#include "llvm/ADT/StringExtras.h"
#include "llvm/Analysis/AliasAnalysis.h"
#include "llvm/Analysis/AliasSetTracker.h"
-#include "llvm/Analysis/AssumptionTracker.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/BlockFrequencyInfo.h"
#include "llvm/Analysis/CodeMetrics.h"
#include "llvm/Analysis/LoopInfo.h"
/// \brief Propagate known metadata from one instruction to another.
static void propagateMetadata(Instruction *To, const Instruction *From) {
- SmallVector<std::pair<unsigned, Value *>, 4> Metadata;
+ SmallVector<std::pair<unsigned, MDNode *>, 4> Metadata;
From->getAllMetadataOtherThanDebugLoc(Metadata);
for (auto M : Metadata) {
LoopVectorizationLegality(Loop *L, ScalarEvolution *SE, const DataLayout *DL,
DominatorTree *DT, TargetLibraryInfo *TLI,
- AliasAnalysis *AA, Function *F)
+ AliasAnalysis *AA, Function *F,
+ const TargetTransformInfo *TTI)
: NumLoads(0), NumStores(0), NumPredStores(0), TheLoop(L), SE(SE), DL(DL),
- DT(DT), TLI(TLI), AA(AA), TheFunction(F), Induction(nullptr),
+ DT(DT), TLI(TLI), AA(AA), TheFunction(F), TTI(TTI), Induction(nullptr),
WidestIndTy(nullptr), HasFunNoNaNAttr(false), MaxSafeDepDistBytes(-1U) {
}
}
SmallPtrSet<Value *, 8>::iterator strides_end() { return StrideSet.end(); }
+ /// Returns true if the target machine supports masked store operation
+ /// for the given \p DataType and kind of access to \p Ptr.
+ bool isLegalMaskedStore(Type *DataType, Value *Ptr) {
+ return TTI->isLegalMaskedStore(DataType, isConsecutivePtr(Ptr));
+ }
+ /// Returns true if the target machine supports masked load operation
+ /// for the given \p DataType and kind of access to \p Ptr.
+ bool isLegalMaskedLoad(Type *DataType, Value *Ptr) {
+ return TTI->isLegalMaskedLoad(DataType, isConsecutivePtr(Ptr));
+ }
+ /// Returns true if vector representation of the instruction \p I
+ /// requires mask.
+ bool isMaskRequired(const Instruction* I) {
+ return (MaskedOp.count(I) != 0);
+ }
private:
/// Check if a single basic block loop is vectorizable.
/// At this point we know that this is a loop with a constant trip count
///
/// Looks for accesses like "a[i * StrideA]" where "StrideA" is loop
/// invariant.
- void collectStridedAcccess(Value *LoadOrStoreInst);
+ void collectStridedAccess(Value *LoadOrStoreInst);
/// Report an analysis message to assist the user in diagnosing loops that are
/// not vectorized.
AliasAnalysis *AA;
/// Parent function
Function *TheFunction;
+ /// Target Transform Info
+ const TargetTransformInfo *TTI;
// --- vectorization state --- //
ValueToValueMap Strides;
SmallPtrSet<Value *, 8> StrideSet;
+
+ /// While vectorizing these instructions we have to generate a
+ /// call to the appropriate masked intrinsic
+ SmallPtrSet<const Instruction*, 8> MaskedOp;
};
/// LoopVectorizationCostModel - estimates the expected speedups due to
LoopVectorizationLegality *Legal,
const TargetTransformInfo &TTI,
const DataLayout *DL, const TargetLibraryInfo *TLI,
- AssumptionTracker *AT, const Function *F,
+ AssumptionCache *AC, const Function *F,
const LoopVectorizeHints *Hints)
: TheLoop(L), SE(SE), LI(LI), Legal(Legal), TTI(TTI), DL(DL), TLI(TLI),
TheFunction(F), Hints(Hints) {
- CodeMetrics::collectEphemeralValues(L, AT, EphValues);
+ CodeMetrics::collectEphemeralValues(L, AC, EphValues);
}
/// Information about vectorization costs
for (unsigned i = 1, ie = LoopID->getNumOperands(); i < ie; ++i) {
const MDString *S = nullptr;
- SmallVector<Value*, 4> Args;
+ SmallVector<Metadata *, 4> Args;
// The expected hint is either a MDString or a MDNode with the first
// operand a MDString.
