mirror of
https://github.com/RPCS3/llvm-mirror.git
synced 2024-11-25 12:12:47 +01:00
Temporarily Revert "[AssumeBundles] Use operand bundles to encode alignment assumptions"
due to the performance bugs filed in https://bugs.llvm.org/show_bug.cgi?id=46753. An SROA change soon may obviate some of these problems. This reverts commit 8d09f20798ac180b1749276bff364682ce0196ab. (cherry picked from commit 7bfaa40086359ed7e41c862ab0a65e0bb1be0aeb)
This commit is contained in:
Eric Christopher
Hans Wennborg
parent
4af794b176
commit
5e978f5f2c
@ -782,11 +782,7 @@ public:
|
||||
|
||||
/// Create an assume intrinsic call that allows the optimizer to
|
||||
/// assume that the provided condition will be true.
|
||||
///
|
||||
/// The optional argument \p OpBundles specifies operand bundles that are
|
||||
/// added to the call instruction.
|
||||
CallInst *CreateAssumption(Value *Cond,
|
||||
ArrayRef<OperandBundleDef> OpBundles = llvm::None);
|
||||
CallInst *CreateAssumption(Value *Cond);
|
||||
|
||||
/// Create a call to the experimental.gc.statepoint intrinsic to
|
||||
/// start a new statepoint sequence.
|
||||
@ -2506,11 +2502,13 @@ public:
|
||||
|
||||
private:
|
||||
/// Helper function that creates an assume intrinsic call that
|
||||
/// represents an alignment assumption on the provided pointer \p PtrValue
|
||||
/// with offset \p OffsetValue and alignment value \p AlignValue.
|
||||
/// represents an alignment assumption on the provided Ptr, Mask, Type
|
||||
/// and Offset. It may be sometimes useful to do some other logic
|
||||
/// based on this alignment check, thus it can be stored into 'TheCheck'.
|
||||
CallInst *CreateAlignmentAssumptionHelper(const DataLayout &DL,
|
||||
Value *PtrValue, Value *AlignValue,
|
||||
Value *OffsetValue);
|
||||
Value *PtrValue, Value *Mask,
|
||||
Type *IntPtrTy, Value *OffsetValue,
|
||||
Value **TheCheck);
|
||||
|
||||
public:
|
||||
/// Create an assume intrinsic call that represents an alignment
|
||||
@ -2519,9 +2517,13 @@ public:
|
||||
/// An optional offset can be provided, and if it is provided, the offset
|
||||
/// must be subtracted from the provided pointer to get the pointer with the
|
||||
/// specified alignment.
|
||||
///
|
||||
/// It may be sometimes useful to do some other logic
|
||||
/// based on this alignment check, thus it can be stored into 'TheCheck'.
|
||||
CallInst *CreateAlignmentAssumption(const DataLayout &DL, Value *PtrValue,
|
||||
unsigned Alignment,
|
||||
Value *OffsetValue = nullptr);
|
||||
Value *OffsetValue = nullptr,
|
||||
Value **TheCheck = nullptr);
|
||||
|
||||
/// Create an assume intrinsic call that represents an alignment
|
||||
/// assumption on the provided pointer.
|
||||
@ -2530,11 +2532,15 @@ public:
|
||||
/// must be subtracted from the provided pointer to get the pointer with the
|
||||
/// specified alignment.
|
||||
///
|
||||
/// It may be sometimes useful to do some other logic
|
||||
/// based on this alignment check, thus it can be stored into 'TheCheck'.
|
||||
///
|
||||
/// This overload handles the condition where the Alignment is dependent
|
||||
/// on an existing value rather than a static value.
|
||||
CallInst *CreateAlignmentAssumption(const DataLayout &DL, Value *PtrValue,
|
||||
Value *Alignment,
|
||||
Value *OffsetValue = nullptr);
|
||||
Value *OffsetValue = nullptr,
|
||||
Value **TheCheck = nullptr);
|
||||
};
|
||||
|
||||
/// This provides a uniform API for creating instructions and inserting
|
||||
|
||||
@ -37,9 +37,9 @@ struct AlignmentFromAssumptionsPass
|
||||
ScalarEvolution *SE = nullptr;
|
||||
DominatorTree *DT = nullptr;
|
||||
|
||||
bool extractAlignmentInfo(CallInst *I, unsigned Idx, Value *&AAPtr,
|
||||
const SCEV *&AlignSCEV, const SCEV *&OffSCEV);
|
||||
bool processAssumption(CallInst *I, unsigned Idx);
|
||||
bool extractAlignmentInfo(CallInst *I, Value *&AAPtr, const SCEV *&AlignSCEV,
|
||||
const SCEV *&OffSCEV);
|
||||
bool processAssumption(CallInst *I);
|
||||
};
|
||||
}
|
||||
|
||||
|
||||
@ -108,17 +108,10 @@ llvm::getKnowledgeFromBundle(CallInst &Assume,
|
||||
Result.AttrKind = Attribute::getAttrKindFromName(BOI.Tag->getKey());
|
||||
if (bundleHasArgument(BOI, ABA_WasOn))
|
||||
Result.WasOn = getValueFromBundleOpInfo(Assume, BOI, ABA_WasOn);
|
||||
auto GetArgOr1 = [&](unsigned Idx) -> unsigned {
|
||||
if (auto *ConstInt = dyn_cast<ConstantInt>(
|
||||
getValueFromBundleOpInfo(Assume, BOI, ABA_Argument + Idx)))
|
||||
return ConstInt->getZExtValue();
|
||||
return 1;
|
||||
};
|
||||
if (BOI.End - BOI.Begin > ABA_Argument)
|
||||
Result.ArgValue = GetArgOr1(0);
|
||||
if (Result.AttrKind == Attribute::Alignment)
|
||||
if (BOI.End - BOI.Begin > ABA_Argument + 1)
|
||||
Result.ArgValue = MinAlign(Result.ArgValue, GetArgOr1(1));
|
||||
Result.ArgValue =
|
||||
cast<ConstantInt>(getValueFromBundleOpInfo(Assume, BOI, ABA_Argument))
|
||||
->getZExtValue();
|
||||
return Result;
|
||||
}
|
||||
|
||||
|
||||
@ -71,9 +71,8 @@ Value *IRBuilderBase::getCastedInt8PtrValue(Value *Ptr) {
|
||||
static CallInst *createCallHelper(Function *Callee, ArrayRef<Value *> Ops,
|
||||
IRBuilderBase *Builder,
|
||||
const Twine &Name = "",
|
||||
Instruction *FMFSource = nullptr,
|
||||
ArrayRef<OperandBundleDef> OpBundles = {}) {
|
||||
CallInst *CI = Builder->CreateCall(Callee, Ops, OpBundles, Name);
|
||||
Instruction *FMFSource = nullptr) {
|
||||
CallInst *CI = Builder->CreateCall(Callee, Ops, Name);
|
||||
if (FMFSource)
|
||||
CI->copyFastMathFlags(FMFSource);
|
||||
return CI;
|
||||
@ -450,16 +449,14 @@ CallInst *IRBuilderBase::CreateInvariantStart(Value *Ptr, ConstantInt *Size) {
|
||||
return createCallHelper(TheFn, Ops, this);
|
||||
}
|
||||
|
||||
CallInst *
|
||||
IRBuilderBase::CreateAssumption(Value *Cond,
|
||||
ArrayRef<OperandBundleDef> OpBundles) {
|
||||
CallInst *IRBuilderBase::CreateAssumption(Value *Cond) {
|
||||
assert(Cond->getType() == getInt1Ty() &&
|
||||
"an assumption condition must be of type i1");
|
||||
|
||||
Value *Ops[] = { Cond };
|
||||
Module *M = BB->getParent()->getParent();
|
||||
Function *FnAssume = Intrinsic::getDeclaration(M, Intrinsic::assume);
|
||||
return createCallHelper(FnAssume, Ops, this, "", nullptr, OpBundles);
|
||||
return createCallHelper(FnAssume, Ops, this);
|
||||
}
|
||||
|
||||
/// Create a call to a Masked Load intrinsic.
