// Variable uniformity checks are inspired by:
// Karrenberg, R. and Hack, S. Whole Function Vectorization.
//
+// The interleaved access vectorization is based on the paper:
+// Dorit Nuzman, Ira Rosen and Ayal Zaks. Auto-Vectorization of Interleaved
+// Data for SIMD
+//
// Other ideas/concepts are from:
// A. Zaks and D. Nuzman. Autovectorization in GCC-two years later.
//
//
//===----------------------------------------------------------------------===//
-#define LV_NAME "loop-vectorize"
-#define DEBUG_TYPE LV_NAME
-
#include "llvm/Transforms/Vectorize.h"
#include "llvm/ADT/DenseMap.h"
-#include "llvm/ADT/EquivalenceClasses.h"
+#include "llvm/ADT/Hashing.h"
#include "llvm/ADT/MapVector.h"
#include "llvm/ADT/SetVector.h"
#include "llvm/ADT/SmallPtrSet.h"
#include "llvm/ADT/SmallSet.h"
#include "llvm/ADT/SmallVector.h"
+#include "llvm/ADT/Statistic.h"
#include "llvm/ADT/StringExtras.h"
#include "llvm/Analysis/AliasAnalysis.h"
-#include "llvm/Analysis/Dominators.h"
+#include "llvm/Analysis/BasicAliasAnalysis.h"
+#include "llvm/Analysis/AliasSetTracker.h"
+#include "llvm/Analysis/AssumptionCache.h"
+#include "llvm/Analysis/BlockFrequencyInfo.h"
+#include "llvm/Analysis/CodeMetrics.h"
+#include "llvm/Analysis/DemandedBits.h"
+#include "llvm/Analysis/GlobalsModRef.h"
+#include "llvm/Analysis/LoopAccessAnalysis.h"
#include "llvm/Analysis/LoopInfo.h"
#include "llvm/Analysis/LoopIterator.h"
#include "llvm/Analysis/LoopPass.h"
#include "llvm/Analysis/ScalarEvolutionExpressions.h"
#include "llvm/Analysis/TargetTransformInfo.h"
#include "llvm/Analysis/ValueTracking.h"
-#include "llvm/Analysis/Verifier.h"
#include "llvm/IR/Constants.h"
#include "llvm/IR/DataLayout.h"
+#include "llvm/IR/DebugInfo.h"
#include "llvm/IR/DerivedTypes.h"
+#include "llvm/IR/DiagnosticInfo.h"
+#include "llvm/IR/Dominators.h"
#include "llvm/IR/Function.h"
#include "llvm/IR/IRBuilder.h"
#include "llvm/IR/Instructions.h"
#include "llvm/IR/IntrinsicInst.h"
#include "llvm/IR/LLVMContext.h"
#include "llvm/IR/Module.h"
+#include "llvm/IR/PatternMatch.h"
#include "llvm/IR/Type.h"
#include "llvm/IR/Value.h"
+#include "llvm/IR/ValueHandle.h"
+#include "llvm/IR/Verifier.h"
#include "llvm/Pass.h"
+#include "llvm/Support/BranchProbability.h"
#include "llvm/Support/CommandLine.h"
#include "llvm/Support/Debug.h"
-#include "llvm/Support/PatternMatch.h"
#include "llvm/Support/raw_ostream.h"
-#include "llvm/Support/ValueHandle.h"
-#include "llvm/Target/TargetLibraryInfo.h"
#include "llvm/Transforms/Scalar.h"
#include "llvm/Transforms/Utils/BasicBlockUtils.h"
#include "llvm/Transforms/Utils/Local.h"
+#include "llvm/Analysis/VectorUtils.h"
+#include "llvm/Transforms/Utils/LoopUtils.h"
#include <algorithm>
+#include <functional>
#include <map>
+#include <tuple>
using namespace llvm;
using namespace llvm::PatternMatch;
-static cl::opt<unsigned>
-VectorizationFactor("force-vector-width", cl::init(0), cl::Hidden,
- cl::desc("Sets the SIMD width. Zero is autoselect."));
+#define LV_NAME "loop-vectorize"
+#define DEBUG_TYPE LV_NAME
-static cl::opt<unsigned>
-VectorizationUnroll("force-vector-unroll", cl::init(0), cl::Hidden,
- cl::desc("Sets the vectorization unroll count. "
- "Zero is autoselect."));
+STATISTIC(LoopsVectorized, "Number of loops vectorized");
+STATISTIC(LoopsAnalyzed, "Number of loops analyzed for vectorization");
static cl::opt<bool>
EnableIfConversion("enable-if-conversion", cl::init(true), cl::Hidden,
"trip count that is smaller than this "
"value."));
-/// We don't unroll loops with a known constant trip count below this number.
-static const unsigned TinyTripCountUnrollThreshold = 128;
+static cl::opt<bool> MaximizeBandwidth(
+ "vectorizer-maximize-bandwidth", cl::init(false), cl::Hidden,
+ cl::desc("Maximize bandwidth when selecting vectorization factor which "
+ "will be determined by the smallest type in loop."));
-/// When performing memory disambiguation checks at runtime do not make more
-/// than this number of comparisons.
-static const unsigned RuntimeMemoryCheckThreshold = 8;
-
-/// Maximum simd width.
-static const unsigned MaxVectorWidth = 64;
-
-/// Maximum vectorization unroll count.
-static const unsigned MaxUnrollFactor = 16;
-
-/// The cost of a loop that is considered 'small' by the unroller.
-static const unsigned SmallLoopCost = 20;
+/// This enables versioning on the strides of symbolically striding memory
+/// accesses in code like the following.
+/// for (i = 0; i < N; ++i)
+/// A[i * Stride1] += B[i * Stride2] ...
+///
+/// Will be roughly translated to
+/// if (Stride1 == 1 && Stride2 == 1) {
+/// for (i = 0; i < N; i+=4)
+/// A[i:i+3] += ...
+/// } else
+/// ...
+static cl::opt<bool> EnableMemAccessVersioning(
+ "enable-mem-access-versioning", cl::init(true), cl::Hidden,
+ cl::desc("Enable symbolic stride memory access versioning"));
+
+static cl::opt<bool> EnableInterleavedMemAccesses(
+ "enable-interleaved-mem-accesses", cl::init(false), cl::Hidden,
+ cl::desc("Enable vectorization on interleaved memory accesses in a loop"));
+
+/// Maximum factor for an interleaved memory access.
+static cl::opt<unsigned> MaxInterleaveGroupFactor(
+ "max-interleave-group-factor", cl::Hidden,
+ cl::desc("Maximum factor for an interleaved access group (default = 8)"),
+ cl::init(8));
+
+/// We don't interleave loops with a known constant trip count below this
+/// number.
+static const unsigned TinyTripCountInterleaveThreshold = 128;
+
+static cl::opt<unsigned> ForceTargetNumScalarRegs(
+ "force-target-num-scalar-regs", cl::init(0), cl::Hidden,
+ cl::desc("A flag that overrides the target's number of scalar registers."));
+
+static cl::opt<unsigned> ForceTargetNumVectorRegs(
+ "force-target-num-vector-regs", cl::init(0), cl::Hidden,
+ cl::desc("A flag that overrides the target's number of vector registers."));
+
+/// Maximum vectorization interleave count.
+static const unsigned MaxInterleaveFactor = 16;
+
+static cl::opt<unsigned> ForceTargetMaxScalarInterleaveFactor(
+ "force-target-max-scalar-interleave", cl::init(0), cl::Hidden,
+ cl::desc("A flag that overrides the target's max interleave factor for "
+ "scalar loops."));
+
+static cl::opt<unsigned> ForceTargetMaxVectorInterleaveFactor(
+ "force-target-max-vector-interleave", cl::init(0), cl::Hidden,
+ cl::desc("A flag that overrides the target's max interleave factor for "
+ "vectorized loops."));
+
+static cl::opt<unsigned> ForceTargetInstructionCost(
+ "force-target-instruction-cost", cl::init(0), cl::Hidden,
+ cl::desc("A flag that overrides the target's expected cost for "
+ "an instruction to a single constant value. Mostly "
+ "useful for getting consistent testing."));
+
+static cl::opt<unsigned> SmallLoopCost(
+ "small-loop-cost", cl::init(20), cl::Hidden,
+ cl::desc(
+ "The cost of a loop that is considered 'small' by the interleaver."));
+
+static cl::opt<bool> LoopVectorizeWithBlockFrequency(
+ "loop-vectorize-with-block-frequency", cl::init(false), cl::Hidden,
+ cl::desc("Enable the use of the block frequency analysis to access PGO "
+ "heuristics minimizing code growth in cold regions and being more "
+ "aggressive in hot regions."));
+
+// Runtime interleave loops for load/store throughput.
+static cl::opt<bool> EnableLoadStoreRuntimeInterleave(
+ "enable-loadstore-runtime-interleave", cl::init(true), cl::Hidden,
+ cl::desc(
+ "Enable runtime interleaving until load/store ports are saturated"));
+
+/// The number of stores in a loop that are allowed to need predication.
+static cl::opt<unsigned> NumberOfStoresToPredicate(
+ "vectorize-num-stores-pred", cl::init(1), cl::Hidden,
+ cl::desc("Max number of stores to be predicated behind an if."));
+
+static cl::opt<bool> EnableIndVarRegisterHeur(
+ "enable-ind-var-reg-heur", cl::init(true), cl::Hidden,
+ cl::desc("Count the induction variable only once when interleaving"));
+
+static cl::opt<bool> EnableCondStoresVectorization(
+ "enable-cond-stores-vec", cl::init(false), cl::Hidden,
+ cl::desc("Enable if predication of stores during vectorization."));
+
+static cl::opt<unsigned> MaxNestedScalarReductionIC(
+ "max-nested-scalar-reduction-interleave", cl::init(2), cl::Hidden,
+ cl::desc("The maximum interleave count to use when interleaving a scalar "
+ "reduction in a nested loop."));
+
+static cl::opt<unsigned> PragmaVectorizeMemoryCheckThreshold(
+ "pragma-vectorize-memory-check-threshold", cl::init(128), cl::Hidden,
+ cl::desc("The maximum allowed number of runtime memory checks with a "
+ "vectorize(enable) pragma."));
+
+static cl::opt<unsigned> VectorizeSCEVCheckThreshold(
+ "vectorize-scev-check-threshold", cl::init(16), cl::Hidden,
+ cl::desc("The maximum number of SCEV checks allowed."));
+
+static cl::opt<unsigned> PragmaVectorizeSCEVCheckThreshold(
+ "pragma-vectorize-scev-check-threshold", cl::init(128), cl::Hidden,
+ cl::desc("The maximum number of SCEV checks allowed with a "
+ "vectorize(enable) pragma"));
namespace {
// Forward declarations.
+class LoopVectorizeHints;
class LoopVectorizationLegality;
class LoopVectorizationCostModel;
+class LoopVectorizationRequirements;
+
+/// \brief This modifies LoopAccessReport to initialize message with
+/// loop-vectorizer-specific part.
+class VectorizationReport : public LoopAccessReport {
+public:
+ VectorizationReport(Instruction *I = nullptr)
+ : LoopAccessReport("loop not vectorized: ", I) {}
+
+ /// \brief This allows promotion of the loop-access analysis report into the
+ /// loop-vectorizer report. It modifies the message to add the
+ /// loop-vectorizer-specific part of the message.
+ explicit VectorizationReport(const LoopAccessReport &R)
+ : LoopAccessReport(Twine("loop not vectorized: ") + R.str(),
+ R.getInstr()) {}
+};
+
+/// A helper function for converting Scalar types to vector types.
+/// If the incoming type is void, we return void. If the VF is 1, we return
+/// the scalar type.
+static Type* ToVectorTy(Type *Scalar, unsigned VF) {
+ if (Scalar->isVoidTy() || VF == 1)
+ return Scalar;
+ return VectorType::get(Scalar, VF);
+}
+
+/// A helper function that returns GEP instruction and knows to skip a
+/// 'bitcast'. The 'bitcast' may be skipped if the source and the destination
+/// pointee types of the 'bitcast' have the same size.
+/// For example:
+/// bitcast double** %var to i64* - can be skipped
+/// bitcast double** %var to i8* - can not
+static GetElementPtrInst *getGEPInstruction(Value *Ptr) {
+
+ if (isa<GetElementPtrInst>(Ptr))
+ return cast<GetElementPtrInst>(Ptr);
+
+ if (isa<BitCastInst>(Ptr) &&
+ isa<GetElementPtrInst>(cast<BitCastInst>(Ptr)->getOperand(0))) {
+ Type *BitcastTy = Ptr->getType();
+ Type *GEPTy = cast<BitCastInst>(Ptr)->getSrcTy();
+ if (!isa<PointerType>(BitcastTy) || !isa<PointerType>(GEPTy))
+ return nullptr;
+ Type *Pointee1Ty = cast<PointerType>(BitcastTy)->getPointerElementType();
+ Type *Pointee2Ty = cast<PointerType>(GEPTy)->getPointerElementType();
+ const DataLayout &DL = cast<BitCastInst>(Ptr)->getModule()->getDataLayout();
+ if (DL.getTypeSizeInBits(Pointee1Ty) == DL.getTypeSizeInBits(Pointee2Ty))
+ return cast<GetElementPtrInst>(cast<BitCastInst>(Ptr)->getOperand(0));
+ }
+ return nullptr;
+}
/// InnerLoopVectorizer vectorizes loops which contain only one basic
/// block to a specified vectorization factor (VF).
/// and reduction variables that were found to a given vectorization factor.
class InnerLoopVectorizer {
public:
- InnerLoopVectorizer(Loop *OrigLoop, ScalarEvolution *SE, LoopInfo *LI,
- DominatorTree *DT, DataLayout *DL,
- const TargetLibraryInfo *TLI, unsigned VecWidth,
+ InnerLoopVectorizer(Loop *OrigLoop, PredicatedScalarEvolution &PSE,
+ LoopInfo *LI, DominatorTree *DT,
+ const TargetLibraryInfo *TLI,
+ const TargetTransformInfo *TTI, unsigned VecWidth,
unsigned UnrollFactor)
- : OrigLoop(OrigLoop), SE(SE), LI(LI), DT(DT), DL(DL), TLI(TLI),
- VF(VecWidth), UF(UnrollFactor), Builder(SE->getContext()), Induction(0),
- OldInduction(0), WidenMap(UnrollFactor) {}
+ : OrigLoop(OrigLoop), PSE(PSE), LI(LI), DT(DT), TLI(TLI), TTI(TTI),
+ VF(VecWidth), UF(UnrollFactor), Builder(PSE.getSE()->getContext()),
+ Induction(nullptr), OldInduction(nullptr), WidenMap(UnrollFactor),
+ TripCount(nullptr), VectorTripCount(nullptr), Legal(nullptr),
+ AddedSafetyChecks(false) {}
// Perform the actual loop widening (vectorization).
- void vectorize(LoopVectorizationLegality *Legal) {
+ // MinimumBitWidths maps scalar integer values to the smallest bitwidth they
+ // can be validly truncated to. The cost model has assumed this truncation
+ // will happen when vectorizing.
+ void vectorize(LoopVectorizationLegality *L,
+ MapVector<Instruction*,uint64_t> MinimumBitWidths) {
+ MinBWs = MinimumBitWidths;
+ Legal = L;
// Create a new empty loop. Unlink the old loop and connect the new one.
- createEmptyLoop(Legal);
+ createEmptyLoop();
// Widen each instruction in the old loop to a new one in the new loop.
// Use the Legality module to find the induction and reduction variables.
- vectorizeLoop(Legal);
- // Register the new loop and update the analysis passes.
- updateAnalysis();
+ vectorizeLoop();
+ }
+
+ // Return true if any runtime check is added.
+ bool IsSafetyChecksAdded() {
+ return AddedSafetyChecks;
}
virtual ~InnerLoopVectorizer() {}
/// originated from one scalar instruction.
typedef SmallVector<Value*, 2> VectorParts;
- // When we if-convert we need create edge masks. We have to cache values so
- // that we don't end up with exponential recursion/IR.
+ // When we if-convert we need to create edge masks. We have to cache values
+ // so that we don't end up with exponential recursion/IR.
typedef DenseMap<std::pair<BasicBlock*, BasicBlock*>,
VectorParts> EdgeMaskCache;
- /// Add code that checks at runtime if the accessed arrays overlap.
- /// Returns the comparator value or NULL if no check is needed.
- Instruction *addRuntimeCheck(LoopVectorizationLegality *Legal,
- Instruction *Loc);
/// Create an empty loop, based on the loop ranges of the old loop.
- void createEmptyLoop(LoopVectorizationLegality *Legal);
+ void createEmptyLoop();
+ /// Create a new induction variable inside L.
+ PHINode *createInductionVariable(Loop *L, Value *Start, Value *End,
+ Value *Step, Instruction *DL);
/// Copy and widen the instructions from the old loop.
- virtual void vectorizeLoop(LoopVectorizationLegality *Legal);
+ virtual void vectorizeLoop();
/// \brief The Loop exit block may have single value PHI nodes where the
/// incoming value is 'Undef'. While vectorizing we only handled real values
/// See PR14725.
void fixLCSSAPHIs();
+ /// Shrinks vector element sizes based on information in "MinBWs".
+ void truncateToMinimalBitwidths();
+
/// A helper function that computes the predicate of the block BB, assuming
/// that the header block of the loop is set to True. It returns the *entry*
/// mask for the block BB.
VectorParts createEdgeMask(BasicBlock *Src, BasicBlock *Dst);
/// A helper function to vectorize a single BB within the innermost loop.
- void vectorizeBlockInLoop(LoopVectorizationLegality *Legal, BasicBlock *BB,
- PhiVector *PV);
-
+ void vectorizeBlockInLoop(BasicBlock *BB, PhiVector *PV);
+
/// Vectorize a single PHINode in a block. This method handles the induction
/// variable canonicalization. It supports both VF = 1 for unrolled loops and
/// arbitrary length vectors.
void widenPHIInstruction(Instruction *PN, VectorParts &Entry,
- LoopVectorizationLegality *Legal,
unsigned UF, unsigned VF, PhiVector *PV);
/// Insert the new loop to the loop hierarchy and pass manager
void updateAnalysis();
/// This instruction is un-vectorizable. Implement it as a sequence
- /// of scalars.
- virtual void scalarizeInstruction(Instruction *Instr);
+ /// of scalars. If \p IfPredicateStore is true we need to 'hide' each
+ /// scalarized instruction behind an if block predicated on the control
+ /// dependence of the instruction.
+ virtual void scalarizeInstruction(Instruction *Instr,
+ bool IfPredicateStore=false);
/// Vectorize Load and Store instructions,
- virtual void vectorizeMemoryInstruction(Instruction *Instr,
- LoopVectorizationLegality *Legal);
+ virtual void vectorizeMemoryInstruction(Instruction *Instr);
/// Create a broadcast instruction. This method generates a broadcast
/// instruction (shuffle) for loop invariant values and for the induction
/// element.
virtual Value *getBroadcastInstrs(Value *V);
- /// This function adds 0, 1, 2 ... to each vector element, starting at zero.
- /// If Negate is set then negative numbers are added e.g. (0, -1, -2, ...).
- /// The sequence starts at StartIndex.
- virtual Value *getConsecutiveVector(Value* Val, int StartIdx, bool Negate);
+ /// This function adds (StartIdx, StartIdx + Step, StartIdx + 2*Step, ...)
+ /// to each vector element of Val. The sequence starts at StartIndex.
+ virtual Value *getStepVector(Value *Val, int StartIdx, Value *Step);
/// When we go over instructions in the basic block we rely on previous
/// values within the current basic block or on loop invariant values.
/// broadcast them into a vector.
VectorParts &getVectorValue(Value *V);
+ /// Try to vectorize the interleaved access group that \p Instr belongs to.
+ void vectorizeInterleaveGroup(Instruction *Instr);
+
/// Generate a shuffle sequence that will reverse the vector Vec.
virtual Value *reverseVector(Value *Vec);
+ /// Returns (and creates if needed) the original loop trip count.
+ Value *getOrCreateTripCount(Loop *NewLoop);
+
+ /// Returns (and creates if needed) the trip count of the widened loop.
+ Value *getOrCreateVectorTripCount(Loop *NewLoop);
+
+ /// Emit a bypass check to see if the trip count would overflow, or we
+ /// wouldn't have enough iterations to execute one vector loop.
+ void emitMinimumIterationCountCheck(Loop *L, BasicBlock *Bypass);
+ /// Emit a bypass check to see if the vector trip count is nonzero.
+ void emitVectorLoopEnteredCheck(Loop *L, BasicBlock *Bypass);
+ /// Emit a bypass check to see if all of the SCEV assumptions we've
+ /// had to make are correct.
+ void emitSCEVChecks(Loop *L, BasicBlock *Bypass);
+ /// Emit bypass checks to check any memory assumptions we may have made.
+ void emitMemRuntimeChecks(Loop *L, BasicBlock *Bypass);
+
/// This is a helper class that holds the vectorizer state. It maps scalar
/// instructions to vector instructions. When the code is 'unrolled' then
/// then a single scalar value is mapped to multiple vector parts. The parts
/// The original loop.
Loop *OrigLoop;
- /// Scev analysis to use.
- ScalarEvolution *SE;
+ /// A wrapper around ScalarEvolution used to add runtime SCEV checks. Applies
+ /// dynamic knowledge to simplify SCEV expressions and converts them to a
+ /// more usable form.
+ PredicatedScalarEvolution &PSE;
/// Loop Info.
LoopInfo *LI;
/// Dominator Tree.
DominatorTree *DT;
- /// Data Layout.
- DataLayout *DL;
+ /// Alias Analysis.
+ AliasAnalysis *AA;
/// Target Library Info.
const TargetLibraryInfo *TLI;
+ /// Target Transform Info.
+ const TargetTransformInfo *TTI;
/// The vectorization SIMD factor to use. Each vector will have this many
/// vector elements.
///The ExitBlock of the scalar loop.
BasicBlock *LoopExitBlock;
///The vector loop body.
- BasicBlock *LoopVectorBody;
+ SmallVector<BasicBlock *, 4> LoopVectorBody;
///The scalar loop body.
BasicBlock *LoopScalarBody;
/// A list of all bypass blocks. The first block is the entry of the loop.
PHINode *Induction;
/// The induction variable of the old basic block.
PHINode *OldInduction;
- /// Holds the extended (to the widest induction type) start index.
- Value *ExtendedIdx;
/// Maps scalars to widened vectors.
ValueMap WidenMap;
+ /// Store instructions that should be predicated, as a pair
+ /// <StoreInst, Predicate>
+ SmallVector<std::pair<StoreInst*,Value*>, 4> PredicatedStores;
EdgeMaskCache MaskCache;
+ /// Trip count of the original loop.
+ Value *TripCount;
+ /// Trip count of the widened loop (TripCount - TripCount % (VF*UF))
+ Value *VectorTripCount;
+
+ /// Map of scalar integer values to the smallest bitwidth they can be legally
+ /// represented as. The vector equivalents of these values should be truncated
+ /// to this type.
+ MapVector<Instruction*,uint64_t> MinBWs;
+ LoopVectorizationLegality *Legal;
+
+ // Record whether runtime check is added.
+ bool AddedSafetyChecks;
};
class InnerLoopUnroller : public InnerLoopVectorizer {
public:
- InnerLoopUnroller(Loop *OrigLoop, ScalarEvolution *SE, LoopInfo *LI,
- DominatorTree *DT, DataLayout *DL,
- const TargetLibraryInfo *TLI, unsigned UnrollFactor) :
- InnerLoopVectorizer(OrigLoop, SE, LI, DT, DL, TLI, 1, UnrollFactor) { }
+ InnerLoopUnroller(Loop *OrigLoop, PredicatedScalarEvolution &PSE,
+ LoopInfo *LI, DominatorTree *DT,
+ const TargetLibraryInfo *TLI,
+ const TargetTransformInfo *TTI, unsigned UnrollFactor)
+ : InnerLoopVectorizer(OrigLoop, PSE, LI, DT, TLI, TTI, 1, UnrollFactor) {}
private:
- virtual void scalarizeInstruction(Instruction *Instr);
- virtual void vectorizeMemoryInstruction(Instruction *Instr,
- LoopVectorizationLegality *Legal);
- virtual Value *getBroadcastInstrs(Value *V);
- virtual Value *getConsecutiveVector(Value* Val, int StartIdx, bool Negate);
- virtual Value *reverseVector(Value *Vec);
+ void scalarizeInstruction(Instruction *Instr,
+ bool IfPredicateStore = false) override;
+ void vectorizeMemoryInstruction(Instruction *Instr) override;
+ Value *getBroadcastInstrs(Value *V) override;
+ Value *getStepVector(Value *Val, int StartIdx, Value *Step) override;
+ Value *reverseVector(Value *Vec) override;
};
/// \brief Look for a meaningful debug location on the instruction or it's
B.SetCurrentDebugLocation(DebugLoc());
}
+#ifndef NDEBUG
+/// \return string containing a file name and a line # for the given loop.
+static std::string getDebugLocString(const Loop *L) {
+ std::string Result;
+ if (L) {
+ raw_string_ostream OS(Result);
+ if (const DebugLoc LoopDbgLoc = L->getStartLoc())
+ LoopDbgLoc.print(OS);
+ else
+ // Just print the module name.
+ OS << L->getHeader()->getParent()->getParent()->getModuleIdentifier();
+ OS.flush();
+ }
+ return Result;
+}
+#endif
+
+/// \brief Propagate known metadata from one instruction to another.
+static void propagateMetadata(Instruction *To, const Instruction *From) {
+ SmallVector<std::pair<unsigned, MDNode *>, 4> Metadata;
+ From->getAllMetadataOtherThanDebugLoc(Metadata);
+
+ for (auto M : Metadata) {
+ unsigned Kind = M.first;
+
+ // These are safe to transfer (this is safe for TBAA, even when we
+ // if-convert, because should that metadata have had a control dependency
+ // on the condition, and thus actually aliased with some other
+ // non-speculated memory access when the condition was false, this would be
+ // caught by the runtime overlap checks).
+ if (Kind != LLVMContext::MD_tbaa &&
+ Kind != LLVMContext::MD_alias_scope &&
+ Kind != LLVMContext::MD_noalias &&
+ Kind != LLVMContext::MD_fpmath &&
+ Kind != LLVMContext::MD_nontemporal)
+ continue;
+
+ To->setMetadata(Kind, M.second);
+ }
+}
+
+/// \brief Propagate known metadata from one instruction to a vector of others.
+static void propagateMetadata(SmallVectorImpl<Value *> &To,
+ const Instruction *From) {
+ for (Value *V : To)
+ if (Instruction *I = dyn_cast<Instruction>(V))
+ propagateMetadata(I, From);
+}
+
+/// \brief The group of interleaved loads/stores sharing the same stride and
+/// close to each other.
+///
+/// Each member in this group has an index starting from 0, and the largest
+/// index should be less than interleaved factor, which is equal to the absolute
+/// value of the access's stride.
+///
+/// E.g. An interleaved load group of factor 4:
+/// for (unsigned i = 0; i < 1024; i+=4) {
+/// a = A[i]; // Member of index 0
+/// b = A[i+1]; // Member of index 1
+/// d = A[i+3]; // Member of index 3
+/// ...
+/// }
+///
+/// An interleaved store group of factor 4:
+/// for (unsigned i = 0; i < 1024; i+=4) {
+/// ...