}
/// Checks string hint with one operand and set value if valid.
- void setHint(StringRef Name, Value *Arg) {
+ void setHint(StringRef Name, Metadata *Arg) {
if (!Name.startswith(Prefix()))
return;
Name = Name.substr(Prefix().size(), StringRef::npos);
- const ConstantInt *C = dyn_cast<ConstantInt>(Arg);
+ const ConstantInt *C = mdconst::dyn_extract<ConstantInt>(Arg);
if (!C) return;
unsigned Val = C->getZExtValue();
/// Create a new hint from name / value pair.
MDNode *createHintMetadata(StringRef Name, unsigned V) const {
LLVMContext &Context = TheLoop->getHeader()->getContext();
- Value *Vals[] = {MDString::get(Context, Name),
- ConstantInt::get(Type::getInt32Ty(Context), V)};
- return MDNode::get(Context, Vals);
+ Metadata *MDs[] = {MDString::get(Context, Name),
+ ConstantAsMetadata::get(
+ ConstantInt::get(Type::getInt32Ty(Context), V))};
+ return MDNode::get(Context, MDs);
}
/// Matches metadata with hint name.
return false;
for (auto H : HintTypes)
- if (Name->getName().endswith(H.Name))
+ if (Name->getString().endswith(H.Name))
return true;
return false;
}
return;
// Reserve the first element to LoopID (see below).
- SmallVector<Value*, 4> Vals(1);
+ SmallVector<Metadata *, 4> MDs(1);
// If the loop already has metadata, then ignore the existing operands.
MDNode *LoopID = TheLoop->getLoopID();
if (LoopID) {
MDNode *Node = cast<MDNode>(LoopID->getOperand(i));
// If node in update list, ignore old value.
if (!matchesHintMetadataName(Node, HintTypes))
- Vals.push_back(Node);
+ MDs.push_back(Node);
}
}
// Now, add the missing hints.
for (auto H : HintTypes)
- Vals.push_back(
- createHintMetadata(Twine(Prefix(), H.Name).str(), H.Value));
+ MDs.push_back(createHintMetadata(Twine(Prefix(), H.Name).str(), H.Value));
// Replace current metadata node with new one.
LLVMContext &Context = TheLoop->getHeader()->getContext();
- MDNode *NewLoopID = MDNode::get(Context, Vals);
+ MDNode *NewLoopID = MDNode::get(Context, MDs);
// Set operand 0 to refer to the loop id itself.
NewLoopID->replaceOperandWith(0, NewLoopID);
TheLoop->setLoopID(NewLoopID);
- if (LoopID)
- LoopID->replaceAllUsesWith(NewLoopID);
- LoopID = NewLoopID;
}
/// The loop these hints belong to.
BlockFrequencyInfo *BFI;
TargetLibraryInfo *TLI;
AliasAnalysis *AA;
- AssumptionTracker *AT;
+ AssumptionCache *AC;
bool DisableUnrolling;
bool AlwaysVectorize;
SE = &getAnalysis<ScalarEvolution>();
DataLayoutPass *DLP = getAnalysisIfAvailable<DataLayoutPass>();
DL = DLP ? &DLP->getDataLayout() : nullptr;
- LI = &getAnalysis<LoopInfo>();
+ LI = &getAnalysis<LoopInfoWrapperPass>().getLoopInfo();
TTI = &getAnalysis<TargetTransformInfo>();
DT = &getAnalysis<DominatorTreeWrapperPass>().getDomTree();
BFI = &getAnalysis<BlockFrequencyInfo>();
- TLI = getAnalysisIfAvailable<TargetLibraryInfo>();
+ auto *TLIP = getAnalysisIfAvailable<TargetLibraryInfoWrapperPass>();
+ TLI = TLIP ? &TLIP->getTLI() : nullptr;
AA = &getAnalysis<AliasAnalysis>();
- AT = &getAnalysis<AssumptionTracker>();
+ AC = &getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F);
// Compute some weights outside of the loop over the loops. Compute this
// using a BranchProbability to re-use its scaling math.
}
// Check if it is legal to vectorize the loop.
- LoopVectorizationLegality LVL(L, SE, DL, DT, TLI, AA, F);
+ LoopVectorizationLegality LVL(L, SE, DL, DT, TLI, AA, F, TTI);
if (!LVL.canVectorize()) {
DEBUG(dbgs() << "LV: Not vectorizing: Cannot prove legality.\n");
emitMissedWarning(F, L, Hints);
}
// Use the cost model.
