[PATCH] D31204: AMDGPU: Fold rcp/rsq of undef to undef
Matt Arsenault via Phabricator via llvm-commits
llvm-commits at lists.llvm.org
Tue Mar 21 13:29:13 PDT 2017
arsenm created this revision.
Herald added subscribers: tpr, dstuttard, tony-tye, yaxunl, wdng.
https://reviews.llvm.org/D31204
Files:
lib/Transforms/InstCombine/InstCombineCalls.cpp
test/Transforms/InstCombine/amdgcn-intrinsics.ll
Index: test/Transforms/InstCombine/amdgcn-intrinsics.ll
===================================================================
--- test/Transforms/InstCombine/amdgcn-intrinsics.ll
+++ test/Transforms/InstCombine/amdgcn-intrinsics.ll
@@ -7,6 +7,12 @@
declare float @llvm.amdgcn.rcp.f32(float) nounwind readnone
declare double @llvm.amdgcn.rcp.f64(double) nounwind readnone
+; CHECK-LABEL: @test_constant_fold_rcp_f32_undef
+; CHECK-NEXT: ret float undef
+define float @test_constant_fold_rcp_f32_undef() nounwind {
+ %val = call float @llvm.amdgcn.rcp.f32(float undef) nounwind readnone
+ ret float %val
+}
; CHECK-LABEL: @test_constant_fold_rcp_f32_1
; CHECK-NEXT: ret float 1.000000e+00
@@ -50,6 +56,18 @@
ret double %val
}
+; --------------------------------------------------------------------
+; llvm.amdgcn.rsq
+; --------------------------------------------------------------------
+
+declare float @llvm.amdgcn.rsq.f32(float) nounwind readnone
+
+; CHECK-LABEL: @test_constant_fold_rsq_f32_undef
+; CHECK-NEXT: ret float undef
+define float @test_constant_fold_rsq_f32_undef() nounwind {
+ %val = call float @llvm.amdgcn.rsq.f32(float undef) nounwind readnone
+ ret float %val
+}
; --------------------------------------------------------------------
; llvm.amdgcn.frexp.mant
Index: lib/Transforms/InstCombine/InstCombineCalls.cpp
===================================================================
--- lib/Transforms/InstCombine/InstCombineCalls.cpp
+++ lib/Transforms/InstCombine/InstCombineCalls.cpp
@@ -3065,9 +3065,14 @@
break;
}
-
case Intrinsic::amdgcn_rcp: {
- if (const ConstantFP *C = dyn_cast<ConstantFP>(II->getArgOperand(0))) {
+ Value *Src = II->getArgOperand(0);
+
+ // TODO: Move to ConstantFolding/InstSimplify?
+ if (isa<UndefValue>(Src))
+ return replaceInstUsesWith(CI, Src);
+
+ if (const ConstantFP *C = dyn_cast<ConstantFP>(Src)) {
const APFloat &ArgVal = C->getValueAPF();
APFloat Val(ArgVal.getSemantics(), 1.0);
APFloat::opStatus Status = Val.divide(ArgVal,
@@ -3080,6 +3085,14 @@
break;
}
+ case Intrinsic::amdgcn_rsq: {
+ Value *Src = II->getArgOperand(0);
+
+ // TODO: Move to ConstantFolding/InstSimplify?
+ if (isa<UndefValue>(Src))
+ return replaceInstUsesWith(CI, Src);
+ break;
+ }
case Intrinsic::amdgcn_frexp_mant:
case Intrinsic::amdgcn_frexp_exp: {
Value *Src = II->getArgOperand(0);
-------------- next part --------------
A non-text attachment was scrubbed...
Name: D31204.92535.patch
Type: text/x-patch
Size: 2442 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20170321/2ed9b483/attachment-0001.bin>
More information about the llvm-commits
mailing list