+/// A[i] = a; // Member of index 0
+/// A[i+1] = b; // Member of index 1
+/// A[i+2] = c; // Member of index 2
+/// A[i+3] = d; // Member of index 3
+/// }
+///
+/// Note: the interleaved load group could have gaps (missing members), but
+/// the interleaved store group doesn't allow gaps.
+class InterleaveGroup {
+public:
+ InterleaveGroup(Instruction *Instr, int Stride, unsigned Align)
+ : Align(Align), SmallestKey(0), LargestKey(0), InsertPos(Instr) {
+ assert(Align && "The alignment should be non-zero");
+
+ Factor = std::abs(Stride);
+ assert(Factor > 1 && "Invalid interleave factor");
+
+ Reverse = Stride < 0;
+ Members[0] = Instr;
+ }
+
+ bool isReverse() const { return Reverse; }
+ unsigned getFactor() const { return Factor; }
+ unsigned getAlignment() const { return Align; }
+ unsigned getNumMembers() const { return Members.size(); }
+
+ /// \brief Try to insert a new member \p Instr with index \p Index and
+ /// alignment \p NewAlign. The index is related to the leader and it could be
+ /// negative if it is the new leader.
+ ///
+ /// \returns false if the instruction doesn't belong to the group.
+ bool insertMember(Instruction *Instr, int Index, unsigned NewAlign) {
+ assert(NewAlign && "The new member's alignment should be non-zero");
+
+ int Key = Index + SmallestKey;
+
+ // Skip if there is already a member with the same index.
+ if (Members.count(Key))
+ return false;
+
+ if (Key > LargestKey) {
+ // The largest index is always less than the interleave factor.
+ if (Index >= static_cast<int>(Factor))
+ return false;
+
+ LargestKey = Key;
+ } else if (Key < SmallestKey) {
+ // The largest index is always less than the interleave factor.
+ if (LargestKey - Key >= static_cast<int>(Factor))
+ return false;
+
+ SmallestKey = Key;
+ }
+
+ // It's always safe to select the minimum alignment.
+ Align = std::min(Align, NewAlign);
+ Members[Key] = Instr;
+ return true;
+ }
+
+ /// \brief Get the member with the given index \p Index
+ ///
+ /// \returns nullptr if contains no such member.
+ Instruction *getMember(unsigned Index) const {
+ int Key = SmallestKey + Index;
+ if (!Members.count(Key))
+ return nullptr;
+
+ return Members.find(Key)->second;
+ }
+
+ /// \brief Get the index for the given member. Unlike the key in the member
+ /// map, the index starts from 0.
+ unsigned getIndex(Instruction *Instr) const {
+ for (auto I : Members)
+ if (I.second == Instr)
+ return I.first - SmallestKey;
+
+ llvm_unreachable("InterleaveGroup contains no such member");
+ }
+
+ Instruction *getInsertPos() const { return InsertPos; }
+ void setInsertPos(Instruction *Inst) { InsertPos = Inst; }
+
+private:
+ unsigned Factor; // Interleave Factor.
+ bool Reverse;
+ unsigned Align;
+ DenseMap<int, Instruction *> Members;
+ int SmallestKey;
+ int LargestKey;
+
+ // To avoid breaking dependences, vectorized instructions of an interleave
+ // group should be inserted at either the first load or the last store in
+ // program order.
+ //
+ // E.g. %even = load i32 // Insert Position
+ // %add = add i32 %even // Use of %even
+ // %odd = load i32
+ //
+ // store i32 %even
+ // %odd = add i32 // Def of %odd
+ // store i32 %odd // Insert Position
+ Instruction *InsertPos;
+};
+
+/// \brief Drive the analysis of interleaved memory accesses in the loop.
+///
+/// Use this class to analyze interleaved accesses only when we can vectorize
+/// a loop. Otherwise it's meaningless to do analysis as the vectorization
+/// on interleaved accesses is unsafe.
+///
+/// The analysis collects interleave groups and records the relationships
+/// between the member and the group in a map.
+class InterleavedAccessInfo {
+public:
+ InterleavedAccessInfo(PredicatedScalarEvolution &PSE, Loop *L,
+ DominatorTree *DT)
+ : PSE(PSE), TheLoop(L), DT(DT) {}
+
+ ~InterleavedAccessInfo() {
+ SmallSet<InterleaveGroup *, 4> DelSet;
+ // Avoid releasing a pointer twice.
+ for (auto &I : InterleaveGroupMap)
+ DelSet.insert(I.second);
+ for (auto *Ptr : DelSet)
+ delete Ptr;
+ }
+
+ /// \brief Analyze the interleaved accesses and collect them in interleave
+ /// groups. Substitute symbolic strides using \p Strides.
+ void analyzeInterleaving(const ValueToValueMap &Strides);
+
+ /// \brief Check if \p Instr belongs to any interleave group.
+ bool isInterleaved(Instruction *Instr) const {
+ return InterleaveGroupMap.count(Instr);
+ }
+
+ /// \brief Get the interleave group that \p Instr belongs to.
+ ///
+ /// \returns nullptr if doesn't have such group.
+ InterleaveGroup *getInterleaveGroup(Instruction *Instr) const {
+ if (InterleaveGroupMap.count(Instr))
+ return InterleaveGroupMap.find(Instr)->second;
+ return nullptr;
+ }
+
+private:
+ /// A wrapper around ScalarEvolution, used to add runtime SCEV checks.
+ /// Simplifies SCEV expressions in the context of existing SCEV assumptions.
+ /// The interleaved access analysis can also add new predicates (for example
+ /// by versioning strides of pointers).
+ PredicatedScalarEvolution &PSE;
+ Loop *TheLoop;
+ DominatorTree *DT;
+
+ /// Holds the relationships between the members and the interleave group.
+ DenseMap<Instruction *, InterleaveGroup *> InterleaveGroupMap;
+
+ /// \brief The descriptor for a strided memory access.
+ struct StrideDescriptor {
+ StrideDescriptor(int Stride, const SCEV *Scev, unsigned Size,
+ unsigned Align)
+ : Stride(Stride), Scev(Scev), Size(Size), Align(Align) {}
+
+ StrideDescriptor() : Stride(0), Scev(nullptr), Size(0), Align(0) {}
+
+ int Stride; // The access's stride. It is negative for a reverse access.
+ const SCEV *Scev; // The scalar expression of this access
+ unsigned Size; // The size of the memory object.
+ unsigned Align; // The alignment of this access.
+ };
+
+ /// \brief Create a new interleave group with the given instruction \p Instr,
+ /// stride \p Stride and alignment \p Align.
+ ///
+ /// \returns the newly created interleave group.
+ InterleaveGroup *createInterleaveGroup(Instruction *Instr, int Stride,
+ unsigned Align) {
+ assert(!InterleaveGroupMap.count(Instr) &&
+ "Already in an interleaved access group");
+ InterleaveGroupMap[Instr] = new InterleaveGroup(Instr, Stride, Align);
+ return InterleaveGroupMap[Instr];
+ }
+
+ /// \brief Release the group and remove all the relationships.
+ void releaseGroup(InterleaveGroup *Group) {
+ for (unsigned i = 0; i < Group->getFactor(); i++)
+ if (Instruction *Member = Group->getMember(i))
+ InterleaveGroupMap.erase(Member);
+
+ delete Group;
+ }
+
+ /// \brief Collect all the accesses with a constant stride in program order.
+ void collectConstStridedAccesses(
+ MapVector<Instruction *, StrideDescriptor> &StrideAccesses,
+ const ValueToValueMap &Strides);
+};
+
+/// Utility class for getting and setting loop vectorizer hints in the form
+/// of loop metadata.
+/// This class keeps a number of loop annotations locally (as member variables)
+/// and can, upon request, write them back as metadata on the loop. It will
+/// initially scan the loop for existing metadata, and will update the local
+/// values based on information in the loop.
+/// We cannot write all values to metadata, as the mere presence of some info,
+/// for example 'force', means a decision has been made. So, we need to be
+/// careful NOT to add them if the user hasn't specifically asked so.
+class LoopVectorizeHints {
+ enum HintKind {
+ HK_WIDTH,
+ HK_UNROLL,
+ HK_FORCE
+ };
+
+ /// Hint - associates name and validation with the hint value.
+ struct Hint {
+ const char * Name;
+ unsigned Value; // This may have to change for non-numeric values.
+ HintKind Kind;
+
+ Hint(const char * Name, unsigned Value, HintKind Kind)
+ : Name(Name), Value(Value), Kind(Kind) { }
+
+ bool validate(unsigned Val) {
+ switch (Kind) {
+ case HK_WIDTH:
+ return isPowerOf2_32(Val) && Val <= VectorizerParams::MaxVectorWidth;
+ case HK_UNROLL:
+ return isPowerOf2_32(Val) && Val <= MaxInterleaveFactor;
+ case HK_FORCE:
+ return (Val <= 1);
+ }
+ return false;
+ }
+ };
+
+ /// Vectorization width.
+ Hint Width;
+ /// Vectorization interleave factor.
+ Hint Interleave;
+ /// Vectorization forced
+ Hint Force;
+
+ /// Return the loop metadata prefix.
+ static StringRef Prefix() { return "llvm.loop."; }
+
+public:
+ enum ForceKind {
+ FK_Undefined = -1, ///< Not selected.
+ FK_Disabled = 0, ///< Forcing disabled.
+ FK_Enabled = 1, ///< Forcing enabled.
+ };
+
+ LoopVectorizeHints(const Loop *L, bool DisableInterleaving)
+ : Width("vectorize.width", VectorizerParams::VectorizationFactor,
+ HK_WIDTH),
+ Interleave("interleave.count", DisableInterleaving, HK_UNROLL),
+ Force("vectorize.enable", FK_Undefined, HK_FORCE),
+ TheLoop(L) {
+ // Populate values with existing loop metadata.
+ getHintsFromMetadata();
+
+ // force-vector-interleave overrides DisableInterleaving.
+ if (VectorizerParams::isInterleaveForced())
+ Interleave.Value = VectorizerParams::VectorizationInterleave;
+
+ DEBUG(if (DisableInterleaving && Interleave.Value == 1) dbgs()
+ << "LV: Interleaving disabled by the pass manager\n");
+ }
+
+ /// Mark the loop L as already vectorized by setting the width to 1.
+ void setAlreadyVectorized() {
+ Width.Value = Interleave.Value = 1;
+ Hint Hints[] = {Width, Interleave};
+ writeHintsToMetadata(Hints);
+ }
+
+ bool allowVectorization(Function *F, Loop *L, bool AlwaysVectorize) const {
+ if (getForce() == LoopVectorizeHints::FK_Disabled) {
+ DEBUG(dbgs() << "LV: Not vectorizing: #pragma vectorize disable.\n");
+ emitOptimizationRemarkAnalysis(F->getContext(),
+ vectorizeAnalysisPassName(), *F,
+ L->getStartLoc(), emitRemark());
+ return false;
+ }
+
+ if (!AlwaysVectorize && getForce() != LoopVectorizeHints::FK_Enabled) {
+ DEBUG(dbgs() << "LV: Not vectorizing: No #pragma vectorize enable.\n");
+ emitOptimizationRemarkAnalysis(F->getContext(),
+ vectorizeAnalysisPassName(), *F,
+ L->getStartLoc(), emitRemark());
+ return false;
+ }
+
+ if (getWidth() == 1 && getInterleave() == 1) {
+ // FIXME: Add a separate metadata to indicate when the loop has already
+ // been vectorized instead of setting width and count to 1.
+ DEBUG(dbgs() << "LV: Not vectorizing: Disabled/already vectorized.\n");
+ // FIXME: Add interleave.disable metadata. This will allow
+ // vectorize.disable to be used without disabling the pass and errors
+ // to differentiate between disabled vectorization and a width of 1.
+ emitOptimizationRemarkAnalysis(
+ F->getContext(), vectorizeAnalysisPassName(), *F, L->getStartLoc(),
+ "loop not vectorized: vectorization and interleaving are explicitly "
+ "disabled, or vectorize width and interleave count are both set to "
+ "1");
+ return false;
+ }
+
+ return true;
+ }
+
+ /// Dumps all the hint information.
+ std::string emitRemark() const {
+ VectorizationReport R;
+ if (Force.Value == LoopVectorizeHints::FK_Disabled)
+ R << "vectorization is explicitly disabled";
+ else {
+ R << "use -Rpass-analysis=loop-vectorize for more info";
+ if (Force.Value == LoopVectorizeHints::FK_Enabled) {
+ R << " (Force=true";
+ if (Width.Value != 0)
+ R << ", Vector Width=" << Width.Value;
+ if (Interleave.Value != 0)
+ R << ", Interleave Count=" << Interleave.Value;
+ R << ")";
+ }
+ }
+
+ return R.str();
+ }
+
+ unsigned getWidth() const { return Width.Value; }
+ unsigned getInterleave() const { return Interleave.Value; }
+ enum ForceKind getForce() const { return (ForceKind)Force.Value; }
+ const char *vectorizeAnalysisPassName() const {
+ // If hints are provided that don't disable vectorization use the
+ // AlwaysPrint pass name to force the frontend to print the diagnostic.
+ if (getWidth() == 1)
+ return LV_NAME;
+ if (getForce() == LoopVectorizeHints::FK_Disabled)
+ return LV_NAME;
+ if (getForce() == LoopVectorizeHints::FK_Undefined && getWidth() == 0)
+ return LV_NAME;
+ return DiagnosticInfo::AlwaysPrint;
+ }
+
+ bool allowReordering() const {
+ // When enabling loop hints are provided we allow the vectorizer to change
+ // the order of operations that is given by the scalar loop. This is not
+ // enabled by default because can be unsafe or inefficient. For example,
+ // reordering floating-point operations will change the way round-off
+ // error accumulates in the loop.
+ return getForce() == LoopVectorizeHints::FK_Enabled || getWidth() > 1;
+ }
+
+private:
+ /// Find hints specified in the loop metadata and update local values.
+ void getHintsFromMetadata() {
+ MDNode *LoopID = TheLoop->getLoopID();
+ if (!LoopID)
+ return;
+
+ // First operand should refer to the loop id itself.
+ assert(LoopID->getNumOperands() > 0 && "requires at least one operand");
+ assert(LoopID->getOperand(0) == LoopID && "invalid loop id");
+
+ for (unsigned i = 1, ie = LoopID->getNumOperands(); i < ie; ++i) {
+ const MDString *S = nullptr;
+ SmallVector<Metadata *, 4> Args;
+
+ // The expected hint is either a MDString or a MDNode with the first
+ // operand a MDString.
+ if (const MDNode *MD = dyn_cast<MDNode>(LoopID->getOperand(i))) {
+ if (!MD || MD->getNumOperands() == 0)
+ continue;
+ S = dyn_cast<MDString>(MD->getOperand(0));
+ for (unsigned i = 1, ie = MD->getNumOperands(); i < ie; ++i)
+ Args.push_back(MD->getOperand(i));
+ } else {
+ S = dyn_cast<MDString>(LoopID->getOperand(i));
+ assert(Args.size() == 0 && "too many arguments for MDString");
+ }
+
+ if (!S)
+ continue;
+
+ // Check if the hint starts with the loop metadata prefix.
+ StringRef Name = S->getString();
+ if (Args.size() == 1)
+ setHint(Name, Args[0]);
+ }
+ }
+
+ /// Checks string hint with one operand and set value if valid.
+ void setHint(StringRef Name, Metadata *Arg) {
+ if (!Name.startswith(Prefix()))
+ return;
+ Name = Name.substr(Prefix().size(), StringRef::npos);
+
+ const ConstantInt *C = mdconst::dyn_extract<ConstantInt>(Arg);
+ if (!C) return;
+ unsigned Val = C->getZExtValue();
+
+ Hint *Hints[] = {&Width, &Interleave, &Force};
+ for (auto H : Hints) {
+ if (Name == H->Name) {
+ if (H->validate(Val))
+ H->Value = Val;
+ else
+ DEBUG(dbgs() << "LV: ignoring invalid hint '" << Name << "'\n");
+ break;
+ }
+ }
+ }
+
+ /// Create a new hint from name / value pair.
+ MDNode *createHintMetadata(StringRef Name, unsigned V) const {
+ LLVMContext &Context = TheLoop->getHeader()->getContext();
+ Metadata *MDs[] = {MDString::get(Context, Name),
+ ConstantAsMetadata::get(
+ ConstantInt::get(Type::getInt32Ty(Context), V))};
+ return MDNode::get(Context, MDs);
+ }
+
+ /// Matches metadata with hint name.
+ bool matchesHintMetadataName(MDNode *Node, ArrayRef<Hint> HintTypes) {
+ MDString* Name = dyn_cast<MDString>(Node->getOperand(0));
+ if (!Name)
+ return false;
+
+ for (auto H : HintTypes)
+ if (Name->getString().endswith(H.Name))
+ return true;
+ return false;
+ }
+
+ /// Sets current hints into loop metadata, keeping other values intact.
+ void writeHintsToMetadata(ArrayRef<Hint> HintTypes) {
+ if (HintTypes.size() == 0)
+ return;
+
+ // Reserve the first element to LoopID (see below).
+ SmallVector<Metadata *, 4> MDs(1);
+ // If the loop already has metadata, then ignore the existing operands.
+ MDNode *LoopID = TheLoop->getLoopID();
+ if (LoopID) {
+ for (unsigned i = 1, ie = LoopID->getNumOperands(); i < ie; ++i) {
+ MDNode *Node = cast<MDNode>(LoopID->getOperand(i));
+ // If node in update list, ignore old value.
+ if (!matchesHintMetadataName(Node, HintTypes))
+ MDs.push_back(Node);
+ }
+ }
+
+ // Now, add the missing hints.
+ for (auto H : HintTypes)
+ 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, MDs);
+ // Set operand 0 to refer to the loop id itself.
+ NewLoopID->replaceOperandWith(0, NewLoopID);
+
+ TheLoop->setLoopID(NewLoopID);
+ }
+
+ /// The loop these hints belong to.
+ const Loop *TheLoop;
+};
+
+static void emitAnalysisDiag(const Function *TheFunction, const Loop *TheLoop,
+ const LoopVectorizeHints &Hints,
+ const LoopAccessReport &Message) {
+ const char *Name = Hints.vectorizeAnalysisPassName();
+ LoopAccessReport::emitAnalysis(Message, TheFunction, TheLoop, Name);
+}
+
+static void emitMissedWarning(Function *F, Loop *L,
+ const LoopVectorizeHints &LH) {
+ emitOptimizationRemarkMissed(F->getContext(), LV_NAME, *F, L->getStartLoc(),
+ LH.emitRemark());
+
+ if (LH.getForce() == LoopVectorizeHints::FK_Enabled) {
+ if (LH.getWidth() != 1)
+ emitLoopVectorizeWarning(
+ F->getContext(), *F, L->getStartLoc(),
+ "failed explicitly specified loop vectorization");
+ else if (LH.getInterleave() != 1)
+ emitLoopInterleaveWarning(
+ F->getContext(), *F, L->getStartLoc(),
+ "failed explicitly specified loop interleaving");
+ }
+}
+
/// LoopVectorizationLegality checks if it is legal to vectorize a loop, and
/// to what vectorization factor.
/// This class does not look at the profitability of vectorization, only the
/// induction variable and the different reduction variables.
class LoopVectorizationLegality {
public:
- LoopVectorizationLegality(Loop *L, ScalarEvolution *SE, DataLayout *DL,
- DominatorTree *DT, TargetLibraryInfo *TLI)
- : TheLoop(L), SE(SE), DL(DL), DT(DT), TLI(TLI),
- Induction(0), WidestIndTy(0), HasFunNoNaNAttr(false),
- MaxSafeDepDistBytes(-1U) {}
-
- /// This enum represents the kinds of reductions that we support.
- enum ReductionKind {
- RK_NoReduction, ///< Not a reduction.
- RK_IntegerAdd, ///< Sum of integers.
- RK_IntegerMult, ///< Product of integers.
- RK_IntegerOr, ///< Bitwise or logical OR of numbers.
- RK_IntegerAnd, ///< Bitwise or logical AND of numbers.
- RK_IntegerXor, ///< Bitwise or logical XOR of numbers.
- RK_IntegerMinMax, ///< Min/max implemented in terms of select(cmp()).
- RK_FloatAdd, ///< Sum of floats.
- RK_FloatMult, ///< Product of floats.
- RK_FloatMinMax ///< Min/max implemented in terms of select(cmp()).
- };
-
- /// This enum represents the kinds of inductions that we support.
- enum InductionKind {
- IK_NoInduction, ///< Not an induction variable.
- IK_IntInduction, ///< Integer induction variable. Step = 1.
- IK_ReverseIntInduction, ///< Reverse int induction variable. Step = -1.
- IK_PtrInduction, ///< Pointer induction var. Step = sizeof(elem).
- IK_ReversePtrInduction ///< Reverse ptr indvar. Step = - sizeof(elem).
- };
-
- // This enum represents the kind of minmax reduction.
- enum MinMaxReductionKind {
- MRK_Invalid,
- MRK_UIntMin,
- MRK_UIntMax,
- MRK_SIntMin,
- MRK_SIntMax,
- MRK_FloatMin,
- MRK_FloatMax
- };
-
- /// This POD struct holds information about reduction variables.
- struct ReductionDescriptor {
- ReductionDescriptor() : StartValue(0), LoopExitInstr(0),
- Kind(RK_NoReduction), MinMaxKind(MRK_Invalid) {}
-
- ReductionDescriptor(Value *Start, Instruction *Exit, ReductionKind K,
- MinMaxReductionKind MK)
- : StartValue(Start), LoopExitInstr(Exit), Kind(K), MinMaxKind(MK) {}
-
- // The starting value of the reduction.
- // It does not have to be zero!
- TrackingVH<Value> StartValue;
- // The instruction who's value is used outside the loop.
- Instruction *LoopExitInstr;
- // The kind of the reduction.
- ReductionKind Kind;
- // If this a min/max reduction the kind of reduction.
- MinMaxReductionKind MinMaxKind;
- };
-
- /// This POD struct holds information about a potential reduction operation.
- struct ReductionInstDesc {
- ReductionInstDesc(bool IsRedux, Instruction *I) :
- IsReduction(IsRedux), PatternLastInst(I), MinMaxKind(MRK_Invalid) {}
-
- ReductionInstDesc(Instruction *I, MinMaxReductionKind K) :
- IsReduction(true), PatternLastInst(I), MinMaxKind(K) {}
-
- // Is this instruction a reduction candidate.
- bool IsReduction;
- // The last instruction in a min/max pattern (select of the select(icmp())
- // pattern), or the current reduction instruction otherwise.
- Instruction *PatternLastInst;
- // If this is a min/max pattern the comparison predicate.
- MinMaxReductionKind MinMaxKind;
- };
-
- // This POD struct holds information about the memory runtime legality
- // check that a group of pointers do not overlap.
- struct RuntimePointerCheck {
- RuntimePointerCheck() : Need(false) {}
-
- /// Reset the state of the pointer runtime information.
- void reset() {
- Need = false;
- Pointers.clear();
- Starts.clear();
- Ends.clear();
- }
-
- /// Insert a pointer and calculate the start and end SCEVs.
- void insert(ScalarEvolution *SE, Loop *Lp, Value *Ptr, bool WritePtr,
- unsigned DepSetId);
-
- /// This flag indicates if we need to add the runtime check.
- bool Need;
- /// Holds the pointers that we need to check.
- SmallVector<TrackingVH<Value>, 2> Pointers;
- /// Holds the pointer value at the beginning of the loop.
- SmallVector<const SCEV*, 2> Starts;
- /// Holds the pointer value at the end of the loop.
- SmallVector<const SCEV*, 2> Ends;
- /// Holds the information if this pointer is used for writing to memory.
- SmallVector<bool, 2> IsWritePtr;
- /// Holds the id of the set of pointers that could be dependent because of a
- /// shared underlying object.
- SmallVector<unsigned, 2> DependencySetId;
- };
-
- /// A POD for saving information about induction variables.
- struct InductionInfo {
- InductionInfo(Value *Start, InductionKind K) : StartValue(Start), IK(K) {}
- InductionInfo() : StartValue(0), IK(IK_NoInduction) {}
- /// Start value.
- TrackingVH<Value> StartValue;
- /// Induction kind.
- InductionKind IK;
- };
+ LoopVectorizationLegality(Loop *L, PredicatedScalarEvolution &PSE,
+ DominatorTree *DT, TargetLibraryInfo *TLI,
+ AliasAnalysis *AA, Function *F,
+ const TargetTransformInfo *TTI,
+ LoopAccessAnalysis *LAA,
+ LoopVectorizationRequirements *R,
+ const LoopVectorizeHints *H)
+ : NumPredStores(0), TheLoop(L), PSE(PSE), TLI(TLI), TheFunction(F),
+ TTI(TTI), DT(DT), LAA(LAA), LAI(nullptr), InterleaveInfo(PSE, L, DT),
+ Induction(nullptr), WidestIndTy(nullptr), HasFunNoNaNAttr(false),
+ Requirements(R), Hints(H) {}
/// ReductionList contains the reduction descriptors for all
/// of the reductions that were found in the loop.
- typedef DenseMap<PHINode*, ReductionDescriptor> ReductionList;
+ typedef DenseMap<PHINode *, RecurrenceDescriptor> ReductionList;
/// InductionList saves induction variables and maps them to the
/// induction descriptor.
- typedef MapVector<PHINode*, InductionInfo> InductionList;
+ typedef MapVector<PHINode*, InductionDescriptor> InductionList;
/// Returns true if it is legal to vectorize this loop.
/// This does not mean that it is profitable to vectorize this
/// Returns True if V is an induction variable in this loop.
bool isInductionVariable(const Value *V);
+ /// Returns True if PN is a reduction variable in this loop.
+ bool isReductionVariable(PHINode *PN) { return Reductions.count(PN); }
+
/// Return true if the block BB needs to be predicated in order for the loop
/// to be vectorized.
bool blockNeedsPredication(BasicBlock *BB);
/// pointer itself is an induction variable.
/// This check allows us to vectorize A[idx] into a wide load/store.
/// Returns:
- /// 0 - Stride is unknown or non consecutive.
+ /// 0 - Stride is unknown or non-consecutive.
/// 1 - Address is consecutive.
/// -1 - Address is consecutive, and decreasing.
int isConsecutivePtr(Value *Ptr);
bool isUniformAfterVectorization(Instruction* I) { return Uniforms.count(I); }
/// Returns the information that we collected about runtime memory check.
- RuntimePointerCheck *getRuntimePointerCheck() { return &PtrRtCheck; }
+ const RuntimePointerChecking *getRuntimePointerChecking() const {
+ return LAI->getRuntimePointerChecking();
+ }
+
+ const LoopAccessInfo *getLAI() const {
+ return LAI;
+ }
+
+ /// \brief Check if \p Instr belongs to any interleaved access group.
+ bool isAccessInterleaved(Instruction *Instr) {
+ return InterleaveInfo.isInterleaved(Instr);
+ }
+
+ /// \brief Get the interleaved access group that \p Instr belongs to.
+ const InterleaveGroup *getInterleavedAccessGroup(Instruction *Instr) {
+ return InterleaveInfo.getInterleaveGroup(Instr);
+ }
- /// This function returns the identity element (or neutral element) for
- /// the operation K.
