[VPlan] Store memory alignment in VPWidenMemoryRecipe. nfc #165255
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Member
|
@llvm/pr-subscribers-vectorizers @llvm/pr-subscribers-llvm-transforms Author: Mel Chen (Mel-Chen) ChangesAdd an member Alignment to VPWidenMemoryRecipe to store memory alignment directly in the recipe. Update constructors, clone(), and relevant methods to use this stored alignment instead of querying the IR instruction. This allows VPWidenLoadRecipe/VPWidenStoreRecipe to be constructed without relying on the original IR instruction in the future. Full diff: https://github.com/llvm/llvm-project/pull/165255.diff 5 Files Affected:
diff --git a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
index facb0fabdf57e..f7968abbe5b6b 100644
--- a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
+++ b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
@@ -7543,12 +7543,13 @@ VPRecipeBuilder::tryToWidenMemory(Instruction *I, ArrayRef<VPValue *> Operands,
}
if (LoadInst *Load = dyn_cast<LoadInst>(I))
return new VPWidenLoadRecipe(*Load, Ptr, Mask, Consecutive, Reverse,
- VPIRMetadata(*Load, LVer), I->getDebugLoc());
+ Load->getAlign(), VPIRMetadata(*Load, LVer),
+ I->getDebugLoc());
StoreInst *Store = cast<StoreInst>(I);
return new VPWidenStoreRecipe(*Store, Ptr, Operands[0], Mask, Consecutive,
- Reverse, VPIRMetadata(*Store, LVer),
- I->getDebugLoc());
+ Reverse, Store->getAlign(),
+ VPIRMetadata(*Store, LVer), I->getDebugLoc());
}
/// Creates a VPWidenIntOrFpInductionRecpipe for \p Phi. If needed, it will also
diff --git a/llvm/lib/Transforms/Vectorize/VPlan.h b/llvm/lib/Transforms/Vectorize/VPlan.h
index 5b9f005e50e47..1f10058ab4a9a 100644
--- a/llvm/lib/Transforms/Vectorize/VPlan.h
+++ b/llvm/lib/Transforms/Vectorize/VPlan.h
@@ -3179,6 +3179,9 @@ class LLVM_ABI_FOR_TEST VPWidenMemoryRecipe : public VPRecipeBase,
protected:
Instruction &Ingredient;
+ /// Alignment information for this memory access.
+ Align Alignment;
+
/// Whether the accessed addresses are consecutive.
bool Consecutive;
@@ -3198,10 +3201,10 @@ class LLVM_ABI_FOR_TEST VPWidenMemoryRecipe : public VPRecipeBase,
VPWidenMemoryRecipe(const char unsigned SC, Instruction &I,
std::initializer_list<VPValue *> Operands,
- bool Consecutive, bool Reverse,
+ bool Consecutive, bool Reverse, Align Alignment,
const VPIRMetadata &Metadata, DebugLoc DL)
: VPRecipeBase(SC, Operands, DL), VPIRMetadata(Metadata), Ingredient(I),
- Consecutive(Consecutive), Reverse(Reverse) {
+ Alignment(Alignment), Consecutive(Consecutive), Reverse(Reverse) {
assert((Consecutive || !Reverse) && "Reverse implies consecutive");
}
@@ -3242,6 +3245,9 @@ class LLVM_ABI_FOR_TEST VPWidenMemoryRecipe : public VPRecipeBase,
return isMasked() ? getOperand(getNumOperands() - 1) : nullptr;
}
+ /// Returns the alignment of the memory access.
