Skip to content

Commit

Permalink
[InstCombine] Remove some uses with replaceUndefsWith() (#89190)
Browse files Browse the repository at this point in the history 
Now that we don't accept undef splat in PatternMatch, we can remove some
uses of replaceUndefsWith(). I believe in all these cases only poison
splats are possible now, in which case no replacement is necessary.
  • Loading branch information
@nikic
nikic committed Apr 19, 2024
1 parent 3f2371e commit eb7ad88
Show file tree
Hide file tree
Showing 11 changed files with 26 additions and 58 deletions.
14 changes: 2 additions & 12 deletions llvm/lib/Transforms/InstCombine/InstCombineAndOrXor.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -4433,23 +4433,13 @@ Instruction *InstCombinerImpl::foldNot(BinaryOperator &I) {
// ~(C >>s Y) --> ~C >>u Y (when inverting the replicated sign bits)
Constant *C;
if (match(NotVal, m_AShr(m_Constant(C), m_Value(Y))) &&
match(C, m_Negative())) {
// We matched a negative constant, so propagating undef is unsafe.
// Clamp undef elements to -1.
Type *EltTy = Ty->getScalarType();
C = Constant::replaceUndefsWith(C, ConstantInt::getAllOnesValue(EltTy));
match(C, m_Negative()))
return BinaryOperator::CreateLShr(ConstantExpr::getNot(C), Y);
}

// ~(C >>u Y) --> ~C >>s Y (when inverting the replicated sign bits)
if (match(NotVal, m_LShr(m_Constant(C), m_Value(Y))) &&
match(C, m_NonNegative())) {
// We matched a non-negative constant, so propagating undef is unsafe.
// Clamp undef elements to 0.
Type *EltTy = Ty->getScalarType();
C = Constant::replaceUndefsWith(C, ConstantInt::getNullValue(EltTy));
match(C, m_NonNegative()))
return BinaryOperator::CreateAShr(ConstantExpr::getNot(C), Y);
}

// ~(X + C) --> ~C - X
if (match(NotVal, m_Add(m_Value(X), m_ImmConstant(C))))
Expand Down
19 changes: 0 additions & 19 deletions llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -4322,25 +4322,6 @@ static Value *foldICmpWithLowBitMaskedVal(ICmpInst::Predicate Pred, Value *Op0,
if (!IsLowBitMask())
return nullptr;

// The mask value may be a vector constant that has undefined elements. But
// it may not be safe to propagate those undefs into the new compare, so
// replace those elements by copying an existing, defined, and safe scalar
// constant.
Type *OpTy = M->getType();
auto *VecC = dyn_cast<Constant>(M);
auto *OpVTy = dyn_cast<FixedVectorType>(OpTy);
if (OpVTy && VecC && VecC->containsUndefOrPoisonElement()) {
Constant *SafeReplacementConstant = nullptr;
for (unsigned i = 0, e = OpVTy->getNumElements(); i != e; ++i) {
if (!isa<UndefValue>(VecC->getAggregateElement(i))) {
SafeReplacementConstant = VecC->getAggregateElement(i);
break;
}
}
assert(SafeReplacementConstant && "Failed to find undef replacement");
M = Constant::replaceUndefsWith(VecC, SafeReplacementConstant);
}