- static Constant *getReductionIdentity(ReductionKind K, Type *Tp);
+ unsigned getMaxSafeDepDistBytes() { return LAI->getMaxSafeDepDistBytes(); }
- unsigned getMaxSafeDepDistBytes() { return MaxSafeDepDistBytes; }
+ bool hasStride(Value *V) { return StrideSet.count(V); }
+ bool mustCheckStrides() { return !StrideSet.empty(); }
+ SmallPtrSet<Value *, 8>::iterator strides_begin() {
+ return StrideSet.begin();
+ }
+ 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 isConsecutivePtr(Ptr) && TTI->isLegalMaskedStore(DataType);
+ }
+ /// 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 isConsecutivePtr(Ptr) && TTI->isLegalMaskedLoad(DataType);
+ }
+ /// Returns true if vector representation of the instruction \p I
+ /// requires mask.
+ bool isMaskRequired(const Instruction* I) {
+ return (MaskedOp.count(I) != 0);
+ }
+ unsigned getNumStores() const {
+ return LAI->getNumStores();
+ }
+ unsigned getNumLoads() const {
+ return LAI->getNumLoads();
+ }
+ unsigned getNumPredStores() const {
+ return NumPredStores;
+ }
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
/// Return true if all of the instructions in the block can be speculatively
/// executed. \p SafePtrs is a list of addresses that are known to be legal
/// and we know that we can read from them without segfault.
- bool blockCanBePredicated(BasicBlock *BB, SmallPtrSet<Value *, 8>& SafePtrs);
-
- /// Returns True, if 'Phi' is the kind of reduction variable for type
- /// 'Kind'. If this is a reduction variable, it adds it to ReductionList.
- bool AddReductionVar(PHINode *Phi, ReductionKind Kind);
- /// Returns a struct describing if the instruction 'I' can be a reduction
- /// variable of type 'Kind'. If the reduction is a min/max pattern of
- /// select(icmp()) this function advances the instruction pointer 'I' from the
- /// compare instruction to the select instruction and stores this pointer in
- /// 'PatternLastInst' member of the returned struct.
- ReductionInstDesc isReductionInstr(Instruction *I, ReductionKind Kind,
- ReductionInstDesc &Desc);
- /// Returns true if the instruction is a Select(ICmp(X, Y), X, Y) instruction
- /// pattern corresponding to a min(X, Y) or max(X, Y).
- static ReductionInstDesc isMinMaxSelectCmpPattern(Instruction *I,
- ReductionInstDesc &Prev);
- /// Returns the induction kind of Phi. This function may return NoInduction
- /// if the PHI is not an induction variable.
- InductionKind isInductionVariable(PHINode *Phi);
+ bool blockCanBePredicated(BasicBlock *BB, SmallPtrSetImpl<Value *> &SafePtrs);
+
+ /// \brief Collect memory access with loop invariant strides.
+ ///
+ /// Looks for accesses like "a[i * StrideA]" where "StrideA" is loop
+ /// invariant.
+ void collectStridedAccess(Value *LoadOrStoreInst);
+
+ /// Report an analysis message to assist the user in diagnosing loops that are
+ /// not vectorized. These are handled as LoopAccessReport rather than
+ /// VectorizationReport because the << operator of VectorizationReport returns
+ /// LoopAccessReport.
+ void emitAnalysis(const LoopAccessReport &Message) const {
+ emitAnalysisDiag(TheFunction, TheLoop, *Hints, Message);
+ }
+
+ unsigned NumPredStores;
/// The loop that we evaluate.
Loop *TheLoop;
- /// Scev analysis.
- ScalarEvolution *SE;
- /// DataLayout analysis.
- DataLayout *DL;
- /// Dominators.
- DominatorTree *DT;
+ /// A wrapper around ScalarEvolution used to add runtime SCEV checks.
+ /// Applies dynamic knowledge to simplify SCEV expressions in the context
+ /// of existing SCEV assumptions. The analysis will also add a minimal set
+ /// of new predicates if this is required to enable vectorization and
+ /// unrolling.
+ PredicatedScalarEvolution &PSE;
/// Target Library Info.
TargetLibraryInfo *TLI;
-
+ /// Parent function
+ Function *TheFunction;
+ /// Target Transform Info
+ const TargetTransformInfo *TTI;
+ /// Dominator Tree.
+ DominatorTree *DT;
+ // LoopAccess analysis.
+ LoopAccessAnalysis *LAA;
+ // And the loop-accesses info corresponding to this loop. This pointer is
+ // null until canVectorizeMemory sets it up.
+ const LoopAccessInfo *LAI;
+
+ /// The interleave access information contains groups of interleaved accesses
+ /// with the same stride and close to each other.
+ InterleavedAccessInfo InterleaveInfo;
+
// --- vectorization state --- //
/// Holds the integer induction variable. This is the counter of the
/// This set holds the variables which are known to be uniform after
/// vectorization.
SmallPtrSet<Instruction*, 4> Uniforms;
- /// We need to check that all of the pointers in this list are disjoint
- /// at runtime.
- RuntimePointerCheck PtrRtCheck;
+
/// Can we assume the absence of NaNs.
bool HasFunNoNaNAttr;
- unsigned MaxSafeDepDistBytes;
+ /// Vectorization requirements that will go through late-evaluation.
+ LoopVectorizationRequirements *Requirements;
+
+ /// Used to emit an analysis of any legality issues.
+ const LoopVectorizeHints *Hints;
+
+ 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
LoopVectorizationCostModel(Loop *L, ScalarEvolution *SE, LoopInfo *LI,
LoopVectorizationLegality *Legal,
const TargetTransformInfo &TTI,
- DataLayout *DL, const TargetLibraryInfo *TLI)
- : TheLoop(L), SE(SE), LI(LI), Legal(Legal), TTI(TTI), DL(DL), TLI(TLI) {}
+ const TargetLibraryInfo *TLI, DemandedBits *DB,
+ AssumptionCache *AC, const Function *F,
+ const LoopVectorizeHints *Hints,
+ SmallPtrSetImpl<const Value *> &ValuesToIgnore)
+ : TheLoop(L), SE(SE), LI(LI), Legal(Legal), TTI(TTI), TLI(TLI), DB(DB),
+ TheFunction(F), Hints(Hints), ValuesToIgnore(ValuesToIgnore) {}
/// Information about vectorization costs
struct VectorizationFactor {
/// This method checks every power of two up to VF. If UserVF is not ZERO
/// then this vectorization factor will be selected if vectorization is
/// possible.
- VectorizationFactor selectVectorizationFactor(bool OptForSize,
- unsigned UserVF);
+ VectorizationFactor selectVectorizationFactor(bool OptForSize);
- /// \return The size (in bits) of the widest type in the code that
- /// needs to be vectorized. We ignore values that remain scalar such as
+ /// \return The size (in bits) of the smallest and widest types in the code
+ /// that needs to be vectorized. We ignore values that remain scalar such as
/// 64 bit loop indices.
- unsigned getWidestType();
+ std::pair<unsigned, unsigned> getSmallestAndWidestTypes();
+
+ /// \return The desired interleave count.
+ /// If interleave count has been specified by metadata it will be returned.
+ /// Otherwise, the interleave count is computed and returned. VF and LoopCost
+ /// are the selected vectorization factor and the cost of the selected VF.
+ unsigned selectInterleaveCount(bool OptForSize, unsigned VF,
+ unsigned LoopCost);
/// \return The most profitable unroll factor.
- /// If UserUF is non-zero then this method finds the best unroll-factor
- /// based on register pressure and other parameters.
- /// VF and LoopCost are the selected vectorization factor and the cost of the
- /// selected VF.
- unsigned selectUnrollFactor(bool OptForSize, unsigned UserUF, unsigned VF,
- unsigned LoopCost);
+ /// This method finds the best unroll-factor based on register pressure and
+ /// other parameters. VF and LoopCost are the selected vectorization factor
+ /// and the cost of the selected VF.
+ unsigned computeInterleaveCount(bool OptForSize, unsigned VF,
+ unsigned LoopCost);
/// \brief A struct that represents some properties of the register usage
/// of a loop.
unsigned NumInstructions;
};
- /// \return information about the register usage of the loop.
- RegisterUsage calculateRegisterUsage();
+ /// \return Returns information about the register usages of the loop for the
+ /// given vectorization factors.
+ SmallVector<RegisterUsage, 8>
+ calculateRegisterUsage(const SmallVector<unsigned, 8> &VFs);
private:
/// Returns the expected execution cost. The unit of the cost does
/// width. Vector width of one means scalar.
unsigned getInstructionCost(Instruction *I, unsigned VF);
- /// A helper function for converting Scalar types to vector types.
- /// If the incoming type is void, we return void. If the VF is 1, we return
- /// the scalar type.
- static Type* ToVectorTy(Type *Scalar, unsigned VF);
-
/// Returns whether the instruction is a load or store and will be a emitted
/// as a vector operation.
bool isConsecutiveLoadOrStore(Instruction *I);
+ /// Report an analysis message to assist the user in diagnosing loops that are
+ /// not vectorized. These are handled as LoopAccessReport rather than
+ /// VectorizationReport because the << operator of VectorizationReport returns
+ /// LoopAccessReport.
+ void emitAnalysis(const LoopAccessReport &Message) const {
+ emitAnalysisDiag(TheFunction, TheLoop, *Hints, Message);
+ }
+
+public:
+ /// Map of scalar integer values to the smallest bitwidth they can be legally
+ /// represented as. The vector equivalents of these values should be truncated
+ /// to this type.
+ MapVector<Instruction*,uint64_t> MinBWs;
+
/// The loop that we evaluate.
Loop *TheLoop;
/// Scev analysis.
LoopVectorizationLegality *Legal;
/// Vector target information.
const TargetTransformInfo &TTI;
- /// Target data layout information.
- DataLayout *DL;
/// Target Library Info.
const TargetLibraryInfo *TLI;
+ /// Demanded bits analysis
+ DemandedBits *DB;
+ const Function *TheFunction;
+ // Loop Vectorize Hint.
+ const LoopVectorizeHints *Hints;
+ // Values to ignore in the cost model.
+ const SmallPtrSetImpl<const Value *> &ValuesToIgnore;
};
-/// Utility class for getting and setting loop vectorizer hints in the form
-/// of loop metadata.
-struct LoopVectorizeHints {
- /// Vectorization width.
- unsigned Width;
- /// Vectorization unroll factor.
- unsigned Unroll;
-
- LoopVectorizeHints(const Loop *L, bool DisableUnrolling)
- : Width(VectorizationFactor)
- , Unroll(DisableUnrolling ? 1 : VectorizationUnroll)
- , LoopID(L->getLoopID()) {
- getHints(L);
- // The command line options override any loop metadata except for when
- // width == 1 which is used to indicate the loop is already vectorized.
- if (VectorizationFactor.getNumOccurrences() > 0 && Width != 1)
- Width = VectorizationFactor;
- if (VectorizationUnroll.getNumOccurrences() > 0)
- Unroll = VectorizationUnroll;
-
- DEBUG(if (DisableUnrolling && Unroll == 1)
- dbgs() << "LV: Unrolling disabled by the pass manager\n");
- }
-
- /// Return the loop vectorizer metadata prefix.
- static StringRef Prefix() { return "llvm.vectorizer."; }
+/// \brief This holds vectorization requirements that must be verified late in
+/// the process. The requirements are set by legalize and costmodel. Once
+/// vectorization has been determined to be possible and profitable the
+/// requirements can be verified by looking for metadata or compiler options.
+/// For example, some loops require FP commutativity which is only allowed if
+/// vectorization is explicitly specified or if the fast-math compiler option
+/// has been provided.
+/// Late evaluation of these requirements allows helpful diagnostics to be
+/// composed that tells the user what need to be done to vectorize the loop. For
+/// example, by specifying #pragma clang loop vectorize or -ffast-math. Late
+/// evaluation should be used only when diagnostics can generated that can be
+/// followed by a non-expert user.
+class LoopVectorizationRequirements {
+public:
+ LoopVectorizationRequirements()
+ : NumRuntimePointerChecks(0), UnsafeAlgebraInst(nullptr) {}
- MDNode *createHint(LLVMContext &Context, StringRef Name, unsigned V) {
- SmallVector<Value*, 2> Vals;
- Vals.push_back(MDString::get(Context, Name));
- Vals.push_back(ConstantInt::get(Type::getInt32Ty(Context), V));
- return MDNode::get(Context, Vals);
+ void addUnsafeAlgebraInst(Instruction *I) {
+ // First unsafe algebra instruction.
+ if (!UnsafeAlgebraInst)
+ UnsafeAlgebraInst = I;
}
- /// Mark the loop L as already vectorized by setting the width to 1.
- void setAlreadyVectorized(Loop *L) {
- LLVMContext &Context = L->getHeader()->getContext();
-
- Width = 1;
-
- // Create a new loop id with one more operand for the already_vectorized
- // hint. If the loop already has a loop id then copy the existing operands.
- SmallVector<Value*, 4> Vals(1);
- if (LoopID)
- for (unsigned i = 1, ie = LoopID->getNumOperands(); i < ie; ++i)
- Vals.push_back(LoopID->getOperand(i));
-
- Vals.push_back(createHint(Context, Twine(Prefix(), "width").str(), Width));
-
- MDNode *NewLoopID = MDNode::get(Context, Vals);
- // Set operand 0 to refer to the loop id itself.
- NewLoopID->replaceOperandWith(0, NewLoopID);
+ void addRuntimePointerChecks(unsigned Num) { NumRuntimePointerChecks = Num; }
+
+ bool doesNotMeet(Function *F, Loop *L, const LoopVectorizeHints &Hints) {
+ const char *Name = Hints.vectorizeAnalysisPassName();
+ bool Failed = false;
+ if (UnsafeAlgebraInst && !Hints.allowReordering()) {
+ emitOptimizationRemarkAnalysisFPCommute(
+ F->getContext(), Name, *F, UnsafeAlgebraInst->getDebugLoc(),
+ VectorizationReport() << "cannot prove it is safe to reorder "
+ "floating-point operations");
+ Failed = true;
+ }
- L->setLoopID(NewLoopID);
- if (LoopID)
- LoopID->replaceAllUsesWith(NewLoopID);
+ // Test if runtime memcheck thresholds are exceeded.
+ bool PragmaThresholdReached =
+ NumRuntimePointerChecks > PragmaVectorizeMemoryCheckThreshold;
+ bool ThresholdReached =
+ NumRuntimePointerChecks > VectorizerParams::RuntimeMemoryCheckThreshold;
+ if ((ThresholdReached && !Hints.allowReordering()) ||
+ PragmaThresholdReached) {
+ emitOptimizationRemarkAnalysisAliasing(
+ F->getContext(), Name, *F, L->getStartLoc(),
+ VectorizationReport()
+ << "cannot prove it is safe to reorder memory operations");
+ DEBUG(dbgs() << "LV: Too many memory checks needed.\n");
+ Failed = true;
+ }
- LoopID = NewLoopID;
+ return Failed;
}
private:
- MDNode *LoopID;
-
- /// Find hints specified in the loop metadata.
- void getHints(const Loop *L) {
- if (!LoopID)
- return;
-
- // First operand should refer to the loop id itself.
- assert(LoopID->getNumOperands() > 0 && "requires at least one operand");
- assert(LoopID->getOperand(0) == LoopID && "invalid loop id");
-
- for (unsigned i = 1, ie = LoopID->getNumOperands(); i < ie; ++i) {
- const MDString *S = 0;
- SmallVector<Value*, 4> Args;
-
- // The expected hint is either a MDString or a MDNode with the first
- // operand a MDString.
- if (const MDNode *MD = dyn_cast<MDNode>(LoopID->getOperand(i))) {
- if (!MD || MD->getNumOperands() == 0)
- continue;
- S = dyn_cast<MDString>(MD->getOperand(0));
- for (unsigned i = 1, ie = MD->getNumOperands(); i < ie; ++i)
- Args.push_back(MD->getOperand(i));
- } else {
- S = dyn_cast<MDString>(LoopID->getOperand(i));
- assert(Args.size() == 0 && "too many arguments for MDString");
- }
-
- if (!S)
- continue;
-
- // Check if the hint starts with the vectorizer prefix.
- StringRef Hint = S->getString();
- if (!Hint.startswith(Prefix()))
- continue;
- // Remove the prefix.
- Hint = Hint.substr(Prefix().size(), StringRef::npos);
-
- if (Args.size() == 1)
- getHint(Hint, Args[0]);
- }
- }
+ unsigned NumRuntimePointerChecks;
+ Instruction *UnsafeAlgebraInst;
+};
- // Check string hint with one operand.
- void getHint(StringRef Hint, Value *Arg) {
- const ConstantInt *C = dyn_cast<ConstantInt>(Arg);
- if (!C) return;
- unsigned Val = C->getZExtValue();
+static void addInnerLoop(Loop &L, SmallVectorImpl<Loop *> &V) {
+ if (L.empty())
+ return V.push_back(&L);
- if (Hint == "width") {
- if (isPowerOf2_32(Val) && Val <= MaxVectorWidth)
- Width = Val;
- else
- DEBUG(dbgs() << "LV: ignoring invalid width hint metadata\n");
- } else if (Hint == "unroll") {
- if (isPowerOf2_32(Val) && Val <= MaxUnrollFactor)
- Unroll = Val;
- else
- DEBUG(dbgs() << "LV: ignoring invalid unroll hint metadata\n");
- } else {
- DEBUG(dbgs() << "LV: ignoring unknown hint " << Hint << '\n');
- }
- }
-};
+ for (Loop *InnerL : L)
+ addInnerLoop(*InnerL, V);
+}
/// The LoopVectorize Pass.
-struct LoopVectorize : public LoopPass {
+struct LoopVectorize : public FunctionPass {
/// Pass identification, replacement for typeid
static char ID;
- explicit LoopVectorize(bool NoUnrolling = false)
- : LoopPass(ID), DisableUnrolling(NoUnrolling) {
+ explicit LoopVectorize(bool NoUnrolling = false, bool AlwaysVectorize = true)
+ : FunctionPass(ID),
+ DisableUnrolling(NoUnrolling),
+ AlwaysVectorize(AlwaysVectorize) {
initializeLoopVectorizePass(*PassRegistry::getPassRegistry());
}
ScalarEvolution *SE;
- DataLayout *DL;
LoopInfo *LI;
TargetTransformInfo *TTI;
DominatorTree *DT;
+ BlockFrequencyInfo *BFI;
TargetLibraryInfo *TLI;
+ DemandedBits *DB;
+ AliasAnalysis *AA;
+ AssumptionCache *AC;
+ LoopAccessAnalysis *LAA;
bool DisableUnrolling;
-
- virtual bool runOnLoop(Loop *L, LPPassManager &LPM) {
- // We only vectorize innermost loops.
- if (!L->empty())
+ bool AlwaysVectorize;
+
+ BlockFrequency ColdEntryFreq;
+
+ bool runOnFunction(Function &F) override {
+ SE = &getAnalysis<ScalarEvolutionWrapperPass>().getSE();
+ LI = &getAnalysis<LoopInfoWrapperPass>().getLoopInfo();
+ TTI = &getAnalysis<TargetTransformInfoWrapperPass>().getTTI(F);
+ DT = &getAnalysis<DominatorTreeWrapperPass>().getDomTree();
+ BFI = &getAnalysis<BlockFrequencyInfoWrapperPass>().getBFI();
+ auto *TLIP = getAnalysisIfAvailable<TargetLibraryInfoWrapperPass>();
+ TLI = TLIP ? &TLIP->getTLI() : nullptr;
+ AA = &getAnalysis<AAResultsWrapperPass>().getAAResults();
+ AC = &getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F);
+ LAA = &getAnalysis<LoopAccessAnalysis>();
+ DB = &getAnalysis<DemandedBits>();
+
+ // Compute some weights outside of the loop over the loops. Compute this
+ // using a BranchProbability to re-use its scaling math.
+ const BranchProbability ColdProb(1, 5); // 20%
+ ColdEntryFreq = BlockFrequency(BFI->getEntryFreq()) * ColdProb;
+
+ // Don't attempt if
+ // 1. the target claims to have no vector registers, and
+ // 2. interleaving won't help ILP.
+ //
+ // The second condition is necessary because, even if the target has no
+ // vector registers, loop vectorization may still enable scalar
+ // interleaving.
+ if (!TTI->getNumberOfRegisters(true) && TTI->getMaxInterleaveFactor(1) < 2)
return false;
- SE = &getAnalysis<ScalarEvolution>();
- DL = getAnalysisIfAvailable<DataLayout>();
- LI = &getAnalysis<LoopInfo>();
- TTI = &getAnalysis<TargetTransformInfo>();
- DT = &getAnalysis<DominatorTree>();
- TLI = getAnalysisIfAvailable<TargetLibraryInfo>();
+ // Build up a worklist of inner-loops to vectorize. This is necessary as
+ // the act of vectorizing or partially unrolling a loop creates new loops
+ // and can invalidate iterators across the loops.
+ SmallVector<Loop *, 8> Worklist;
- // If the target claims to have no vector registers don't attempt
- // vectorization.
- if (!TTI->getNumberOfRegisters(true))
- return false;
+ for (Loop *L : *LI)
+ addInnerLoop(*L, Worklist);
- if (DL == NULL) {
- DEBUG(dbgs() << "LV: Not vectorizing because of missing data layout\n");
- return false;
+ LoopsAnalyzed += Worklist.size();
+
+ // Now walk the identified inner loops.
+ bool Changed = false;
+ while (!Worklist.empty())
+ Changed |= processLoop(Worklist.pop_back_val());
+
+ // Process each loop nest in the function.
+ return Changed;
+ }
+
+ static void AddRuntimeUnrollDisableMetaData(Loop *L) {
+ SmallVector<Metadata *, 4> MDs;
+ // Reserve first location for self reference to the LoopID metadata node.
+ MDs.push_back(nullptr);
+ bool IsUnrollMetadata = false;
+ MDNode *LoopID = L->getLoopID();
+ if (LoopID) {
+ // First find existing loop unrolling disable metadata.
+ for (unsigned i = 1, ie = LoopID->getNumOperands(); i < ie; ++i) {
+ MDNode *MD = dyn_cast<MDNode>(LoopID->getOperand(i));
+ if (MD) {
+ const MDString *S = dyn_cast<MDString>(MD->getOperand(0));
+ IsUnrollMetadata =
+ S && S->getString().startswith("llvm.loop.unroll.disable");
+ }
+ MDs.push_back(LoopID->getOperand(i));
+ }
+ }
+
+ if (!IsUnrollMetadata) {
+ // Add runtime unroll disable metadata.
+ LLVMContext &Context = L->getHeader()->getContext();
+ SmallVector<Metadata *, 1> DisableOperands;
+ DisableOperands.push_back(
+ MDString::get(Context, "llvm.loop.unroll.runtime.disable"));
+ MDNode *DisableNode = MDNode::get(Context, DisableOperands);
+ MDs.push_back(DisableNode);
+ MDNode *NewLoopID = MDNode::get(Context, MDs);
+ // Set operand 0 to refer to the loop id itself.
+ NewLoopID->replaceOperandWith(0, NewLoopID);
+ L->setLoopID(NewLoopID);
}
+ }
- DEBUG(dbgs() << "LV: Checking a loop in \"" <<
- L->getHeader()->getParent()->getName() << "\"\n");
+ bool processLoop(Loop *L) {
+ assert(L->empty() && "Only process inner loops.");
+
+#ifndef NDEBUG
+ const std::string DebugLocStr = getDebugLocString(L);
+#endif /* NDEBUG */
+
+ DEBUG(dbgs() << "\nLV: Checking a loop in \""
+ << L->getHeader()->getParent()->getName() << "\" from "
+ << DebugLocStr << "\n");
LoopVectorizeHints Hints(L, DisableUnrolling);
- if (Hints.Width == 1 && Hints.Unroll == 1) {
- DEBUG(dbgs() << "LV: Not vectorizing.\n");
+ DEBUG(dbgs() << "LV: Loop hints:"
+ << " force="
+ << (Hints.getForce() == LoopVectorizeHints::FK_Disabled
+ ? "disabled"
+ : (Hints.getForce() == LoopVectorizeHints::FK_Enabled
+ ? "enabled"
+ : "?")) << " width=" << Hints.getWidth()
+ << " unroll=" << Hints.getInterleave() << "\n");
+
+ // Function containing loop
+ Function *F = L->getHeader()->getParent();
+
+ // Looking at the diagnostic output is the only way to determine if a loop
+ // was vectorized (other than looking at the IR or machine code), so it
+ // is important to generate an optimization remark for each loop. Most of
+ // these messages are generated by emitOptimizationRemarkAnalysis. Remarks
+ // generated by emitOptimizationRemark and emitOptimizationRemarkMissed are
+ // less verbose reporting vectorized loops and unvectorized loops that may
+ // benefit from vectorization, respectively.
+
+ if (!Hints.allowVectorization(F, L, AlwaysVectorize)) {
+ DEBUG(dbgs() << "LV: Loop hints prevent vectorization.\n");
return false;
}
+ // Check the loop for a trip count threshold:
+ // do not vectorize loops with a tiny trip count.
+ const unsigned TC = SE->getSmallConstantTripCount(L);
+ if (TC > 0u && TC < TinyTripCountVectorThreshold) {
+ DEBUG(dbgs() << "LV: Found a loop with a very small trip count. "
+ << "This loop is not worth vectorizing.");
+ if (Hints.getForce() == LoopVectorizeHints::FK_Enabled)
+ DEBUG(dbgs() << " But vectorizing was explicitly forced.\n");
+ else {
+ DEBUG(dbgs() << "\n");
+ emitAnalysisDiag(F, L, Hints, VectorizationReport()
+ << "vectorization is not beneficial "
+ "and is not explicitly forced");
+ return false;
+ }
+ }
+
+ PredicatedScalarEvolution PSE(*SE);
+
// Check if it is legal to vectorize the loop.
- LoopVectorizationLegality LVL(L, SE, DL, DT, TLI);
+ LoopVectorizationRequirements Requirements;
+ LoopVectorizationLegality LVL(L, PSE, DT, TLI, AA, F, TTI, LAA,
+ &Requirements, &Hints);
if (!LVL.canVectorize()) {
- DEBUG(dbgs() << "LV: Not vectorizing.\n");
+ DEBUG(dbgs() << "LV: Not vectorizing: Cannot prove legality.\n");
+ emitMissedWarning(F, L, Hints);
return false;
}
+ // Collect values we want to ignore in the cost model. This includes
+ // type-promoting instructions we identified during reduction detection.
+ SmallPtrSet<const Value *, 32> ValuesToIgnore;
+ CodeMetrics::collectEphemeralValues(L, AC, ValuesToIgnore);
+ for (auto &Reduction : *LVL.getReductionVars()) {
+ RecurrenceDescriptor &RedDes = Reduction.second;
+ SmallPtrSetImpl<Instruction *> &Casts = RedDes.getCastInsts();
+ ValuesToIgnore.insert(Casts.begin(), Casts.end());
+ }
+
// Use the cost model.
- LoopVectorizationCostModel CM(L, SE, LI, &LVL, *TTI, DL, TLI);
+ LoopVectorizationCostModel CM(L, PSE.getSE(), LI, &LVL, *TTI, TLI, DB, AC,
+ F, &Hints, ValuesToIgnore);
// Check the function attributes to find out if this function should be
// optimized for size.
- Function *F = L->getHeader()->getParent();
- Attribute::AttrKind SzAttr = Attribute::OptimizeForSize;
- Attribute::AttrKind FlAttr = Attribute::NoImplicitFloat;
- unsigned FnIndex = AttributeSet::FunctionIndex;
- bool OptForSize = F->getAttributes().hasAttribute(FnIndex, SzAttr);
- bool NoFloat = F->getAttributes().hasAttribute(FnIndex, FlAttr);
+ bool OptForSize = Hints.getForce() != LoopVectorizeHints::FK_Enabled &&
+ F->optForSize();
+
+ // Compute the weighted frequency of this loop being executed and see if it
+ // is less than 20% of the function entry baseline frequency. Note that we
+ // always have a canonical loop here because we think we *can* vectorize.
