Skip to content

[InstCombine] Handle scalable splats of constants in getMinimumFPType #132960

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 4 commits into from
Mar 27, 2025
Merged

[InstCombine] Handle scalable splats of constants in getMinimumFPType #132960

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
4b9c243
Precommit tests
lukel97 Mar 25, 2025
01be82c
[InstCombine] Handle scalable splats of constants in getMinimumFPType
lukel97 Mar 25, 2025
d26e672
Add fixed vector test
lukel97 Mar 27, 2025
8c7f8c6
Also handle fixed-length splats, as a sort of fast-path. Should be NFC?
lukel97 Mar 27, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 6 additions & 6 deletions llvm/lib/Transforms/InstCombine/InstCombineCasts.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1684,12 +1684,12 @@ static Type *getMinimumFPType(Value *V, bool PreferBFloat) {
if (Type *T = shrinkFPConstant(CFP, PreferBFloat))
return T;

// We can only correctly find a minimum type for a scalable vector when it is
// a splat. For splats of constant values the fpext is wrapped up as a
// ConstantExpr.
if (auto *FPCExt = dyn_cast<ConstantExpr>(V))
if (FPCExt->getOpcode() == Instruction::FPExt)
return FPCExt->getOperand(0)->getType();
// Try to shrink scalable and fixed splat vectors.
if (auto *FPC = dyn_cast<Constant>(V))
if (isa<VectorType>(V->getType()))
if (auto *Splat = dyn_cast_or_null<ConstantFP>(FPC->getSplatValue()))
if (Type *T = shrinkFPConstant(Splat, PreferBFloat))
return T;

// Try to shrink a vector of FP constants. This returns nullptr on scalable
// vectors
Expand Down
11 changes: 11 additions & 0 deletions llvm/test/Transforms/InstCombine/fpextend.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -448,3 +448,14 @@ define bfloat @bf16_frem(bfloat %x) {
%t3 = fptrunc float %t2 to bfloat
ret bfloat %t3
}

define <4 x float> @v4f32_fadd(<4 x float> %a) {
; CHECK-LABEL: @v4f32_fadd(
; CHECK-NEXT: [[TMP1:%.*]] = fadd <4 x float> [[A:%.*]], splat (float -1.000000e+00)
; CHECK-NEXT: ret <4 x float> [[TMP1]]
;
%2 = fpext <4 x float> %a to <4 x double>
%4 = fadd <4 x double> %2, splat (double -1.000000e+00)
%5 = fptrunc <4 x double> %4 to <4 x float>
ret <4 x float> %5
}
13 changes: 13 additions & 0 deletions llvm/test/Transforms/InstCombine/scalable-const-fp-splat.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,3 +13,16 @@ define <vscale x 2 x float> @shrink_splat_scalable_extend(<vscale x 2 x float> %
%5 = fptrunc <vscale x 2 x double> %4 to <vscale x 2 x float>
ret <vscale x 2 x float> %5
}

define <vscale x 2 x float> @shrink_splat_scalable_extend_rhs_constexpr(<vscale x 2 x float> %a) {
; CHECK-LABEL: define <vscale x 2 x float> @shrink_splat_scalable_extend_rhs_constexpr(
Copy link
Contributor

@david-arm david-arm Mar 26, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Would be good to have a fixed-width variant of this too, if there isn't one already?

Copy link
Contributor Author

@lukel97 lukel97 Mar 27, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Added to fpextend.ll in d26e672

; CHECK-SAME: <vscale x 2 x float> [[A:%.*]]) {
; CHECK-NEXT: [[TMP1:%.*]] = fptrunc <vscale x 2 x double> splat (double -1.000000e+00) to <vscale x 2 x float>
; CHECK-NEXT: [[TMP3:%.*]] = fadd <vscale x 2 x float> [[A]], [[TMP1]]
; CHECK-NEXT: ret <vscale x 2 x float> [[TMP3]]
;
%2 = fpext <vscale x 2 x float> %a to <vscale x 2 x double>
%4 = fadd <vscale x 2 x double> %2, splat (double -1.000000e+00)
%5 = fptrunc <vscale x 2 x double> %4 to <vscale x 2 x float>
ret <vscale x 2 x float> %5
}