- LoopVectorizationCostModel CM(L, SE, LI, &LVL, *TTI, DL, TLI, AT, F,
+ LoopVectorizationCostModel CM(L, SE, LI, &LVL, *TTI, DL, TLI, AC, F,
&Hints);
// Check the function attributes to find out if this function should be
}
void getAnalysisUsage(AnalysisUsage &AU) const override {
- AU.addRequired<AssumptionTracker>();
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequiredID(LoopSimplifyID);
AU.addRequiredID(LCSSAID);
AU.addRequired<BlockFrequencyInfo>();
AU.addRequired<DominatorTreeWrapperPass>();
- AU.addRequired<LoopInfo>();
+ AU.addRequired<LoopInfoWrapperPass>();
AU.addRequired<ScalarEvolution>();
AU.addRequired<TargetTransformInfo>();
AU.addRequired<AliasAnalysis>();
- AU.addPreserved<LoopInfo>();
+ AU.addPreserved<LoopInfoWrapperPass>();
AU.addPreserved<DominatorTreeWrapperPass>();
AU.addPreserved<AliasAnalysis>();
}
unsigned ScalarAllocatedSize = DL->getTypeAllocSize(ScalarDataTy);
unsigned VectorElementSize = DL->getTypeStoreSize(DataTy)/VF;
- if (SI && Legal->blockNeedsPredication(SI->getParent()))
+ if (SI && Legal->blockNeedsPredication(SI->getParent()) &&
+ !Legal->isMaskRequired(SI))
return scalarizeInstruction(Instr, true);
if (ScalarAllocatedSize != VectorElementSize)
Ptr = Builder.CreateExtractElement(PtrVal[0], Zero);
}
+ VectorParts Mask = createBlockInMask(Instr->getParent());
// Handle Stores:
if (SI) {
assert(!Legal->isUniform(SI->getPointerOperand()) &&
// We don't want to update the value in the map as it might be used in
// another expression. So don't use a reference type for "StoredVal".
VectorParts StoredVal = getVectorValue(SI->getValueOperand());
-
+
for (unsigned Part = 0; Part < UF; ++Part) {
// Calculate the pointer for the specific unroll-part.
Value *PartPtr = Builder.CreateGEP(Ptr, Builder.getInt32(Part * VF));
Value *VecPtr = Builder.CreateBitCast(PartPtr,
DataTy->getPointerTo(AddressSpace));
- StoreInst *NewSI =
- Builder.CreateAlignedStore(StoredVal[Part], VecPtr, Alignment);
+
+ Instruction *NewSI;
+ if (Legal->isMaskRequired(SI))
+ NewSI = Builder.CreateMaskedStore(StoredVal[Part], VecPtr, Alignment,
+ Mask[Part]);
+ else
+ NewSI = Builder.CreateAlignedStore(StoredVal[Part], VecPtr, Alignment);
propagateMetadata(NewSI, SI);
}
return;
if (Reverse) {
// If the address is consecutive but reversed, then the
- // wide store needs to start at the last vector element.
+ // wide load needs to start at the last vector element.
PartPtr = Builder.CreateGEP(Ptr, Builder.getInt32(-Part * VF));
PartPtr = Builder.CreateGEP(PartPtr, Builder.getInt32(1 - VF));
}
+ Instruction* NewLI;
Value *VecPtr = Builder.CreateBitCast(PartPtr,
DataTy->getPointerTo(AddressSpace));
- LoadInst *NewLI = Builder.CreateAlignedLoad(VecPtr, Alignment, "wide.load");
+ if (Legal->isMaskRequired(LI))
+ NewLI = Builder.CreateMaskedLoad(VecPtr, Alignment, Mask[Part],
+ UndefValue::get(DataTy),
+ "wide.masked.load");
+ else
+ NewLI = Builder.CreateAlignedLoad(VecPtr, Alignment, "wide.load");
propagateMetadata(NewLI, LI);
Entry[Part] = Reverse ? reverseVector(NewLI) : NewLI;
}
Cmp = Builder.CreateICmp(ICmpInst::ICMP_EQ, Cmp, ConstantInt::get(Cmp->getType(), 1));
CondBlock = IfBlock->splitBasicBlock(InsertPt, "cond.store");
LoopVectorBody.push_back(CondBlock);
- VectorLp->addBasicBlockToLoop(CondBlock, LI->getBase());
+ VectorLp->addBasicBlockToLoop(CondBlock, *LI);
// Update Builder with newly created basic block.