+ // FIXME: This is hidden behind a flag due to pervasive problems with
+ // exactly what block frequency models.
+ if (LoopVectorizeWithBlockFrequency) {
+ BlockFrequency LoopEntryFreq = BFI->getBlockFreq(L->getLoopPreheader());
+ if (Hints.getForce() != LoopVectorizeHints::FK_Enabled &&
+ LoopEntryFreq < ColdEntryFreq)
+ OptForSize = true;
+ }
- if (NoFloat) {
+ // Check the function attributes to see if implicit floats are allowed.
+ // FIXME: This check doesn't seem possibly correct -- what if the loop is
+ // an integer loop and the vector instructions selected are purely integer
+ // vector instructions?
+ if (F->hasFnAttribute(Attribute::NoImplicitFloat)) {
DEBUG(dbgs() << "LV: Can't vectorize when the NoImplicitFloat"
"attribute is used.\n");
+ emitAnalysisDiag(
+ F, L, Hints,
+ VectorizationReport()
+ << "loop not vectorized due to NoImplicitFloat attribute");
+ emitMissedWarning(F, L, Hints);
return false;
}
// Select the optimal vectorization factor.
- LoopVectorizationCostModel::VectorizationFactor VF;
- VF = CM.selectVectorizationFactor(OptForSize, Hints.Width);
- // Select the unroll factor.
- unsigned UF = CM.selectUnrollFactor(OptForSize, Hints.Unroll, VF.Width,
- VF.Cost);
+ const LoopVectorizationCostModel::VectorizationFactor VF =
+ CM.selectVectorizationFactor(OptForSize);
+
+ // Select the interleave count.
+ unsigned IC = CM.selectInterleaveCount(OptForSize, VF.Width, VF.Cost);
+
+ // Get user interleave count.
+ unsigned UserIC = Hints.getInterleave();
+
+ // Identify the diagnostic messages that should be produced.
+ std::string VecDiagMsg, IntDiagMsg;
+ bool VectorizeLoop = true, InterleaveLoop = true;
+
+ if (Requirements.doesNotMeet(F, L, Hints)) {
+ DEBUG(dbgs() << "LV: Not vectorizing: loop did not meet vectorization "
+ "requirements.\n");
+ emitMissedWarning(F, L, Hints);
+ return false;
+ }
if (VF.Width == 1) {
DEBUG(dbgs() << "LV: Vectorization is possible but not beneficial.\n");
+ VecDiagMsg =
+ "the cost-model indicates that vectorization is not beneficial";
+ VectorizeLoop = false;
}
- DEBUG(dbgs() << "LV: Found a vectorizable loop ("<< VF.Width << ") in "<<
- F->getParent()->getModuleIdentifier() << '\n');
- DEBUG(dbgs() << "LV: Unroll Factor is " << UF << '\n');
+ if (IC == 1 && UserIC <= 1) {
+ // Tell the user interleaving is not beneficial.
+ DEBUG(dbgs() << "LV: Interleaving is not beneficial.\n");
+ IntDiagMsg =
+ "the cost-model indicates that interleaving is not beneficial";
+ InterleaveLoop = false;
+ if (UserIC == 1)
+ IntDiagMsg +=
+ " and is explicitly disabled or interleave count is set to 1";
+ } else if (IC > 1 && UserIC == 1) {
+ // Tell the user interleaving is beneficial, but it explicitly disabled.
+ DEBUG(dbgs()
+ << "LV: Interleaving is beneficial but is explicitly disabled.");
+ IntDiagMsg = "the cost-model indicates that interleaving is beneficial "
+ "but is explicitly disabled or interleave count is set to 1";
+ InterleaveLoop = false;
+ }
- if (VF.Width == 1) {
- if (UF == 1)
- return false;
- // We decided not to vectorize, but we may want to unroll.
- InnerLoopUnroller Unroller(L, SE, LI, DT, DL, TLI, UF);
- Unroller.vectorize(&LVL);
+ // Override IC if user provided an interleave count.
+ IC = UserIC > 0 ? UserIC : IC;
+
+ // Emit diagnostic messages, if any.
+ const char *VAPassName = Hints.vectorizeAnalysisPassName();
+ if (!VectorizeLoop && !InterleaveLoop) {
+ // Do not vectorize or interleaving the loop.
+ emitOptimizationRemarkAnalysis(F->getContext(), VAPassName, *F,
+ L->getStartLoc(), VecDiagMsg);
+ emitOptimizationRemarkAnalysis(F->getContext(), LV_NAME, *F,
+ L->getStartLoc(), IntDiagMsg);
+ return false;
+ } else if (!VectorizeLoop && InterleaveLoop) {
+ DEBUG(dbgs() << "LV: Interleave Count is " << IC << '\n');
+ emitOptimizationRemarkAnalysis(F->getContext(), VAPassName, *F,
+ L->getStartLoc(), VecDiagMsg);
+ } else if (VectorizeLoop && !InterleaveLoop) {
+ DEBUG(dbgs() << "LV: Found a vectorizable loop (" << VF.Width << ") in "
+ << DebugLocStr << '\n');
+ emitOptimizationRemarkAnalysis(F->getContext(), LV_NAME, *F,
+ L->getStartLoc(), IntDiagMsg);
+ } else if (VectorizeLoop && InterleaveLoop) {
+ DEBUG(dbgs() << "LV: Found a vectorizable loop (" << VF.Width << ") in "
+ << DebugLocStr << '\n');
+ DEBUG(dbgs() << "LV: Interleave Count is " << IC << '\n');
+ }
+
+ if (!VectorizeLoop) {
+ assert(IC > 1 && "interleave count should not be 1 or 0");
+ // If we decided that it is not legal to vectorize the loop then
+ // interleave it.
+ InnerLoopUnroller Unroller(L, PSE, LI, DT, TLI, TTI, IC);
+ Unroller.vectorize(&LVL, CM.MinBWs);
+
+ emitOptimizationRemark(F->getContext(), LV_NAME, *F, L->getStartLoc(),
+ Twine("interleaved loop (interleaved count: ") +
+ Twine(IC) + ")");
} else {
// If we decided that it is *legal* to vectorize the loop then do it.
- InnerLoopVectorizer LB(L, SE, LI, DT, DL, TLI, VF.Width, UF);
- LB.vectorize(&LVL);
+ InnerLoopVectorizer LB(L, PSE, LI, DT, TLI, TTI, VF.Width, IC);
+ LB.vectorize(&LVL, CM.MinBWs);
+ ++LoopsVectorized;
+
+ // Add metadata to disable runtime unrolling scalar loop when there's no
+ // runtime check about strides and memory. Because at this situation,
+ // scalar loop is rarely used not worthy to be unrolled.
+ if (!LB.IsSafetyChecksAdded())
+ AddRuntimeUnrollDisableMetaData(L);
+
+ // Report the vectorization decision.
+ emitOptimizationRemark(F->getContext(), LV_NAME, *F, L->getStartLoc(),
+ Twine("vectorized loop (vectorization width: ") +
+ Twine(VF.Width) + ", interleaved count: " +
+ Twine(IC) + ")");
}
// Mark the loop as already vectorized to avoid vectorizing again.
- Hints.setAlreadyVectorized(L);
+ Hints.setAlreadyVectorized();
DEBUG(verifyFunction(*L->getHeader()->getParent()));
return true;
}
- virtual void getAnalysisUsage(AnalysisUsage &AU) const {
- LoopPass::getAnalysisUsage(AU);
+ void getAnalysisUsage(AnalysisUsage &AU) const override {
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequiredID(LoopSimplifyID);
AU.addRequiredID(LCSSAID);
- AU.addRequired<DominatorTree>();
- AU.addRequired<LoopInfo>();
- AU.addRequired<ScalarEvolution>();
- AU.addRequired<TargetTransformInfo>();
- AU.addPreserved<LoopInfo>();
- AU.addPreserved<DominatorTree>();
+ AU.addRequired<BlockFrequencyInfoWrapperPass>();
+ AU.addRequired<DominatorTreeWrapperPass>();
+ AU.addRequired<LoopInfoWrapperPass>();
+ AU.addRequired<ScalarEvolutionWrapperPass>();
+ AU.addRequired<TargetTransformInfoWrapperPass>();
+ AU.addRequired<AAResultsWrapperPass>();
+ AU.addRequired<LoopAccessAnalysis>();
+ AU.addRequired<DemandedBits>();
+ AU.addPreserved<LoopInfoWrapperPass>();
+ AU.addPreserved<DominatorTreeWrapperPass>();
+ AU.addPreserved<BasicAAWrapperPass>();
+ AU.addPreserved<AAResultsWrapperPass>();
+ AU.addPreserved<GlobalsAAWrapperPass>();
}
};
// LoopVectorizationCostModel.
//===----------------------------------------------------------------------===//
-void
-LoopVectorizationLegality::RuntimePointerCheck::insert(ScalarEvolution *SE,
- Loop *Lp, Value *Ptr,
- bool WritePtr,
- unsigned DepSetId) {
- const SCEV *Sc = SE->getSCEV(Ptr);
- const SCEVAddRecExpr *AR = dyn_cast<SCEVAddRecExpr>(Sc);
- assert(AR && "Invalid addrec expression");
- const SCEV *Ex = SE->getBackedgeTakenCount(Lp);
- const SCEV *ScEnd = AR->evaluateAtIteration(Ex, *SE);
- Pointers.push_back(Ptr);
- Starts.push_back(AR->getStart());
- Ends.push_back(ScEnd);
- IsWritePtr.push_back(WritePtr);
- DependencySetId.push_back(DepSetId);
-}
-
Value *InnerLoopVectorizer::getBroadcastInstrs(Value *V) {
// We need to place the broadcast of invariant variables outside the loop.
Instruction *Instr = dyn_cast<Instruction>(V);
- bool NewInstr = (Instr && Instr->getParent() == LoopVectorBody);
+ bool NewInstr =
+ (Instr && std::find(LoopVectorBody.begin(), LoopVectorBody.end(),
+ Instr->getParent()) != LoopVectorBody.end());
bool Invariant = OrigLoop->isLoopInvariant(V) && !NewInstr;
// Place the code for broadcasting invariant variables in the new preheader.
return Shuf;
}
-Value *InnerLoopVectorizer::getConsecutiveVector(Value* Val, int StartIdx,
- bool Negate) {
+Value *InnerLoopVectorizer::getStepVector(Value *Val, int StartIdx,
+ Value *Step) {
assert(Val->getType()->isVectorTy() && "Must be a vector");
assert(Val->getType()->getScalarType()->isIntegerTy() &&
"Elem must be an integer");
+ assert(Step->getType() == Val->getType()->getScalarType() &&
+ "Step has wrong type");
// Create the types.
Type *ITy = Val->getType()->getScalarType();
VectorType *Ty = cast<VectorType>(Val->getType());
SmallVector<Constant*, 8> Indices;
// Create a vector of consecutive numbers from zero to VF.
- for (int i = 0; i < VLen; ++i) {
- int64_t Idx = Negate ? (-i) : i;
- Indices.push_back(ConstantInt::get(ITy, StartIdx + Idx, Negate));
- }
+ for (int i = 0; i < VLen; ++i)
+ Indices.push_back(ConstantInt::get(ITy, StartIdx + i));
// Add the consecutive indices to the vector value.
Constant *Cv = ConstantVector::get(Indices);
assert(Cv->getType() == Val->getType() && "Invalid consecutive vec");
- return Builder.CreateAdd(Val, Cv, "induction");
+ Step = Builder.CreateVectorSplat(VLen, Step);
+ assert(Step->getType() == Val->getType() && "Invalid step vec");
+ // FIXME: The newly created binary instructions should contain nsw/nuw flags,
+ // which can be found from the original scalar operations.
+ Step = Builder.CreateMul(Cv, Step);
+ return Builder.CreateAdd(Val, Step, "induction");
}
int LoopVectorizationLegality::isConsecutivePtr(Value *Ptr) {
- assert(Ptr->getType()->isPointerTy() && "Unexpected non ptr");
+ assert(Ptr->getType()->isPointerTy() && "Unexpected non-ptr");
+ auto *SE = PSE.getSE();
// Make sure that the pointer does not point to structs.
- if (cast<PointerType>(Ptr->getType())->getElementType()->isAggregateType())
+ if (Ptr->getType()->getPointerElementType()->isAggregateType())
return 0;
// If this value is a pointer induction variable we know it is consecutive.
PHINode *Phi = dyn_cast_or_null<PHINode>(Ptr);
if (Phi && Inductions.count(Phi)) {
- InductionInfo II = Inductions[Phi];
- if (IK_PtrInduction == II.IK)
- return 1;
- else if (IK_ReversePtrInduction == II.IK)
- return -1;
+ InductionDescriptor II = Inductions[Phi];
+ return II.getConsecutiveDirection();
}
- GetElementPtrInst *Gep = dyn_cast_or_null<GetElementPtrInst>(Ptr);
+ GetElementPtrInst *Gep = getGEPInstruction(Ptr);
if (!Gep)
return 0;
unsigned NumOperands = Gep->getNumOperands();
- Value *LastIndex = Gep->getOperand(NumOperands - 1);
-
Value *GpPtr = Gep->getPointerOperand();
// If this GEP value is a consecutive pointer induction variable and all of
// the indices are constant then we know it is consecutive. We can
// Make sure that all of the index operands are loop invariant.
for (unsigned i = 1; i < NumOperands; ++i)
- if (!SE->isLoopInvariant(SE->getSCEV(Gep->getOperand(i)), TheLoop))
+ if (!SE->isLoopInvariant(PSE.getSCEV(Gep->getOperand(i)), TheLoop))
return 0;
- InductionInfo II = Inductions[Phi];
- if (IK_PtrInduction == II.IK)
- return 1;
- else if (IK_ReversePtrInduction == II.IK)
- return -1;
+ InductionDescriptor II = Inductions[Phi];
+ return II.getConsecutiveDirection();
}
- // Check that all of the gep indices are uniform except for the last.
- for (unsigned i = 0; i < NumOperands - 1; ++i)
- if (!SE->isLoopInvariant(SE->getSCEV(Gep->getOperand(i)), TheLoop))
+ unsigned InductionOperand = getGEPInductionOperand(Gep);
+
+ // Check that all of the gep indices are uniform except for our induction
+ // operand.
+ for (unsigned i = 0; i != NumOperands; ++i)
+ if (i != InductionOperand &&
+ !SE->isLoopInvariant(PSE.getSCEV(Gep->getOperand(i)), TheLoop))
return 0;
- // We can emit wide load/stores only if the last index is the induction
- // variable.
- const SCEV *Last = SE->getSCEV(LastIndex);
+ // We can emit wide load/stores only if the last non-zero index is the
+ // induction variable.
+ const SCEV *Last = nullptr;
+ if (!Strides.count(Gep))
+ Last = PSE.getSCEV(Gep->getOperand(InductionOperand));
+ else {
+ // Because of the multiplication by a stride we can have a s/zext cast.
+ // We are going to replace this stride by 1 so the cast is safe to ignore.
+ //
+ // %indvars.iv = phi i64 [ 0, %entry ], [ %indvars.iv.next, %for.body ]
+ // %0 = trunc i64 %indvars.iv to i32
+ // %mul = mul i32 %0, %Stride1
+ // %idxprom = zext i32 %mul to i64 << Safe cast.
+ // %arrayidx = getelementptr inbounds i32* %B, i64 %idxprom
+ //
+ Last = replaceSymbolicStrideSCEV(PSE, Strides,
+ Gep->getOperand(InductionOperand), Gep);
+ if (const SCEVCastExpr *C = dyn_cast<SCEVCastExpr>(Last))
+ Last =
+ (C->getSCEVType() == scSignExtend || C->getSCEVType() == scZeroExtend)
+ ? C->getOperand()
+ : Last;
+ }
if (const SCEVAddRecExpr *AR = dyn_cast<SCEVAddRecExpr>(Last)) {
const SCEV *Step = AR->getStepRecurrence(*SE);
}
bool LoopVectorizationLegality::isUniform(Value *V) {
- return (SE->isLoopInvariant(SE->getSCEV(V), TheLoop));
+ return LAI->isUniform(V);
}
InnerLoopVectorizer::VectorParts&
assert(V != Induction && "The new induction variable should not be used.");
assert(!V->getType()->isVectorTy() && "Can't widen a vector");
+ // If we have a stride that is replaced by one, do it here.
+ if (Legal->hasStride(V))
+ V = ConstantInt::get(V->getType(), 1);
+
// If we have this scalar in the map, return it.
if (WidenMap.has(V))
return WidenMap.get(V);
"reverse");
}
+// Get a mask to interleave \p NumVec vectors into a wide vector.
+// I.e. <0, VF, VF*2, ..., VF*(NumVec-1), 1, VF+1, VF*2+1, ...>
+// E.g. For 2 interleaved vectors, if VF is 4, the mask is:
+// <0, 4, 1, 5, 2, 6, 3, 7>
+static Constant *getInterleavedMask(IRBuilder<> &Builder, unsigned VF,
+ unsigned NumVec) {
+ SmallVector<Constant *, 16> Mask;
+ for (unsigned i = 0; i < VF; i++)
+ for (unsigned j = 0; j < NumVec; j++)
+ Mask.push_back(Builder.getInt32(j * VF + i));
+
+ return ConstantVector::get(Mask);
+}
+
+// Get the strided mask starting from index \p Start.
+// I.e. <Start, Start + Stride, ..., Start + Stride*(VF-1)>
+static Constant *getStridedMask(IRBuilder<> &Builder, unsigned Start,
+ unsigned Stride, unsigned VF) {
+ SmallVector<Constant *, 16> Mask;
+ for (unsigned i = 0; i < VF; i++)
+ Mask.push_back(Builder.getInt32(Start + i * Stride));
+
+ return ConstantVector::get(Mask);
+}
+
+// Get a mask of two parts: The first part consists of sequential integers
+// starting from 0, The second part consists of UNDEFs.
+// I.e. <0, 1, 2, ..., NumInt - 1, undef, ..., undef>
+static Constant *getSequentialMask(IRBuilder<> &Builder, unsigned NumInt,
+ unsigned NumUndef) {
+ SmallVector<Constant *, 16> Mask;
+ for (unsigned i = 0; i < NumInt; i++)
+ Mask.push_back(Builder.getInt32(i));
+
+ Constant *Undef = UndefValue::get(Builder.getInt32Ty());
+ for (unsigned i = 0; i < NumUndef; i++)
+ Mask.push_back(Undef);
+
+ return ConstantVector::get(Mask);
+}
+
+// Concatenate two vectors with the same element type. The 2nd vector should
+// not have more elements than the 1st vector. If the 2nd vector has less
+// elements, extend it with UNDEFs.
+static Value *ConcatenateTwoVectors(IRBuilder<> &Builder, Value *V1,
+ Value *V2) {
+ VectorType *VecTy1 = dyn_cast<VectorType>(V1->getType());
+ VectorType *VecTy2 = dyn_cast<VectorType>(V2->getType());
+ assert(VecTy1 && VecTy2 &&
+ VecTy1->getScalarType() == VecTy2->getScalarType() &&
+ "Expect two vectors with the same element type");
+
+ unsigned NumElts1 = VecTy1->getNumElements();
+ unsigned NumElts2 = VecTy2->getNumElements();
+ assert(NumElts1 >= NumElts2 && "Unexpect the first vector has less elements");
+
+ if (NumElts1 > NumElts2) {
+ // Extend with UNDEFs.
+ Constant *ExtMask =
+ getSequentialMask(Builder, NumElts2, NumElts1 - NumElts2);
+ V2 = Builder.CreateShuffleVector(V2, UndefValue::get(VecTy2), ExtMask);
+ }
+
+ Constant *Mask = getSequentialMask(Builder, NumElts1 + NumElts2, 0);
+ return Builder.CreateShuffleVector(V1, V2, Mask);
+}
+
+// Concatenate vectors in the given list. All vectors have the same type.
+static Value *ConcatenateVectors(IRBuilder<> &Builder,
+ ArrayRef<Value *> InputList) {
+ unsigned NumVec = InputList.size();
+ assert(NumVec > 1 && "Should be at least two vectors");
+
+ SmallVector<Value *, 8> ResList;
+ ResList.append(InputList.begin(), InputList.end());
+ do {
+ SmallVector<Value *, 8> TmpList;
+ for (unsigned i = 0; i < NumVec - 1; i += 2) {
+ Value *V0 = ResList[i], *V1 = ResList[i + 1];
+ assert((V0->getType() == V1->getType() || i == NumVec - 2) &&
+ "Only the last vector may have a different type");
+
+ TmpList.push_back(ConcatenateTwoVectors(Builder, V0, V1));
+ }
+
+ // Push the last vector if the total number of vectors is odd.
+ if (NumVec % 2 != 0)
+ TmpList.push_back(ResList[NumVec - 1]);
+
+ ResList = TmpList;
+ NumVec = ResList.size();
+ } while (NumVec > 1);
+
+ return ResList[0];
+}
+
+// Try to vectorize the interleave group that \p Instr belongs to.
+//
+// E.g. Translate following interleaved load group (factor = 3):
+// for (i = 0; i < N; i+=3) {
+// R = Pic[i]; // Member of index 0
+// G = Pic[i+1]; // Member of index 1
+// B = Pic[i+2]; // Member of index 2
+// ... // do something to R, G, B
+// }
+// To:
+// %wide.vec = load <12 x i32> ; Read 4 tuples of R,G,B
+// %R.vec = shuffle %wide.vec, undef, <0, 3, 6, 9> ; R elements
+// %G.vec = shuffle %wide.vec, undef, <1, 4, 7, 10> ; G elements
+// %B.vec = shuffle %wide.vec, undef, <2, 5, 8, 11> ; B elements
+//
+// Or translate following interleaved store group (factor = 3):
+// for (i = 0; i < N; i+=3) {
+// ... do something to R, G, B
+// Pic[i] = R; // Member of index 0
+// Pic[i+1] = G; // Member of index 1
+// Pic[i+2] = B; // Member of index 2
+// }
+// To:
+// %R_G.vec = shuffle %R.vec, %G.vec, <0, 1, 2, ..., 7>
+// %B_U.vec = shuffle %B.vec, undef, <0, 1, 2, 3, u, u, u, u>
+// %interleaved.vec = shuffle %R_G.vec, %B_U.vec,
+// <0, 4, 8, 1, 5, 9, 2, 6, 10, 3, 7, 11> ; Interleave R,G,B elements
+// store <12 x i32> %interleaved.vec ; Write 4 tuples of R,G,B
+void InnerLoopVectorizer::vectorizeInterleaveGroup(Instruction *Instr) {
+ const InterleaveGroup *Group = Legal->getInterleavedAccessGroup(Instr);
+ assert(Group && "Fail to get an interleaved access group.");
+
+ // Skip if current instruction is not the insert position.
+ if (Instr != Group->getInsertPos())
+ return;
+
+ LoadInst *LI = dyn_cast<LoadInst>(Instr);
+ StoreInst *SI = dyn_cast<StoreInst>(Instr);
+ Value *Ptr = LI ? LI->getPointerOperand() : SI->getPointerOperand();
+
+ // Prepare for the vector type of the interleaved load/store.
+ Type *ScalarTy = LI ? LI->getType() : SI->getValueOperand()->getType();
+ unsigned InterleaveFactor = Group->getFactor();
+ Type *VecTy = VectorType::get(ScalarTy, InterleaveFactor * VF);
+ Type *PtrTy = VecTy->getPointerTo(Ptr->getType()->getPointerAddressSpace());
+
+ // Prepare for the new pointers.
+ setDebugLocFromInst(Builder, Ptr);
+ VectorParts &PtrParts = getVectorValue(Ptr);
+ SmallVector<Value *, 2> NewPtrs;
+ unsigned Index = Group->getIndex(Instr);
+ for (unsigned Part = 0; Part < UF; Part++) {
+ // Extract the pointer for current instruction from the pointer vector. A
+ // reverse access uses the pointer in the last lane.
+ Value *NewPtr = Builder.CreateExtractElement(
+ PtrParts[Part],
+ Group->isReverse() ? Builder.getInt32(VF - 1) : Builder.getInt32(0));
+
+ // Notice current instruction could be any index. Need to adjust the address
+ // to the member of index 0.
+ //
+ // E.g. a = A[i+1]; // Member of index 1 (Current instruction)
+ // b = A[i]; // Member of index 0
+ // Current pointer is pointed to A[i+1], adjust it to A[i].
+ //
+ // E.g. A[i+1] = a; // Member of index 1
+ // A[i] = b; // Member of index 0
+ // A[i+2] = c; // Member of index 2 (Current instruction)
+ // Current pointer is pointed to A[i+2], adjust it to A[i].
+ NewPtr = Builder.CreateGEP(NewPtr, Builder.getInt32(-Index));
+
+ // Cast to the vector pointer type.
+ NewPtrs.push_back(Builder.CreateBitCast(NewPtr, PtrTy));
+ }
+
+ setDebugLocFromInst(Builder, Instr);
+ Value *UndefVec = UndefValue::get(VecTy);
+
+ // Vectorize the interleaved load group.
+ if (LI) {
+ for (unsigned Part = 0; Part < UF; Part++) {
+ Instruction *NewLoadInstr = Builder.CreateAlignedLoad(
+ NewPtrs[Part], Group->getAlignment(), "wide.vec");
+
+ for (unsigned i = 0; i < InterleaveFactor; i++) {
+ Instruction *Member = Group->getMember(i);
+
+ // Skip the gaps in the group.
+ if (!Member)
+ continue;
+
+ Constant *StrideMask = getStridedMask(Builder, i, InterleaveFactor, VF);
+ Value *StridedVec = Builder.CreateShuffleVector(
+ NewLoadInstr, UndefVec, StrideMask, "strided.vec");
+
+ // If this member has different type, cast the result type.
+ if (Member->getType() != ScalarTy) {
+ VectorType *OtherVTy = VectorType::get(Member->getType(), VF);
+ StridedVec = Builder.CreateBitOrPointerCast(StridedVec, OtherVTy);
+ }
+
+ VectorParts &Entry = WidenMap.get(Member);
+ Entry[Part] =
+ Group->isReverse() ? reverseVector(StridedVec) : StridedVec;
+ }
+
+ propagateMetadata(NewLoadInstr, Instr);
+ }
+ return;
+ }
+
+ // The sub vector type for current instruction.
+ VectorType *SubVT = VectorType::get(ScalarTy, VF);
+
+ // Vectorize the interleaved store group.
+ for (unsigned Part = 0; Part < UF; Part++) {
+ // Collect the stored vector from each member.
+ SmallVector<Value *, 4> StoredVecs;
+ for (unsigned i = 0; i < InterleaveFactor; i++) {
+ // Interleaved store group doesn't allow a gap, so each index has a member
+ Instruction *Member = Group->getMember(i);
+ assert(Member && "Fail to get a member from an interleaved store group");
+
+ Value *StoredVec =
+ getVectorValue(dyn_cast<StoreInst>(Member)->getValueOperand())[Part];
+ if (Group->isReverse())
+ StoredVec = reverseVector(StoredVec);
+
+ // If this member has different type, cast it to an unified type.
+ if (StoredVec->getType() != SubVT)
+ StoredVec = Builder.CreateBitOrPointerCast(StoredVec, SubVT);
+
+ StoredVecs.push_back(StoredVec);
+ }
+
+ // Concatenate all vectors into a wide vector.