+ Align getAlign() const { return Alignment; }
+
/// Generate the wide load/store.
void execute(VPTransformState &State) override {
llvm_unreachable("VPWidenMemoryRecipe should not be instantiated.");
@@ -3259,18 +3265,18 @@ class LLVM_ABI_FOR_TEST VPWidenMemoryRecipe : public VPRecipeBase,
struct LLVM_ABI_FOR_TEST VPWidenLoadRecipe final : public VPWidenMemoryRecipe,
public VPValue {
VPWidenLoadRecipe(LoadInst &Load, VPValue *Addr, VPValue *Mask,
- bool Consecutive, bool Reverse,
+ bool Consecutive, bool Reverse, Align Alignment,
const VPIRMetadata &Metadata, DebugLoc DL)
: VPWidenMemoryRecipe(VPDef::VPWidenLoadSC, Load, {Addr}, Consecutive,
- Reverse, Metadata, DL),
+ Reverse, Alignment, Metadata, DL),
VPValue(this, &Load) {
setMask(Mask);
}
VPWidenLoadRecipe *clone() override {
return new VPWidenLoadRecipe(cast<LoadInst>(Ingredient), getAddr(),
- getMask(), Consecutive, Reverse, *this,
- getDebugLoc());
+ getMask(), Consecutive, Reverse, getAlign(),
+ *this, getDebugLoc());
}
VP_CLASSOF_IMPL(VPDef::VPWidenLoadSC);
@@ -3301,8 +3307,8 @@ struct VPWidenLoadEVLRecipe final : public VPWidenMemoryRecipe, public VPValue {
VPWidenLoadEVLRecipe(VPWidenLoadRecipe &L, VPValue *Addr, VPValue &EVL,
VPValue *Mask)
: VPWidenMemoryRecipe(VPDef::VPWidenLoadEVLSC, L.getIngredient(),
- {Addr, &EVL}, L.isConsecutive(), L.isReverse(), L,
- L.getDebugLoc()),
+ {Addr, &EVL}, L.isConsecutive(), L.isReverse(),
+ L.getAlign(), L, L.getDebugLoc()),
VPValue(this, &getIngredient()) {
setMask(Mask);
}
@@ -3340,16 +3346,16 @@ struct VPWidenLoadEVLRecipe final : public VPWidenMemoryRecipe, public VPValue {
struct LLVM_ABI_FOR_TEST VPWidenStoreRecipe final : public VPWidenMemoryRecipe {
VPWidenStoreRecipe(StoreInst &Store, VPValue *Addr, VPValue *StoredVal,
VPValue *Mask, bool Consecutive, bool Reverse,
- const VPIRMetadata &Metadata, DebugLoc DL)
+ Align Alignment, const VPIRMetadata &Metadata, DebugLoc DL)
: VPWidenMemoryRecipe(VPDef::VPWidenStoreSC, Store, {Addr, StoredVal},
- Consecutive, Reverse, Metadata, DL) {
+ Consecutive, Reverse, Alignment, Metadata, DL) {
setMask(Mask);
}
VPWidenStoreRecipe *clone() override {
return new VPWidenStoreRecipe(cast<StoreInst>(Ingredient), getAddr(),
getStoredValue(), getMask(), Consecutive,
- Reverse, *this, getDebugLoc());
+ Reverse, getAlign(), *this, getDebugLoc());
}
VP_CLASSOF_IMPL(VPDef::VPWidenStoreSC);
@@ -3384,7 +3390,7 @@ struct VPWidenStoreEVLRecipe final : public VPWidenMemoryRecipe {
VPValue *Mask)
: VPWidenMemoryRecipe(VPDef::VPWidenStoreEVLSC, S.getIngredient(),
{Addr, S.getStoredValue(), &EVL}, S.isConsecutive(),
- S.isReverse(), S, S.getDebugLoc()) {
+ S.isReverse(), S.getAlign(), S, S.getDebugLoc()) {
setMask(Mask);
}
diff --git a/llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp b/llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp
index 931a5b7582c4e..fac710a8e7420 100644
--- a/llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp
+++ b/llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp
@@ -3525,7 +3525,6 @@ void VPPredInstPHIRecipe::print(raw_ostream &O, const Twine &Indent,
InstructionCost VPWidenMemoryRecipe::computeCost(ElementCount VF,
VPCostContext &Ctx) const {
Type *Ty = toVectorTy(getLoadStoreType(&Ingredient), VF);
- const Align Alignment = getLoadStoreAlignment(&Ingredient);
unsigned AS = cast<PointerType>(Ctx.Types.inferScalarType(getAddr()))
->getAddressSpace();
unsigned Opcode = isa<VPWidenLoadRecipe, VPWidenLoadEVLRecipe>(this)
@@ -3575,7 +3574,6 @@ InstructionCost VPWidenMemoryRecipe::computeCost(ElementCount VF,
void VPWidenLoadRecipe::execute(VPTransformState &State) {
Type *ScalarDataTy = getLoadStoreType(&Ingredient);
auto *DataTy = VectorType::get(ScalarDataTy, State.VF);
- const Align Alignment = getLoadStoreAlignment(&Ingredient);
bool CreateGather = !isConsecutive();
auto &Builder = State.Builder;
@@ -3630,7 +3628,6 @@ static Instruction *createReverseEVL(IRBuilderBase &Builder, Value *Operand,
void VPWidenLoadEVLRecipe::execute(VPTransformState &State) {
Type *ScalarDataTy = getLoadStoreType(&Ingredient);
auto *DataTy = VectorType::get(ScalarDataTy, State.VF);
- const Align Alignment = getLoadStoreAlignment(&Ingredient);
bool CreateGather = !isConsecutive();
auto &Builder = State.Builder;
@@ -3674,7 +3671,6 @@ InstructionCost VPWidenLoadEVLRecipe::computeCost(ElementCount VF,
// TODO: Using getMemoryOpCost() instead of getMaskedMemoryOpCost when we
// don't need to compare to the legacy cost model.
Type *Ty = toVectorTy(getLoadStoreType(&Ingredient), VF);
- const Align Alignment = getLoadStoreAlignment(&Ingredient);
unsigned AS = getLoadStoreAddressSpace(&Ingredient);
InstructionCost Cost = Ctx.TTI.getMaskedMemoryOpCost(
Instruction::Load, Ty, Alignment, AS, Ctx.CostKind);
@@ -3699,7 +3695,6 @@ void VPWidenLoadEVLRecipe::print(raw_ostream &O, const Twine &Indent,
void VPWidenStoreRecipe::execute(VPTransformState &State) {
VPValue *StoredVPValue = getStoredValue();
bool CreateScatter = !isConsecutive();
- const Align Alignment = getLoadStoreAlignment(&Ingredient);
auto &Builder = State.Builder;
@@ -3742,7 +3737,6 @@ void VPWidenStoreRecipe::print(raw_ostream &O, const Twine &Indent,
void VPWidenStoreEVLRecipe::execute(VPTransformState &State) {
VPValue *StoredValue = getStoredValue();
bool CreateScatter = !isConsecutive();
- const Align Alignment = getLoadStoreAlignment(&Ingredient);
auto &Builder = State.Builder;
@@ -3785,7 +3779,6 @@ InstructionCost VPWidenStoreEVLRecipe::computeCost(ElementCount VF,
// TODO: Using getMemoryOpCost() instead of getMaskedMemoryOpCost when we
// don't need to compare to the legacy cost model.
