Skip to content

release/20.x: [LoopVectorize] Fix cost model assert when vectorising calls (#125716) #126209

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 2 commits into from
Feb 7, 2025
Merged

release/20.x: [LoopVectorize] Fix cost model assert when vectorising calls (#125716) #126209

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
3542150
[LoopVectorize] Fix cost model assert when vectorising calls (#125716)
david-arm Feb 7, 2025
b5f41cc
[LoopVectorize] Fix build error (#126218)
david-arm Feb 7, 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
15 changes: 1 addition & 14 deletions llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6262,7 +6262,6 @@ void LoopVectorizationCostModel::setVectorizedCallDecision(ElementCount VF) {

// Find the cost of vectorizing the call, if we can find a suitable
// vector variant of the function.
bool UsesMask = false;
VFInfo FuncInfo;
Function *VecFunc = nullptr;
// Search through any available variants for one we can use at this VF.
Expand Down Expand Up @@ -6314,7 +6313,6 @@ void LoopVectorizationCostModel::setVectorizedCallDecision(ElementCount VF) {
break;
}
case VFParamKind::GlobalPredicate:
UsesMask = true;
break;
default:
ParamsOk = false;
Expand All @@ -6331,19 +6329,8 @@ void LoopVectorizationCostModel::setVectorizedCallDecision(ElementCount VF) {
break;
}

// Add in the cost of synthesizing a mask if one wasn't required.
InstructionCost MaskCost = 0;
if (VecFunc && UsesMask && !MaskRequired)
MaskCost = TTI.getShuffleCost(
TargetTransformInfo::SK_Broadcast,
VectorType::get(IntegerType::getInt1Ty(
VecFunc->getFunctionType()->getContext()),
VF),
{}, CostKind);

if (TLI && VecFunc && !CI->isNoBuiltin())
VectorCost =
TTI.getCallInstrCost(nullptr, RetTy, Tys, CostKind) + MaskCost;
VectorCost = TTI.getCallInstrCost(nullptr, RetTy, Tys, CostKind);

// Find the cost of an intrinsic; some targets may have instructions that
// perform the operation without needing an actual call.
Expand Down
Loading