+ Value *WideVec = ConcatenateVectors(Builder, StoredVecs);
+
+ // Interleave the elements in the wide vector.
+ Constant *IMask = getInterleavedMask(Builder, VF, InterleaveFactor);
+ Value *IVec = Builder.CreateShuffleVector(WideVec, UndefVec, IMask,
+ "interleaved.vec");
+
+ Instruction *NewStoreInstr =
+ Builder.CreateAlignedStore(IVec, NewPtrs[Part], Group->getAlignment());
+ propagateMetadata(NewStoreInstr, Instr);
+ }
+}
-void InnerLoopVectorizer::vectorizeMemoryInstruction(Instruction *Instr,
- LoopVectorizationLegality *Legal) {
+void InnerLoopVectorizer::vectorizeMemoryInstruction(Instruction *Instr) {
// Attempt to issue a wide load.
LoadInst *LI = dyn_cast<LoadInst>(Instr);
StoreInst *SI = dyn_cast<StoreInst>(Instr);
assert((LI || SI) && "Invalid Load/Store instruction");
+ // Try to vectorize the interleave group if this access is interleaved.
+ if (Legal->isAccessInterleaved(Instr))
+ return vectorizeInterleaveGroup(Instr);
+
Type *ScalarDataTy = LI ? LI->getType() : SI->getValueOperand()->getType();
Type *DataTy = VectorType::get(ScalarDataTy, VF);
Value *Ptr = LI ? LI->getPointerOperand() : SI->getPointerOperand();
unsigned Alignment = LI ? LI->getAlignment() : SI->getAlignment();
+ // An alignment of 0 means target abi alignment. We need to use the scalar's
+ // target abi alignment in such a case.
+ const DataLayout &DL = Instr->getModule()->getDataLayout();
+ if (!Alignment)
+ Alignment = DL.getABITypeAlignment(ScalarDataTy);
unsigned AddressSpace = Ptr->getType()->getPointerAddressSpace();
- unsigned ScalarAllocatedSize = DL->getTypeAllocSize(ScalarDataTy);
- unsigned VectorElementSize = DL->getTypeStoreSize(DataTy)/VF;
+ unsigned ScalarAllocatedSize = DL.getTypeAllocSize(ScalarDataTy);
+ unsigned VectorElementSize = DL.getTypeStoreSize(DataTy) / VF;
+
+ if (SI && Legal->blockNeedsPredication(SI->getParent()) &&
+ !Legal->isMaskRequired(SI))
+ return scalarizeInstruction(Instr, true);
if (ScalarAllocatedSize != VectorElementSize)
return scalarizeInstruction(Instr);
- // If the pointer is loop invariant or if it is non consecutive,
+ // If the pointer is loop invariant or if it is non-consecutive,
// scalarize the load.
int ConsecutiveStride = Legal->isConsecutivePtr(Ptr);
bool Reverse = ConsecutiveStride < 0;
VectorParts &Entry = WidenMap.get(Instr);
// Handle consecutive loads/stores.
- GetElementPtrInst *Gep = dyn_cast<GetElementPtrInst>(Ptr);
+ GetElementPtrInst *Gep = getGEPInstruction(Ptr);
if (Gep && Legal->isInductionVariable(Gep->getPointerOperand())) {
setDebugLocFromInst(Builder, Gep);
Value *PtrOperand = Gep->getPointerOperand();
Ptr = Builder.Insert(Gep2);
} else if (Gep) {
setDebugLocFromInst(Builder, Gep);
- assert(SE->isLoopInvariant(SE->getSCEV(Gep->getPointerOperand()),
- OrigLoop) && "Base ptr must be invariant");
+ assert(PSE.getSE()->isLoopInvariant(PSE.getSCEV(Gep->getPointerOperand()),
+ OrigLoop) &&
+ "Base ptr must be invariant");
// The last index does not have to be the induction. It can be
// consecutive and be a function of the index. For example A[I+1];
unsigned NumOperands = Gep->getNumOperands();
- unsigned LastOperand = NumOperands - 1;
+ unsigned InductionOperand = getGEPInductionOperand(Gep);
// Create the new GEP with the new induction variable.
GetElementPtrInst *Gep2 = cast<GetElementPtrInst>(Gep->clone());
Instruction *GepOperandInst = dyn_cast<Instruction>(GepOperand);
// Update last index or loop invariant instruction anchored in loop.
- if (i == LastOperand ||
+ if (i == InductionOperand ||
(GepOperandInst && OrigLoop->contains(GepOperandInst))) {
- assert((i == LastOperand ||
- SE->isLoopInvariant(SE->getSCEV(GepOperandInst), OrigLoop)) &&
+ assert((i == InductionOperand ||
+ PSE.getSE()->isLoopInvariant(PSE.getSCEV(GepOperandInst),
+ OrigLoop)) &&
"Must be last index or loop invariant");
VectorParts &GEPParts = getVectorValue(GepOperand);
Ptr = Builder.CreateExtractElement(PtrVal[0], Zero);
}
+ VectorParts Mask = createBlockInMask(Instr->getParent());
// Handle Stores:
if (SI) {
assert(!Legal->isUniform(SI->getPointerOperand()) &&
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 *PartPtr =
+ Builder.CreateGEP(nullptr, Ptr, Builder.getInt32(Part * VF));
if (Reverse) {
- // If we store to reverse consecutive memory locations then we need
+ // If we store to reverse consecutive memory locations, then we need
// to reverse the order of elements in the stored value.
StoredVal[Part] = reverseVector(StoredVal[Part]);
// If the address is consecutive but reversed, then the
// wide store needs to start at the last vector element.
- PartPtr = Builder.CreateGEP(Ptr, Builder.getInt32(-Part * VF));
- PartPtr = Builder.CreateGEP(PartPtr, Builder.getInt32(1 - VF));
+ PartPtr = Builder.CreateGEP(nullptr, Ptr, Builder.getInt32(-Part * VF));
+ PartPtr = Builder.CreateGEP(nullptr, PartPtr, Builder.getInt32(1 - VF));
+ Mask[Part] = reverseVector(Mask[Part]);
}
Value *VecPtr = Builder.CreateBitCast(PartPtr,
DataTy->getPointerTo(AddressSpace));
- Builder.CreateStore(StoredVal[Part], VecPtr)->setAlignment(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;
}
setDebugLocFromInst(Builder, LI);
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 *PartPtr =
+ Builder.CreateGEP(nullptr, Ptr, Builder.getInt32(Part * VF));
if (Reverse) {
// If the address is consecutive but reversed, then the
- // wide store needs to start at the last vector element.
- PartPtr = Builder.CreateGEP(Ptr, Builder.getInt32(-Part * VF));
- PartPtr = Builder.CreateGEP(PartPtr, Builder.getInt32(1 - VF));
+ // wide load needs to start at the last vector element.
+ PartPtr = Builder.CreateGEP(nullptr, Ptr, Builder.getInt32(-Part * VF));
+ PartPtr = Builder.CreateGEP(nullptr, PartPtr, Builder.getInt32(1 - VF));
+ Mask[Part] = reverseVector(Mask[Part]);
}
+ Instruction* NewLI;
Value *VecPtr = Builder.CreateBitCast(PartPtr,
DataTy->getPointerTo(AddressSpace));
- Value *LI = Builder.CreateLoad(VecPtr, "wide.load");
- cast<LoadInst>(LI)->setAlignment(Alignment);
- Entry[Part] = Reverse ? reverseVector(LI) : LI;
+ 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;
}
}
-void InnerLoopVectorizer::scalarizeInstruction(Instruction *Instr) {
+void InnerLoopVectorizer::scalarizeInstruction(Instruction *Instr,
+ bool IfPredicateStore) {
assert(!Instr->getType()->isAggregateType() && "Can't handle vectors");
// Holds vector parameters or scalars, in case of uniform vals.
SmallVector<VectorParts, 4> Params;
// Try using previously calculated values.
Instruction *SrcInst = dyn_cast<Instruction>(SrcOp);
- // If the src is an instruction that appeared earlier in the basic block
+ // If the src is an instruction that appeared earlier in the basic block,
// then it should already be vectorized.
if (SrcInst && OrigLoop->contains(SrcInst)) {
assert(WidenMap.has(SrcInst) && "Source operand is unavailable");
// Does this instruction return a value ?
bool IsVoidRetTy = Instr->getType()->isVoidTy();
- Value *UndefVec = IsVoidRetTy ? 0 :
+ Value *UndefVec = IsVoidRetTy ? nullptr :
UndefValue::get(VectorType::get(Instr->getType(), VF));
// Create a new entry in the WidenMap and initialize it to Undef or Null.
VectorParts &VecResults = WidenMap.splat(Instr, UndefVec);
+ VectorParts Cond;
+ if (IfPredicateStore) {
+ assert(Instr->getParent()->getSinglePredecessor() &&
+ "Only support single predecessor blocks");
+ Cond = createEdgeMask(Instr->getParent()->getSinglePredecessor(),
+ Instr->getParent());
+ }
+
// For each vector unroll 'part':
for (unsigned Part = 0; Part < UF; ++Part) {
// For each scalar that we create:
for (unsigned Width = 0; Width < VF; ++Width) {
+
+ // Start if-block.
+ Value *Cmp = nullptr;
+ if (IfPredicateStore) {
+ Cmp = Builder.CreateExtractElement(Cond[Part], Builder.getInt32(Width));
+ Cmp = Builder.CreateICmp(ICmpInst::ICMP_EQ, Cmp,
+ ConstantInt::get(Cmp->getType(), 1));
+ }
+
Instruction *Cloned = Instr->clone();
if (!IsVoidRetTy)
Cloned->setName(Instr->getName() + ".cloned");
if (!IsVoidRetTy)
VecResults[Part] = Builder.CreateInsertElement(VecResults[Part], Cloned,
Builder.getInt32(Width));
+ // End if-block.
+ if (IfPredicateStore)
+ PredicatedStores.push_back(std::make_pair(cast<StoreInst>(Cloned),
+ Cmp));
}
}
}
-Instruction *
-InnerLoopVectorizer::addRuntimeCheck(LoopVectorizationLegality *Legal,
- Instruction *Loc) {
- LoopVectorizationLegality::RuntimePointerCheck *PtrRtCheck =
- Legal->getRuntimePointerCheck();
+PHINode *InnerLoopVectorizer::createInductionVariable(Loop *L, Value *Start,
+ Value *End, Value *Step,
+ Instruction *DL) {
+ BasicBlock *Header = L->getHeader();
+ BasicBlock *Latch = L->getLoopLatch();
+ // As we're just creating this loop, it's possible no latch exists
+ // yet. If so, use the header as this will be a single block loop.
+ if (!Latch)
+ Latch = Header;
+
+ IRBuilder<> Builder(&*Header->getFirstInsertionPt());
+ setDebugLocFromInst(Builder, getDebugLocFromInstOrOperands(OldInduction));
+ auto *Induction = Builder.CreatePHI(Start->getType(), 2, "index");
+
+ Builder.SetInsertPoint(Latch->getTerminator());
+
+ // Create i+1 and fill the PHINode.
+ Value *Next = Builder.CreateAdd(Induction, Step, "index.next");
+ Induction->addIncoming(Start, L->getLoopPreheader());
+ Induction->addIncoming(Next, Latch);
+ // Create the compare.
+ Value *ICmp = Builder.CreateICmpEQ(Next, End);
+ Builder.CreateCondBr(ICmp, L->getExitBlock(), Header);
+
+ // Now we have two terminators. Remove the old one from the block.
+ Latch->getTerminator()->eraseFromParent();
+
+ return Induction;
+}
+
+Value *InnerLoopVectorizer::getOrCreateTripCount(Loop *L) {
+ if (TripCount)
+ return TripCount;
- if (!PtrRtCheck->Need)
- return NULL;
+ IRBuilder<> Builder(L->getLoopPreheader()->getTerminator());
+ // Find the loop boundaries.
+ ScalarEvolution *SE = PSE.getSE();
+ const SCEV *BackedgeTakenCount = SE->getBackedgeTakenCount(OrigLoop);
+ assert(BackedgeTakenCount != SE->getCouldNotCompute() &&
+ "Invalid loop count");
- unsigned NumPointers = PtrRtCheck->Pointers.size();
- SmallVector<TrackingVH<Value> , 2> Starts;
- SmallVector<TrackingVH<Value> , 2> Ends;
+ Type *IdxTy = Legal->getWidestInductionType();
+
+ // The exit count might have the type of i64 while the phi is i32. This can
+ // happen if we have an induction variable that is sign extended before the
+ // compare. The only way that we get a backedge taken count is that the
+ // induction variable was signed and as such will not overflow. In such a case
+ // truncation is legal.
+ if (BackedgeTakenCount->getType()->getPrimitiveSizeInBits() >
+ IdxTy->getPrimitiveSizeInBits())
+ BackedgeTakenCount = SE->getTruncateOrNoop(BackedgeTakenCount, IdxTy);
+ BackedgeTakenCount = SE->getNoopOrZeroExtend(BackedgeTakenCount, IdxTy);
+
+ // Get the total trip count from the count by adding 1.
+ const SCEV *ExitCount = SE->getAddExpr(
+ BackedgeTakenCount, SE->getOne(BackedgeTakenCount->getType()));
- LLVMContext &Ctx = Loc->getContext();
- SCEVExpander Exp(*SE, "induction");
+ const DataLayout &DL = L->getHeader()->getModule()->getDataLayout();
- for (unsigned i = 0; i < NumPointers; ++i) {
- Value *Ptr = PtrRtCheck->Pointers[i];
- const SCEV *Sc = SE->getSCEV(Ptr);
+ // Expand the trip count and place the new instructions in the preheader.
+ // Notice that the pre-header does not change, only the loop body.
+ SCEVExpander Exp(*SE, DL, "induction");
- if (SE->isLoopInvariant(Sc, OrigLoop)) {
- DEBUG(dbgs() << "LV: Adding RT check for a loop invariant ptr:" <<
- *Ptr <<"\n");
- Starts.push_back(Ptr);
- Ends.push_back(Ptr);
- } else {
- DEBUG(dbgs() << "LV: Adding RT check for range:" << *Ptr << '\n');
- unsigned AS = Ptr->getType()->getPointerAddressSpace();
+ // Count holds the overall loop count (N).
+ TripCount = Exp.expandCodeFor(ExitCount, ExitCount->getType(),
+ L->getLoopPreheader()->getTerminator());
- // Use this type for pointer arithmetic.
- Type *PtrArithTy = Type::getInt8PtrTy(Ctx, AS);
+ if (TripCount->getType()->isPointerTy())
+ TripCount =
+ CastInst::CreatePointerCast(TripCount, IdxTy,
+ "exitcount.ptrcnt.to.int",
+ L->getLoopPreheader()->getTerminator());
- Value *Start = Exp.expandCodeFor(PtrRtCheck->Starts[i], PtrArithTy, Loc);
- Value *End = Exp.expandCodeFor(PtrRtCheck->Ends[i], PtrArithTy, Loc);
- Starts.push_back(Start);
- Ends.push_back(End);
- }
- }
+ return TripCount;
+}
- IRBuilder<> ChkBuilder(Loc);
- // Our instructions might fold to a constant.
- Value *MemoryRuntimeCheck = 0;
- for (unsigned i = 0; i < NumPointers; ++i) {
- for (unsigned j = i+1; j < NumPointers; ++j) {
- // No need to check if two readonly pointers intersect.
- if (!PtrRtCheck->IsWritePtr[i] && !PtrRtCheck->IsWritePtr[j])
- continue;
+Value *InnerLoopVectorizer::getOrCreateVectorTripCount(Loop *L) {
+ if (VectorTripCount)
+ return VectorTripCount;
+
+ Value *TC = getOrCreateTripCount(L);
+ IRBuilder<> Builder(L->getLoopPreheader()->getTerminator());
+
+ // Now we need to generate the expression for N - (N % VF), which is
+ // the part that the vectorized body will execute.
+ // The loop step is equal to the vectorization factor (num of SIMD elements)
+ // times the unroll factor (num of SIMD instructions).
+ Constant *Step = ConstantInt::get(TC->getType(), VF * UF);
+ Value *R = Builder.CreateURem(TC, Step, "n.mod.vf");
+ VectorTripCount = Builder.CreateSub(TC, R, "n.vec");
+
+ return VectorTripCount;
+}
+
+void InnerLoopVectorizer::emitMinimumIterationCountCheck(Loop *L,
+ BasicBlock *Bypass) {
+ Value *Count = getOrCreateTripCount(L);
+ BasicBlock *BB = L->getLoopPreheader();
+ IRBuilder<> Builder(BB->getTerminator());
+
+ // Generate code to check that the loop's trip count that we computed by
+ // adding one to the backedge-taken count will not overflow.
+ Value *CheckMinIters =
+ Builder.CreateICmpULT(Count,
+ ConstantInt::get(Count->getType(), VF * UF),
+ "min.iters.check");
+
+ BasicBlock *NewBB = BB->splitBasicBlock(BB->getTerminator(),
+ "min.iters.checked");
+ if (L->getParentLoop())
+ L->getParentLoop()->addBasicBlockToLoop(NewBB, *LI);
+ ReplaceInstWithInst(BB->getTerminator(),
+ BranchInst::Create(Bypass, NewBB, CheckMinIters));
+ LoopBypassBlocks.push_back(BB);
+}
- // Only need to check pointers between two different dependency sets.
- if (PtrRtCheck->DependencySetId[i] == PtrRtCheck->DependencySetId[j])
- continue;
+void InnerLoopVectorizer::emitVectorLoopEnteredCheck(Loop *L,
+ BasicBlock *Bypass) {
+ Value *TC = getOrCreateVectorTripCount(L);
+ BasicBlock *BB = L->getLoopPreheader();
+ IRBuilder<> Builder(BB->getTerminator());
+
+ // Now, compare the new count to zero. If it is zero skip the vector loop and
+ // jump to the scalar loop.
+ Value *Cmp = Builder.CreateICmpEQ(TC, Constant::getNullValue(TC->getType()),
+ "cmp.zero");
+
+ // Generate code to check that the loop's trip count that we computed by
+ // adding one to the backedge-taken count will not overflow.
+ BasicBlock *NewBB = BB->splitBasicBlock(BB->getTerminator(),
+ "vector.ph");
+ if (L->getParentLoop())
+ L->getParentLoop()->addBasicBlockToLoop(NewBB, *LI);
+ ReplaceInstWithInst(BB->getTerminator(),
+ BranchInst::Create(Bypass, NewBB, Cmp));
+ LoopBypassBlocks.push_back(BB);
+}
+
+void InnerLoopVectorizer::emitSCEVChecks(Loop *L, BasicBlock *Bypass) {
+ BasicBlock *BB = L->getLoopPreheader();
- unsigned AS0 = Starts[i]->getType()->getPointerAddressSpace();
- unsigned AS1 = Starts[j]->getType()->getPointerAddressSpace();
+ // Generate the code to check that the SCEV assumptions that we made.
+ // We want the new basic block to start at the first instruction in a
+ // sequence of instructions that form a check.
+ SCEVExpander Exp(*PSE.getSE(), Bypass->getModule()->getDataLayout(),
+ "scev.check");
+ Value *SCEVCheck =
+ Exp.expandCodeForPredicate(&PSE.getUnionPredicate(), BB->getTerminator());
- assert((AS0 == Ends[j]->getType()->getPointerAddressSpace()) &&
- (AS1 == Ends[i]->getType()->getPointerAddressSpace()) &&
- "Trying to bounds check pointers with different address spaces");
+ if (auto *C = dyn_cast<ConstantInt>(SCEVCheck))
+ if (C->isZero())
+ return;
- Type *PtrArithTy0 = Type::getInt8PtrTy(Ctx, AS0);
- Type *PtrArithTy1 = Type::getInt8PtrTy(Ctx, AS1);
+ // Create a new block containing the stride check.
+ BB->setName("vector.scevcheck");
+ auto *NewBB = BB->splitBasicBlock(BB->getTerminator(), "vector.ph");
+ if (L->getParentLoop())
+ L->getParentLoop()->addBasicBlockToLoop(NewBB, *LI);
+ ReplaceInstWithInst(BB->getTerminator(),
+ BranchInst::Create(Bypass, NewBB, SCEVCheck));
+ LoopBypassBlocks.push_back(BB);
+ AddedSafetyChecks = true;
+}
- Value *Start0 = ChkBuilder.CreateBitCast(Starts[i], PtrArithTy0, "bc");
- Value *Start1 = ChkBuilder.CreateBitCast(Starts[j], PtrArithTy1, "bc");
- Value *End0 = ChkBuilder.CreateBitCast(Ends[i], PtrArithTy1, "bc");
- Value *End1 = ChkBuilder.CreateBitCast(Ends[j], PtrArithTy0, "bc");
+void InnerLoopVectorizer::emitMemRuntimeChecks(Loop *L,
+ BasicBlock *Bypass) {
+ BasicBlock *BB = L->getLoopPreheader();
- Value *Cmp0 = ChkBuilder.CreateICmpULE(Start0, End1, "bound0");
- Value *Cmp1 = ChkBuilder.CreateICmpULE(Start1, End0, "bound1");
- Value *IsConflict = ChkBuilder.CreateAnd(Cmp0, Cmp1, "found.conflict");
- if (MemoryRuntimeCheck)
- IsConflict = ChkBuilder.CreateOr(MemoryRuntimeCheck, IsConflict,
- "conflict.rdx");
- MemoryRuntimeCheck = IsConflict;
- }
- }
+ // Generate the code that checks in runtime if arrays overlap. We put the
+ // checks into a separate block to make the more common case of few elements
+ // faster.
+ Instruction *FirstCheckInst;
+ Instruction *MemRuntimeCheck;
+ std::tie(FirstCheckInst, MemRuntimeCheck) =
+ Legal->getLAI()->addRuntimeChecks(BB->getTerminator());
+ if (!MemRuntimeCheck)
+ return;
- // We have to do this trickery because the IRBuilder might fold the check to a
- // constant expression in which case there is no Instruction anchored in a
- // the block.
- Instruction *Check = BinaryOperator::CreateAnd(MemoryRuntimeCheck,
- ConstantInt::getTrue(Ctx));
- ChkBuilder.Insert(Check, "memcheck.conflict");
- return Check;
+ // Create a new block containing the memory check.
+ BB->setName("vector.memcheck");
+ auto *NewBB = BB->splitBasicBlock(BB->getTerminator(), "vector.ph");
+ if (L->getParentLoop())
+ L->getParentLoop()->addBasicBlockToLoop(NewBB, *LI);
+ ReplaceInstWithInst(BB->getTerminator(),
+ BranchInst::Create(Bypass, NewBB, MemRuntimeCheck));
+ LoopBypassBlocks.push_back(BB);
+ AddedSafetyChecks = true;
}
-void
-InnerLoopVectorizer::createEmptyLoop(LoopVectorizationLegality *Legal) {
+
+void InnerLoopVectorizer::createEmptyLoop() {
/*
In this function we generate a new loop. The new loop will contain
the vectorized instructions while the old loop will continue to run the
scalar remainder.
- [ ] <-- vector loop bypass (may consist of multiple blocks).
- / |
- / v
- | [ ] <-- vector pre header.
- | |
- | v
- | [ ] \
- | [ ]_| <-- vector loop.
- | |
- \ v
- >[ ] <--- middle-block.
- / |
- / v
- | [ ] <--- new preheader.
+ [ ] <-- loop iteration number check.
+ / |
+ / v
+ | [ ] <-- vector loop bypass (may consist of multiple blocks).
+ | / |
+ | / v
+ || [ ] <-- vector pre header.
+ |/ |
+ | v
+ | [ ] \
+ | [ ]_| <-- vector loop.
+ | |
+ | v
+ | -[ ] <--- middle-block.
+ | / |
+ | / v
+ -|- >[ ] <--- new preheader.
| |
| v
| [ ] \
*/
BasicBlock *OldBasicBlock = OrigLoop->getHeader();
- BasicBlock *BypassBlock = OrigLoop->getLoopPreheader();
+ BasicBlock *VectorPH = OrigLoop->getLoopPreheader();
BasicBlock *ExitBlock = OrigLoop->getExitBlock();
+ assert(VectorPH && "Invalid loop structure");
assert(ExitBlock && "Must have an exit block");
// Some loops have a single integer induction variable, while other loops
// don't. One example is c++ iterators that often have multiple pointer
// induction variables. In the code below we also support a case where we
// don't have a single induction variable.
+ //
+ // We try to obtain an induction variable from the original loop as hard
+ // as possible. However if we don't find one that:
+ // - is an integer
+ // - counts from zero, stepping by one
+ // - is the size of the widest induction variable type
+ // then we create a new one.
OldInduction = Legal->getInduction();
Type *IdxTy = Legal->getWidestInductionType();
- // Find the loop boundaries.
- const SCEV *ExitCount = SE->getBackedgeTakenCount(OrigLoop);
- assert(ExitCount != SE->getCouldNotCompute() && "Invalid loop count");
-
- // Get the total trip count from the count by adding 1.
- ExitCount = SE->getAddExpr(ExitCount,
- SE->getConstant(ExitCount->getType(), 1));
-
- // Expand the trip count and place the new instructions in the preheader.
- // Notice that the pre-header does not change, only the loop body.
- SCEVExpander Exp(*SE, "induction");
-
- // Count holds the overall loop count (N).
- Value *Count = Exp.expandCodeFor(ExitCount, ExitCount->getType(),
- BypassBlock->getTerminator());
-
- // The loop index does not have to start at Zero. Find the original start
- // value from the induction PHI node. If we don't have an induction variable
- // then we know that it starts at zero.
- Builder.SetInsertPoint(BypassBlock->getTerminator());
- Value *StartIdx = ExtendedIdx = OldInduction ?
- Builder.CreateZExt(OldInduction->getIncomingValueForBlock(BypassBlock),
- IdxTy):
- ConstantInt::get(IdxTy, 0);
-
- assert(BypassBlock && "Invalid loop structure");
- LoopBypassBlocks.push_back(BypassBlock);
-
// Split the single block loop into the two loop structure described above.
- BasicBlock *VectorPH =
- BypassBlock->splitBasicBlock(BypassBlock->getTerminator(), "vector.ph");
BasicBlock *VecBody =
- VectorPH->splitBasicBlock(VectorPH->getTerminator(), "vector.body");
+ VectorPH->splitBasicBlock(VectorPH->getTerminator(), "vector.body");
BasicBlock *MiddleBlock =
VecBody->splitBasicBlock(VecBody->getTerminator(), "middle.block");
BasicBlock *ScalarPH =
// 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(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.
- Builder.SetInsertPoint(VecBody->getFirstNonPHI());
+ // Find the loop boundaries.
+ Value *Count = getOrCreateTripCount(Lp);
+
+ Value *StartIdx = ConstantInt::get(IdxTy, 0);
+
+ // We need to test whether the backedge-taken count is uint##_max. Adding one
+ // to it will cause overflow and an incorrect loop trip count in the vector
+ // body. In case of overflow we want to directly jump to the scalar remainder
+ // loop.
+ emitMinimumIterationCountCheck(Lp, ScalarPH);
+ // Now, compare the new count to zero. If it is zero skip the vector loop and
+ // jump to the scalar loop.
+ emitVectorLoopEnteredCheck(Lp, ScalarPH);
+ // Generate the code to check any assumptions that we've made for SCEV
+ // expressions.