Builder.SetInsertPoint(InsertPt);
}
if (IfPredicateStore) {
BasicBlock *NewIfBlock = CondBlock->splitBasicBlock(InsertPt, "else");
LoopVectorBody.push_back(NewIfBlock);
- VectorLp->addBasicBlockToLoop(NewIfBlock, LI->getBase());
+ VectorLp->addBasicBlockToLoop(NewIfBlock, *LI);
Builder.SetInsertPoint(InsertPt);
Instruction *OldBr = IfBlock->getTerminator();
BranchInst::Create(CondBlock, NewIfBlock, Cmp, OldBr);
// before calling any utilities such as SCEV that require valid LoopInfo.
if (ParentLoop) {
ParentLoop->addChildLoop(Lp);
- ParentLoop->addBasicBlockToLoop(ScalarPH, LI->getBase());
- ParentLoop->addBasicBlockToLoop(VectorPH, LI->getBase());
- ParentLoop->addBasicBlockToLoop(MiddleBlock, LI->getBase());
+ ParentLoop->addBasicBlockToLoop(ScalarPH, *LI);
+ ParentLoop->addBasicBlockToLoop(VectorPH, *LI);
+ ParentLoop->addBasicBlockToLoop(MiddleBlock, *LI);
} else {
LI->addTopLevelLoop(Lp);
}
- Lp->addBasicBlockToLoop(VecBody, LI->getBase());
+ Lp->addBasicBlockToLoop(VecBody, *LI);
// Use this IR builder to create the loop instructions (Phi, Br, Cmp)
// inside the loop.
BasicBlock *CheckBlock =
LastBypassBlock->splitBasicBlock(PastOverflowCheck, "overflow.checked");
if (ParentLoop)
- ParentLoop->addBasicBlockToLoop(CheckBlock, LI->getBase());
+ ParentLoop->addBasicBlockToLoop(CheckBlock, *LI);
LoopBypassBlocks.push_back(CheckBlock);
Instruction *OldTerm = LastBypassBlock->getTerminator();
BranchInst::Create(ScalarPH, CheckBlock, CheckBCOverflow, OldTerm);
BasicBlock *CheckBlock =
LastBypassBlock->splitBasicBlock(FirstCheckInst, "vector.stridecheck");
if (ParentLoop)
- ParentLoop->addBasicBlockToLoop(CheckBlock, LI->getBase());
+ ParentLoop->addBasicBlockToLoop(CheckBlock, *LI);
LoopBypassBlocks.push_back(CheckBlock);
// Replace the branch into the memory check block with a conditional branch
BasicBlock *CheckBlock =
LastBypassBlock->splitBasicBlock(MemRuntimeCheck, "vector.memcheck");
if (ParentLoop)
- ParentLoop->addBasicBlockToLoop(CheckBlock, LI->getBase());
+ ParentLoop->addBasicBlockToLoop(CheckBlock, *LI);
LoopBypassBlocks.push_back(CheckBlock);
// Replace the branch into the memory check block with a conditional branch
}
// Fix the vector-loop phi.
- // We created the induction variable so we know that the
- // preheader is the first entry.
- BasicBlock *VecPreheader = Induction->getIncomingBlock(0);
// Reductions do not have to start at zero. They can start with
// any loop invariant values.
// Make sure to add the reduction stat value only to the
// first unroll part.
Value *StartVal = (part == 0) ? VectorStart : Identity;
- cast<PHINode>(VecRdxPhi[part])->addIncoming(StartVal, VecPreheader);
+ cast<PHINode>(VecRdxPhi[part])->addIncoming(StartVal,
+ LoopVectorPreHeader);
cast<PHINode>(VecRdxPhi[part])->addIncoming(Val[part],
LoopVectorBody.back());
}
}
// We can only vectorize innermost loops.
- if (TheLoop->getSubLoopsVector().size()) {
+ if (!TheLoop->getSubLoopsVector().empty()) {
emitAnalysis(Report() << "loop is not the innermost loop");
return false;
}
return false;
}
+ // We only handle bottom-tested loops, i.e. loop in which the condition is
+ // checked at the end of each iteration. With that we can assume that all
+ // instructions in the loop are executed the same number of times.