Type *Ty = toVectorTy(getLoadStoreType(&Ingredient), VF);
- const Align Alignment = getLoadStoreAlignment(&Ingredient);
unsigned AS = getLoadStoreAddressSpace(&Ingredient);
InstructionCost Cost = Ctx.TTI.getMaskedMemoryOpCost(
Instruction::Store, Ty, Alignment, AS, Ctx.CostKind);
diff --git a/llvm/lib/Transforms/Vectorize/VPlanTransforms.cpp b/llvm/lib/Transforms/Vectorize/VPlanTransforms.cpp
index 84817d78a077a..079aa189475df 100644
--- a/llvm/lib/Transforms/Vectorize/VPlanTransforms.cpp
+++ b/llvm/lib/Transforms/Vectorize/VPlanTransforms.cpp
@@ -91,13 +91,14 @@ bool VPlanTransforms::tryToConvertVPInstructionsToVPRecipes(
if (LoadInst *Load = dyn_cast<LoadInst>(Inst)) {
NewRecipe = new VPWidenLoadRecipe(
*Load, Ingredient.getOperand(0), nullptr /*Mask*/,
- false /*Consecutive*/, false /*Reverse*/, VPIRMetadata(*Load),
- Ingredient.getDebugLoc());
+ false /*Consecutive*/, false /*Reverse*/, Load->getAlign(),
+ VPIRMetadata(*Load), Ingredient.getDebugLoc());
} else if (StoreInst *Store = dyn_cast<StoreInst>(Inst)) {
NewRecipe = new VPWidenStoreRecipe(
*Store, Ingredient.getOperand(1), Ingredient.getOperand(0),
nullptr /*Mask*/, false /*Consecutive*/, false /*Reverse*/,
- VPIRMetadata(*Store), Ingredient.getDebugLoc());
+ Store->getAlign(), VPIRMetadata(*Store),
+ Ingredient.getDebugLoc());
} else if (GetElementPtrInst *GEP = dyn_cast<GetElementPtrInst>(Inst)) {
NewRecipe = new VPWidenGEPRecipe(GEP, Ingredient.operands());
} else if (CallInst *CI = dyn_cast<CallInst>(Inst)) {
@@ -4234,10 +4235,11 @@ void VPlanTransforms::narrowInterleaveGroups(VPlan &Plan, ElementCount VF,
if (auto *LoadGroup = dyn_cast<VPInterleaveRecipe>(R)) {
// Narrow interleave group to wide load, as transformed VPlan will only
// process one original iteration.
+ auto *LI =
+ cast<LoadInst>(LoadGroup->getInterleaveGroup()->getInsertPos());
auto *L = new VPWidenLoadRecipe(
- *cast<LoadInst>(LoadGroup->getInterleaveGroup()->getInsertPos()),
- LoadGroup->getAddr(), LoadGroup->getMask(), /*Consecutive=*/true,
- /*Reverse=*/false, {}, LoadGroup->getDebugLoc());
+ *LI, LoadGroup->getAddr(), LoadGroup->getMask(), /*Consecutive=*/true,
+ /*Reverse=*/false, LI->getAlign(), {}, LoadGroup->getDebugLoc());
L->insertBefore(LoadGroup);
NarrowedOps.insert(L);
return L;
@@ -4280,10 +4282,11 @@ void VPlanTransforms::narrowInterleaveGroups(VPlan &Plan, ElementCount VF,
Res = NarrowOp(Member0);
}
+ auto *SI =
+ cast<StoreInst>(StoreGroup->getInterleaveGroup()->getInsertPos());
auto *S = new VPWidenStoreRecipe(
- *cast<StoreInst>(StoreGroup->getInterleaveGroup()->getInsertPos()),
- StoreGroup->getAddr(), Res, nullptr, /*Consecutive=*/true,
- /*Reverse=*/false, {}, StoreGroup->getDebugLoc());
+ *SI, StoreGroup->getAddr(), Res, nullptr, /*Consecutive=*/true,
+ /*Reverse=*/false, SI->getAlign(), {}, StoreGroup->getDebugLoc());
S->insertBefore(StoreGroup);
StoreGroup->eraseFromParent();
}
diff --git a/llvm/unittests/Transforms/Vectorize/VPlanTest.cpp b/llvm/unittests/Transforms/Vectorize/VPlanTest.cpp
index c1791dfa5b761..59a9ea1a720b3 100644
--- a/llvm/unittests/Transforms/Vectorize/VPlanTest.cpp
+++ b/llvm/unittests/Transforms/Vectorize/VPlanTest.cpp
@@ -1132,7 +1132,8 @@ TEST_F(VPRecipeTest, CastVPWidenMemoryRecipeToVPUserAndVPDef) {
new LoadInst(Int32, PoisonValue::get(Int32Ptr), "", false, Align(1));
VPValue *Addr = Plan.getOrAddLiveIn(ConstantInt::get(Int32, 1));