return IC.Builder.CreateICmp(DstPred, X, M);
}

Expand Down
3 changes: 0 additions & 3 deletions llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -842,9 +842,6 @@ Instruction *InstCombinerImpl::visitFMul(BinaryOperator &I) {
isKnownNeverNaN(&I, /*Depth=*/0, SQ.getWithInstruction(&I)))) {
if (FPC->isNegative())
Op0 = Builder.CreateFNegFMF(Op0, &I);
Op1 = Constant::replaceUndefsWith(
cast<Constant>(Op1),
ConstantFP::get(Op1->getType()->getScalarType(), *FPC));
CallInst *CopySign = Builder.CreateIntrinsic(Intrinsic::copysign,
{I.getType()}, {Op1, Op0}, &I);
return replaceInstUsesWith(I, CopySign);
Expand Down
8 changes: 4 additions & 4 deletions llvm/test/Transforms/InstCombine/binop-itofp.ll
View file
Original file line number Diff line number Diff line change
Expand Up @@ -1070,7 +1070,7 @@ define <2 x float> @nonzero_check_on_constant_for_si_fmul_vec_w_poison(i1 %c, i1
; CHECK-NEXT: [[CONV_I_V:%.*]] = insertelement <2 x i16> poison, i16 [[CONV_I_S]], i64 0
; CHECK-NEXT: [[CONV_I:%.*]] = shufflevector <2 x i16> [[CONV_I_V]], <2 x i16> poison, <2 x i32> zeroinitializer
; CHECK-NEXT: [[CONV1_I:%.*]] = sitofp <2 x i16> [[CONV_I]] to <2 x float>
; CHECK-NEXT: [[MUL3_I_I:%.*]] = call <2 x float> @llvm.copysign.v2f32(<2 x float> zeroinitializer, <2 x float> [[CONV1_I]])
; CHECK-NEXT: [[MUL3_I_I:%.*]] = call <2 x float> @llvm.copysign.v2f32(<2 x float> <float poison, float 0.000000e+00>, <2 x float> [[CONV1_I]])
; CHECK-NEXT: store i32 [[SEL]], ptr [[G_2345:%.*]], align 4
; CHECK-NEXT: ret <2 x float> [[MUL3_I_I]]
;
Expand All @@ -1090,9 +1090,9 @@ define <2 x float> @nonzero_check_on_constant_for_si_fmul_nz_vec_w_poison(i1 %c,
; CHECK-NEXT: [[CONV_I_S:%.*]] = trunc nuw i32 [[SEL]] to i16
; CHECK-NEXT: [[CONV_I_V:%.*]] = insertelement <2 x i16> poison, i16 [[CONV_I_S]], i64 0
; CHECK-NEXT: [[CONV_I:%.*]] = shufflevector <2 x i16> [[CONV_I_V]], <2 x i16> poison, <2 x i32> zeroinitializer
; CHECK-NEXT: [[CONV1_I:%.*]] = sitofp <2 x i16> [[CONV_I]] to <2 x float>
; CHECK-NEXT: [[MUL3_I_I:%.*]] = sitofp <2 x i16> [[CONV_I]] to <2 x float>
; CHECK-NEXT: store i32 [[SEL]], ptr [[G_2345:%.*]], align 4
; CHECK-NEXT: ret <2 x float> [[CONV1_I]]
; CHECK-NEXT: ret <2 x float> [[MUL3_I_I]]
;
%sel = select i1 %c, i32 65529, i32 53264
%conv.i.s = trunc i32 %sel to i16
Expand All @@ -1112,7 +1112,7 @@ define <2 x float> @nonzero_check_on_constant_for_si_fmul_negz_vec_w_poison(i1 %
; CHECK-NEXT: [[CONV_I:%.*]] = shufflevector <2 x i16> [[CONV_I_V]], <2 x i16> poison, <2 x i32> zeroinitializer
; CHECK-NEXT: [[CONV1_I:%.*]] = sitofp <2 x i16> [[CONV_I]] to <2 x float>
; CHECK-NEXT: [[TMP1:%.*]] = fneg <2 x float> [[CONV1_I]]
; CHECK-NEXT: [[MUL3_I_I:%.*]] = call <2 x float> @llvm.copysign.v2f32(<2 x float> zeroinitializer, <2 x float> [[TMP1]])
; CHECK-NEXT: [[MUL3_I_I:%.*]] = call <2 x float> @llvm.copysign.v2f32(<2 x float> <float poison, float -0.000000e+00>, <2 x float> [[TMP1]])
; CHECK-NEXT: store i32 [[SEL]], ptr [[G_2345:%.*]], align 4
; CHECK-NEXT: ret <2 x float> [[MUL3_I_I]]
;
Expand Down
4 changes: 2 additions & 2 deletions ...test/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-ne-to-icmp-ugt.ll
View file
Original file line number Diff line number Diff line change
Expand Up @@ -81,7 +81,7 @@ define <2 x i1> @p2_vec_nonsplat_edgecase1(<2 x i8> %x) {

define <3 x i1> @p3_vec_splat_poison(<3 x i8> %x) {
; CHECK-LABEL: @p3_vec_splat_poison(
; CHECK-NEXT: [[RET:%.*]] = icmp ugt <3 x i8> [[X:%.*]], <i8 3, i8 3, i8 3>
; CHECK-NEXT: [[RET:%.*]] = icmp ugt <3 x i8> [[X:%.*]], <i8 3, i8 poison, i8 3>
; CHECK-NEXT: ret <3 x i1> [[RET]]
;
%tmp0 = and <3 x i8> %x, <i8 3, i8 poison, i8 3>
Expand All @@ -91,7 +91,7 @@ define <3 x i1> @p3_vec_splat_poison(<3 x i8> %x) {

