Identify and simplify idempotent intrinsics. Test case included.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@174650 91177308-0d34-0410-b5e6-96231b3b80d8
This commit is contained in:
Michael Ilseman 2013-02-07 19:26:05 +00:00
parent 96a6555b57
commit f89de816ae
2 changed files with 87 additions and 0 deletions

View File

@ -2925,6 +2925,37 @@ Value *llvm::SimplifyCmpInst(unsigned Predicate, Value *LHS, Value *RHS,
RecursionLimit); RecursionLimit);
} }
static bool IsIdempotent(Intrinsic::ID ID) {
switch (ID) {
default: return false;
// Unary idempotent: f(f(x)) = f(x)
case Intrinsic::fabs:
case Intrinsic::floor:
case Intrinsic::ceil:
case Intrinsic::trunc:
case Intrinsic::rint:
case Intrinsic::nearbyint:
return true;
}
}
template <typename IterTy>
static Value *SimplifyIntrinsic(Intrinsic::ID IID, IterTy ArgBegin, IterTy ArgEnd,
const Query &Q, unsigned MaxRecurse) {
// Perform idempotent optimizations
if (!IsIdempotent(IID))
return 0;
// Unary Ops
if (std::distance(ArgBegin, ArgEnd) == 1)
if (IntrinsicInst *II = dyn_cast<IntrinsicInst>(*ArgBegin))
if (II->getIntrinsicID() == IID)
return II;
return 0;
}
template <typename IterTy> template <typename IterTy>
static Value *SimplifyCall(Value *V, IterTy ArgBegin, IterTy ArgEnd, static Value *SimplifyCall(Value *V, IterTy ArgBegin, IterTy ArgEnd,
const Query &Q, unsigned MaxRecurse) { const Query &Q, unsigned MaxRecurse) {
@ -2941,6 +2972,11 @@ static Value *SimplifyCall(Value *V, IterTy ArgBegin, IterTy ArgEnd,
if (!F) if (!F)
return 0; return 0;
if (unsigned IID = F->getIntrinsicID())
if (Value *Ret =
SimplifyIntrinsic((Intrinsic::ID) IID, ArgBegin, ArgEnd, Q, MaxRecurse))
return Ret;
if (!canConstantFoldCallTo(F)) if (!canConstantFoldCallTo(F))
return 0; return 0;

View File

@ -50,3 +50,54 @@ define float @test_fabs_libcall() {
ret float %x ret float %x
; CHECK-NEXT: ret float 4.2{{0+}}e+01 ; CHECK-NEXT: ret float 4.2{{0+}}e+01
} }
declare float @llvm.fabs.f32(float) nounwind readnone
declare float @llvm.floor.f32(float) nounwind readnone
declare float @llvm.ceil.f32(float) nounwind readnone
declare float @llvm.trunc.f32(float) nounwind readnone
declare float @llvm.rint.f32(float) nounwind readnone
declare float @llvm.nearbyint.f32(float) nounwind readnone
; Test idempotent intrinsics
define float @test_idempotence(float %a) {
; CHECK: @test_idempotence
; CHECK: fabs
; CHECK-NOT: fabs
%a0 = call float @llvm.fabs.f32(float %a)
%a1 = call float @llvm.fabs.f32(float %a0)
; CHECK: floor
; CHECK-NOT: floor
%b0 = call float @llvm.floor.f32(float %a)
%b1 = call float @llvm.floor.f32(float %b0)
; CHECK: ceil
; CHECK-NOT: ceil
%c0 = call float @llvm.ceil.f32(float %a)
%c1 = call float @llvm.ceil.f32(float %c0)
; CHECK: trunc
; CHECK-NOT: trunc
%d0 = call float @llvm.trunc.f32(float %a)
%d1 = call float @llvm.trunc.f32(float %d0)
; CHECK: rint
; CHECK-NOT: rint
%e0 = call float @llvm.rint.f32(float %a)
%e1 = call float @llvm.rint.f32(float %e0)
; CHECK: nearbyint
; CHECK-NOT: nearbyint
%f0 = call float @llvm.nearbyint.f32(float %a)
%f1 = call float @llvm.nearbyint.f32(float %f0)
%r0 = fadd float %a1, %b1
%r1 = fadd float %r0, %c1
%r2 = fadd float %r1, %d1
%r3 = fadd float %r2, %e1
%r4 = fadd float %r3, %f1
ret float %r4
}