VPValue *Mask = Plan.getOrAddLiveIn(ConstantInt::get(Int32, 2));
- VPWidenLoadRecipe Recipe(*Load, Addr, Mask, true, false, {}, {});
+ VPWidenLoadRecipe Recipe(*Load, Addr, Mask, true, false, Load->getAlign(), {},
+ {});
EXPECT_TRUE(isa<VPUser>(&Recipe));
VPRecipeBase *BaseR = &Recipe;
EXPECT_TRUE(isa<VPUser>(BaseR));
@@ -1249,7 +1250,8 @@ TEST_F(VPRecipeTest, MayHaveSideEffectsAndMayReadWriteMemory) {
new LoadInst(Int32, PoisonValue::get(Int32Ptr), "", false, Align(1));
VPValue *Mask = Plan.getOrAddLiveIn(ConstantInt::get(Int32, 1));
VPValue *Addr = Plan.getOrAddLiveIn(ConstantInt::get(Int32, 2));
- VPWidenLoadRecipe Recipe(*Load, Addr, Mask, true, false, {}, {});
+ VPWidenLoadRecipe Recipe(*Load, Addr, Mask, true, false, Load->getAlign(),
+ {}, {});
EXPECT_FALSE(Recipe.mayHaveSideEffects());
EXPECT_TRUE(Recipe.mayReadFromMemory());
EXPECT_FALSE(Recipe.mayWriteToMemory());
@@ -1263,8 +1265,8 @@ TEST_F(VPRecipeTest, MayHaveSideEffectsAndMayReadWriteMemory) {
VPValue *Mask = Plan.getOrAddLiveIn(ConstantInt::get(Int32, 1));
VPValue *Addr = Plan.getOrAddLiveIn(ConstantInt::get(Int32, 2));
VPValue *StoredV = Plan.getOrAddLiveIn(ConstantInt::get(Int32, 3));
- VPWidenStoreRecipe Recipe(*Store, Addr, StoredV, Mask, false, false, {},
- {});
+ VPWidenStoreRecipe Recipe(*Store, Addr, StoredV, Mask, false, false,
+ Store->getAlign(), {}, {});
EXPECT_TRUE(Recipe.mayHaveSideEffects());
EXPECT_FALSE(Recipe.mayReadFromMemory());
EXPECT_TRUE(Recipe.mayWriteToMemory());
|
Add an member Alignment to VPWidenMemoryRecipe to store memory alignment directly in the recipe. Update constructors, clone(), and relevant methods to use this stored alignment instead of querying the IR instruction. This allows VPWidenLoadRecipe/VPWidenStoreRecipe to be constructed without relying on the original IR instruction in the future.
Mel-Chen
force-pushed
the
nfc-alignment
branch
from
03b6e2f to
e4dc7ce
Compare
Collaborator
|
Post-commit nit: constructors that have a Load/Store/Instruction parameter can retrieve the alignment from it rather than requesting that it be passed as an additional parameter. |
Lukacma
pushed a commit
to Lukacma/llvm-project
that referenced
this pull request
Oct 29, 2025
Add an member Alignment to VPWidenMemoryRecipe to store memory alignment directly in the recipe. Update constructors, clone(), and relevant methods to use this stored alignment instead of querying the IR instruction. This allows VPWidenLoadRecipe/VPWidenStoreRecipe to be constructed without relying on the original IR instruction in the future.
Contributor
Author
aokblast
pushed a commit
to aokblast/llvm-project
that referenced
this pull request
Oct 30, 2025
Add an member Alignment to VPWidenMemoryRecipe to store memory alignment directly in the recipe. Update constructors, clone(), and relevant methods to use this stored alignment instead of querying the IR instruction. This allows VPWidenLoadRecipe/VPWidenStoreRecipe to be constructed without relying on the original IR instruction in the future.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Add an member Alignment to VPWidenMemoryRecipe to store memory alignment directly in the recipe. Update constructors, clone(), and relevant methods to use this stored alignment instead of querying the IR instruction. This allows VPWidenLoadRecipe/VPWidenStoreRecipe to be constructed without relying on the original IR instruction in the future.