[llvm] [VPlan] Extract reverse operation for reverse accesses (PR #146525)
Florian Hahn via llvm-commits
llvm-commits at lists.llvm.org
Thu Sep 11 06:24:32 PDT 2025
================
@@ -2281,21 +2295,37 @@ InstructionCost VPWidenCastRecipe::computeCost(ElementCount VF,
return TTI::CastContextHint::Normal;
};
+ using namespace llvm::VPlanPatternMatch;
VPValue *Operand = getOperand(0);
TTI::CastContextHint CCH = TTI::CastContextHint::None;
// For Trunc/FPTrunc, get the context from the only user.
- if ((Opcode == Instruction::Trunc || Opcode == Instruction::FPTrunc) &&
- !hasMoreThanOneUniqueUser() && getNumUsers() > 0) {
- if (auto *StoreRecipe = dyn_cast<VPRecipeBase>(*user_begin()))
- CCH = ComputeCCH(StoreRecipe);
+ if (Opcode == Instruction::Trunc || Opcode == Instruction::FPTrunc) {
+ auto GetOnlyUser = [](const VPSingleDefRecipe *R) -> VPRecipeBase * {
+ if (R->getNumUsers() == 0 || R->hasMoreThanOneUniqueUser())
+ return nullptr;
+ return dyn_cast<VPRecipeBase>(*R->user_begin());
+ };
+
+ if (VPRecipeBase *Recipe = GetOnlyUser(this)) {
+ if (match(Recipe, m_VPInstruction<VPInstruction::Reverse>(m_VPValue())))
+ Recipe = GetOnlyUser(cast<VPInstruction>(Recipe));
+ if (Recipe)
+ CCH = ComputeCCH(Recipe);
----------------
fhahn wrote:
Hmm, if we have a shuffle inbetween the load and a cast for example, can the cast still be folded into the load in most cases? Curious if this may have surfaced an in-accuracy of the current cost modeling.
https://github.com/llvm/llvm-project/pull/146525
More information about the llvm-commits
mailing list