|
||||
@ -1110,37 +1107,63 @@ Value *IRBuilderBase::CreatePreserveStructAccessIndex(
|
||||
return Fn;
|
||||
}
|
||||
|
||||
CallInst *IRBuilderBase::CreateAlignmentAssumptionHelper(const DataLayout &DL,
|
||||
Value *PtrValue,
|
||||
Value *AlignValue,
|
||||
Value *OffsetValue) {
|
||||
SmallVector<Value *, 4> Vals({PtrValue, AlignValue});
|
||||
if (OffsetValue)
|
||||
Vals.push_back(OffsetValue);
|
||||
OperandBundleDefT<Value *> AlignOpB("align", Vals);
|
||||
return CreateAssumption(ConstantInt::getTrue(getContext()), {AlignOpB});
|
||||
CallInst *IRBuilderBase::CreateAlignmentAssumptionHelper(
|
||||
const DataLayout &DL, Value *PtrValue, Value *Mask, Type *IntPtrTy,
|
||||
Value *OffsetValue, Value **TheCheck) {
|
||||
Value *PtrIntValue = CreatePtrToInt(PtrValue, IntPtrTy, "ptrint");
|
||||
|
||||
if (OffsetValue) {
|
||||
bool IsOffsetZero = false;
|
||||
if (const auto *CI = dyn_cast<ConstantInt>(OffsetValue))
|
||||
IsOffsetZero = CI->isZero();
|
||||
|
||||
if (!IsOffsetZero) {
|
||||
if (OffsetValue->getType() != IntPtrTy)
|
||||
OffsetValue = CreateIntCast(OffsetValue, IntPtrTy, /*isSigned*/ true,
|
||||
"offsetcast");
|
||||
PtrIntValue = CreateSub(PtrIntValue, OffsetValue, "offsetptr");
|
||||
}
|
||||
}
|
||||
|
||||
Value *Zero = ConstantInt::get(IntPtrTy, 0);
|
||||
Value *MaskedPtr = CreateAnd(PtrIntValue, Mask, "maskedptr");
|
||||
Value *InvCond = CreateICmpEQ(MaskedPtr, Zero, "maskcond");
|
||||
if (TheCheck)
|
||||
*TheCheck = InvCond;
|
||||
|
||||
return CreateAssumption(InvCond);
|
||||
}
|
||||
|
||||
CallInst *IRBuilderBase::CreateAlignmentAssumption(const DataLayout &DL,
|
||||
Value *PtrValue,
|
||||
unsigned Alignment,
|
||||
Value *OffsetValue) {
|
||||
CallInst *IRBuilderBase::CreateAlignmentAssumption(
|
||||
const DataLayout &DL, Value *PtrValue, unsigned Alignment,
|
||||
Value *OffsetValue, Value **TheCheck) {
|
||||
assert(isa<PointerType>(PtrValue->getType()) &&
|
||||
"trying to create an alignment assumption on a non-pointer?");
|
||||
assert(Alignment != 0 && "Invalid Alignment");
|
||||
auto *PtrTy = cast<PointerType>(PtrValue->getType());
|
||||
Type *IntPtrTy = getIntPtrTy(DL, PtrTy->getAddressSpace());
|
||||
Value *AlignValue = ConstantInt::get(IntPtrTy, Alignment);
|
||||
return CreateAlignmentAssumptionHelper(DL, PtrValue, AlignValue, OffsetValue);
|
||||
|
||||
Value *Mask = ConstantInt::get(IntPtrTy, Alignment - 1);
|
||||
return CreateAlignmentAssumptionHelper(DL, PtrValue, Mask, IntPtrTy,
|
||||
OffsetValue, TheCheck);
|
||||
}
|
||||
|
||||
CallInst *IRBuilderBase::CreateAlignmentAssumption(const DataLayout &DL,
|
||||
Value *PtrValue,
|
||||
Value *Alignment,
|
||||
Value *OffsetValue) {
|
||||
CallInst *IRBuilderBase::CreateAlignmentAssumption(
|
||||
const DataLayout &DL, Value *PtrValue, Value *Alignment,
|
||||
Value *OffsetValue, Value **TheCheck) {
|
||||
assert(isa<PointerType>(PtrValue->getType()) &&
|
||||
"trying to create an alignment assumption on a non-pointer?");
|
||||
return CreateAlignmentAssumptionHelper(DL, PtrValue, Alignment, OffsetValue);
|
||||
auto *PtrTy = cast<PointerType>(PtrValue->getType());
|
||||
Type *IntPtrTy = getIntPtrTy(DL, PtrTy->getAddressSpace());
|
||||
|
||||
if (Alignment->getType() != IntPtrTy)
|
||||
Alignment = CreateIntCast(Alignment, IntPtrTy, /*isSigned*/ false,
|
||||
"alignmentcast");
|
||||
|
||||
Value *Mask = CreateSub(Alignment, ConstantInt::get(IntPtrTy, 1), "mask");
|
||||
|
||||
return CreateAlignmentAssumptionHelper(DL, PtrValue, Mask, IntPtrTy,
|
||||
OffsetValue, TheCheck);
|
||||
}
|
||||
|
||||
IRBuilderDefaultInserter::~IRBuilderDefaultInserter() {}
|
||||
|
||||
@ -4449,32 +4449,21 @@ void Verifier::visitIntrinsicCall(Intrinsic::ID ID, CallBase &Call) {
|
||||
Assert(Elem.Tag->getKey() == "ignore" ||
|
||||
Attribute::isExistingAttribute(Elem.Tag->getKey()),
|
||||
"tags must be valid attribute names");
|
||||
Assert(Elem.End - Elem.Begin <= 2, "to many arguments");
|
||||
Attribute::AttrKind Kind =
|
||||
Attribute::getAttrKindFromName(Elem.Tag->getKey());
|
||||
unsigned ArgCount = Elem.End - Elem.Begin;
|
||||
if (Kind == Attribute::Alignment) {
|
||||
Assert(ArgCount <= 3 && ArgCount >= 2,
|
||||
"alignment assumptions should have 2 or 3 arguments");
|
||||
Assert(Call.getOperand(Elem.Begin)->getType()->isPointerTy(),
|
||||
"first argument should be a pointer");
|
||||
Assert(Call.getOperand(Elem.Begin + 1)->getType()->isIntegerTy(),
|
||||
"second argument should be an integer");
|
||||
if (ArgCount == 3)
|
||||
Assert(Call.getOperand(Elem.Begin + 2)->getType()->isIntegerTy(),
|
||||
"third argument should be an integer if present");
|
||||
return;
|
||||
}
|
||||
Assert(ArgCount <= 2, "to many arguments");
|
||||
if (Kind == Attribute::None)
|
||||
break;
|
||||
if (Attribute::doesAttrKindHaveArgument(Kind)) {
|
||||
Assert(ArgCount == 2, "this attribute should have 2 arguments");
|
||||
Assert(Elem.End - Elem.Begin == 2,
|
||||
"this attribute should have 2 arguments");
|
||||
Assert(isa<ConstantInt>(Call.getOperand(Elem.Begin + 1)),
|
||||
"the second argument should be a constant integral value");
|
||||
} else if (isFuncOnlyAttr(Kind)) {
|
||||
Assert((ArgCount) == 0, "this attribute has no argument");
|
||||
Assert((Elem.End - Elem.Begin) == 0, "this attribute has no argument");
|
||||
} else if (!isFuncOrArgAttr(Kind)) {
|
||||
Assert((ArgCount) == 1, "this attribute should have one argument");
|
||||
Assert((Elem.End - Elem.Begin) == 1,
|
||||
"this attribute should have one argument");
|
||||
}
|
||||
}
|
||||
break;
|
||||
|
||||
@ -4220,16 +4220,11 @@ Instruction *InstCombiner::visitCallInst(CallInst &CI) {
|
||||
break;
|
||||
case Intrinsic::assume: {
|
||||
Value *IIOperand = II->getArgOperand(0);
|
||||
SmallVector<OperandBundleDef, 4> OpBundles;
|
||||
II->getOperandBundlesAsDefs(OpBundles);
|
||||
bool HasOpBundles = !OpBundles.empty();
|
||||
// Remove an assume if it is followed by an identical assume.
|
||||
// TODO: Do we need this? Unless there are conflicting assumptions, the
|
||||
// computeKnownBits(IIOperand) below here eliminates redundant assumes.
|
||||
Instruction *Next = II->getNextNonDebugInstruction();
|
||||
if (HasOpBundles &&
|
||||
match(Next, m_Intrinsic<Intrinsic::assume>(m_Specific(IIOperand))) &&
|
||||
!cast<IntrinsicInst>(Next)->hasOperandBundles())
|
||||
if (match(Next, m_Intrinsic<Intrinsic::assume>(m_Specific(IIOperand))))
|
||||
return eraseInstFromFunction(CI);
|
||||
|
||||
// Canonicalize assume(a && b) -> assume(a); assume(b);
|
||||
@ -4239,15 +4234,14 @@ Instruction *InstCombiner::visitCallInst(CallInst &CI) {
|
||||
Value *AssumeIntrinsic = II->getCalledOperand();
|
||||
Value *A, *B;
|
||||
if (match(IIOperand, m_And(m_Value(A), m_Value(B)))) {
|
||||
Builder.CreateCall(AssumeIntrinsicTy, AssumeIntrinsic, A, OpBundles,
|
||||
II->getName());
|
||||
Builder.CreateCall(AssumeIntrinsicTy, AssumeIntrinsic, A, II->getName());
|
||||
Builder.CreateCall(AssumeIntrinsicTy, AssumeIntrinsic, B, II->getName());
|
||||
return eraseInstFromFunction(*II);
|
||||
}
|
||||
// assume(!(a || b)) -> assume(!a); assume(!b);
|
||||
if (match(IIOperand, m_Not(m_Or(m_Value(A), m_Value(B))))) {
|
||||
Builder.CreateCall(AssumeIntrinsicTy, AssumeIntrinsic,
|
||||
Builder.CreateNot(A), OpBundles, II->getName());
|
||||
Builder.CreateNot(A), II->getName());
|
||||
Builder.CreateCall(AssumeIntrinsicTy, AssumeIntrinsic,
|
||||
Builder.CreateNot(B), II->getName());
|
||||
return eraseInstFromFunction(*II);
|
||||
@ -4263,8 +4257,7 @@ Instruction *InstCombiner::visitCallInst(CallInst &CI) {
|
||||
isValidAssumeForContext(II, LHS, &DT)) {
|
||||
MDNode *MD = MDNode::get(II->getContext(), None);
|
||||
LHS->setMetadata(LLVMContext::MD_nonnull, MD);
|
||||
if (!HasOpBundles)
|
||||
return eraseInstFromFunction(*II);
|
||||
return eraseInstFromFunction(*II);
|
||||
|
||||
// TODO: apply nonnull return attributes to calls and invokes
|
||||
// TODO: apply range metadata for range check patterns?