define <3 x i1> @p3_vec_nonsplat_poison(<3 x i8> %x) {
; CHECK-LABEL: @p3_vec_nonsplat_poison(
; CHECK-NEXT: [[RET:%.*]] = icmp ugt <3 x i8> [[X:%.*]], <i8 -1, i8 -1, i8 3>
; CHECK-NEXT: [[RET:%.*]] = icmp ugt <3 x i8> [[X:%.*]], <i8 -1, i8 poison, i8 3>
; CHECK-NEXT: ret <3 x i1> [[RET]]
;
%tmp0 = and <3 x i8> %x, <i8 -1, i8 poison, i8 3>
Expand Down
4 changes: 2 additions & 2 deletions ...est/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-sgt-to-icmp-sgt.ll
View file
Original file line number Diff line number Diff line change
Expand Up @@ -75,7 +75,7 @@ define <2 x i1> @p2_vec_nonsplat_edgecase() {
define <3 x i1> @p3_vec_splat_poison() {
; CHECK-LABEL: @p3_vec_splat_poison(
; CHECK-NEXT: [[X:%.*]] = call <3 x i8> @gen3x8()
; CHECK-NEXT: [[RET:%.*]] = icmp sgt <3 x i8> [[X]], <i8 3, i8 3, i8 3>
; CHECK-NEXT: [[RET:%.*]] = icmp sgt <3 x i8> [[X]], <i8 3, i8 poison, i8 3>
; CHECK-NEXT: ret <3 x i1> [[RET]]
;
%x = call <3 x i8> @gen3x8()
Expand All @@ -87,7 +87,7 @@ define <3 x i1> @p3_vec_splat_poison() {
define <3 x i1> @p3_vec_nonsplat_poison() {
; CHECK-LABEL: @p3_vec_nonsplat_poison(
; CHECK-NEXT: [[X:%.*]] = call <3 x i8> @gen3x8()
; CHECK-NEXT: [[RET:%.*]] = icmp sgt <3 x i8> [[X]], <i8 15, i8 3, i8 15>
; CHECK-NEXT: [[RET:%.*]] = icmp sgt <3 x i8> [[X]], <i8 15, i8 3, i8 poison>
; CHECK-NEXT: ret <3 x i1> [[RET]]
;
%x = call <3 x i8> @gen3x8()
Expand Down
4 changes: 2 additions & 2 deletions ...est/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-slt-to-icmp-sgt.ll
View file
Original file line number Diff line number Diff line change
Expand Up @@ -60,7 +60,7 @@ define <2 x i1> @p2_vec_nonsplat_edgecase(<2 x i8> %x) {

define <3 x i1> @p3_vec_splat_poison(<3 x i8> %x) {
; CHECK-LABEL: @p3_vec_splat_poison(
; CHECK-NEXT: [[RET:%.*]] = icmp sgt <3 x i8> [[X:%.*]], <i8 3, i8 3, i8 3>
; CHECK-NEXT: [[RET:%.*]] = icmp sgt <3 x i8> [[X:%.*]], <i8 3, i8 poison, i8 3>
; CHECK-NEXT: ret <3 x i1> [[RET]]
;
%tmp0 = and <3 x i8> %x, <i8 3, i8 poison, i8 3>
Expand All @@ -70,7 +70,7 @@ define <3 x i1> @p3_vec_splat_poison(<3 x i8> %x) {

