RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-11-25 04:02:41 +01:00
Code Issues Projects Releases Wiki Activity

[SDAG] add check to sext-of-setcc fold to bypass changing a legal op

Browse Source 
I accidentaly pushed a draft of D103280 that was discussed
during the review, but it was not supposed to be the final
version.

Rather than revert and recommit, I'm updating the existing
code. This way we have a record of the codegen diff that
would result if we decide to remove this predicate in the
future.
This commit is contained in:
Sanjay Patel 2021-05-31 08:53:09 -04:00
parent 471f397407
commit 4f81b668e8
2 changed files with 10 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
lib/CodeGen/SelectionDAG/DAGCombiner.cpp
View File

@ -10941,8 +10941,8 @@ SDValue DAGCombiner::visitSIGN_EXTEND(SDNode *N) {
// Try to eliminate the sext of a setcc by zexting the compare operands.
// TODO: Handle signed compare by sexting the ops.
if (!ISD::isSignedIntSetCC(CC) && N0.hasOneUse() &&
TLI.isOperationLegalOrCustom(ISD::SETCC, VT) /*&&
!TLI.isOperationLegalOrCustom(ISD::SETCC, SVT)*/) {
TLI.isOperationLegalOrCustom(ISD::SETCC, VT) &&
!TLI.isOperationLegalOrCustom(ISD::SETCC, SVT)) {
// We have an unsupported narrow vector compare op that would be legal
// if extended to the destination type. See if the compare operands
// can be freely extended to the destination type.

17
test/CodeGen/X86/sext-vsetcc.ll
View File

@ -417,20 +417,19 @@ define <8 x i32> @cmp_ne_zextload_from_legal_op(<8 x i16>* %x, <8 x i16>* %y) {
;
; AVX2-LABEL: cmp_ne_zextload_from_legal_op:
; AVX2: # %bb.0:
; AVX2-NEXT: vpmovzxwd {{.*#+}} ymm0 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
; AVX2-NEXT: vpmovzxwd {{.*#+}} ymm1 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
; AVX2-NEXT: vpcmpeqd %ymm1, %ymm0, %ymm0
; AVX2-NEXT: vpcmpeqd %ymm1, %ymm1, %ymm1
; AVX2-NEXT: vpxor %ymm1, %ymm0, %ymm0
; AVX2-NEXT: vmovdqa (%rdi), %xmm0
; AVX2-NEXT: vpcmpeqw (%rsi), %xmm0, %xmm0
; AVX2-NEXT: vpcmpeqd %xmm1, %xmm1, %xmm1
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpmovsxwd %xmm0, %ymm0
; AVX2-NEXT: retq
;
; AVX512-LABEL: cmp_ne_zextload_from_legal_op:
; AVX512: # %bb.0:
; AVX512-NEXT: vpmovzxwd {{.*#+}} ymm0 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
; AVX512-NEXT: vpmovzxwd {{.*#+}} ymm1 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
; AVX512-NEXT: vpcmpeqd %ymm1, %ymm0, %ymm0
; AVX512-NEXT: vmovdqa (%rdi), %xmm0
; AVX512-NEXT: vpcmpeqw (%rsi), %xmm0, %xmm0
; AVX512-NEXT: vpternlogq $15, %zmm0, %zmm0, %zmm0
; AVX512-NEXT: # kill: def $ymm0 killed $ymm0 killed $zmm0
; AVX512-NEXT: vpmovsxwd %xmm0, %ymm0
; AVX512-NEXT: retq
%loadx = load <8 x i16>, <8 x i16>* %x
%loady = load <8 x i16>, <8 x i16>* %y