|
||||
|
||||
@ -15,7 +15,6 @@
|
||||
//
|
||||
//===----------------------------------------------------------------------===//
|
||||
|
||||
#include "llvm/IR/Instructions.h"
|
||||
#include "llvm/InitializePasses.h"
|
||||
#define AA_NAME "alignment-from-assumptions"
|
||||
#define DEBUG_TYPE AA_NAME
|
||||
@ -204,33 +203,103 @@ static Align getNewAlignment(const SCEV *AASCEV, const SCEV *AlignSCEV,
|
||||
}
|
||||
|
||||
bool AlignmentFromAssumptionsPass::extractAlignmentInfo(CallInst *I,
|
||||
unsigned Idx,
|
||||
Value *&AAPtr,
|
||||
const SCEV *&AlignSCEV,
|
||||
const SCEV *&OffSCEV) {
|
||||
Type *Int64Ty = Type::getInt64Ty(I->getContext());
|
||||
OperandBundleUse AlignOB = I->getOperandBundleAt(Idx);
|
||||
if (AlignOB.getTagName() != "align")
|
||||
// An alignment assume must be a statement about the least-significant
|
||||
// bits of the pointer being zero, possibly with some offset.
|
||||
ICmpInst *ICI = dyn_cast<ICmpInst>(I->getArgOperand(0));
|
||||
if (!ICI)
|
||||
return false;
|
||||
assert(AlignOB.Inputs.size() >= 2);
|
||||
AAPtr = AlignOB.Inputs[0].get();
|
||||
// TODO: Consider accumulating the offset to the base.
|
||||
AAPtr = AAPtr->stripPointerCastsSameRepresentation();
|
||||
AlignSCEV = SE->getSCEV(AlignOB.Inputs[1].get());
|
||||
AlignSCEV = SE->getTruncateOrZeroExtend(AlignSCEV, Int64Ty);
|
||||
if (AlignOB.Inputs.size() == 3)
|
||||
OffSCEV = SE->getSCEV(AlignOB.Inputs[2].get());
|
||||
else
|
||||
|
||||
// This must be an expression of the form: x & m == 0.
|
||||
if (ICI->getPredicate() != ICmpInst::ICMP_EQ)
|
||||
return false;
|
||||
|
||||
// Swap things around so that the RHS is 0.
|
||||
Value *CmpLHS = ICI->getOperand(0);
|
||||
Value *CmpRHS = ICI->getOperand(1);
|
||||
const SCEV *CmpLHSSCEV = SE->getSCEV(CmpLHS);
|
||||
const SCEV *CmpRHSSCEV = SE->getSCEV(CmpRHS);
|
||||
if (CmpLHSSCEV->isZero())
|
||||
std::swap(CmpLHS, CmpRHS);
|
||||
else if (!CmpRHSSCEV->isZero())
|
||||
return false;
|
||||
|
||||
BinaryOperator *CmpBO = dyn_cast<BinaryOperator>(CmpLHS);
|
||||
if (!CmpBO || CmpBO->getOpcode() != Instruction::And)
|
||||
return false;
|
||||
|
||||
// Swap things around so that the right operand of the and is a constant
|
||||
// (the mask); we cannot deal with variable masks.
|
||||
Value *AndLHS = CmpBO->getOperand(0);
|
||||
Value *AndRHS = CmpBO->getOperand(1);
|
||||
const SCEV *AndLHSSCEV = SE->getSCEV(AndLHS);
|
||||
const SCEV *AndRHSSCEV = SE->getSCEV(AndRHS);
|
||||
if (isa<SCEVConstant>(AndLHSSCEV)) {
|
||||
std::swap(AndLHS, AndRHS);
|
||||
std::swap(AndLHSSCEV, AndRHSSCEV);
|
||||
}
|
||||
|
||||
const SCEVConstant *MaskSCEV = dyn_cast<SCEVConstant>(AndRHSSCEV);
|
||||
if (!MaskSCEV)
|
||||
return false;
|
||||
|
||||
// The mask must have some trailing ones (otherwise the condition is
|
||||
// trivial and tells us nothing about the alignment of the left operand).
|
||||
unsigned TrailingOnes = MaskSCEV->getAPInt().countTrailingOnes();
|
||||
if (!TrailingOnes)
|
||||
return false;
|
||||
|
||||
// Cap the alignment at the maximum with which LLVM can deal (and make sure
|
||||
// we don't overflow the shift).
|
||||
uint64_t Alignment;
|
||||
TrailingOnes = std::min(TrailingOnes,
|
||||
unsigned(sizeof(unsigned) * CHAR_BIT - 1));
|
||||
Alignment = std::min(1u << TrailingOnes, +Value::MaximumAlignment);
|
||||
|
||||
Type *Int64Ty = Type::getInt64Ty(I->getParent()->getParent()->getContext());
|
||||
AlignSCEV = SE->getConstant(Int64Ty, Alignment);
|
||||
|
||||
// The LHS might be a ptrtoint instruction, or it might be the pointer
|
||||
// with an offset.
|
||||
AAPtr = nullptr;
|
||||
OffSCEV = nullptr;
|
||||
if (PtrToIntInst *PToI = dyn_cast<PtrToIntInst>(AndLHS)) {
|
||||
AAPtr = PToI->getPointerOperand();
|
||||
OffSCEV = SE->getZero(Int64Ty);
|
||||
OffSCEV = SE->getTruncateOrZeroExtend(OffSCEV, Int64Ty);
|
||||
} else if (const SCEVAddExpr* AndLHSAddSCEV =
|
||||
dyn_cast<SCEVAddExpr>(AndLHSSCEV)) {
|
||||
// Try to find the ptrtoint; subtract it and the rest is the offset.
|
||||
for (SCEVAddExpr::op_iterator J = AndLHSAddSCEV->op_begin(),
|
||||
JE = AndLHSAddSCEV->op_end(); J != JE; ++J)
|
||||
if (const SCEVUnknown *OpUnk = dyn_cast<SCEVUnknown>(*J))
|
||||
if (PtrToIntInst *PToI = dyn_cast<PtrToIntInst>(OpUnk->getValue())) {
|
||||
AAPtr = PToI->getPointerOperand();
|
||||
OffSCEV = SE->getMinusSCEV(AndLHSAddSCEV, *J);
|
||||
break;
|
||||
}
|
||||
}
|
||||
|
||||
if (!AAPtr)
|
||||
return false;
|
||||
|
||||
// Sign extend the offset to 64 bits (so that it is like all of the other
|
||||
// expressions).