define <3 x i1> @p3_vec_nonsplat_poison(<3 x i8> %x) {
; CHECK-LABEL: @p3_vec_nonsplat_poison(
; CHECK-NEXT: [[RET:%.*]] = icmp sgt <3 x i8> [[X:%.*]], <i8 15, i8 15, i8 3>
; CHECK-NEXT: [[RET:%.*]] = icmp sgt <3 x i8> [[X:%.*]], <i8 poison, i8 15, i8 3>
; CHECK-NEXT: ret <3 x i1> [[RET]]
;
%tmp0 = and <3 x i8> %x, <i8 poison, i8 15, i8 3>
Expand Down
4 changes: 2 additions & 2 deletions ...est/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-ugt-to-icmp-ugt.ll
View file
Original file line number Diff line number Diff line change
Expand Up @@ -98,7 +98,7 @@ define <2 x i1> @p2_vec_nonsplat_edgecase1() {
define <3 x i1> @p3_vec_splat_poison() {
; CHECK-LABEL: @p3_vec_splat_poison(
; CHECK-NEXT: [[X:%.*]] = call <3 x i8> @gen3x8()
; CHECK-NEXT: [[RET:%.*]] = icmp ugt <3 x i8> [[X]], <i8 3, i8 3, i8 3>
; CHECK-NEXT: [[RET:%.*]] = icmp ugt <3 x i8> [[X]], <i8 3, i8 poison, i8 3>
; CHECK-NEXT: ret <3 x i1> [[RET]]
;
%x = call <3 x i8> @gen3x8()
Expand All @@ -110,7 +110,7 @@ define <3 x i1> @p3_vec_splat_poison() {
define <3 x i1> @p3_vec_nonsplat_poison() {
; CHECK-LABEL: @p3_vec_nonsplat_poison(
; CHECK-NEXT: [[X:%.*]] = call <3 x i8> @gen3x8()
; CHECK-NEXT: [[RET:%.*]] = icmp ugt <3 x i8> [[X]], <i8 3, i8 3, i8 15>
; CHECK-NEXT: [[RET:%.*]] = icmp ugt <3 x i8> [[X]], <i8 3, i8 poison, i8 15>
; CHECK-NEXT: ret <3 x i1> [[RET]]
;
%x = call <3 x i8> @gen3x8()
Expand Down
4 changes: 2 additions & 2 deletions ...est/Transforms/InstCombine/canonicalize-constant-low-bit-mask-and-icmp-ult-to-icmp-ugt.ll
View file
Original file line number Diff line number Diff line change
Expand Up @@ -82,7 +82,7 @@ define <2 x i1> @p2_vec_nonsplat_edgecase1(<2 x i8> %x) {

define <3 x i1> @p3_vec_splat_poison(<3 x i8> %x) {
; CHECK-LABEL: @p3_vec_splat_poison(
; CHECK-NEXT: [[RET:%.*]] = icmp ugt <3 x i8> [[X:%.*]], <i8 3, i8 3, i8 3>
; CHECK-NEXT: [[RET:%.*]] = icmp ugt <3 x i8> [[X:%.*]], <i8 3, i8 poison, i8 3>
; CHECK-NEXT: ret <3 x i1> [[RET]]
;
%tmp0 = and <3 x i8> %x, <i8 3, i8 poison, i8 3>
Expand All @@ -92,7 +92,7 @@ define <3 x i1> @p3_vec_splat_poison(<3 x i8> %x) {

define <3 x i1> @p3_vec_nonsplat_poison(<3 x i8> %x) {
; CHECK-LABEL: @p3_vec_nonsplat_poison(
; CHECK-NEXT: [[RET:%.*]] = icmp ugt <3 x i8> [[X:%.*]], <i8 7, i8 31, i8 7>
; CHECK-NEXT: [[RET:%.*]] = icmp ugt <3 x i8> [[X:%.*]], <i8 7, i8 31, i8 poison>
; CHECK-NEXT: ret <3 x i1> [[RET]]
;
%tmp0 = and <3 x i8> %x, <i8 7, i8 31, i8 poison>
Expand Down
16 changes: 8 additions & 8 deletions llvm/test/Transforms/InstCombine/fmul.ll
View file
Original file line number Diff line number Diff line change
Expand Up @@ -633,11 +633,11 @@ define float @log2half(float %x, float %y) {

define float @log2half_commute(float %x1, float %y) {
; CHECK-LABEL: @log2half_commute(
; CHECK-NEXT: [[X1:%.*]] = fmul fast float [[X2:%.*]], 0x3FC24924A0000000
; CHECK-NEXT: [[X:%.*]] = fmul fast float [[X1:%.*]], 0x3FC24924A0000000
; CHECK-NEXT: [[TMP1:%.*]] = call fast float @llvm.log2.f32(float [[Y:%.*]])
; CHECK-NEXT: [[TMP2:%.*]] = fmul fast float [[TMP1]], [[X1]]
; CHECK-NEXT: [[TMP3:%.*]] = fsub fast float [[TMP2]], [[X1]]
; CHECK-NEXT: ret float [[TMP3]]
; CHECK-NEXT: [[TMP2:%.*]] = fmul fast float [[TMP1]], [[X]]
; CHECK-NEXT: [[MUL:%.*]] = fsub fast float [[TMP2]], [[X]]
; CHECK-NEXT: ret float [[MUL]]
;
%x = fdiv fast float %x1, 7.0 ; thwart complexity-based canonicalization
%halfy = fmul fast float %y, 0.5
Expand Down Expand Up @@ -687,8 +687,8 @@ define float @fdiv_constant_numerator_fmul_fast(float %x) {
define float @fdiv_constant_numerator_fmul_precdiv(float %x) {
; CHECK-LABEL: @fdiv_constant_numerator_fmul_precdiv(
; CHECK-NEXT: [[T1:%.*]] = fdiv float 2.000000e+03, [[X:%.*]]
; CHECK-NEXT: [[T4:%.*]] = fmul reassoc float [[T1]], 6.000000e+03
; CHECK-NEXT: ret float [[T4]]
; CHECK-NEXT: [[T3:%.*]] = fmul reassoc float [[T1]], 6.000000e+03
; CHECK-NEXT: ret float [[T3]]
;
%t1 = fdiv float 2.0e+3, %x
%t3 = fmul reassoc float %t1, 6.0e+3
Expand Down Expand Up @@ -1288,7 +1288,7 @@ define half @mul_zero_nnan(half %x) {

