SDValue BC = peekThroughBitcasts(Vec);
     // Ensure MOVMSK was testing every signbit of BC.

diff --git a/llvm/test/CodeGen/X86/vector-compare-all_of.ll b/llvm/test/CodeGen/X86/vector-compare-all_of.ll
@@ -881,7 +881,7 @@ define i8 @test_v32i8_sext(<32 x i8> %a0, <32 x i8> %a1) {
   ret i8 %11
 }
 
-; FIXME: Should not "MOVMSK(PCMPEQ(..)) -> PTESTZ(..)" when cmp result has muti-uses.
+; Should not "MOVMSK(PCMPEQ(..)) -> PTESTZ(..)" when cmp result has muti-uses.
 define i32 @test_v32i8_muti_uses(<32 x i8> %x, <32 x i8>%y, i32 %z) {
 ; SSE-LABEL: test_v32i8_muti_uses:
 ; SSE:       # %bb.0:
@@ -914,10 +914,9 @@ define i32 @test_v32i8_muti_uses(<32 x i8> %x, <32 x i8>%y, i32 %z) {
 ;
 ; AVX2-LABEL: test_v32i8_muti_uses:
 ; AVX2:       # %bb.0:
-; AVX2-NEXT:    vpcmpeqb %ymm1, %ymm0, %ymm2
-; AVX2-NEXT:    vpmovmskb %ymm2, %ecx
-; AVX2-NEXT:    vpsubb %ymm1, %ymm0, %ymm0
-; AVX2-NEXT:    vptest %ymm0, %ymm0
+; AVX2-NEXT:    vpcmpeqb %ymm1, %ymm0, %ymm0
+; AVX2-NEXT:    vpmovmskb %ymm0, %ecx
+; AVX2-NEXT:    cmpl $-1, %ecx
 ; AVX2-NEXT:    movl $16, %eax
 ; AVX2-NEXT:    cmovnel %ecx, %eax
 ; AVX2-NEXT:    vzeroupper

diff --git a/llvm/test/CodeGen/X86/vector-compare-any_of.ll b/llvm/test/CodeGen/X86/vector-compare-any_of.ll
@@ -1358,11 +1358,9 @@ define {i32, i1} @test_v16i8_muti_uses(<16 x i8> %x, <16 x i8>%y, <16 x i8> %z)
 ; AVX2:       # %bb.0:
 ; AVX2-NEXT:    vpcmpeqb %xmm1, %xmm0, %xmm0
 ; AVX2-NEXT:    vpcmpeqb %xmm1, %xmm2, %xmm1
-; AVX2-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm2
-; AVX2-NEXT:    vpmovmskb %ymm2, %eax
-; AVX2-NEXT:    vpor %xmm1, %xmm0, %xmm0
-; AVX2-NEXT:    vpmovmskb %xmm0, %ecx
-; AVX2-NEXT:    testl %ecx, %ecx
+; AVX2-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-NEXT:    vpmovmskb %ymm0, %eax
+; AVX2-NEXT:    testl %eax, %eax
 ; AVX2-NEXT:    sete %dl
 ; AVX2-NEXT:    vzeroupper
 ; AVX2-NEXT:    retq