|
||||
unsigned OffSCEVBits = OffSCEV->getType()->getPrimitiveSizeInBits();
|
||||
if (OffSCEVBits < 64)
|
||||
OffSCEV = SE->getSignExtendExpr(OffSCEV, Int64Ty);
|
||||
else if (OffSCEVBits > 64)
|
||||
return false;
|
||||
|
||||
AAPtr = AAPtr->stripPointerCasts();
|
||||
return true;
|
||||
}
|
||||
|
||||
bool AlignmentFromAssumptionsPass::processAssumption(CallInst *ACall,
|
||||
unsigned Idx) {
|
||||
bool AlignmentFromAssumptionsPass::processAssumption(CallInst *ACall) {
|
||||
Value *AAPtr;
|
||||
const SCEV *AlignSCEV, *OffSCEV;
|
||||
if (!extractAlignmentInfo(ACall, Idx, AAPtr, AlignSCEV, OffSCEV))
|
||||
if (!extractAlignmentInfo(ACall, AAPtr, AlignSCEV, OffSCEV))
|
||||
return false;
|
||||
|
||||
// Skip ConstantPointerNull and UndefValue. Assumptions on these shouldn't
|
||||
@ -248,14 +317,13 @@ bool AlignmentFromAssumptionsPass::processAssumption(CallInst *ACall,
|
||||
continue;
|
||||
|
||||
if (Instruction *K = dyn_cast<Instruction>(J))
|
||||
if (isValidAssumeForContext(ACall, K, DT))
|
||||
WorkList.push_back(K);
|
||||
}
|
||||
|
||||
while (!WorkList.empty()) {
|
||||
Instruction *J = WorkList.pop_back_val();
|
||||
if (LoadInst *LI = dyn_cast<LoadInst>(J)) {
|
||||
if (!isValidAssumeForContext(ACall, J, DT))
|
||||
continue;
|
||||
Align NewAlignment = getNewAlignment(AASCEV, AlignSCEV, OffSCEV,
|
||||
LI->getPointerOperand(), SE);
|
||||
if (NewAlignment > LI->getAlign()) {
|
||||
@ -263,8 +331,6 @@ bool AlignmentFromAssumptionsPass::processAssumption(CallInst *ACall,
|
||||
++NumLoadAlignChanged;
|
||||
}
|
||||
} else if (StoreInst *SI = dyn_cast<StoreInst>(J)) {
|
||||
if (!isValidAssumeForContext(ACall, J, DT))
|
||||
continue;
|
||||
Align NewAlignment = getNewAlignment(AASCEV, AlignSCEV, OffSCEV,
|
||||
SI->getPointerOperand(), SE);
|
||||
if (NewAlignment > SI->getAlign()) {
|
||||
@ -272,8 +338,6 @@ bool AlignmentFromAssumptionsPass::processAssumption(CallInst *ACall,
|
||||
++NumStoreAlignChanged;
|
||||
}
|
||||
} else if (MemIntrinsic *MI = dyn_cast<MemIntrinsic>(J)) {
|
||||
if (!isValidAssumeForContext(ACall, J, DT))
|
||||
continue;
|
||||
Align NewDestAlignment =
|
||||
getNewAlignment(AASCEV, AlignSCEV, OffSCEV, MI->getDest(), SE);
|
||||
|
||||
@ -305,7 +369,7 @@ bool AlignmentFromAssumptionsPass::processAssumption(CallInst *ACall,
|
||||
Visited.insert(J);
|
||||
for (User *UJ : J->users()) {
|
||||
Instruction *K = cast<Instruction>(UJ);
|
||||
if (!Visited.count(K))
|
||||
if (!Visited.count(K) && isValidAssumeForContext(ACall, K, DT))
|
||||
WorkList.push_back(K);
|
||||
}
|
||||
}
|
||||
@ -332,11 +396,8 @@ bool AlignmentFromAssumptionsPass::runImpl(Function &F, AssumptionCache &AC,
|
||||
|
||||
bool Changed = false;
|
||||
for (auto &AssumeVH : AC.assumptions())
|
||||
if (AssumeVH) {
|
||||
CallInst *Call = cast<CallInst>(AssumeVH);
|
||||
for (unsigned Idx = 0; Idx < Call->getNumOperandBundles(); Idx++)
|
||||
Changed |= processAssumption(Call, Idx);
|
||||
}
|
||||
if (AssumeVH)
|
||||
Changed |= processAssumption(cast<CallInst>(AssumeVH));
|
||||
|
||||
return Changed;
|
||||
}
|
||||
|
||||
@ -4,7 +4,10 @@ target datalayout = "e-i64:64-f80:128-n8:16:32:64-S128"
|
||||
|
||||
define i32 @foo(i32* nocapture %a) nounwind uwtable readonly {
|
||||
entry:
|
||||
tail call void @llvm.assume(i1 true) ["align"(i32* %a, i32 32)]
|
||||
%ptrint = ptrtoint i32* %a to i64
|
||||
%maskedptr = and i64 %ptrint, 31
|
||||
%maskcond = icmp eq i64 %maskedptr, 0
|
||||
tail call void @llvm.assume(i1 %maskcond)
|
||||
%0 = load i32, i32* %a, align 4
|
||||
ret i32 %0
|
||||
|
||||
@ -15,7 +18,11 @@ entry:
|
||||
|
||||
define i32 @foo2(i32* nocapture %a) nounwind uwtable readonly {
|
||||
entry:
|
||||
tail call void @llvm.assume(i1 true) ["align"(i32* %a, i32 32, i32 24)]
|
||||
%ptrint = ptrtoint i32* %a to i64
|
||||
%offsetptr = add i64 %ptrint, 24
|
||||
%maskedptr = and i64 %offsetptr, 31
|
||||
%maskcond = icmp eq i64 %maskedptr, 0
|
||||
tail call void @llvm.assume(i1 %maskcond)
|
||||
%arrayidx = getelementptr inbounds i32, i32* %a, i64 2
|
||||
%0 = load i32, i32* %arrayidx, align 4
|
||||
ret i32 %0
|
||||
@ -27,7 +34,11 @@ entry:
|
||||
|
||||
define i32 @foo2a(i32* nocapture %a) nounwind uwtable readonly {
|
||||
entry:
|
||||
tail call void @llvm.assume(i1 true) ["align"(i32* %a, i32 32, i32 28)]
|
||||
%ptrint = ptrtoint i32* %a to i64
|
||||
%offsetptr = add i64 %ptrint, 28
|
||||
%maskedptr = and i64 %offsetptr, 31
|
||||
%maskcond = icmp eq i64 %maskedptr, 0
|
||||
tail call void @llvm.assume(i1 %maskcond)
|
||||
%arrayidx = getelementptr inbounds i32, i32* %a, i64 -1
|
||||
%0 = load i32, i32* %arrayidx, align 4
|
||||
ret i32 %0
|
||||
@ -39,7 +50,10 @@ entry:
|
||||
|
||||
define i32 @goo(i32* nocapture %a) nounwind uwtable readonly {
|
||||
entry:
|
||||
tail call void @llvm.assume(i1 true) ["align"(i32* %a, i32 32, i32 0)]
|
||||
%ptrint = ptrtoint i32* %a to i64
|
||||
%maskedptr = and i64 %ptrint, 31
|
||||
%maskcond = icmp eq i64 %maskedptr, 0
|
||||
tail call void @llvm.assume(i1 %maskcond)
|
||||
%0 = load i32, i32* %a, align 4
|
||||
ret i32 %0
|
||||
|
||||
@ -50,7 +64,10 @@ entry:
|
||||
|
||||
define i32 @hoo(i32* nocapture %a) nounwind uwtable readonly {
|
||||
entry:
|
||||
tail call void @llvm.assume(i1 true) ["align"(i32* %a, i64 32, i32 0)]
|
||||
%ptrint = ptrtoint i32* %a to i64
|
||||
%maskedptr = and i64 %ptrint, 31
|
||||
%maskcond = icmp eq i64 %maskedptr, 0
|
||||
tail call void @llvm.assume(i1 %maskcond)
|
||||
br label %for.body
|
||||
|
||||
for.body: ; preds = %entry, %for.body
|
||||
@ -81,7 +98,10 @@ for.end: ; preds = %for.body
|
||||
; load(a, i0+i1+i2+32)
|
||||
define void @hoo2(i32* nocapture %a, i64 %id, i64 %num) nounwind uwtable readonly {
|
||||
entry:
|
||||
tail call void @llvm.assume(i1 true) ["align"(i32* %a, i8 32, i64 0)]
|
||||
%ptrint = ptrtoint i32* %a to i64
|
||||
%maskedptr = and i64 %ptrint, 31
|
||||
%maskcond = icmp eq i64 %maskedptr, 0
|
||||
tail call void @llvm.assume(i1 %maskcond)
|
||||
%id.mul = shl nsw i64 %id, 6
|
||||
%num.mul = shl nsw i64 %num, 6
|
||||
br label %for0.body
|
||||
@ -127,7 +147,10 @@ return:
|
||||
|
||||
define i32 @joo(i32* nocapture %a) nounwind uwtable readonly {
|
||||
entry:
|
||||
tail call void @llvm.assume(i1 true) ["align"(i32* %a, i8 32, i8 0)]
|
||||
%ptrint = ptrtoint i32* %a to i64
|
||||
%maskedptr = and i64 %ptrint, 31
|
||||
%maskcond = icmp eq i64 %maskedptr, 0
|
||||
tail call void @llvm.assume(i1 %maskcond)
|
||||
br label %for.body
|
||||
|
||||
for.body: ; preds = %entry, %for.body
|
||||
@ -152,13 +175,16 @@ for.end: ; preds = %for.body
|
||||
|
||||
define i32 @koo(i32* nocapture %a) nounwind uwtable readonly {
|
||||
entry:
|
||||
%ptrint = ptrtoint i32* %a to i64
|
||||
%maskedptr = and i64 %ptrint, 31