define <2 x float> @mul_zero_nnan_vec_poison(<2 x float> %x) {
; CHECK-LABEL: @mul_zero_nnan_vec_poison(
; CHECK-NEXT: [[R:%.*]] = call nnan <2 x float> @llvm.copysign.v2f32(<2 x float> zeroinitializer, <2 x float> [[X:%.*]])
; CHECK-NEXT: [[R:%.*]] = call nnan <2 x float> @llvm.copysign.v2f32(<2 x float> <float 0.000000e+00, float poison>, <2 x float> [[X:%.*]])
; CHECK-NEXT: ret <2 x float> [[R]]
;
%r = fmul nnan <2 x float> %x, <float 0.0, float poison>
Expand Down Expand Up @@ -1389,7 +1389,7 @@ define <3 x float> @mul_neg_zero_nnan_ninf_vec(<3 x float> nofpclass(inf nan) %a
; CHECK-LABEL: @mul_neg_zero_nnan_ninf_vec(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[TMP0:%.*]] = fneg <3 x float> [[A:%.*]]
; CHECK-NEXT: [[RET:%.*]] = call <3 x float> @llvm.copysign.v3f32(<3 x float> zeroinitializer, <3 x float> [[TMP0]])
; CHECK-NEXT: [[RET:%.*]] = call <3 x float> @llvm.copysign.v3f32(<3 x float> <float -0.000000e+00, float poison, float poison>, <3 x float> [[TMP0]])
; CHECK-NEXT: ret <3 x float> [[RET]]
;
entry:
Expand Down
4 changes: 2 additions & 2 deletions llvm/test/Transforms/InstCombine/vector-xor.ll
View file
Original file line number Diff line number Diff line change
Expand Up @@ -140,7 +140,7 @@ define <4 x i32> @test_v4i32_not_ashr_negative_const(<4 x i32> %a0) {

define <4 x i32> @test_v4i32_not_ashr_negative_const_poison(<4 x i32> %a0) {
; CHECK-LABEL: @test_v4i32_not_ashr_negative_const_poison(
; CHECK-NEXT: [[TMP1:%.*]] = lshr <4 x i32> <i32 2, i32 4, i32 0, i32 8>, [[A0:%.*]]
; CHECK-NEXT: [[TMP1:%.*]] = lshr <4 x i32> <i32 2, i32 4, i32 poison, i32 8>, [[A0:%.*]]
; CHECK-NEXT: ret <4 x i32> [[TMP1]]
;
%1 = ashr <4 x i32> <i32 -3, i32 -5, i32 poison, i32 -9>, %a0
Expand Down Expand Up @@ -172,7 +172,7 @@ define <4 x i32> @test_v4i32_not_lshr_nonnegative_const(<4 x i32> %a0) {

define <4 x i32> @test_v4i32_not_lshr_nonnegative_const_poison(<4 x i32> %a0) {
; CHECK-LABEL: @test_v4i32_not_lshr_nonnegative_const_poison(
; CHECK-NEXT: [[TMP1:%.*]] = ashr <4 x i32> <i32 -4, i32 -6, i32 -1, i32 -10>, [[A0:%.*]]
; CHECK-NEXT: [[TMP1:%.*]] = ashr <4 x i32> <i32 -4, i32 -6, i32 poison, i32 -10>, [[A0:%.*]]
; CHECK-NEXT: ret <4 x i32> [[TMP1]]
;
%1 = lshr <4 x i32> <i32 3, i32 5, i32 poison, i32 9>, %a0
Expand Down

0 comments on commit eb7ad88

Please sign in to comment.