+ if (TheLoop->getExitingBlock() != TheLoop->getLoopLatch()) {
+ emitAnalysis(
+ Report() << "loop control flow is not understood by vectorizer");
+ return false;
+ }
+
// We need to have a loop header.
DEBUG(dbgs() << "LV: Found a loop: " <<
TheLoop->getHeader()->getName() << '\n');
return false;
}
- // We only allow if-converted PHIs with more than two incoming values.
+ // We only allow if-converted PHIs with exactly two incoming values.
if (Phi->getNumIncomingValues() != 2) {
emitAnalysis(Report(it)
<< "control flow not understood by vectorizer");
return false;
}
if (EnableMemAccessVersioning)
- collectStridedAcccess(ST);
+ collectStridedAccess(ST);
}
if (EnableMemAccessVersioning)
if (LoadInst *LI = dyn_cast<LoadInst>(it))
- collectStridedAcccess(LI);
+ collectStridedAccess(LI);
// Reduction instructions are allowed to have exit users.
// All other instructions must not have external users.
return Stride;
}
-void LoopVectorizationLegality::collectStridedAcccess(Value *MemAccess) {
+void LoopVectorizationLegality::collectStridedAccess(Value *MemAccess) {
Value *Ptr = nullptr;
if (LoadInst *LI = dyn_cast<LoadInst>(MemAccess))
Ptr = LI->getPointerOperand();
if (I->getType()->isPointerTy() && isConsecutivePtr(I))
Worklist.insert(Worklist.end(), I->op_begin(), I->op_end());
- while (Worklist.size()) {
+ while (!Worklist.empty()) {
Instruction *I = dyn_cast<Instruction>(Worklist.back());
Worklist.pop_back();
bool UseDeferred = SetIteration > 0;
PtrAccessSet &S = UseDeferred ? DeferredAccesses : Accesses;
- for (auto A : AS) {
- Value *Ptr = A.getValue();
- bool IsWrite = S.count(MemAccessInfo(Ptr, true));
+ for (auto AV : AS) {
+ Value *Ptr = AV.getValue();
- // If we're using the deferred access set, then it contains only reads.
- bool IsReadOnlyPtr = ReadOnlyPtr.count(Ptr) && !IsWrite;
- if (UseDeferred && !IsReadOnlyPtr)
- continue;
- // Otherwise, the pointer must be in the PtrAccessSet, either as a read
- // or a write.
- assert(((IsReadOnlyPtr && UseDeferred) || IsWrite ||
- S.count(MemAccessInfo(Ptr, false))) &&
- "Alias-set pointer not in the access set?");
-
- MemAccessInfo Access(Ptr, IsWrite);
- DepCands.insert(Access);
-
- // Memorize read-only pointers for later processing and skip them in the
- // first round (they need to be checked after we have seen all write
- // pointers). Note: we also mark pointer that are not consecutive as
- // "read-only" pointers (so that we check "a[b[i]] +="). Hence, we need
- // the second check for "!IsWrite".
- if (!UseDeferred && IsReadOnlyPtr) {
- DeferredAccesses.insert(Access);
- continue;
- }
+ // For a single memory access in AliasSetTracker, Accesses may contain
+ // both read and write, and they both need to be handled for CheckDeps.
+ for (auto AC : S) {
+ if (AC.getPointer() != Ptr)
+ continue;
- // If this is a write - check other reads and writes for conflicts. If
- // this is a read only check other writes for conflicts (but only if
- // there is no other write to the ptr - this is an optimization to
- // catch "a[i] = a[i] + " without having to do a dependence check).
- if ((IsWrite || IsReadOnlyPtr) && SetHasWrite) {
- CheckDeps.insert(Access);
- IsRTCheckNeeded = true;
- }
+ bool IsWrite = AC.getInt();
- if (IsWrite)
- SetHasWrite = true;
-
- // Create sets of pointers connected by a shared alias set and
- // underlying object.
- typedef SmallVector<Value *, 16> ValueVector;
- ValueVector TempObjects;
- GetUnderlyingObjects(Ptr, TempObjects, DL);
- for (Value *UnderlyingObj : TempObjects) {
- UnderlyingObjToAccessMap::iterator Prev =
- ObjToLastAccess.find(UnderlyingObj);
- if (Prev != ObjToLastAccess.end())
- DepCands.unionSets(Access, Prev->second);
-
- ObjToLastAccess[UnderlyingObj] = Access;
+ // If we're using the deferred access set, then it contains only
+ // reads.