|
||||
%maskcond = icmp eq i64 %maskedptr, 0
|
||||
tail call void @llvm.assume(i1 %maskcond)
|
||||
br label %for.body
|
||||
|
||||
for.body: ; preds = %entry, %for.body
|
||||
%indvars.iv = phi i64 [ 0, %entry ], [ %indvars.iv.next, %for.body ]
|
||||
%r.06 = phi i32 [ 0, %entry ], [ %add, %for.body ]
|
||||
%arrayidx = getelementptr inbounds i32, i32* %a, i64 %indvars.iv
|
||||
tail call void @llvm.assume(i1 true) ["align"(i32* %a, i8 32, i8 0)]
|
||||
%0 = load i32, i32* %arrayidx, align 4
|
||||
%add = add nsw i32 %0, %r.06
|
||||
%indvars.iv.next = add i64 %indvars.iv, 4
|
||||
@ -177,7 +203,10 @@ for.end: ; preds = %for.body
|
||||
|
||||
define i32 @koo2(i32* nocapture %a) nounwind uwtable readonly {
|
||||
entry:
|
||||
tail call void @llvm.assume(i1 true) ["align"(i32* %a, i128 32, i128 0)]
|
||||
%ptrint = ptrtoint i32* %a to i64
|
||||
%maskedptr = and i64 %ptrint, 31
|
||||
%maskcond = icmp eq i64 %maskedptr, 0
|
||||
tail call void @llvm.assume(i1 %maskcond)
|
||||
br label %for.body
|
||||
|
||||
for.body: ; preds = %entry, %for.body
|
||||
@ -202,7 +231,10 @@ for.end: ; preds = %for.body
|
||||
|
||||
define i32 @moo(i32* nocapture %a) nounwind uwtable {
|
||||
entry:
|
||||
tail call void @llvm.assume(i1 true) ["align"(i32* %a, i16 32)]
|
||||
%ptrint = ptrtoint i32* %a to i64
|
||||
%maskedptr = and i64 %ptrint, 31
|
||||
%maskcond = icmp eq i64 %maskedptr, 0
|
||||
tail call void @llvm.assume(i1 %maskcond)
|
||||
%0 = bitcast i32* %a to i8*
|
||||
tail call void @llvm.memset.p0i8.i64(i8* align 4 %0, i8 0, i64 64, i1 false)
|
||||
ret i32 undef
|
||||
@ -214,9 +246,15 @@ entry:
|
||||
|
||||
define i32 @moo2(i32* nocapture %a, i32* nocapture %b) nounwind uwtable {
|
||||
entry:
|
||||
tail call void @llvm.assume(i1 true) ["align"(i32* %b, i32 128)]
|
||||
%ptrint = ptrtoint i32* %a to i64
|
||||
%maskedptr = and i64 %ptrint, 31
|
||||
%maskcond = icmp eq i64 %maskedptr, 0
|
||||
tail call void @llvm.assume(i1 %maskcond)
|
||||
%ptrint1 = ptrtoint i32* %b to i64
|
||||
%maskedptr3 = and i64 %ptrint1, 127
|
||||
%maskcond4 = icmp eq i64 %maskedptr3, 0
|
||||
tail call void @llvm.assume(i1 %maskcond4)
|
||||
%0 = bitcast i32* %a to i8*
|
||||
tail call void @llvm.assume(i1 true) ["align"(i8* %0, i16 32)]
|
||||
%1 = bitcast i32* %b to i8*
|
||||
tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 %0, i8* align 4 %1, i64 64, i1 false)
|
||||
ret i32 undef
|
||||
@ -226,19 +264,6 @@ entry:
|
||||
; CHECK: ret i32 undef
|
||||
}
|
||||
|
||||
define i32 @moo3(i32* nocapture %a, i32* nocapture %b) nounwind uwtable {
|
||||
entry:
|
||||
%0 = bitcast i32* %a to i8*
|
||||
tail call void @llvm.assume(i1 true) ["align"(i8* %0, i16 32), "align"(i32* %b, i32 128)]
|
||||
%1 = bitcast i32* %b to i8*
|
||||
tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 %0, i8* align 4 %1, i64 64, i1 false)
|
||||
ret i32 undef
|
||||
|
||||
; CHECK-LABEL: @moo3
|
||||
; CHECK: @llvm.memcpy.p0i8.p0i8.i64(i8* align 32 %0, i8* align 128 %1, i64 64, i1 false)
|
||||
; CHECK: ret i32 undef
|
||||
}
|
||||
|
||||
declare void @llvm.assume(i1) nounwind
|
||||
|
||||
declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) nounwind
|
||||
|
||||
@ -7,12 +7,18 @@ define i32 @foo(i32* nocapture %a) nounwind uwtable readonly {
|
||||
; CHECK-LABEL: define {{[^@]+}}@foo
|
||||
; CHECK-SAME: (i32* nocapture [[A:%.*]]) #0
|
||||
; CHECK-NEXT: entry:
|
||||
; CHECK-NEXT: call void @llvm.assume(i1 true) [ "align"(i32* [[A]], i64 32) ]
|
||||
; CHECK-NEXT: [[PTRINT:%.*]] = ptrtoint i32* [[A]] to i64
|
||||
; CHECK-NEXT: [[MASKEDPTR:%.*]] = and i64 [[PTRINT]], 31
|
||||
; CHECK-NEXT: [[MASKCOND:%.*]] = icmp eq i64 [[MASKEDPTR]], 0
|
||||
; CHECK-NEXT: tail call void @llvm.assume(i1 [[MASKCOND]])
|
||||
; CHECK-NEXT: [[TMP0:%.*]] = load i32, i32* [[A]], align 32
|
||||
; CHECK-NEXT: ret i32 [[TMP0]]
|
||||
;
|
||||
entry:
|
||||
call void @llvm.assume(i1 true) ["align"(i32* %a, i64 32)]
|
||||
%ptrint = ptrtoint i32* %a to i64
|
||||
%maskedptr = and i64 %ptrint, 31
|
||||
%maskcond = icmp eq i64 %maskedptr, 0
|
||||
tail call void @llvm.assume(i1 %maskcond)
|
||||
%0 = load i32, i32* %a, align 4
|
||||
ret i32 %0
|
||||
|
||||
@ -22,13 +28,21 @@ define i32 @foo2(i32* nocapture %a) nounwind uwtable readonly {
|
||||
; CHECK-LABEL: define {{[^@]+}}@foo2
|
||||
; CHECK-SAME: (i32* nocapture [[A:%.*]]) #0
|
||||
; CHECK-NEXT: entry:
|
||||
; CHECK-NEXT: call void @llvm.assume(i1 true) [ "align"(i32* [[A]], i64 32, i64 24) ]
|
||||
; CHECK-NEXT: [[PTRINT:%.*]] = ptrtoint i32* [[A]] to i64
|
||||
; CHECK-NEXT: [[OFFSETPTR:%.*]] = add i64 [[PTRINT]], 24
|
||||
; CHECK-NEXT: [[MASKEDPTR:%.*]] = and i64 [[OFFSETPTR]], 31
|
||||
; CHECK-NEXT: [[MASKCOND:%.*]] = icmp eq i64 [[MASKEDPTR]], 0
|
||||
; CHECK-NEXT: tail call void @llvm.assume(i1 [[MASKCOND]])
|
||||
; CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i32, i32* [[A]], i64 2
|
||||
; CHECK-NEXT: [[TMP0:%.*]] = load i32, i32* [[ARRAYIDX]], align 16
|
||||
; CHECK-NEXT: ret i32 [[TMP0]]
|
||||
;
|
||||
entry:
|
||||
call void @llvm.assume(i1 true) ["align"(i32* %a, i64 32, i64 24)]
|
||||
%ptrint = ptrtoint i32* %a to i64
|
||||
%offsetptr = add i64 %ptrint, 24
|
||||
%maskedptr = and i64 %offsetptr, 31
|
||||
%maskcond = icmp eq i64 %maskedptr, 0
|
||||
tail call void @llvm.assume(i1 %maskcond)
|
||||
%arrayidx = getelementptr inbounds i32, i32* %a, i64 2
|
||||
%0 = load i32, i32* %arrayidx, align 4
|
||||
ret i32 %0
|
||||
@ -39,13 +53,21 @@ define i32 @foo2a(i32* nocapture %a) nounwind uwtable readonly {
|
||||
; CHECK-LABEL: define {{[^@]+}}@foo2a
|
||||
; CHECK-SAME: (i32* nocapture [[A:%.*]]) #0
|
||||
; CHECK-NEXT: entry:
|
||||
; CHECK-NEXT: call void @llvm.assume(i1 true) [ "align"(i32* [[A]], i64 32, i64 28) ]
|
||||
; CHECK-NEXT: [[PTRINT:%.*]] = ptrtoint i32* [[A]] to i64
|
||||
; CHECK-NEXT: [[OFFSETPTR:%.*]] = add i64 [[PTRINT]], 28
|
||||
; CHECK-NEXT: [[MASKEDPTR:%.*]] = and i64 [[OFFSETPTR]], 31
|
||||
; CHECK-NEXT: [[MASKCOND:%.*]] = icmp eq i64 [[MASKEDPTR]], 0
|
||||
; CHECK-NEXT: tail call void @llvm.assume(i1 [[MASKCOND]])
|
||||
; CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i32, i32* [[A]], i64 -1
|
||||
; CHECK-NEXT: [[TMP0:%.*]] = load i32, i32* [[ARRAYIDX]], align 32
|
||||
; CHECK-NEXT: ret i32 [[TMP0]]
|
||||
;
|
||||
entry:
|
||||
call void @llvm.assume(i1 true) ["align"(i32* %a, i64 32, i64 28)]
|
||||
%ptrint = ptrtoint i32* %a to i64
|
||||
%offsetptr = add i64 %ptrint, 28
|
||||
%maskedptr = and i64 %offsetptr, 31
|
||||
%maskcond = icmp eq i64 %maskedptr, 0
|
||||
tail call void @llvm.assume(i1 %maskcond)
|
||||
%arrayidx = getelementptr inbounds i32, i32* %a, i64 -1
|
||||
%0 = load i32, i32* %arrayidx, align 4
|
||||
ret i32 %0
|
||||