+ emitSCEVChecks(Lp, ScalarPH);
+ // Generate the code that checks in runtime if arrays overlap. We put the
+ // checks into a separate block to make the more common case of few elements
+ // faster.
+ emitMemRuntimeChecks(Lp, ScalarPH);
+
// Generate the induction variable.
- setDebugLocFromInst(Builder, getDebugLocFromInstOrOperands(OldInduction));
- Induction = Builder.CreatePHI(IdxTy, 2, "index");
// The loop step is equal to the vectorization factor (num of SIMD elements)
// times the unroll factor (num of SIMD instructions).
+ Value *CountRoundDown = getOrCreateVectorTripCount(Lp);
Constant *Step = ConstantInt::get(IdxTy, VF * UF);
-
- // This is the IR builder that we use to add all of the logic for bypassing
- // the new vector loop.
- IRBuilder<> BypassBuilder(BypassBlock->getTerminator());
- setDebugLocFromInst(BypassBuilder,
- getDebugLocFromInstOrOperands(OldInduction));
-
- // We may need to extend the index in case there is a type mismatch.
- // We know that the count starts at zero and does not overflow.
- if (Count->getType() != IdxTy) {
- // The exit count can be of pointer type. Convert it to the correct
- // integer type.
- if (ExitCount->getType()->isPointerTy())
- Count = BypassBuilder.CreatePointerCast(Count, IdxTy, "ptrcnt.to.int");
- else
- Count = BypassBuilder.CreateZExtOrTrunc(Count, IdxTy, "cnt.cast");
- }
-
- // Add the start index to the loop count to get the new end index.
- Value *IdxEnd = BypassBuilder.CreateAdd(Count, StartIdx, "end.idx");
-
- // Now we need to generate the expression for N - (N % VF), which is
- // the part that the vectorized body will execute.
- Value *R = BypassBuilder.CreateURem(Count, Step, "n.mod.vf");
- Value *CountRoundDown = BypassBuilder.CreateSub(Count, R, "n.vec");
- Value *IdxEndRoundDown = BypassBuilder.CreateAdd(CountRoundDown, StartIdx,
- "end.idx.rnd.down");
-
- // Now, compare the new count to zero. If it is zero skip the vector loop and
- // jump to the scalar loop.
- Value *Cmp = BypassBuilder.CreateICmpEQ(IdxEndRoundDown, StartIdx,
- "cmp.zero");
-
- BasicBlock *LastBypassBlock = BypassBlock;
-
- // Generate the code that checks in runtime if arrays overlap. We put the
- // checks into a separate block to make the more common case of few elements
- // faster.
- Instruction *MemRuntimeCheck = addRuntimeCheck(Legal,
- BypassBlock->getTerminator());
- if (MemRuntimeCheck) {
- // Create a new block containing the memory check.
- BasicBlock *CheckBlock = BypassBlock->splitBasicBlock(MemRuntimeCheck,
- "vector.memcheck");
- if (ParentLoop)
- ParentLoop->addBasicBlockToLoop(CheckBlock, LI->getBase());
- LoopBypassBlocks.push_back(CheckBlock);
-
- // Replace the branch into the memory check block with a conditional branch
- // for the "few elements case".
- Instruction *OldTerm = BypassBlock->getTerminator();
- BranchInst::Create(MiddleBlock, CheckBlock, Cmp, OldTerm);
- OldTerm->eraseFromParent();
-
- Cmp = MemRuntimeCheck;
- LastBypassBlock = CheckBlock;
- }
-
- LastBypassBlock->getTerminator()->eraseFromParent();
- BranchInst::Create(MiddleBlock, VectorPH, Cmp,
- LastBypassBlock);
+ Induction =
+ createInductionVariable(Lp, StartIdx, CountRoundDown, Step,
+ getDebugLocFromInstOrOperands(OldInduction));
// We are going to resume the execution of the scalar loop.
// Go over all of the induction variables that we found and fix the
// If we come from a bypass edge then we need to start from the original
// start value.
- // This variable saves the new starting index for the scalar loop.
- PHINode *ResumeIndex = 0;
+ // This variable saves the new starting index for the scalar loop. It is used
+ // to test if there are any tail iterations left once the vector loop has
+ // completed.
LoopVectorizationLegality::InductionList::iterator I, E;
LoopVectorizationLegality::InductionList *List = Legal->getInductionVars();
- // Set builder to point to last bypass block.
- BypassBuilder.SetInsertPoint(LoopBypassBlocks.back()->getTerminator());
for (I = List->begin(), E = List->end(); I != E; ++I) {
PHINode *OrigPhi = I->first;
- LoopVectorizationLegality::InductionInfo II = I->second;
-
- Type *ResumeValTy = (OrigPhi == OldInduction) ? IdxTy : OrigPhi->getType();
- PHINode *ResumeVal = PHINode::Create(ResumeValTy, 2, "resume.val",
- MiddleBlock->getTerminator());
- // We might have extended the type of the induction variable but we need a
- // truncated version for the scalar loop.
- PHINode *TruncResumeVal = (OrigPhi == OldInduction) ?
- PHINode::Create(OrigPhi->getType(), 2, "trunc.resume.val",
- MiddleBlock->getTerminator()) : 0;
-
- Value *EndValue = 0;
- switch (II.IK) {
- case LoopVectorizationLegality::IK_NoInduction:
- llvm_unreachable("Unknown induction");
- case LoopVectorizationLegality::IK_IntInduction: {
- // Handle the integer induction counter.
- assert(OrigPhi->getType()->isIntegerTy() && "Invalid type");
-
- // We have the canonical induction variable.
- if (OrigPhi == OldInduction) {
- // Create a truncated version of the resume value for the scalar loop,
- // we might have promoted the type to a larger width.
- EndValue =
- BypassBuilder.CreateTrunc(IdxEndRoundDown, OrigPhi->getType());
- // The new PHI merges the original incoming value, in case of a bypass,
- // or the value at the end of the vectorized loop.
- for (unsigned I = 0, E = LoopBypassBlocks.size(); I != E; ++I)
- TruncResumeVal->addIncoming(II.StartValue, LoopBypassBlocks[I]);
- TruncResumeVal->addIncoming(EndValue, VecBody);
-
- // We know what the end value is.
- EndValue = IdxEndRoundDown;
- // We also know which PHI node holds it.
- ResumeIndex = ResumeVal;
- break;
- }
-
- // Not the canonical induction variable - add the vector loop count to the
- // start value.
- Value *CRD = BypassBuilder.CreateSExtOrTrunc(CountRoundDown,
- II.StartValue->getType(),
- "cast.crd");
- EndValue = BypassBuilder.CreateAdd(CRD, II.StartValue , "ind.end");
- break;
- }
- case LoopVectorizationLegality::IK_ReverseIntInduction: {
- // Convert the CountRoundDown variable to the PHI size.
- Value *CRD = BypassBuilder.CreateSExtOrTrunc(CountRoundDown,
- II.StartValue->getType(),
- "cast.crd");
- // Handle reverse integer induction counter.
- EndValue = BypassBuilder.CreateSub(II.StartValue, CRD, "rev.ind.end");
- break;
- }
- case LoopVectorizationLegality::IK_PtrInduction: {
- // For pointer induction variables, calculate the offset using
- // the end index.
- EndValue = BypassBuilder.CreateGEP(II.StartValue, CountRoundDown,
- "ptr.ind.end");
- break;
- }
- case LoopVectorizationLegality::IK_ReversePtrInduction: {
- // The value at the end of the loop for the reverse pointer is calculated
- // by creating a GEP with a negative index starting from the start value.
- Value *Zero = ConstantInt::get(CountRoundDown->getType(), 0);
- Value *NegIdx = BypassBuilder.CreateSub(Zero, CountRoundDown,
- "rev.ind.end");
- EndValue = BypassBuilder.CreateGEP(II.StartValue, NegIdx,
- "rev.ptr.ind.end");
- break;
+ InductionDescriptor II = I->second;
+
+ // Create phi nodes to merge from the backedge-taken check block.
+ PHINode *BCResumeVal = PHINode::Create(OrigPhi->getType(), 3,
+ "bc.resume.val",
+ ScalarPH->getTerminator());
+ Value *EndValue;
+ if (OrigPhi == OldInduction) {
+ // We know what the end value is.
+ EndValue = CountRoundDown;
+ } else {
+ IRBuilder<> B(LoopBypassBlocks.back()->getTerminator());
+ Value *CRD = B.CreateSExtOrTrunc(CountRoundDown,
+ II.getStepValue()->getType(),
+ "cast.crd");
+ EndValue = II.transform(B, CRD);
+ EndValue->setName("ind.end");
}
- }// end of case
// The new PHI merges the original incoming value, in case of a bypass,
// or the value at the end of the vectorized loop.
- for (unsigned I = 0, E = LoopBypassBlocks.size(); I != E; ++I) {
- if (OrigPhi == OldInduction)
- ResumeVal->addIncoming(StartIdx, LoopBypassBlocks[I]);
- else
- ResumeVal->addIncoming(II.StartValue, LoopBypassBlocks[I]);
- }
- ResumeVal->addIncoming(EndValue, VecBody);
+ BCResumeVal->addIncoming(EndValue, MiddleBlock);
// Fix the scalar body counter (PHI node).
unsigned BlockIdx = OrigPhi->getBasicBlockIndex(ScalarPH);
- // The old inductions phi node in the scalar body needs the truncated value.
- if (OrigPhi == OldInduction)
- OrigPhi->setIncomingValue(BlockIdx, TruncResumeVal);
- else
- OrigPhi->setIncomingValue(BlockIdx, ResumeVal);
- }
- // If we are generating a new induction variable then we also need to
- // generate the code that calculates the exit value. This value is not
- // simply the end of the counter because we may skip the vectorized body
- // in case of a runtime check.
- if (!OldInduction){
- assert(!ResumeIndex && "Unexpected resume value found");
- ResumeIndex = PHINode::Create(IdxTy, 2, "new.indc.resume.val",
- MiddleBlock->getTerminator());
+ // The old induction's phi node in the scalar body needs the truncated
+ // value.
for (unsigned I = 0, E = LoopBypassBlocks.size(); I != E; ++I)
- ResumeIndex->addIncoming(StartIdx, LoopBypassBlocks[I]);
- ResumeIndex->addIncoming(IdxEndRoundDown, VecBody);
+ BCResumeVal->addIncoming(II.getStartValue(), LoopBypassBlocks[I]);
+ OrigPhi->setIncomingValue(BlockIdx, BCResumeVal);
}
- // Make sure that we found the index where scalar loop needs to continue.
- assert(ResumeIndex && ResumeIndex->getType()->isIntegerTy() &&
- "Invalid resume Index");
-
// Add a check in the middle block to see if we have completed
// all of the iterations in the first vector loop.
// If (N - N%VF) == N, then we *don't* need to run the remainder.
- Value *CmpN = CmpInst::Create(Instruction::ICmp, CmpInst::ICMP_EQ, IdxEnd,
- ResumeIndex, "cmp.n",
+ Value *CmpN = CmpInst::Create(Instruction::ICmp, CmpInst::ICMP_EQ, Count,
+ CountRoundDown, "cmp.n",
MiddleBlock->getTerminator());
-
- BranchInst::Create(ExitBlock, ScalarPH, CmpN, MiddleBlock->getTerminator());
- // Remove the old terminator.
- MiddleBlock->getTerminator()->eraseFromParent();
-
- // Create i+1 and fill the PHINode.
- Value *NextIdx = Builder.CreateAdd(Induction, Step, "index.next");
- Induction->addIncoming(StartIdx, VectorPH);
- Induction->addIncoming(NextIdx, VecBody);
- // Create the compare.
- Value *ICmp = Builder.CreateICmpEQ(NextIdx, IdxEndRoundDown);
- Builder.CreateCondBr(ICmp, MiddleBlock, VecBody);
-
- // Now we have two terminators. Remove the old one from the block.
- VecBody->getTerminator()->eraseFromParent();
+ ReplaceInstWithInst(MiddleBlock->getTerminator(),
+ BranchInst::Create(ExitBlock, ScalarPH, CmpN));
// Get ready to start creating new instructions into the vectorized body.
- Builder.SetInsertPoint(VecBody->getFirstInsertionPt());
+ Builder.SetInsertPoint(&*VecBody->getFirstInsertionPt());
// Save the state.
- LoopVectorPreHeader = VectorPH;
+ LoopVectorPreHeader = Lp->getLoopPreheader();
LoopScalarPreHeader = ScalarPH;
LoopMiddleBlock = MiddleBlock;
LoopExitBlock = ExitBlock;
- LoopVectorBody = VecBody;
+ LoopVectorBody.push_back(VecBody);
LoopScalarBody = OldBasicBlock;
+
+ LoopVectorizeHints Hints(Lp, true);
+ Hints.setAlreadyVectorized();
}
-/// This function returns the identity element (or neutral element) for
-/// the operation K.
-Constant*
-LoopVectorizationLegality::getReductionIdentity(ReductionKind K, Type *Tp) {
- switch (K) {
- case RK_IntegerXor:
- case RK_IntegerAdd:
- case RK_IntegerOr:
- // Adding, Xoring, Oring zero to a number does not change it.
- return ConstantInt::get(Tp, 0);
- case RK_IntegerMult:
- // Multiplying a number by 1 does not change it.
- return ConstantInt::get(Tp, 1);
- case RK_IntegerAnd:
- // AND-ing a number with an all-1 value does not change it.
- return ConstantInt::get(Tp, -1, true);
- case RK_FloatMult:
- // Multiplying a number by 1 does not change it.
- return ConstantFP::get(Tp, 1.0L);
- case RK_FloatAdd:
- // Adding zero to a number does not change it.
- return ConstantFP::get(Tp, 0.0L);
- default:
- llvm_unreachable("Unknown reduction kind");
+namespace {
+struct CSEDenseMapInfo {
+ static bool canHandle(Instruction *I) {
+ return isa<InsertElementInst>(I) || isa<ExtractElementInst>(I) ||
+ isa<ShuffleVectorInst>(I) || isa<GetElementPtrInst>(I);
+ }
+ static inline Instruction *getEmptyKey() {
+ return DenseMapInfo<Instruction *>::getEmptyKey();
+ }
+ static inline Instruction *getTombstoneKey() {
+ return DenseMapInfo<Instruction *>::getTombstoneKey();
+ }
+ static unsigned getHashValue(Instruction *I) {
+ assert(canHandle(I) && "Unknown instruction!");
+ return hash_combine(I->getOpcode(), hash_combine_range(I->value_op_begin(),
+ I->value_op_end()));
}
+ static bool isEqual(Instruction *LHS, Instruction *RHS) {
+ if (LHS == getEmptyKey() || RHS == getEmptyKey() ||
+ LHS == getTombstoneKey() || RHS == getTombstoneKey())
+ return LHS == RHS;
+ return LHS->isIdenticalTo(RHS);
+ }
+};
+}
+
+/// \brief Check whether this block is a predicated block.
+/// Due to if predication of stores we might create a sequence of "if(pred) a[i]
+/// = ...; " blocks. We start with one vectorized basic block. For every
+/// conditional block we split this vectorized block. Therefore, every second
+/// block will be a predicated one.
+static bool isPredicatedBlock(unsigned BlockNum) {
+ return BlockNum % 2;
}
-static Intrinsic::ID checkUnaryFloatSignature(const CallInst &I,
- Intrinsic::ID ValidIntrinsicID) {
- if (I.getNumArgOperands() != 1 ||
- !I.getArgOperand(0)->getType()->isFloatingPointTy() ||
- I.getType() != I.getArgOperand(0)->getType() ||
- !I.onlyReadsMemory())
- return Intrinsic::not_intrinsic;
+///\brief Perform cse of induction variable instructions.
+static void cse(SmallVector<BasicBlock *, 4> &BBs) {
+ // Perform simple cse.
+ SmallDenseMap<Instruction *, Instruction *, 4, CSEDenseMapInfo> CSEMap;
+ for (unsigned i = 0, e = BBs.size(); i != e; ++i) {
+ BasicBlock *BB = BBs[i];
+ for (BasicBlock::iterator I = BB->begin(), E = BB->end(); I != E;) {
+ Instruction *In = &*I++;
+
+ if (!CSEDenseMapInfo::canHandle(In))
+ continue;
+
+ // Check if we can replace this instruction with any of the
+ // visited instructions.
+ if (Instruction *V = CSEMap.lookup(In)) {
+ In->replaceAllUsesWith(V);
+ In->eraseFromParent();
+ continue;
+ }
+ // Ignore instructions in conditional blocks. We create "if (pred) a[i] =
+ // ...;" blocks for predicated stores. Every second block is a predicated
+ // block.
+ if (isPredicatedBlock(i))
+ continue;
- return ValidIntrinsicID;
+ CSEMap[In] = In;
+ }
+ }
}
-static Intrinsic::ID checkBinaryFloatSignature(const CallInst &I,
- Intrinsic::ID ValidIntrinsicID) {
- if (I.getNumArgOperands() != 2 ||
- !I.getArgOperand(0)->getType()->isFloatingPointTy() ||
- !I.getArgOperand(1)->getType()->isFloatingPointTy() ||
- I.getType() != I.getArgOperand(0)->getType() ||
- I.getType() != I.getArgOperand(1)->getType() ||
- !I.onlyReadsMemory())
- return Intrinsic::not_intrinsic;
-
- return ValidIntrinsicID;
+/// \brief Adds a 'fast' flag to floating point operations.
+static Value *addFastMathFlag(Value *V) {
+ if (isa<FPMathOperator>(V)){
+ FastMathFlags Flags;
+ Flags.setUnsafeAlgebra();
+ cast<Instruction>(V)->setFastMathFlags(Flags);
+ }
+ return V;
}
+/// Estimate the overhead of scalarizing a value. Insert and Extract are set if
+/// the result needs to be inserted and/or extracted from vectors.
+static unsigned getScalarizationOverhead(Type *Ty, bool Insert, bool Extract,
+ const TargetTransformInfo &TTI) {
+ if (Ty->isVoidTy())
+ return 0;
-static Intrinsic::ID
-getIntrinsicIDForCall(CallInst *CI, const TargetLibraryInfo *TLI) {
- // If we have an intrinsic call, check if it is trivially vectorizable.
- if (IntrinsicInst *II = dyn_cast<IntrinsicInst>(CI)) {
- switch (II->getIntrinsicID()) {
- case Intrinsic::sqrt:
- case Intrinsic::sin:
- case Intrinsic::cos:
- case Intrinsic::exp:
- case Intrinsic::exp2:
- case Intrinsic::log:
- case Intrinsic::log10:
- case Intrinsic::log2:
- case Intrinsic::fabs:
- case Intrinsic::copysign:
- case Intrinsic::floor:
- case Intrinsic::ceil:
- case Intrinsic::trunc:
- case Intrinsic::rint:
- case Intrinsic::nearbyint:
- case Intrinsic::round:
- case Intrinsic::pow:
- case Intrinsic::fma:
- case Intrinsic::fmuladd:
- case Intrinsic::lifetime_start:
- case Intrinsic::lifetime_end:
- return II->getIntrinsicID();
- default:
- return Intrinsic::not_intrinsic;
- }
+ assert(Ty->isVectorTy() && "Can only scalarize vectors");
+ unsigned Cost = 0;
+
+ for (int i = 0, e = Ty->getVectorNumElements(); i < e; ++i) {
+ if (Insert)
+ Cost += TTI.getVectorInstrCost(Instruction::InsertElement, Ty, i);
+ if (Extract)
+ Cost += TTI.getVectorInstrCost(Instruction::ExtractElement, Ty, i);
}
- if (!TLI)
- return Intrinsic::not_intrinsic;
+ return Cost;
+}
- LibFunc::Func Func;
+// Estimate cost of a call instruction CI if it were vectorized with factor VF.
+// Return the cost of the instruction, including scalarization overhead if it's
+// needed. The flag NeedToScalarize shows if the call needs to be scalarized -
+// i.e. either vector version isn't available, or is too expensive.
+static unsigned getVectorCallCost(CallInst *CI, unsigned VF,
+ const TargetTransformInfo &TTI,
+ const TargetLibraryInfo *TLI,
+ bool &NeedToScalarize) {
Function *F = CI->getCalledFunction();
- // We're going to make assumptions on the semantics of the functions, check
- // that the target knows that it's available in this environment and it does
- // not have local linkage.
- if (!F || F->hasLocalLinkage() || !TLI->getLibFunc(F->getName(), Func))
- return Intrinsic::not_intrinsic;
-
- // Otherwise check if we have a call to a function that can be turned into a
- // vector intrinsic.
- switch (Func) {
- default:
- break;
- case LibFunc::sin:
- case LibFunc::sinf:
- case LibFunc::sinl:
- return checkUnaryFloatSignature(*CI, Intrinsic::sin);
- case LibFunc::cos:
- case LibFunc::cosf:
- case LibFunc::cosl:
- return checkUnaryFloatSignature(*CI, Intrinsic::cos);
- case LibFunc::exp:
- case LibFunc::expf:
- case LibFunc::expl:
- return checkUnaryFloatSignature(*CI, Intrinsic::exp);
- case LibFunc::exp2:
- case LibFunc::exp2f:
- case LibFunc::exp2l:
- return checkUnaryFloatSignature(*CI, Intrinsic::exp2);
- case LibFunc::log:
- case LibFunc::logf:
- case LibFunc::logl:
- return checkUnaryFloatSignature(*CI, Intrinsic::log);
- case LibFunc::log10:
- case LibFunc::log10f:
- case LibFunc::log10l:
- return checkUnaryFloatSignature(*CI, Intrinsic::log10);
- case LibFunc::log2:
- case LibFunc::log2f:
- case LibFunc::log2l:
- return checkUnaryFloatSignature(*CI, Intrinsic::log2);
- case LibFunc::fabs:
- case LibFunc::fabsf:
- case LibFunc::fabsl:
- return checkUnaryFloatSignature(*CI, Intrinsic::fabs);
- case LibFunc::copysign:
- case LibFunc::copysignf:
- case LibFunc::copysignl:
- return checkBinaryFloatSignature(*CI, Intrinsic::copysign);
- case LibFunc::floor:
- case LibFunc::floorf:
- case LibFunc::floorl:
- return checkUnaryFloatSignature(*CI, Intrinsic::floor);
- case LibFunc::ceil:
- case LibFunc::ceilf:
- case LibFunc::ceill:
- return checkUnaryFloatSignature(*CI, Intrinsic::ceil);
- case LibFunc::trunc:
- case LibFunc::truncf:
- case LibFunc::truncl:
- return checkUnaryFloatSignature(*CI, Intrinsic::trunc);
- case LibFunc::rint:
- case LibFunc::rintf:
- case LibFunc::rintl:
- return checkUnaryFloatSignature(*CI, Intrinsic::rint);
- case LibFunc::nearbyint:
- case LibFunc::nearbyintf:
- case LibFunc::nearbyintl:
- return checkUnaryFloatSignature(*CI, Intrinsic::nearbyint);
- case LibFunc::round:
- case LibFunc::roundf:
- case LibFunc::roundl:
- return checkUnaryFloatSignature(*CI, Intrinsic::round);
- case LibFunc::pow:
- case LibFunc::powf:
- case LibFunc::powl:
- return checkBinaryFloatSignature(*CI, Intrinsic::pow);
- }
-
- return Intrinsic::not_intrinsic;
-}
+ StringRef FnName = CI->getCalledFunction()->getName();
+ Type *ScalarRetTy = CI->getType();
+ SmallVector<Type *, 4> Tys, ScalarTys;
+ for (auto &ArgOp : CI->arg_operands())
+ ScalarTys.push_back(ArgOp->getType());
+
+ // Estimate cost of scalarized vector call. The source operands are assumed
+ // to be vectors, so we need to extract individual elements from there,
+ // execute VF scalar calls, and then gather the result into the vector return
+ // value.
+ unsigned ScalarCallCost = TTI.getCallInstrCost(F, ScalarRetTy, ScalarTys);
+ if (VF == 1)
+ return ScalarCallCost;
+
+ // Compute corresponding vector type for return value and arguments.
+ Type *RetTy = ToVectorTy(ScalarRetTy, VF);
+ for (unsigned i = 0, ie = ScalarTys.size(); i != ie; ++i)
+ Tys.push_back(ToVectorTy(ScalarTys[i], VF));
+
+ // Compute costs of unpacking argument values for the scalar calls and
+ // packing the return values to a vector.
+ unsigned ScalarizationCost =
+ getScalarizationOverhead(RetTy, true, false, TTI);
+ for (unsigned i = 0, ie = Tys.size(); i != ie; ++i)
+ ScalarizationCost += getScalarizationOverhead(Tys[i], false, true, TTI);
+
+ unsigned Cost = ScalarCallCost * VF + ScalarizationCost;
+
+ // If we can't emit a vector call for this function, then the currently found
+ // cost is the cost we need to return.
+ NeedToScalarize = true;
+ if (!TLI || !TLI->isFunctionVectorizable(FnName, VF) || CI->isNoBuiltin())
+ return Cost;
-/// This function translates the reduction kind to an LLVM binary operator.
-static unsigned
-getReductionBinOp(LoopVectorizationLegality::ReductionKind Kind) {
- switch (Kind) {
- case LoopVectorizationLegality::RK_IntegerAdd:
- return Instruction::Add;
- case LoopVectorizationLegality::RK_IntegerMult:
- return Instruction::Mul;
- case LoopVectorizationLegality::RK_IntegerOr:
- return Instruction::Or;
- case LoopVectorizationLegality::RK_IntegerAnd:
- return Instruction::And;
- case LoopVectorizationLegality::RK_IntegerXor:
- return Instruction::Xor;
- case LoopVectorizationLegality::RK_FloatMult:
- return Instruction::FMul;
- case LoopVectorizationLegality::RK_FloatAdd:
- return Instruction::FAdd;
- case LoopVectorizationLegality::RK_IntegerMinMax:
- return Instruction::ICmp;
- case LoopVectorizationLegality::RK_FloatMinMax:
- return Instruction::FCmp;
- default:
- llvm_unreachable("Unknown reduction operation");
+ // If the corresponding vector cost is cheaper, return its cost.
+ unsigned VectorCallCost = TTI.getCallInstrCost(nullptr, RetTy, Tys);
+ if (VectorCallCost < Cost) {
+ NeedToScalarize = false;
+ return VectorCallCost;
}
+ return Cost;
}
-Value *createMinMaxOp(IRBuilder<> &Builder,
- LoopVectorizationLegality::MinMaxReductionKind RK,
- Value *Left,
- Value *Right) {
- CmpInst::Predicate P = CmpInst::ICMP_NE;
- switch (RK) {
- default:
- llvm_unreachable("Unknown min/max reduction kind");
- case LoopVectorizationLegality::MRK_UIntMin:
- P = CmpInst::ICMP_ULT;
- break;
- case LoopVectorizationLegality::MRK_UIntMax:
- P = CmpInst::ICMP_UGT;
- break;
- case LoopVectorizationLegality::MRK_SIntMin:
- P = CmpInst::ICMP_SLT;
- break;
- case LoopVectorizationLegality::MRK_SIntMax:
- P = CmpInst::ICMP_SGT;
- break;
- case LoopVectorizationLegality::MRK_FloatMin:
- P = CmpInst::FCMP_OLT;
- break;
- case LoopVectorizationLegality::MRK_FloatMax:
- P = CmpInst::FCMP_OGT;
- break;
- }
-
- Value *Cmp;
- if (RK == LoopVectorizationLegality::MRK_FloatMin ||
- RK == LoopVectorizationLegality::MRK_FloatMax)
- Cmp = Builder.CreateFCmp(P, Left, Right, "rdx.minmax.cmp");
- else
- Cmp = Builder.CreateICmp(P, Left, Right, "rdx.minmax.cmp");
+// Estimate cost of an intrinsic call instruction CI if it were vectorized with
+// factor VF. Return the cost of the instruction, including scalarization
+// overhead if it's needed.