+ bool IsReadOnlyPtr = ReadOnlyPtr.count(Ptr) && !IsWrite;
+ if (UseDeferred && !IsReadOnlyPtr)
+ continue;
+ // Otherwise, the pointer must be in the PtrAccessSet, either as a
+ // read or a write.
+ assert(((IsReadOnlyPtr && UseDeferred) || IsWrite ||
+ S.count(MemAccessInfo(Ptr, false))) &&
+ "Alias-set pointer not in the access set?");
+
+ MemAccessInfo Access(Ptr, IsWrite);
+ DepCands.insert(Access);
+
+ // Memorize read-only pointers for later processing and skip them in
+ // the first round (they need to be checked after we have seen all
+ // write pointers). Note: we also mark pointer that are not
+ // consecutive as "read-only" pointers (so that we check
+ // "a[b[i]] +="). Hence, we need the second check for "!IsWrite".
+ if (!UseDeferred && IsReadOnlyPtr) {
+ DeferredAccesses.insert(Access);
+ continue;
+ }
+
+ // If this is a write - check other reads and writes for conflicts. If
+ // this is a read only check other writes for conflicts (but only if
+ // there is no other write to the ptr - this is an optimization to
+ // catch "a[i] = a[i] + " without having to do a dependence check).
+ if ((IsWrite || IsReadOnlyPtr) && SetHasWrite) {
+ CheckDeps.insert(Access);
+ IsRTCheckNeeded = true;
+ }
+
+ if (IsWrite)
+ SetHasWrite = true;
+
+ // Create sets of pointers connected by a shared alias set and
+ // underlying object.
+ typedef SmallVector<Value *, 16> ValueVector;
+ ValueVector TempObjects;
+ GetUnderlyingObjects(Ptr, TempObjects, DL);
+ for (Value *UnderlyingObj : TempObjects) {
+ UnderlyingObjToAccessMap::iterator Prev =
+ ObjToLastAccess.find(UnderlyingObj);
+ if (Prev != ObjToLastAccess.end())
+ DepCands.unionSets(Access, Prev->second);
+
+ ObjToLastAccess[UnderlyingObj] = Access;
+ }
}
}
}
// If we did *not* see this pointer before, insert it to the read-write
// list. At this phase it is only a 'write' list.
- if (Seen.insert(Ptr)) {
+ if (Seen.insert(Ptr).second) {
++NumReadWrites;
AliasAnalysis::Location Loc = AA->getLocation(ST);
// read a few words, modify, and write a few words, and some of the
// words may be written to the same address.
bool IsReadOnlyPtr = false;
- if (Seen.insert(Ptr) || !isStridedPtr(SE, DL, Ptr, TheLoop, Strides)) {
+ if (Seen.insert(Ptr).second ||
+ !isStridedPtr(SE, DL, Ptr, TheLoop, Strides)) {
++NumReads;
IsReadOnlyPtr = true;
}
// value must only be used once, except by phi nodes and min/max
// reductions which are represented as a cmp followed by a select.
ReductionInstDesc IgnoredVal(false, nullptr);
- if (VisitedInsts.insert(UI)) {
+ if (VisitedInsts.insert(UI).second) {
if (isa<PHINode>(UI))
PHIs.push_back(UI);
else
return IK_NoInduction;
assert(PhiTy->isPointerTy() && "The PHI must be a pointer");
- uint64_t Size = DL->getTypeAllocSize(PhiTy->getPointerElementType());
+ Type *PointerElementType = PhiTy->getPointerElementType();
+ // The pointer stride cannot be determined if the pointer element type is not
+ // sized.
+ if (!PointerElementType->isSized())
+ return IK_NoInduction;
+
+ uint64_t Size = DL->getTypeAllocSize(PointerElementType);
if (C->getValue()->equalsInt(Size))
return IK_PtrInduction;
else if (C->getValue()->equalsInt(0 - Size))
bool LoopVectorizationLegality::blockCanBePredicated(BasicBlock *BB,
SmallPtrSetImpl<Value *> &SafePtrs) {
+
for (BasicBlock::iterator it = BB->begin(), e = BB->end(); it != e; ++it) {
+ // Check that we don't have a constant expression that can trap as operand.