@ -56,12 +78,18 @@ define i32 @goo(i32* nocapture %a) nounwind uwtable readonly {
|
||||
; CHECK-LABEL: define {{[^@]+}}@goo
|
||||
; CHECK-SAME: (i32* nocapture [[A:%.*]]) #0
|
||||
; CHECK-NEXT: entry:
|
||||
; CHECK-NEXT: call void @llvm.assume(i1 true) [ "align"(i32* [[A]], i64 32) ]
|
||||
; CHECK-NEXT: [[PTRINT:%.*]] = ptrtoint i32* [[A]] to i64
|
||||
; CHECK-NEXT: [[MASKEDPTR:%.*]] = and i64 [[PTRINT]], 31
|
||||
; CHECK-NEXT: [[MASKCOND:%.*]] = icmp eq i64 [[MASKEDPTR]], 0
|
||||
; CHECK-NEXT: tail call void @llvm.assume(i1 [[MASKCOND]])
|
||||
; CHECK-NEXT: [[TMP0:%.*]] = load i32, i32* [[A]], align 32
|
||||
; CHECK-NEXT: ret i32 [[TMP0]]
|
||||
;
|
||||
entry:
|
||||
call void @llvm.assume(i1 true) ["align"(i32* %a, i64 32)]
|
||||
%ptrint = ptrtoint i32* %a to i64
|
||||
%maskedptr = and i64 %ptrint, 31
|
||||
%maskcond = icmp eq i64 %maskedptr, 0
|
||||
tail call void @llvm.assume(i1 %maskcond)
|
||||
%0 = load i32, i32* %a, align 4
|
||||
ret i32 %0
|
||||
|
||||
@ -71,7 +99,10 @@ define i32 @hoo(i32* nocapture %a) nounwind uwtable readonly {
|
||||
; CHECK-LABEL: define {{[^@]+}}@hoo
|
||||
; CHECK-SAME: (i32* nocapture [[A:%.*]]) #0
|
||||
; CHECK-NEXT: entry:
|
||||
; CHECK-NEXT: call void @llvm.assume(i1 true) [ "align"(i32* [[A]], i64 32) ]
|
||||
; CHECK-NEXT: [[PTRINT:%.*]] = ptrtoint i32* [[A]] to i64
|
||||
; CHECK-NEXT: [[MASKEDPTR:%.*]] = and i64 [[PTRINT]], 31
|
||||
; CHECK-NEXT: [[MASKCOND:%.*]] = icmp eq i64 [[MASKEDPTR]], 0
|
||||
; CHECK-NEXT: tail call void @llvm.assume(i1 [[MASKCOND]])
|
||||
; CHECK-NEXT: br label [[FOR_BODY:%.*]]
|
||||
; CHECK: for.body:
|
||||
; CHECK-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ 0, [[ENTRY:%.*]] ], [ [[INDVARS_IV_NEXT:%.*]], [[FOR_BODY]] ]
|
||||
@ -88,7 +119,10 @@ define i32 @hoo(i32* nocapture %a) nounwind uwtable readonly {
|
||||
; CHECK-NEXT: ret i32 [[ADD_LCSSA]]
|
||||
;
|
||||
entry:
|
||||
call void @llvm.assume(i1 true) ["align"(i32* %a, i64 32)]
|
||||
%ptrint = ptrtoint i32* %a to i64
|
||||
%maskedptr = and i64 %ptrint, 31
|
||||
%maskcond = icmp eq i64 %maskedptr, 0
|
||||
tail call void @llvm.assume(i1 %maskcond)
|
||||
br label %for.body
|
||||
|
||||
for.body: ; preds = %entry, %for.body
|
||||
@ -112,7 +146,10 @@ define i32 @joo(i32* nocapture %a) nounwind uwtable readonly {
|
||||
; CHECK-LABEL: define {{[^@]+}}@joo
|
||||
; CHECK-SAME: (i32* nocapture [[A:%.*]]) #0
|
||||
; CHECK-NEXT: entry:
|
||||
; CHECK-NEXT: call void @llvm.assume(i1 true) [ "align"(i32* [[A]], i64 32) ]
|
||||
; CHECK-NEXT: [[PTRINT:%.*]] = ptrtoint i32* [[A]] to i64
|
||||
; CHECK-NEXT: [[MASKEDPTR:%.*]] = and i64 [[PTRINT]], 31
|
||||
; CHECK-NEXT: [[MASKCOND:%.*]] = icmp eq i64 [[MASKEDPTR]], 0
|
||||
; CHECK-NEXT: tail call void @llvm.assume(i1 [[MASKCOND]])
|
||||
; CHECK-NEXT: br label [[FOR_BODY:%.*]]
|
||||
; CHECK: for.body:
|
||||
; CHECK-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ 4, [[ENTRY:%.*]] ], [ [[INDVARS_IV_NEXT:%.*]], [[FOR_BODY]] ]
|
||||
@ -129,7 +166,10 @@ define i32 @joo(i32* nocapture %a) nounwind uwtable readonly {
|
||||
; CHECK-NEXT: ret i32 [[ADD_LCSSA]]
|
||||
;
|
||||
entry:
|
||||
call void @llvm.assume(i1 true) ["align"(i32* %a, i64 32)]
|
||||
%ptrint = ptrtoint i32* %a to i64
|
||||
%maskedptr = and i64 %ptrint, 31
|
||||
%maskcond = icmp eq i64 %maskedptr, 0
|
||||
tail call void @llvm.assume(i1 %maskcond)
|
||||
br label %for.body
|
||||
|
||||
for.body: ; preds = %entry, %for.body
|
||||
@ -153,7 +193,10 @@ define i32 @koo(i32* nocapture %a) nounwind uwtable readonly {
|
||||
; CHECK-LABEL: define {{[^@]+}}@koo
|
||||
; CHECK-SAME: (i32* nocapture [[A:%.*]]) #0
|
||||
; CHECK-NEXT: entry:
|
||||
; CHECK-NEXT: call void @llvm.assume(i1 true) [ "align"(i32* [[A]], i64 32) ]
|
||||
; CHECK-NEXT: [[PTRINT:%.*]] = ptrtoint i32* [[A]] to i64
|
||||
; CHECK-NEXT: [[MASKEDPTR:%.*]] = and i64 [[PTRINT]], 31
|
||||
; CHECK-NEXT: [[MASKCOND:%.*]] = icmp eq i64 [[MASKEDPTR]], 0
|
||||
; CHECK-NEXT: tail call void @llvm.assume(i1 [[MASKCOND]])
|
||||
; CHECK-NEXT: br label [[FOR_BODY:%.*]]
|
||||
; CHECK: for.body:
|
||||
; CHECK-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ 0, [[ENTRY:%.*]] ], [ [[INDVARS_IV_NEXT:%.*]], [[FOR_BODY]] ]
|
||||
@ -170,7 +213,10 @@ define i32 @koo(i32* nocapture %a) nounwind uwtable readonly {
|
||||
; CHECK-NEXT: ret i32 [[ADD_LCSSA]]
|
||||
;
|
||||
entry:
|
||||
call void @llvm.assume(i1 true) ["align"(i32* %a, i64 32)]
|
||||
%ptrint = ptrtoint i32* %a to i64
|
||||
%maskedptr = and i64 %ptrint, 31
|
||||
%maskcond = icmp eq i64 %maskedptr, 0
|
||||
tail call void @llvm.assume(i1 %maskcond)
|
||||
br label %for.body
|
||||
|
||||
for.body: ; preds = %entry, %for.body
|
||||
@ -194,7 +240,10 @@ define i32 @koo2(i32* nocapture %a) nounwind uwtable readonly {
|
||||
; CHECK-LABEL: define {{[^@]+}}@koo2
|
||||
; CHECK-SAME: (i32* nocapture [[A:%.*]]) #0
|
||||
; CHECK-NEXT: entry:
|
||||
; CHECK-NEXT: call void @llvm.assume(i1 true) [ "align"(i32* [[A]], i64 32) ]
|
||||
; CHECK-NEXT: [[PTRINT:%.*]] = ptrtoint i32* [[A]] to i64
|
||||
; CHECK-NEXT: [[MASKEDPTR:%.*]] = and i64 [[PTRINT]], 31
|
||||
; CHECK-NEXT: [[MASKCOND:%.*]] = icmp eq i64 [[MASKEDPTR]], 0
|
||||
; CHECK-NEXT: tail call void @llvm.assume(i1 [[MASKCOND]])
|
||||
; CHECK-NEXT: br label [[FOR_BODY:%.*]]
|
||||
; CHECK: for.body:
|
||||
; CHECK-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ -4, [[ENTRY:%.*]] ], [ [[INDVARS_IV_NEXT:%.*]], [[FOR_BODY]] ]
|
||||
@ -211,7 +260,10 @@ define i32 @koo2(i32* nocapture %a) nounwind uwtable readonly {
|
||||
; CHECK-NEXT: ret i32 [[ADD_LCSSA]]
|
||||
;
|
||||
entry:
|
||||
call void @llvm.assume(i1 true) ["align"(i32* %a, i64 32)]
|
||||
%ptrint = ptrtoint i32* %a to i64
|
||||
%maskedptr = and i64 %ptrint, 31
|
||||
%maskcond = icmp eq i64 %maskedptr, 0
|
||||
tail call void @llvm.assume(i1 %maskcond)
|
||||
br label %for.body
|
||||
|
||||
for.body: ; preds = %entry, %for.body
|
||||
@ -235,13 +287,19 @@ define i32 @moo(i32* nocapture %a) nounwind uwtable {
|
||||
; CHECK-LABEL: define {{[^@]+}}@moo
|
||||
; CHECK-SAME: (i32* nocapture [[A:%.*]]) #1
|
||||
; CHECK-NEXT: entry:
|
||||
; CHECK-NEXT: call void @llvm.assume(i1 true) [ "align"(i32* [[A]], i64 32) ]
|
||||
; CHECK-NEXT: [[PTRINT:%.*]] = ptrtoint i32* [[A]] to i64
|
||||
; CHECK-NEXT: [[MASKEDPTR:%.*]] = and i64 [[PTRINT]], 31
|
||||
; CHECK-NEXT: [[MASKCOND:%.*]] = icmp eq i64 [[MASKEDPTR]], 0
|
||||
; CHECK-NEXT: tail call void @llvm.assume(i1 [[MASKCOND]])
|
||||
; CHECK-NEXT: [[TMP0:%.*]] = bitcast i32* [[A]] to i8*
|
||||
; CHECK-NEXT: tail call void @llvm.memset.p0i8.i64(i8* align 32 [[TMP0]], i8 0, i64 64, i1 false)
|
||||
; CHECK-NEXT: ret i32 undef
|
||||
;
|
||||
entry:
|
||||
call void @llvm.assume(i1 true) ["align"(i32* %a, i64 32)]
|
||||
%ptrint = ptrtoint i32* %a to i64