+static unsigned getVectorIntrinsicCost(CallInst *CI, unsigned VF,
+ const TargetTransformInfo &TTI,
+ const TargetLibraryInfo *TLI) {
+ Intrinsic::ID ID = getIntrinsicIDForCall(CI, TLI);
+ assert(ID && "Expected intrinsic call!");
+
+ Type *RetTy = ToVectorTy(CI->getType(), VF);
+ SmallVector<Type *, 4> Tys;
+ for (unsigned i = 0, ie = CI->getNumArgOperands(); i != ie; ++i)
+ Tys.push_back(ToVectorTy(CI->getArgOperand(i)->getType(), VF));
+
+ return TTI.getIntrinsicInstrCost(ID, RetTy, Tys);
+}
+
+static Type *smallestIntegerVectorType(Type *T1, Type *T2) {
+ IntegerType *I1 = cast<IntegerType>(T1->getVectorElementType());
+ IntegerType *I2 = cast<IntegerType>(T2->getVectorElementType());
+ return I1->getBitWidth() < I2->getBitWidth() ? T1 : T2;
+}
+static Type *largestIntegerVectorType(Type *T1, Type *T2) {
+ IntegerType *I1 = cast<IntegerType>(T1->getVectorElementType());
+ IntegerType *I2 = cast<IntegerType>(T2->getVectorElementType());
+ return I1->getBitWidth() > I2->getBitWidth() ? T1 : T2;
+}
+
+void InnerLoopVectorizer::truncateToMinimalBitwidths() {
+ // For every instruction `I` in MinBWs, truncate the operands, create a
+ // truncated version of `I` and reextend its result. InstCombine runs
+ // later and will remove any ext/trunc pairs.
+ //
+ for (auto &KV : MinBWs) {
+ VectorParts &Parts = WidenMap.get(KV.first);
+ for (Value *&I : Parts) {
+ if (I->use_empty())
+ continue;
+ Type *OriginalTy = I->getType();
+ Type *ScalarTruncatedTy = IntegerType::get(OriginalTy->getContext(),
+ KV.second);
+ Type *TruncatedTy = VectorType::get(ScalarTruncatedTy,
+ OriginalTy->getVectorNumElements());
+ if (TruncatedTy == OriginalTy)
+ continue;
+
+ IRBuilder<> B(cast<Instruction>(I));
+ auto ShrinkOperand = [&](Value *V) -> Value* {
+ if (auto *ZI = dyn_cast<ZExtInst>(V))
+ if (ZI->getSrcTy() == TruncatedTy)
+ return ZI->getOperand(0);
+ return B.CreateZExtOrTrunc(V, TruncatedTy);
+ };
+
+ // The actual instruction modification depends on the instruction type,
+ // unfortunately.
+ Value *NewI = nullptr;
+ if (BinaryOperator *BO = dyn_cast<BinaryOperator>(I)) {
+ NewI = B.CreateBinOp(BO->getOpcode(),
+ ShrinkOperand(BO->getOperand(0)),
+ ShrinkOperand(BO->getOperand(1)));
+ cast<BinaryOperator>(NewI)->copyIRFlags(I);
+ } else if (ICmpInst *CI = dyn_cast<ICmpInst>(I)) {
+ NewI = B.CreateICmp(CI->getPredicate(),
+ ShrinkOperand(CI->getOperand(0)),
+ ShrinkOperand(CI->getOperand(1)));
+ } else if (SelectInst *SI = dyn_cast<SelectInst>(I)) {
+ NewI = B.CreateSelect(SI->getCondition(),
+ ShrinkOperand(SI->getTrueValue()),
+ ShrinkOperand(SI->getFalseValue()));
+ } else if (CastInst *CI = dyn_cast<CastInst>(I)) {
+ switch (CI->getOpcode()) {
+ default: llvm_unreachable("Unhandled cast!");
+ case Instruction::Trunc:
+ NewI = ShrinkOperand(CI->getOperand(0));
+ break;
+ case Instruction::SExt:
+ NewI = B.CreateSExtOrTrunc(CI->getOperand(0),
+ smallestIntegerVectorType(OriginalTy,
+ TruncatedTy));
+ break;
+ case Instruction::ZExt:
+ NewI = B.CreateZExtOrTrunc(CI->getOperand(0),
+ smallestIntegerVectorType(OriginalTy,
+ TruncatedTy));
+ break;
+ }
+ } else if (ShuffleVectorInst *SI = dyn_cast<ShuffleVectorInst>(I)) {
+ auto Elements0 = SI->getOperand(0)->getType()->getVectorNumElements();
+ auto *O0 =
+ B.CreateZExtOrTrunc(SI->getOperand(0),
+ VectorType::get(ScalarTruncatedTy, Elements0));
+ auto Elements1 = SI->getOperand(1)->getType()->getVectorNumElements();
+ auto *O1 =
+ B.CreateZExtOrTrunc(SI->getOperand(1),
+ VectorType::get(ScalarTruncatedTy, Elements1));
+
+ NewI = B.CreateShuffleVector(O0, O1, SI->getMask());
+ } else if (isa<LoadInst>(I)) {
+ // Don't do anything with the operands, just extend the result.
+ continue;
+ } else {
+ llvm_unreachable("Unhandled instruction type!");
+ }
+
+ // Lastly, extend the result.
+ NewI->takeName(cast<Instruction>(I));
+ Value *Res = B.CreateZExtOrTrunc(NewI, OriginalTy);
+ I->replaceAllUsesWith(Res);
+ cast<Instruction>(I)->eraseFromParent();
+ I = Res;
+ }
+ }
- Value *Select = Builder.CreateSelect(Cmp, Left, Right, "rdx.minmax.select");
- return Select;
+ // We'll have created a bunch of ZExts that are now parentless. Clean up.
+ for (auto &KV : MinBWs) {
+ VectorParts &Parts = WidenMap.get(KV.first);
+ for (Value *&I : Parts) {
+ ZExtInst *Inst = dyn_cast<ZExtInst>(I);
+ if (Inst && Inst->use_empty()) {
+ Value *NewI = Inst->getOperand(0);
+ Inst->eraseFromParent();
+ I = NewI;
+ }
+ }
+ }
}
-void
-InnerLoopVectorizer::vectorizeLoop(LoopVectorizationLegality *Legal) {
+void InnerLoopVectorizer::vectorizeLoop() {
//===------------------------------------------------===//
//
// Notice: any optimization or new instruction that go
// Vectorize all of the blocks in the original loop.
for (LoopBlocksDFS::RPOIterator bb = DFS.beginRPO(),
be = DFS.endRPO(); bb != be; ++bb)
- vectorizeBlockInLoop(Legal, *bb, &RdxPHIsToFix);
+ vectorizeBlockInLoop(*bb, &RdxPHIsToFix);
+ // Insert truncates and extends for any truncated instructions as hints to
+ // InstCombine.
+ if (VF > 1)
+ truncateToMinimalBitwidths();
+
// At this point every instruction in the original loop is widened to
// a vector form. We are almost done. Now, we need to fix the PHI nodes
// that we vectorized. The PHI nodes are currently empty because we did
assert(RdxPhi && "Unable to recover vectorized PHI");
// Find the reduction variable descriptor.
- assert(Legal->getReductionVars()->count(RdxPhi) &&
+ assert(Legal->isReductionVariable(RdxPhi) &&
"Unable to find the reduction variable");
- LoopVectorizationLegality::ReductionDescriptor RdxDesc =
- (*Legal->getReductionVars())[RdxPhi];
+ RecurrenceDescriptor RdxDesc = (*Legal->getReductionVars())[RdxPhi];
- setDebugLocFromInst(Builder, RdxDesc.StartValue);
+ RecurrenceDescriptor::RecurrenceKind RK = RdxDesc.getRecurrenceKind();
+ TrackingVH<Value> ReductionStartValue = RdxDesc.getRecurrenceStartValue();
+ Instruction *LoopExitInst = RdxDesc.getLoopExitInstr();
+ RecurrenceDescriptor::MinMaxRecurrenceKind MinMaxKind =
+ RdxDesc.getMinMaxRecurrenceKind();
+ setDebugLocFromInst(Builder, ReductionStartValue);
// We need to generate a reduction vector from the incoming scalar.
- // To do so, we need to generate the 'identity' vector and overide
+ // To do so, we need to generate the 'identity' vector and override
// one of the elements with the incoming scalar reduction. We need
// to do it in the vector-loop preheader.
- Builder.SetInsertPoint(LoopBypassBlocks.front()->getTerminator());
+ Builder.SetInsertPoint(LoopBypassBlocks[1]->getTerminator());
// This is the vector-clone of the value that leaves the loop.
- VectorParts &VectorExit = getVectorValue(RdxDesc.LoopExitInstr);
+ VectorParts &VectorExit = getVectorValue(LoopExitInst);
Type *VecTy = VectorExit[0]->getType();
// Find the reduction identity variable. Zero for addition, or, xor,
// one for multiplication, -1 for And.
Value *Identity;
Value *VectorStart;
- if (RdxDesc.Kind == LoopVectorizationLegality::RK_IntegerMinMax ||
- RdxDesc.Kind == LoopVectorizationLegality::RK_FloatMinMax) {
+ if (RK == RecurrenceDescriptor::RK_IntegerMinMax ||
+ RK == RecurrenceDescriptor::RK_FloatMinMax) {
// MinMax reduction have the start value as their identify.
if (VF == 1) {
- VectorStart = Identity = RdxDesc.StartValue;
+ VectorStart = Identity = ReductionStartValue;
} else {
- VectorStart = Identity = Builder.CreateVectorSplat(VF,
- RdxDesc.StartValue,
- "minmax.ident");
+ VectorStart = Identity =
+ Builder.CreateVectorSplat(VF, ReductionStartValue, "minmax.ident");
}
} else {
// Handle other reduction kinds:
- Constant *Iden =
- LoopVectorizationLegality::getReductionIdentity(RdxDesc.Kind,
- VecTy->getScalarType());
+ Constant *Iden = RecurrenceDescriptor::getRecurrenceIdentity(
+ RK, VecTy->getScalarType());
if (VF == 1) {
Identity = Iden;
// This vector is the Identity vector where the first element is the
// incoming scalar reduction.
- VectorStart = RdxDesc.StartValue;
+ VectorStart = ReductionStartValue;
} else {
Identity = ConstantVector::getSplat(VF, Iden);
// This vector is the Identity vector where the first element is the
// incoming scalar reduction.
- VectorStart = Builder.CreateInsertElement(Identity,
- RdxDesc.StartValue, Zero);
+ VectorStart =
+ Builder.CreateInsertElement(Identity, ReductionStartValue, Zero);
}
}
// 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(Val[part], LoopVectorBody);
+ cast<PHINode>(VecRdxPhi[part])->addIncoming(StartVal,
+ LoopVectorPreHeader);
+ cast<PHINode>(VecRdxPhi[part])->addIncoming(Val[part],
+ LoopVectorBody.back());
}
// Before each round, move the insertion point right between
// the PHIs and the values we are going to write.
// This allows us to write both PHINodes and the extractelement
// instructions.
- Builder.SetInsertPoint(LoopMiddleBlock->getFirstInsertionPt());
+ Builder.SetInsertPoint(&*LoopMiddleBlock->getFirstInsertionPt());
- VectorParts RdxParts;
- setDebugLocFromInst(Builder, RdxDesc.LoopExitInstr);
- for (unsigned part = 0; part < UF; ++part) {
- // This PHINode contains the vectorized reduction variable, or
- // the initial value vector, if we bypass the vector loop.
- VectorParts &RdxExitVal = getVectorValue(RdxDesc.LoopExitInstr);
- PHINode *NewPhi = Builder.CreatePHI(VecTy, 2, "rdx.vec.exit.phi");
- Value *StartVal = (part == 0) ? VectorStart : Identity;
- for (unsigned I = 0, E = LoopBypassBlocks.size(); I != E; ++I)
- NewPhi->addIncoming(StartVal, LoopBypassBlocks[I]);
- NewPhi->addIncoming(RdxExitVal[part], LoopVectorBody);
- RdxParts.push_back(NewPhi);
+ VectorParts RdxParts = getVectorValue(LoopExitInst);
+ setDebugLocFromInst(Builder, LoopExitInst);
+
+ // If the vector reduction can be performed in a smaller type, we truncate
+ // then extend the loop exit value to enable InstCombine to evaluate the
+ // entire expression in the smaller type.
+ if (VF > 1 && RdxPhi->getType() != RdxDesc.getRecurrenceType()) {
+ Type *RdxVecTy = VectorType::get(RdxDesc.getRecurrenceType(), VF);
+ Builder.SetInsertPoint(LoopVectorBody.back()->getTerminator());
+ for (unsigned part = 0; part < UF; ++part) {
+ Value *Trunc = Builder.CreateTrunc(RdxParts[part], RdxVecTy);
+ Value *Extnd = RdxDesc.isSigned() ? Builder.CreateSExt(Trunc, VecTy)
+ : Builder.CreateZExt(Trunc, VecTy);
+ for (Value::user_iterator UI = RdxParts[part]->user_begin();
+ UI != RdxParts[part]->user_end();)
+ if (*UI != Trunc) {
+ (*UI++)->replaceUsesOfWith(RdxParts[part], Extnd);
+ RdxParts[part] = Extnd;
+ } else {
+ ++UI;
+ }
+ }
+ Builder.SetInsertPoint(&*LoopMiddleBlock->getFirstInsertionPt());
+ for (unsigned part = 0; part < UF; ++part)
+ RdxParts[part] = Builder.CreateTrunc(RdxParts[part], RdxVecTy);
}
// Reduce all of the unrolled parts into a single vector.
Value *ReducedPartRdx = RdxParts[0];
- unsigned Op = getReductionBinOp(RdxDesc.Kind);
+ unsigned Op = RecurrenceDescriptor::getRecurrenceBinOp(RK);
setDebugLocFromInst(Builder, ReducedPartRdx);
for (unsigned part = 1; part < UF; ++part) {
if (Op != Instruction::ICmp && Op != Instruction::FCmp)
- ReducedPartRdx = Builder.CreateBinOp((Instruction::BinaryOps)Op,
- RdxParts[part], ReducedPartRdx,
- "bin.rdx");
+ // Floating point operations had to be 'fast' to enable the reduction.
+ ReducedPartRdx = addFastMathFlag(
+ Builder.CreateBinOp((Instruction::BinaryOps)Op, RdxParts[part],
+ ReducedPartRdx, "bin.rdx"));
else
- ReducedPartRdx = createMinMaxOp(Builder, RdxDesc.MinMaxKind,
- ReducedPartRdx, RdxParts[part]);
+ ReducedPartRdx = RecurrenceDescriptor::createMinMaxOp(
+ Builder, MinMaxKind, ReducedPartRdx, RdxParts[part]);
}
if (VF > 1) {
assert(isPowerOf2_32(VF) &&
"Reduction emission only supported for pow2 vectors!");
Value *TmpVec = ReducedPartRdx;
- SmallVector<Constant*, 32> ShuffleMask(VF, 0);
+ SmallVector<Constant*, 32> ShuffleMask(VF, nullptr);
for (unsigned i = VF; i != 1; i >>= 1) {
// Move the upper half of the vector to the lower half.
for (unsigned j = 0; j != i/2; ++j)
"rdx.shuf");
if (Op != Instruction::ICmp && Op != Instruction::FCmp)
- TmpVec = Builder.CreateBinOp((Instruction::BinaryOps)Op, TmpVec, Shuf,
- "bin.rdx");
+ // Floating point operations had to be 'fast' to enable the reduction.
+ TmpVec = addFastMathFlag(Builder.CreateBinOp(
+ (Instruction::BinaryOps)Op, TmpVec, Shuf, "bin.rdx"));
else
- TmpVec = createMinMaxOp(Builder, RdxDesc.MinMaxKind, TmpVec, Shuf);
+ TmpVec = RecurrenceDescriptor::createMinMaxOp(Builder, MinMaxKind,
+ TmpVec, Shuf);
}
// The result is in the first element of the vector.
ReducedPartRdx = Builder.CreateExtractElement(TmpVec,
Builder.getInt32(0));
+
+ // If the reduction can be performed in a smaller type, we need to extend
+ // the reduction to the wider type before we branch to the original loop.
+ if (RdxPhi->getType() != RdxDesc.getRecurrenceType())
+ ReducedPartRdx =
+ RdxDesc.isSigned()
+ ? Builder.CreateSExt(ReducedPartRdx, RdxPhi->getType())
+ : Builder.CreateZExt(ReducedPartRdx, RdxPhi->getType());
}
+ // Create a phi node that merges control-flow from the backedge-taken check
+ // block and the middle block.
+ PHINode *BCBlockPhi = PHINode::Create(RdxPhi->getType(), 2, "bc.merge.rdx",
+ LoopScalarPreHeader->getTerminator());
+ for (unsigned I = 0, E = LoopBypassBlocks.size(); I != E; ++I)
+ BCBlockPhi->addIncoming(ReductionStartValue, LoopBypassBlocks[I]);
+ BCBlockPhi->addIncoming(ReducedPartRdx, LoopMiddleBlock);
+
// Now, we need to fix the users of the reduction variable
// inside and outside of the scalar remainder loop.
// We know that the loop is in LCSSA form. We need to update the
// We found our reduction value exit-PHI. Update it with the
// incoming bypass edge.
- if (LCSSAPhi->getIncomingValue(0) == RdxDesc.LoopExitInstr) {
+ if (LCSSAPhi->getIncomingValue(0) == LoopExitInst) {
// Add an edge coming from the bypass.
LCSSAPhi->addIncoming(ReducedPartRdx, LoopMiddleBlock);
break;
assert(IncomingEdgeBlockIdx >= 0 && "Invalid block index");
// Pick the other block.
int SelfEdgeBlockIdx = (IncomingEdgeBlockIdx ? 0 : 1);
- (RdxPhi)->setIncomingValue(SelfEdgeBlockIdx, ReducedPartRdx);
- (RdxPhi)->setIncomingValue(IncomingEdgeBlockIdx, RdxDesc.LoopExitInstr);
+ (RdxPhi)->setIncomingValue(SelfEdgeBlockIdx, BCBlockPhi);
+ (RdxPhi)->setIncomingValue(IncomingEdgeBlockIdx, LoopExitInst);
}// end of for each redux variable.
-
+
fixLCSSAPHIs();
+
+ // Make sure DomTree is updated.
+ updateAnalysis();
+
+ // Predicate any stores.
+ for (auto KV : PredicatedStores) {
+ BasicBlock::iterator I(KV.first);
+ auto *BB = SplitBlock(I->getParent(), &*std::next(I), DT, LI);
+ auto *T = SplitBlockAndInsertIfThen(KV.second, &*I, /*Unreachable=*/false,
+ /*BranchWeights=*/nullptr, DT);
+ I->moveBefore(T);
+ I->getParent()->setName("pred.store.if");
+ BB->setName("pred.store.continue");
+ }
+ DEBUG(DT->verifyDomTree());
+ // Remove redundant induction instructions.
+ cse(LoopVectorBody);
}
void InnerLoopVectorizer::fixLCSSAPHIs() {
LCSSAPhi->addIncoming(UndefValue::get(LCSSAPhi->getType()),
LoopMiddleBlock);
}
-}
+}
InnerLoopVectorizer::VectorParts
InnerLoopVectorizer::createEdgeMask(BasicBlock *Src, BasicBlock *Dst) {
return BlockMask;
}
-void InnerLoopVectorizer::widenPHIInstruction(Instruction *PN,
- InnerLoopVectorizer::VectorParts &Entry,
- LoopVectorizationLegality *Legal,
- unsigned UF, unsigned VF, PhiVector *PV) {
+void InnerLoopVectorizer::widenPHIInstruction(
+ Instruction *PN, InnerLoopVectorizer::VectorParts &Entry, unsigned UF,
+ unsigned VF, PhiVector *PV) {
PHINode* P = cast<PHINode>(PN);
// Handle reduction variables:
- if (Legal->getReductionVars()->count(P)) {
+ if (Legal->isReductionVariable(P)) {
for (unsigned part = 0; part < UF; ++part) {
// This is phase one of vectorizing PHIs.
Type *VecTy = (VF == 1) ? PN->getType() :
VectorType::get(PN->getType(), VF);
- Entry[part] = PHINode::Create(VecTy, 2, "vec.phi",
- LoopVectorBody-> getFirstInsertionPt());
+ Entry[part] = PHINode::Create(
+ VecTy, 2, "vec.phi", &*LoopVectorBody.back()->getFirstInsertionPt());
}
PV->push_back(P);
return;
setDebugLocFromInst(Builder, P);
// Check for PHI nodes that are lowered to vector selects.
if (P->getParent() != OrigLoop->getHeader()) {
- // We know that all PHIs in non header blocks are converted into
+ // We know that all PHIs in non-header blocks are converted into
// selects, so we don't have to worry about the insertion order and we
// can just use the builder.
// At this point we generate the predication tree. There may be
assert(Legal->getInductionVars()->count(P) &&
"Not an induction variable");
- LoopVectorizationLegality::InductionInfo II =
- Legal->getInductionVars()->lookup(P);
+ InductionDescriptor II = Legal->getInductionVars()->lookup(P);
- switch (II.IK) {
- case LoopVectorizationLegality::IK_NoInduction:
+ // FIXME: The newly created binary instructions should contain nsw/nuw flags,
+ // which can be found from the original scalar operations.
+ switch (II.getKind()) {
+ case InductionDescriptor::IK_NoInduction:
llvm_unreachable("Unknown induction");
- case LoopVectorizationLegality::IK_IntInduction: {
- assert(P->getType() == II.StartValue->getType() && "Types must match");
- Type *PhiTy = P->getType();
- Value *Broadcasted;
- if (P == OldInduction) {
- // Handle the canonical induction variable. We might have had to
- // extend the type.
- Broadcasted = Builder.CreateTrunc(Induction, PhiTy);
- } else {
- // Handle other induction variables that are now based on the
- // canonical one.
- Value *NormalizedIdx = Builder.CreateSub(Induction, ExtendedIdx,
- "normalized.idx");
- NormalizedIdx = Builder.CreateSExtOrTrunc(NormalizedIdx, PhiTy);
- Broadcasted = Builder.CreateAdd(II.StartValue, NormalizedIdx,
- "offset.idx");
+ case InductionDescriptor::IK_IntInduction: {
+ assert(P->getType() == II.getStartValue()->getType() &&
+ "Types must match");
+ // Handle other induction variables that are now based on the
+ // canonical one.
+ Value *V = Induction;
+ if (P != OldInduction) {
+ V = Builder.CreateSExtOrTrunc(Induction, P->getType());
+ V = II.transform(Builder, V);
+ V->setName("offset.idx");
}
- Broadcasted = getBroadcastInstrs(Broadcasted);
+ Value *Broadcasted = getBroadcastInstrs(V);
// After broadcasting the induction variable we need to make the vector
// consecutive by adding 0, 1, 2, etc.
for (unsigned part = 0; part < UF; ++part)
- Entry[part] = getConsecutiveVector(Broadcasted, VF * part, false);
+ Entry[part] = getStepVector(Broadcasted, VF * part, II.getStepValue());
return;
}
- case LoopVectorizationLegality::IK_ReverseIntInduction:
- case LoopVectorizationLegality::IK_PtrInduction:
- case LoopVectorizationLegality::IK_ReversePtrInduction:
- // Handle reverse integer and pointer inductions.
- Value *StartIdx = ExtendedIdx;
- // This is the normalized GEP that starts counting at zero.
- Value *NormalizedIdx = Builder.CreateSub(Induction, StartIdx,
- "normalized.idx");
-
- // Handle the reverse integer induction variable case.
- if (LoopVectorizationLegality::IK_ReverseIntInduction == II.IK) {
- IntegerType *DstTy = cast<IntegerType>(II.StartValue->getType());
- Value *CNI = Builder.CreateSExtOrTrunc(NormalizedIdx, DstTy,
- "resize.norm.idx");
- Value *ReverseInd = Builder.CreateSub(II.StartValue, CNI,
- "reverse.idx");
-
- // This is a new value so do not hoist it out.
- Value *Broadcasted = getBroadcastInstrs(ReverseInd);
- // After broadcasting the induction variable we need to make the
- // vector consecutive by adding ... -3, -2, -1, 0.
- for (unsigned part = 0; part < UF; ++part)
- Entry[part] = getConsecutiveVector(Broadcasted, -(int)VF * part,
- true);
- return;
- }
-
+ case InductionDescriptor::IK_PtrInduction:
// Handle the pointer induction variable case.
assert(P->getType()->isPointerTy() && "Unexpected type.");
-
- // Is this a reverse induction ptr or a consecutive induction ptr.
- bool Reverse = (LoopVectorizationLegality::IK_ReversePtrInduction ==
- II.IK);
-
+ // This is the normalized GEP that starts counting at zero.
+ Value *PtrInd = Induction;
+ PtrInd = Builder.CreateSExtOrTrunc(PtrInd, II.getStepValue()->getType());
// This is the vector of results. Notice that we don't generate
// vector geps because scalar geps result in better code.
for (unsigned part = 0; part < UF; ++part) {
if (VF == 1) {
- int EltIndex = (part) * (Reverse ? -1 : 1);
- Constant *Idx = ConstantInt::get(Induction->getType(), EltIndex);
- Value *GlobalIdx;
- if (Reverse)
- GlobalIdx = Builder.CreateSub(Idx, NormalizedIdx, "gep.ridx");
- else
- GlobalIdx = Builder.CreateAdd(NormalizedIdx, Idx, "gep.idx");
-
- Value *SclrGep = Builder.CreateGEP(II.StartValue, GlobalIdx,
- "next.gep");
+ int EltIndex = part;
+ Constant *Idx = ConstantInt::get(PtrInd->getType(), EltIndex);
+ Value *GlobalIdx = Builder.CreateAdd(PtrInd, Idx);
+ Value *SclrGep = II.transform(Builder, GlobalIdx);
+ SclrGep->setName("next.gep");
Entry[part] = SclrGep;
continue;
}
Value *VecVal = UndefValue::get(VectorType::get(P->getType(), VF));
for (unsigned int i = 0; i < VF; ++i) {
- int EltIndex = (i + part * VF) * (Reverse ? -1 : 1);
- Constant *Idx = ConstantInt::get(Induction->getType(), EltIndex);
- Value *GlobalIdx;
- if (!Reverse)
- GlobalIdx = Builder.CreateAdd(NormalizedIdx, Idx, "gep.idx");
- else
- GlobalIdx = Builder.CreateSub(Idx, NormalizedIdx, "gep.ridx");
-
- Value *SclrGep = Builder.CreateGEP(II.StartValue, GlobalIdx,
- "next.gep");
+ int EltIndex = i + part * VF;
+ Constant *Idx = ConstantInt::get(PtrInd->getType(), EltIndex);
+ Value *GlobalIdx = Builder.CreateAdd(PtrInd, Idx);
+ Value *SclrGep = II.transform(Builder, GlobalIdx);
+ SclrGep->setName("next.gep");
VecVal = Builder.CreateInsertElement(VecVal, SclrGep,
Builder.getInt32(i),
"insert.gep");
}
}
-void
-InnerLoopVectorizer::vectorizeBlockInLoop(LoopVectorizationLegality *Legal,
- BasicBlock *BB, PhiVector *PV) {
+void InnerLoopVectorizer::vectorizeBlockInLoop(BasicBlock *BB, PhiVector *PV) {
// For each instruction in the old loop.
for (BasicBlock::iterator it = BB->begin(), e = BB->end(); it != e; ++it) {
- VectorParts &Entry = WidenMap.get(it);
+ VectorParts &Entry = WidenMap.get(&*it);
+
switch (it->getOpcode()) {
case Instruction::Br:
// Nothing to do for PHIs and BR, since we already took care of the
// loop control flow instructions.
continue;
- case Instruction::PHI:{
+ case Instruction::PHI: {
// Vectorize PHINodes.