+ for (Instruction::op_iterator OI = it->op_begin(), OE = it->op_end();
+ OI != OE; ++OI) {
+ if (Constant *C = dyn_cast<Constant>(*OI))
+ if (C->canTrap())
+ return false;
+ }
// We might be able to hoist the load.
if (it->mayReadFromMemory()) {
LoadInst *LI = dyn_cast<LoadInst>(it);
- if (!LI || !SafePtrs.count(LI->getPointerOperand()))
+ if (!LI)
return false;
+ if (!SafePtrs.count(LI->getPointerOperand())) {
+ if (isLegalMaskedLoad(LI->getType(), LI->getPointerOperand())) {
+ MaskedOp.insert(LI);
+ continue;
+ }
+ return false;
+ }
}
// We don't predicate stores at the moment.
StoreInst *SI = dyn_cast<StoreInst>(it);
// We only support predication of stores in basic blocks with one
// predecessor.
- if (!SI || ++NumPredStores > NumberOfStoresToPredicate ||
- !SafePtrs.count(SI->getPointerOperand()) ||
- !SI->getParent()->getSinglePredecessor())
+ if (!SI)
return false;
+
+ bool isSafePtr = (SafePtrs.count(SI->getPointerOperand()) != 0);
+ bool isSinglePredecessor = SI->getParent()->getSinglePredecessor();
+
+ if (++NumPredStores > NumberOfStoresToPredicate || !isSafePtr ||
+ !isSinglePredecessor) {
+ // Build a masked store if it is legal for the target, otherwise scalarize
+ // the block.
+ bool isLegalMaskedOp =
+ isLegalMaskedStore(SI->getValueOperand()->getType(),
+ SI->getPointerOperand());
+ if (isLegalMaskedOp) {
+ --NumPredStores;
+ MaskedOp.insert(SI);
+ continue;
+ }
+ return false;
+ }
}
if (it->mayThrow())
return false;
- // Check that we don't have a constant expression that can trap as operand.
- for (Instruction::op_iterator OI = it->op_begin(), OE = it->op_end();
- OI != OE; ++OI) {
- if (Constant *C = dyn_cast<Constant>(*OI))
- if (C->canTrap())
- return false;
- }
-
// The instructions below can trap.
switch (it->getOpcode()) {
default: continue;
case Instruction::SDiv:
case Instruction::URem:
case Instruction::SRem:
- return false;
+ return false;
}
}
MaxVectorSize = 1;
}
- assert(MaxVectorSize <= 32 && "Did not expect to pack so many elements"
+ assert(MaxVectorSize <= 64 && "Did not expect to pack so many elements"
" into one vector!");
unsigned VF = MaxVectorSize;
INITIALIZE_PASS_BEGIN(LoopVectorize, LV_NAME, lv_name, false, false)
INITIALIZE_AG_DEPENDENCY(TargetTransformInfo)
INITIALIZE_AG_DEPENDENCY(AliasAnalysis)
-INITIALIZE_PASS_DEPENDENCY(AssumptionTracker)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(BlockFrequencyInfo)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
INITIALIZE_PASS_DEPENDENCY(ScalarEvolution)
INITIALIZE_PASS_DEPENDENCY(LCSSA)
-INITIALIZE_PASS_DEPENDENCY(LoopInfo)
+INITIALIZE_PASS_DEPENDENCY(LoopInfoWrapperPass)
INITIALIZE_PASS_DEPENDENCY(LoopSimplify)
INITIALIZE_PASS_END(LoopVectorize, LV_NAME, lv_name, false, false)
ConstantInt::get(Cond[Part]->getType(), 1));
CondBlock = IfBlock->splitBasicBlock(InsertPt, "cond.store");
LoopVectorBody.push_back(CondBlock);
- VectorLp->addBasicBlockToLoop(CondBlock, LI->getBase());
+ VectorLp->addBasicBlockToLoop(CondBlock, *LI);
// Update Builder with newly created basic block.
Builder.SetInsertPoint(InsertPt);
}
if (IfPredicateStore) {
BasicBlock *NewIfBlock = CondBlock->splitBasicBlock(InsertPt, "else");
LoopVectorBody.push_back(NewIfBlock);
- VectorLp->addBasicBlockToLoop(NewIfBlock, LI->getBase());
+ VectorLp->addBasicBlockToLoop(NewIfBlock, *LI);
Builder.SetInsertPoint(InsertPt);
Instruction *OldBr = IfBlock->getTerminator();
BranchInst::Create(CondBlock, NewIfBlock, Cmp, OldBr);
projects / oota-llvm.git / blobdiff