|
||||
%maskedptr = and i64 %ptrint, 31
|
||||
%maskcond = icmp eq i64 %maskedptr, 0
|
||||
tail call void @llvm.assume(i1 %maskcond)
|
||||
%0 = bitcast i32* %a to i8*
|
||||
tail call void @llvm.memset.p0i8.i64(i8* align 4 %0, i8 0, i64 64, i1 false)
|
||||
ret i32 undef
|
||||
@ -252,16 +310,28 @@ define i32 @moo2(i32* nocapture %a, i32* nocapture %b) nounwind uwtable {
|
||||
; CHECK-LABEL: define {{[^@]+}}@moo2
|
||||
; CHECK-SAME: (i32* nocapture [[A:%.*]], i32* nocapture [[B:%.*]]) #1
|
||||
; CHECK-NEXT: entry:
|
||||
; CHECK-NEXT: call void @llvm.assume(i1 true) [ "align"(i32* [[A]], i64 32) ]
|
||||
; CHECK-NEXT: call void @llvm.assume(i1 true) [ "align"(i32* [[B]], i64 128) ]
|
||||
; CHECK-NEXT: [[PTRINT:%.*]] = ptrtoint i32* [[A]] to i64
|
||||
; CHECK-NEXT: [[MASKEDPTR:%.*]] = and i64 [[PTRINT]], 31
|
||||
; CHECK-NEXT: [[MASKCOND:%.*]] = icmp eq i64 [[MASKEDPTR]], 0
|
||||
; CHECK-NEXT: tail call void @llvm.assume(i1 [[MASKCOND]])
|
||||
; CHECK-NEXT: [[PTRINT1:%.*]] = ptrtoint i32* [[B]] to i64
|
||||
; CHECK-NEXT: [[MASKEDPTR3:%.*]] = and i64 [[PTRINT1]], 127
|
||||
; CHECK-NEXT: [[MASKCOND4:%.*]] = icmp eq i64 [[MASKEDPTR3]], 0
|
||||
; CHECK-NEXT: tail call void @llvm.assume(i1 [[MASKCOND4]])
|
||||
; CHECK-NEXT: [[TMP0:%.*]] = bitcast i32* [[A]] to i8*
|
||||
; CHECK-NEXT: [[TMP1:%.*]] = bitcast i32* [[B]] to i8*
|
||||
; CHECK-NEXT: tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 32 [[TMP0]], i8* align 128 [[TMP1]], i64 64, i1 false)
|
||||
; CHECK-NEXT: ret i32 undef
|
||||
;
|
||||
entry:
|
||||
call void @llvm.assume(i1 true) ["align"(i32* %a, i64 32)]
|
||||
call void @llvm.assume(i1 true) ["align"(i32* %b, i64 128)]
|
||||
%ptrint = ptrtoint i32* %a to i64
|
||||
%maskedptr = and i64 %ptrint, 31
|
||||
%maskcond = icmp eq i64 %maskedptr, 0
|
||||
tail call void @llvm.assume(i1 %maskcond)
|
||||
%ptrint1 = ptrtoint i32* %b to i64
|
||||
%maskedptr3 = and i64 %ptrint1, 127
|
||||
%maskcond4 = icmp eq i64 %maskedptr3, 0
|
||||
tail call void @llvm.assume(i1 %maskcond4)
|
||||
%0 = bitcast i32* %a to i8*
|
||||
%1 = bitcast i32* %b to i8*
|
||||
tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 %0, i8* align 4 %1, i64 64, i1 false)
|
||||
|
||||
@ -23,7 +23,10 @@ define void @foo(float* nocapture %a, float* nocapture readonly %c) #0 {
|
||||
; CHECK-LABEL: define {{[^@]+}}@foo
|
||||
; CHECK-SAME: (float* nocapture [[A:%.*]], float* nocapture readonly [[C:%.*]]) #0
|
||||
; CHECK-NEXT: entry:
|
||||
; CHECK-NEXT: call void @llvm.assume(i1 true) [ "align"(float* [[A]], i64 128) ]
|
||||
; CHECK-NEXT: [[PTRINT:%.*]] = ptrtoint float* [[A]] to i64
|
||||
; CHECK-NEXT: [[MASKEDPTR:%.*]] = and i64 [[PTRINT]], 127
|
||||
; CHECK-NEXT: [[MASKCOND:%.*]] = icmp eq i64 [[MASKEDPTR]], 0
|
||||
; CHECK-NEXT: call void @llvm.assume(i1 [[MASKCOND]])
|
||||
; CHECK-NEXT: [[TMP0:%.*]] = load float, float* [[C]], align 4
|
||||
; CHECK-NEXT: [[ARRAYIDX_I:%.*]] = getelementptr inbounds float, float* [[A]], i64 5
|
||||
; CHECK-NEXT: store float [[TMP0]], float* [[ARRAYIDX_I]], align 4
|
||||
@ -84,8 +87,14 @@ define void @foo2(float* nocapture %a, float* nocapture %b, float* nocapture rea
|
||||
; CHECK-LABEL: define {{[^@]+}}@foo2
|
||||
; CHECK-SAME: (float* nocapture [[A:%.*]], float* nocapture [[B:%.*]], float* nocapture readonly [[C:%.*]]) #0
|
||||
; CHECK-NEXT: entry:
|
||||
; CHECK-NEXT: call void @llvm.assume(i1 true) [ "align"(float* [[A]], i64 128) ]
|
||||
; CHECK-NEXT: call void @llvm.assume(i1 true) [ "align"(float* [[B]], i64 128) ]
|
||||
; CHECK-NEXT: [[PTRINT:%.*]] = ptrtoint float* [[A]] to i64
|
||||
; CHECK-NEXT: [[MASKEDPTR:%.*]] = and i64 [[PTRINT]], 127
|
||||
; CHECK-NEXT: [[MASKCOND:%.*]] = icmp eq i64 [[MASKEDPTR]], 0
|
||||
; CHECK-NEXT: call void @llvm.assume(i1 [[MASKCOND]])
|
||||
; CHECK-NEXT: [[PTRINT1:%.*]] = ptrtoint float* [[B]] to i64
|
||||
; CHECK-NEXT: [[MASKEDPTR2:%.*]] = and i64 [[PTRINT1]], 127
|
||||
; CHECK-NEXT: [[MASKCOND3:%.*]] = icmp eq i64 [[MASKEDPTR2]], 0
|
||||
; CHECK-NEXT: call void @llvm.assume(i1 [[MASKCOND3]])
|
||||
; CHECK-NEXT: [[TMP0:%.*]] = load float, float* [[C]], align 4
|
||||
; CHECK-NEXT: [[ARRAYIDX_I:%.*]] = getelementptr inbounds float, float* [[A]], i64 5
|
||||
; CHECK-NEXT: store float [[TMP0]], float* [[ARRAYIDX_I]], align 4
|
||||
|
||||
@ -377,7 +377,6 @@ define i32 @assumption_conflicts_with_known_bits(i32 %a, i32 %b) {
|
||||
define void @debug_interference(i8 %x) {
|
||||
; CHECK-LABEL: @debug_interference(
|
||||
; CHECK-NEXT: [[CMP2:%.*]] = icmp ne i8 [[X:%.*]], 0
|
||||
; CHECK-NEXT: tail call void @llvm.assume(i1 false)
|
||||
; CHECK-NEXT: tail call void @llvm.dbg.value(metadata i32 5, metadata !7, metadata !DIExpression()), !dbg !9
|
||||
; CHECK-NEXT: tail call void @llvm.assume(i1 false)
|
||||
; CHECK-NEXT: tail call void @llvm.dbg.value(metadata i32 5, metadata !7, metadata !DIExpression()), !dbg !9
|
||||
|
||||
@ -41,7 +41,10 @@ define void @caller1(i1 %c, i64* align 1 %ptr) {
|
||||
; ASSUMPTIONS-ON-NEXT: br i1 [[C:%.*]], label [[TRUE2_CRITEDGE:%.*]], label [[FALSE1:%.*]]
|
||||
; ASSUMPTIONS-ON: false1:
|
||||
; ASSUMPTIONS-ON-NEXT: store volatile i64 1, i64* [[PTR:%.*]], align 8
|
||||
; ASSUMPTIONS-ON-NEXT: call void @llvm.assume(i1 true) [ "align"(i64* [[PTR]], i64 8) ]
|
||||
; ASSUMPTIONS-ON-NEXT: [[PTRINT:%.*]] = ptrtoint i64* [[PTR]] to i64
|
||||
; ASSUMPTIONS-ON-NEXT: [[MASKEDPTR:%.*]] = and i64 [[PTRINT]], 7
|
||||
; ASSUMPTIONS-ON-NEXT: [[MASKCOND:%.*]] = icmp eq i64 [[MASKEDPTR]], 0
|
||||
; ASSUMPTIONS-ON-NEXT: tail call void @llvm.assume(i1 [[MASKCOND]])
|
||||
; ASSUMPTIONS-ON-NEXT: store volatile i64 0, i64* [[PTR]], align 8
|
||||
; ASSUMPTIONS-ON-NEXT: store volatile i64 -1, i64* [[PTR]], align 8
|
||||
; ASSUMPTIONS-ON-NEXT: store volatile i64 -1, i64* [[PTR]], align 8
|
||||
@ -51,7 +54,10 @@ define void @caller1(i1 %c, i64* align 1 %ptr) {
|
||||
; ASSUMPTIONS-ON-NEXT: store volatile i64 3, i64* [[PTR]], align 8
|
||||
; ASSUMPTIONS-ON-NEXT: ret void
|
||||
; ASSUMPTIONS-ON: true2.critedge:
|
||||
; ASSUMPTIONS-ON-NEXT: call void @llvm.assume(i1 true) [ "align"(i64* [[PTR]], i64 8) ]
|
||||
; ASSUMPTIONS-ON-NEXT: [[PTRINT_C:%.*]] = ptrtoint i64* [[PTR]] to i64
|
||||
; ASSUMPTIONS-ON-NEXT: [[MASKEDPTR_C:%.*]] = and i64 [[PTRINT_C]], 7
|
||||
; ASSUMPTIONS-ON-NEXT: [[MASKCOND_C:%.*]] = icmp eq i64 [[MASKEDPTR_C]], 0
|
||||
; ASSUMPTIONS-ON-NEXT: tail call void @llvm.assume(i1 [[MASKCOND_C]])
|
||||
; ASSUMPTIONS-ON-NEXT: store volatile i64 0, i64* [[PTR]], align 8
|
||||
; ASSUMPTIONS-ON-NEXT: store volatile i64 -1, i64* [[PTR]], align 8
|
||||
; ASSUMPTIONS-ON-NEXT: store volatile i64 -1, i64* [[PTR]], align 8
|
||||
@ -88,17 +94,26 @@ false2:
|
||||
; This test checks that alignment assumptions do not prevent SROA.
|
||||
; See PR45763.
|
||||
|
||||
define internal void @callee2(i64* noalias sret align 32 %arg) {
|
||||
define internal void @callee2(i64* noalias sret align 8 %arg) {
|
||||
store i64 0, i64* %arg, align 8
|
||||
ret void
|
||||
}
|
||||
|
||||
define amdgpu_kernel void @caller2() {
|
||||
; CHECK-LABEL: @caller2(
|
||||
; CHECK-NEXT: ret void
|
||||
; ASSUMPTIONS-OFF-LABEL: @caller2(
|
||||
; ASSUMPTIONS-OFF-NEXT: ret void
|
||||
;
|
||||
; ASSUMPTIONS-ON-LABEL: @caller2(
|
||||
; ASSUMPTIONS-ON-NEXT: [[ALLOCA:%.*]] = alloca i64, align 8, addrspace(5)
|
||||
; ASSUMPTIONS-ON-NEXT: [[CAST:%.*]] = addrspacecast i64 addrspace(5)* [[ALLOCA]] to i64*
|
||||
; ASSUMPTIONS-ON-NEXT: [[PTRINT:%.*]] = ptrtoint i64* [[CAST]] to i64
|
||||
; ASSUMPTIONS-ON-NEXT: [[MASKEDPTR:%.*]] = and i64 [[PTRINT]], 7
|
||||
; ASSUMPTIONS-ON-NEXT: [[MASKCOND:%.*]] = icmp eq i64 [[MASKEDPTR]], 0
|
||||
; ASSUMPTIONS-ON-NEXT: call void @llvm.assume(i1 [[MASKCOND]])
|
||||
; ASSUMPTIONS-ON-NEXT: ret void
|
||||
;
|
||||
%alloca = alloca i64, align 8, addrspace(5)
|
||||
%cast = addrspacecast i64 addrspace(5)* %alloca to i64*
|
||||
call void @callee2(i64* sret align 32 %cast)
|
||||
call void @callee2(i64* sret align 8 %cast)
|
||||
ret void
|
||||
}
|
||||
|
||||
@ -1,4 +1,3 @@
|
||||
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
|
||||
; RUN: not opt -verify < %s 2>&1 | FileCheck %s
|
||||
|
||||
declare void @llvm.assume(i1)
|
||||
@ -7,21 +6,14 @@ define void @func(i32* %P, i32 %P1, i32* %P2, i32* %P3) {
|
||||
; CHECK: tags must be valid attribute names
|
||||
call void @llvm.assume(i1 true) ["adazdazd"()]
|
||||
; CHECK: the second argument should be a constant integral value
|
||||
call void @llvm.assume(i1 true) ["dereferenceable"(i32* %P, i32 %P1)]
|
||||
call void @llvm.assume(i1 true) ["align"(i32* %P, i32 %P1)]
|
||||
; CHECK: to many arguments
|
||||
call void @llvm.assume(i1 true) ["dereferenceable"(i32* %P, i32 8, i32 8)]
|
||||
call void @llvm.assume(i1 true) ["align"(i32* %P, i32 8, i32 8)]
|
||||
; CHECK: this attribute should have 2 arguments
|
||||
call void @llvm.assume(i1 true) ["dereferenceable"(i32* %P)]
|
||||
call void @llvm.assume(i1 true) ["align"(i32* %P)]
|
||||
; CHECK: this attribute has no argument
|
||||
call void @llvm.assume(i1 true) ["dereferenceable"(i32* %P, i32 4), "cold"(i32* %P)]
|
||||
call void @llvm.assume(i1 true) ["align"(i32* %P, i32 4), "cold"(i32* %P)]
|
||||
; CHECK: this attribute should have one argument
|
||||
call void @llvm.assume(i1 true) ["noalias"()]
|
||||
call void @llvm.assume(i1 true) ["align"(i32* %P, i32 %P1, i32 4)]
|
||||
; CHECK: alignment assumptions should have 2 or 3 arguments
|
||||
call void @llvm.assume(i1 true) ["align"(i32* %P, i32 %P1, i32 4, i32 4)]
|
||||
; CHECK: second argument should be an integer
|
||||
call void @llvm.assume(i1 true) ["align"(i32* %P, i32* %P2)]
|
||||
; CHECK: third argument should be an integer if present
|
||||
call void @llvm.assume(i1 true) ["align"(i32* %P, i32 %P1, i32* %P2)]
|
||||
ret void
|
||||
}
|
||||
|
||||
@ -546,41 +546,3 @@ TEST(AssumeQueryAPI, AssumptionCache) {
|
||||
ASSERT_EQ(AR[0].Index, 1u);
|
||||
ASSERT_EQ(AR[0].Assume, &*First);
|
||||
}
|
||||
|
||||
TEST(AssumeQueryAPI, Alignment) {
|
||||
LLVMContext C;
|
||||
SMDiagnostic Err;
|
||||
std::unique_ptr<Module> Mod = parseAssemblyString(
|
||||
"declare void @llvm.assume(i1)\n"
|
||||
"define void @test(i32* %P, i32* %P1, i32* %P2, i32 %I3, i1 %B) {\n"
|
||||
"call void @llvm.assume(i1 true) [\"align\"(i32* %P, i32 8, i32 %I3)]\n"
|
||||
"call void @llvm.assume(i1 true) [\"align\"(i32* %P1, i32 %I3, i32 "
|
||||
"%I3)]\n"
|
||||
"call void @llvm.assume(i1 true) [\"align\"(i32* %P2, i32 16, i32 8)]\n"
|
||||
"ret void\n}\n",
|
||||
Err, C);
|
||||
if (!Mod)
|
||||
Err.print("AssumeQueryAPI", errs());
|
||||
|
||||
Function *F = Mod->getFunction("test");
|
||||
BasicBlock::iterator Start = F->begin()->begin();
|
||||
IntrinsicInst *II;
|
||||
RetainedKnowledge RK;
|
||||
II = cast<IntrinsicInst>(&*Start);
|
||||
RK = getKnowledgeFromBundle(*II, II->bundle_op_info_begin()[0]);
|
||||
ASSERT_EQ(RK.AttrKind, Attribute::Alignment);
|
||||
ASSERT_EQ(RK.WasOn, F->getArg(0));
|
||||
ASSERT_EQ(RK.ArgValue, 1u);
|
||||
Start++;
|
||||
II = cast<IntrinsicInst>(&*Start);
|
||||
RK = getKnowledgeFromBundle(*II, II->bundle_op_info_begin()[0]);
|
||||
ASSERT_EQ(RK.AttrKind, Attribute::Alignment);
|
||||
ASSERT_EQ(RK.WasOn, F->getArg(1));
|
||||
ASSERT_EQ(RK.ArgValue, 1u);
|
||||
Start++;
|
||||
II = cast<IntrinsicInst>(&*Start);
|
||||
RK = getKnowledgeFromBundle(*II, II->bundle_op_info_begin()[0]);
|
||||
ASSERT_EQ(RK.AttrKind, Attribute::Alignment);
|
||||
ASSERT_EQ(RK.WasOn, F->getArg(2));
|
||||
ASSERT_EQ(RK.ArgValue, 8u);
|
||||
}
|
||||
|
||||
Loading…
Reference in New Issue
Block a user