- widenPHIInstruction(it, Entry, Legal, UF, VF, PV);
+ widenPHIInstruction(&*it, Entry, UF, VF, PV);
continue;
}// End of PHI.
for (unsigned Part = 0; Part < UF; ++Part) {
Value *V = Builder.CreateBinOp(BinOp->getOpcode(), A[Part], B[Part]);
- // Update the NSW, NUW and Exact flags. Notice: V can be an Undef.
- BinaryOperator *VecOp = dyn_cast<BinaryOperator>(V);
- if (VecOp && isa<OverflowingBinaryOperator>(BinOp)) {
- VecOp->setHasNoSignedWrap(BinOp->hasNoSignedWrap());
- VecOp->setHasNoUnsignedWrap(BinOp->hasNoUnsignedWrap());
- }
- if (VecOp && isa<PossiblyExactOperator>(VecOp))
- VecOp->setIsExact(BinOp->isExact());
+ if (BinaryOperator *VecOp = dyn_cast<BinaryOperator>(V))
+ VecOp->copyIRFlags(BinOp);
Entry[Part] = V;
}
+
+ propagateMetadata(Entry, &*it);
break;
}
case Instruction::Select: {
// Widen selects.
// If the selector is loop invariant we can create a select
// instruction with a scalar condition. Otherwise, use vector-select.
- bool InvariantCond = SE->isLoopInvariant(SE->getSCEV(it->getOperand(0)),
- OrigLoop);
- setDebugLocFromInst(Builder, it);
+ auto *SE = PSE.getSE();
+ bool InvariantCond =
+ SE->isLoopInvariant(PSE.getSCEV(it->getOperand(0)), OrigLoop);
+ setDebugLocFromInst(Builder, &*it);
// The condition can be loop invariant but still defined inside the
// loop. This means that we can't just use the original 'cond' value.
VectorParts &Cond = getVectorValue(it->getOperand(0));
VectorParts &Op0 = getVectorValue(it->getOperand(1));
VectorParts &Op1 = getVectorValue(it->getOperand(2));
-
+
Value *ScalarCond = (VF == 1) ? Cond[0] :
Builder.CreateExtractElement(Cond[0], Builder.getInt32(0));
Op0[Part],
Op1[Part]);
}
+
+ propagateMetadata(Entry, &*it);
break;
}
// Widen compares. Generate vector compares.
bool FCmp = (it->getOpcode() == Instruction::FCmp);
CmpInst *Cmp = dyn_cast<CmpInst>(it);
- setDebugLocFromInst(Builder, it);
+ setDebugLocFromInst(Builder, &*it);
VectorParts &A = getVectorValue(it->getOperand(0));
VectorParts &B = getVectorValue(it->getOperand(1));
for (unsigned Part = 0; Part < UF; ++Part) {
- Value *C = 0;
- if (FCmp)
+ Value *C = nullptr;
+ if (FCmp) {
C = Builder.CreateFCmp(Cmp->getPredicate(), A[Part], B[Part]);
- else
+ cast<FCmpInst>(C)->copyFastMathFlags(&*it);
+ } else {
C = Builder.CreateICmp(Cmp->getPredicate(), A[Part], B[Part]);
+ }
Entry[Part] = C;
}
+
+ propagateMetadata(Entry, &*it);
break;
}
case Instruction::Store:
case Instruction::Load:
- vectorizeMemoryInstruction(it, Legal);
+ vectorizeMemoryInstruction(&*it);
break;
case Instruction::ZExt:
case Instruction::SExt:
case Instruction::FPTrunc:
case Instruction::BitCast: {
CastInst *CI = dyn_cast<CastInst>(it);
- setDebugLocFromInst(Builder, it);
+ setDebugLocFromInst(Builder, &*it);
/// Optimize the special case where the source is the induction
/// variable. Notice that we can only optimize the 'trunc' case
/// because: a. FP conversions lose precision, b. sext/zext may wrap,
Value *ScalarCast = Builder.CreateCast(CI->getOpcode(), Induction,
CI->getType());
Value *Broadcasted = getBroadcastInstrs(ScalarCast);
+ InductionDescriptor II =
+ Legal->getInductionVars()->lookup(OldInduction);
+ Constant *Step = ConstantInt::getSigned(
+ CI->getType(), II.getStepValue()->getSExtValue());
for (unsigned Part = 0; Part < UF; ++Part)
- Entry[Part] = getConsecutiveVector(Broadcasted, VF * Part, false);
+ Entry[Part] = getStepVector(Broadcasted, VF * Part, Step);
+ propagateMetadata(Entry, &*it);
break;
}
/// Vectorize casts.
VectorParts &A = getVectorValue(it->getOperand(0));
for (unsigned Part = 0; Part < UF; ++Part)
Entry[Part] = Builder.CreateCast(CI->getOpcode(), A[Part], DestTy);
+ propagateMetadata(Entry, &*it);
break;
}
// Ignore dbg intrinsics.
if (isa<DbgInfoIntrinsic>(it))
break;
- setDebugLocFromInst(Builder, it);
+ setDebugLocFromInst(Builder, &*it);
Module *M = BB->getParent()->getParent();
CallInst *CI = cast<CallInst>(it);
+
+ StringRef FnName = CI->getCalledFunction()->getName();
+ Function *F = CI->getCalledFunction();
+ Type *RetTy = ToVectorTy(CI->getType(), VF);
+ SmallVector<Type *, 4> Tys;
+ for (unsigned i = 0, ie = CI->getNumArgOperands(); i != ie; ++i)
+ Tys.push_back(ToVectorTy(CI->getArgOperand(i)->getType(), VF));
+
Intrinsic::ID ID = getIntrinsicIDForCall(CI, TLI);
- assert(ID && "Not an intrinsic call!");
- switch (ID) {
- case Intrinsic::lifetime_end:
- case Intrinsic::lifetime_start:
- scalarizeInstruction(it);
+ if (ID &&
+ (ID == Intrinsic::assume || ID == Intrinsic::lifetime_end ||
+ ID == Intrinsic::lifetime_start)) {
+ scalarizeInstruction(&*it);
+ break;
+ }
+ // The flag shows whether we use Intrinsic or a usual Call for vectorized
+ // version of the instruction.
+ // Is it beneficial to perform intrinsic call compared to lib call?
+ bool NeedToScalarize;
+ unsigned CallCost = getVectorCallCost(CI, VF, *TTI, TLI, NeedToScalarize);
+ bool UseVectorIntrinsic =
+ ID && getVectorIntrinsicCost(CI, VF, *TTI, TLI) <= CallCost;
+ if (!UseVectorIntrinsic && NeedToScalarize) {
+ scalarizeInstruction(&*it);
break;
- default:
- for (unsigned Part = 0; Part < UF; ++Part) {
- SmallVector<Value *, 4> Args;
- for (unsigned i = 0, ie = CI->getNumArgOperands(); i != ie; ++i) {
- VectorParts &Arg = getVectorValue(CI->getArgOperand(i));
- Args.push_back(Arg[Part]);
+ }
+
+ for (unsigned Part = 0; Part < UF; ++Part) {
+ SmallVector<Value *, 4> Args;
+ for (unsigned i = 0, ie = CI->getNumArgOperands(); i != ie; ++i) {
+ Value *Arg = CI->getArgOperand(i);
+ // Some intrinsics have a scalar argument - don't replace it with a
+ // vector.
+ if (!UseVectorIntrinsic || !hasVectorInstrinsicScalarOpd(ID, i)) {
+ VectorParts &VectorArg = getVectorValue(CI->getArgOperand(i));
+ Arg = VectorArg[Part];
}
- Type *Tys[] = {CI->getType()};
- if (VF > 1)
- Tys[0] = VectorType::get(CI->getType()->getScalarType(), VF);
+ Args.push_back(Arg);
+ }
- Function *F = Intrinsic::getDeclaration(M, ID, Tys);
- Entry[Part] = Builder.CreateCall(F, Args);
+ Function *VectorF;
+ if (UseVectorIntrinsic) {
+ // Use vector version of the intrinsic.
+ Type *TysForDecl[] = {CI->getType()};
+ if (VF > 1)
+ TysForDecl[0] = VectorType::get(CI->getType()->getScalarType(), VF);
+ VectorF = Intrinsic::getDeclaration(M, ID, TysForDecl);
+ } else {
+ // Use vector version of the library call.
+ StringRef VFnName = TLI->getVectorizedFunction(FnName, VF);
+ assert(!VFnName.empty() && "Vector function name is empty.");
+ VectorF = M->getFunction(VFnName);
+ if (!VectorF) {
+ // Generate a declaration
+ FunctionType *FTy = FunctionType::get(RetTy, Tys, false);
+ VectorF =
+ Function::Create(FTy, Function::ExternalLinkage, VFnName, M);
+ VectorF->copyAttributesFrom(F);
+ }
}
- break;
+ assert(VectorF && "Can't create vector function.");
+ Entry[Part] = Builder.CreateCall(VectorF, Args);
}
+
+ propagateMetadata(Entry, &*it);
break;
}
default:
// All other instructions are unsupported. Scalarize them.
- scalarizeInstruction(it);
+ scalarizeInstruction(&*it);
break;
}// end of switch.
}// end of for_each instr.
void InnerLoopVectorizer::updateAnalysis() {
// Forget the original basic block.
- SE->forgetLoop(OrigLoop);
+ PSE.getSE()->forgetLoop(OrigLoop);
// Update the dominator tree information.
assert(DT->properlyDominates(LoopBypassBlocks.front(), LoopExitBlock) &&
for (unsigned I = 1, E = LoopBypassBlocks.size(); I != E; ++I)
DT->addNewBlock(LoopBypassBlocks[I], LoopBypassBlocks[I-1]);
DT->addNewBlock(LoopVectorPreHeader, LoopBypassBlocks.back());
- DT->addNewBlock(LoopVectorBody, LoopVectorPreHeader);
- DT->addNewBlock(LoopMiddleBlock, LoopBypassBlocks.front());
- DT->addNewBlock(LoopScalarPreHeader, LoopMiddleBlock);
+
+ // We don't predicate stores by this point, so the vector body should be a
+ // single loop.
+ assert(LoopVectorBody.size() == 1 && "Expected single block loop!");
+ DT->addNewBlock(LoopVectorBody[0], LoopVectorPreHeader);
+
+ DT->addNewBlock(LoopMiddleBlock, LoopVectorBody.back());
+ DT->addNewBlock(LoopScalarPreHeader, LoopBypassBlocks[0]);
DT->changeImmediateDominator(LoopScalarBody, LoopScalarPreHeader);
- DT->changeImmediateDominator(LoopExitBlock, LoopMiddleBlock);
+ DT->changeImmediateDominator(LoopExitBlock, LoopBypassBlocks[0]);
+
+ DEBUG(DT->verifyDomTree());
+}
- DEBUG(DT->verifyAnalysis());
+/// \brief Check whether it is safe to if-convert this phi node.
+///
+/// Phi nodes with constant expressions that can trap are not safe to if
+/// convert.
+static bool canIfConvertPHINodes(BasicBlock *BB) {
+ for (BasicBlock::iterator I = BB->begin(), E = BB->end(); I != E; ++I) {
+ PHINode *Phi = dyn_cast<PHINode>(I);
+ if (!Phi)
+ return true;
+ for (unsigned p = 0, e = Phi->getNumIncomingValues(); p != e; ++p)
+ if (Constant *C = dyn_cast<Constant>(Phi->getIncomingValue(p)))
+ if (C->canTrap())
+ return false;
+ }
+ return true;
}
bool LoopVectorizationLegality::canVectorizeWithIfConvert() {
- if (!EnableIfConversion)
+ if (!EnableIfConversion) {
+ emitAnalysis(VectorizationReport() << "if-conversion is disabled");
return false;
+ }
assert(TheLoop->getNumBlocks() > 1 && "Single block loops are vectorizable");
- std::vector<BasicBlock*> &LoopBlocks = TheLoop->getBlocksVector();
// A list of pointers that we can safely read and write to.
SmallPtrSet<Value *, 8> SafePointes;
// Collect safe addresses.
- for (unsigned i = 0, e = LoopBlocks.size(); i < e; ++i) {
- BasicBlock *BB = LoopBlocks[i];
+ for (Loop::block_iterator BI = TheLoop->block_begin(),
+ BE = TheLoop->block_end(); BI != BE; ++BI) {
+ BasicBlock *BB = *BI;
if (blockNeedsPredication(BB))
continue;
}
// Collect the blocks that need predication.
- for (unsigned i = 0, e = LoopBlocks.size(); i < e; ++i) {
- BasicBlock *BB = LoopBlocks[i];
+ BasicBlock *Header = TheLoop->getHeader();
+ for (Loop::block_iterator BI = TheLoop->block_begin(),
+ BE = TheLoop->block_end(); BI != BE; ++BI) {
+ BasicBlock *BB = *BI;
// We don't support switch statements inside loops.
- if (!isa<BranchInst>(BB->getTerminator()))
+ if (!isa<BranchInst>(BB->getTerminator())) {
+ emitAnalysis(VectorizationReport(BB->getTerminator())
+ << "loop contains a switch statement");
return false;
+ }
// We must be able to predicate all blocks that need to be predicated.
- if (blockNeedsPredication(BB) && !blockCanBePredicated(BB, SafePointes))
+ if (blockNeedsPredication(BB)) {
+ if (!blockCanBePredicated(BB, SafePointes)) {
+ emitAnalysis(VectorizationReport(BB->getTerminator())
+ << "control flow cannot be substituted for a select");
+ return false;
+ }
+ } else if (BB != Header && !canIfConvertPHINodes(BB)) {
+ emitAnalysis(VectorizationReport(BB->getTerminator())
+ << "control flow cannot be substituted for a select");
return false;
+ }
}
// We can if-convert this loop.
bool LoopVectorizationLegality::canVectorize() {
// We must have a loop in canonical form. Loops with indirectbr in them cannot
// be canonicalized.
- if (!TheLoop->getLoopPreheader())
+ if (!TheLoop->getLoopPreheader()) {
+ emitAnalysis(
+ VectorizationReport() <<
+ "loop control flow is not understood by vectorizer");
return false;
+ }
// We can only vectorize innermost loops.
- if (TheLoop->getSubLoopsVector().size())
+ if (!TheLoop->empty()) {
+ emitAnalysis(VectorizationReport() << "loop is not the innermost loop");
return false;
+ }
// We must have a single backedge.
- if (TheLoop->getNumBackEdges() != 1)
+ if (TheLoop->getNumBackEdges() != 1) {
+ emitAnalysis(
+ VectorizationReport() <<
+ "loop control flow is not understood by vectorizer");
return false;
+ }
// We must have a single exiting block.
- if (!TheLoop->getExitingBlock())
+ if (!TheLoop->getExitingBlock()) {
+ emitAnalysis(
+ VectorizationReport() <<
+ "loop control flow is not understood by vectorizer");
+ 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(
+ VectorizationReport() <<
+ "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');
- // Check if we can if-convert non single-bb loops.
+ // Check if we can if-convert non-single-bb loops.
unsigned NumBlocks = TheLoop->getNumBlocks();
if (NumBlocks != 1 && !canVectorizeWithIfConvert()) {
DEBUG(dbgs() << "LV: Can't if-convert the loop.\n");
}
// ScalarEvolution needs to be able to find the exit count.
- const SCEV *ExitCount = SE->getBackedgeTakenCount(TheLoop);
- if (ExitCount == SE->getCouldNotCompute()) {
+ const SCEV *ExitCount = PSE.getSE()->getBackedgeTakenCount(TheLoop);
+ if (ExitCount == PSE.getSE()->getCouldNotCompute()) {
+ emitAnalysis(VectorizationReport()
+ << "could not determine number of loop iterations");
DEBUG(dbgs() << "LV: SCEV could not compute the loop exit count.\n");
return false;
}
- // Do not loop-vectorize loops with a tiny trip count.
- BasicBlock *Latch = TheLoop->getLoopLatch();
- unsigned TC = SE->getSmallConstantTripCount(TheLoop, Latch);
- if (TC > 0u && TC < TinyTripCountVectorThreshold) {
- DEBUG(dbgs() << "LV: Found a loop with a very small trip count. " <<
- "This loop is not worth vectorizing.\n");
- return false;
- }
-
// Check if we can vectorize the instructions and CFG in this loop.
if (!canVectorizeInstrs()) {
DEBUG(dbgs() << "LV: Can't vectorize the instructions or CFG\n");
// Collect all of the variables that remain uniform after vectorization.
collectLoopUniforms();
- DEBUG(dbgs() << "LV: We can vectorize this loop" <<
- (PtrRtCheck.Need ? " (with a runtime bound check)" : "")
- <<"!\n");
+ DEBUG(dbgs() << "LV: We can vectorize this loop"
+ << (LAI->getRuntimePointerChecking()->Need
+ ? " (with a runtime bound check)"
+ : "")
+ << "!\n");
+
+ bool UseInterleaved = TTI->enableInterleavedAccessVectorization();
+
+ // If an override option has been passed in for interleaved accesses, use it.
+ if (EnableInterleavedMemAccesses.getNumOccurrences() > 0)
+ UseInterleaved = EnableInterleavedMemAccesses;
+
+ // Analyze interleaved memory accesses.
+ if (UseInterleaved)
+ InterleaveInfo.analyzeInterleaving(Strides);
+
+ unsigned SCEVThreshold = VectorizeSCEVCheckThreshold;
+ if (Hints->getForce() == LoopVectorizeHints::FK_Enabled)
+ SCEVThreshold = PragmaVectorizeSCEVCheckThreshold;
+
+ if (PSE.getUnionPredicate().getComplexity() > SCEVThreshold) {
+ emitAnalysis(VectorizationReport()
+ << "Too many SCEV assumptions need to be made and checked "
+ << "at runtime");
+ DEBUG(dbgs() << "LV: Too many SCEV checks needed.\n");
+ return false;
+ }
// Okay! We can vectorize. At this point we don't have any other mem analysis
// which may limit our maximum vectorization factor, so just return true with
return true;
}
-static Type *convertPointerToIntegerType(DataLayout &DL, Type *Ty) {
+static Type *convertPointerToIntegerType(const DataLayout &DL, Type *Ty) {
if (Ty->isPointerTy())
return DL.getIntPtrType(Ty);
+ // It is possible that char's or short's overflow when we ask for the loop's
+ // trip count, work around this by changing the type size.
+ if (Ty->getScalarSizeInBits() < 32)
+ return Type::getInt32Ty(Ty->getContext());
+
return Ty;
}
-static Type* getWiderType(DataLayout &DL, Type *Ty0, Type *Ty1) {
+static Type* getWiderType(const DataLayout &DL, Type *Ty0, Type *Ty1) {
Ty0 = convertPointerToIntegerType(DL, Ty0);
Ty1 = convertPointerToIntegerType(DL, Ty1);
if (Ty0->getScalarSizeInBits() > Ty1->getScalarSizeInBits())
/// \brief Check that the instruction has outside loop users and is not an
/// identified reduction variable.
static bool hasOutsideLoopUser(const Loop *TheLoop, Instruction *Inst,
- SmallPtrSet<Value *, 4> &Reductions) {
+ SmallPtrSetImpl<Value *> &Reductions) {
// Reduction instructions are allowed to have exit users. All other
// instructions must not have external users.
if (!Reductions.count(Inst))
//Check that all of the users of the loop are inside the BB.
- for (Value::use_iterator I = Inst->use_begin(), E = Inst->use_end();
- I != E; ++I) {
- Instruction *U = cast<Instruction>(*I);
+ for (User *U : Inst->users()) {
+ Instruction *UI = cast<Instruction>(U);
// This user may be a reduction exit value.
- if (!TheLoop->contains(U)) {
- DEBUG(dbgs() << "LV: Found an outside user for : " << *U << '\n');
+ if (!TheLoop->contains(UI)) {
+ DEBUG(dbgs() << "LV: Found an outside user for : " << *UI << '\n');
return true;
}
}
}
bool LoopVectorizationLegality::canVectorizeInstrs() {
- BasicBlock *PreHeader = TheLoop->getLoopPreheader();
BasicBlock *Header = TheLoop->getHeader();
// Look for the attribute signaling the absence of NaNs.
Function &F = *Header->getParent();
+ const DataLayout &DL = F.getParent()->getDataLayout();
if (F.hasFnAttribute("no-nans-fp-math"))
- HasFunNoNaNAttr = F.getAttributes().getAttribute(
- AttributeSet::FunctionIndex,
- "no-nans-fp-math").getValueAsString() == "true";
+ HasFunNoNaNAttr =
+ F.getFnAttribute("no-nans-fp-math").getValueAsString() == "true";
// For each block in the loop.
for (Loop::block_iterator bb = TheLoop->block_begin(),
if (!PhiTy->isIntegerTy() &&
!PhiTy->isFloatingPointTy() &&
!PhiTy->isPointerTy()) {
+ emitAnalysis(VectorizationReport(&*it)
+ << "loop control flow is not understood by vectorizer");
DEBUG(dbgs() << "LV: Found an non-int non-pointer PHI.\n");
return false;
}
if (*bb != Header) {
// Check that this instruction has no outside users or is an
// identified reduction value with an outside user.
- if(!hasOutsideLoopUser(TheLoop, it, AllowedExit))
+ if (!hasOutsideLoopUser(TheLoop, &*it, AllowedExit))
continue;
+ emitAnalysis(VectorizationReport(&*it) <<
+ "value could not be identified as "
+ "an induction or reduction variable");
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(VectorizationReport(&*it)
+ << "control flow not understood by vectorizer");
DEBUG(dbgs() << "LV: Found an invalid PHI.\n");
return false;
}
- // This is the value coming from the preheader.
- Value *StartValue = Phi->getIncomingValueForBlock(PreHeader);
- // Check if this is an induction variable.
- InductionKind IK = isInductionVariable(Phi);
-
- if (IK_NoInduction != IK) {
+ InductionDescriptor ID;
+ if (InductionDescriptor::isInductionPHI(Phi, PSE.getSE(), ID)) {
+ Inductions[Phi] = ID;
// Get the widest type.
if (!WidestIndTy)
- WidestIndTy = convertPointerToIntegerType(*DL, PhiTy);
+ WidestIndTy = convertPointerToIntegerType(DL, PhiTy);
else
- WidestIndTy = getWiderType(*DL, PhiTy, WidestIndTy);
+ WidestIndTy = getWiderType(DL, PhiTy, WidestIndTy);
// Int inductions are special because we only allow one IV.
- if (IK == IK_IntInduction) {
+ if (ID.getKind() == InductionDescriptor::IK_IntInduction &&
+ ID.getStepValue()->isOne() &&
+ isa<Constant>(ID.getStartValue()) &&
+ cast<Constant>(ID.getStartValue())->isNullValue()) {
// Use the phi node with the widest type as induction. Use the last
// one if there are multiple (no good reason for doing this other
- // than it is expedient).
+ // than it is expedient). We've checked that it begins at zero and
+ // steps by one, so this is a canonical induction variable.
if (!Induction || PhiTy == WidestIndTy)
Induction = Phi;
}
- DEBUG(dbgs() << "LV: Found an induction variable.\n");
- Inductions[Phi] = InductionInfo(StartValue, IK);
-
- // Until we explicitly handle the case of an induction variable with
- // an outside loop user we have to give up vectorizing this loop.
- if (hasOutsideLoopUser(TheLoop, it, AllowedExit))
- return false;
-
- continue;
- }
-
- if (AddReductionVar(Phi, RK_IntegerAdd)) {
- DEBUG(dbgs() << "LV: Found an ADD reduction PHI."<< *Phi <<"\n");
- continue;
- }
- if (AddReductionVar(Phi, RK_IntegerMult)) {
- DEBUG(dbgs() << "LV: Found a MUL reduction PHI."<< *Phi <<"\n");
- continue;
- }
- if (AddReductionVar(Phi, RK_IntegerOr)) {
- DEBUG(dbgs() << "LV: Found an OR reduction PHI."<< *Phi <<"\n");
- continue;
- }
- if (AddReductionVar(Phi, RK_IntegerAnd)) {
- DEBUG(dbgs() << "LV: Found an AND reduction PHI."<< *Phi <<"\n");
- continue;
- }
- if (AddReductionVar(Phi, RK_IntegerXor)) {
- DEBUG(dbgs() << "LV: Found a XOR reduction PHI."<< *Phi <<"\n");
- continue;
- }
- if (AddReductionVar(Phi, RK_IntegerMinMax)) {
- DEBUG(dbgs() << "LV: Found a MINMAX reduction PHI."<< *Phi <<"\n");
- continue;
- }
- if (AddReductionVar(Phi, RK_FloatMult)) {
- DEBUG(dbgs() << "LV: Found an FMult reduction PHI."<< *Phi <<"\n");
- continue;
- }
- if (AddReductionVar(Phi, RK_FloatAdd)) {
- DEBUG(dbgs() << "LV: Found an FAdd reduction PHI."<< *Phi <<"\n");
- continue;
- }
- if (AddReductionVar(Phi, RK_FloatMinMax)) {
- DEBUG(dbgs() << "LV: Found an float MINMAX reduction PHI."<< *Phi <<
- "\n");
- continue;
- }
-
- DEBUG(dbgs() << "LV: Found an unidentified PHI."<< *Phi <<"\n");
- return false;
- }// end of PHI handling
-
- // We still don't handle functions. However, we can ignore dbg intrinsic
- // calls and we do handle certain intrinsic and libm functions.
- CallInst *CI = dyn_cast<CallInst>(it);
- if (CI && !getIntrinsicIDForCall(CI, TLI) && !isa<DbgInfoIntrinsic>(CI)) {
- DEBUG(dbgs() << "LV: Found a call site.\n");
- return false;
- }
-
- // Check that the instruction return type is vectorizable.
- if (!VectorType::isValidElementType(it->getType()) &&
- !it->getType()->isVoidTy()) {
- DEBUG(dbgs() << "LV: Found unvectorizable type.\n");
- return false;
- }
-
- // Check that the stored type is vectorizable.
- if (StoreInst *ST = dyn_cast<StoreInst>(it)) {
- Type *T = ST->getValueOperand()->getType();
- if (!VectorType::isValidElementType(T))
- return false;
- }
-
- // Reduction instructions are allowed to have exit users.
- // All other instructions must not have external users.
- if (hasOutsideLoopUser(TheLoop, it, AllowedExit))
- return false;
-
- } // next instr.
-
- }
-
- if (!Induction) {
- DEBUG(dbgs() << "LV: Did not find one integer induction var.\n");
- if (Inductions.empty())
- return false;
- }
-
- return true;
-}
-
-void LoopVectorizationLegality::collectLoopUniforms() {
- // We now know that the loop is vectorizable!
- // Collect variables that will remain uniform after vectorization.
- std::vector<Value*> Worklist;
- BasicBlock *Latch = TheLoop->getLoopLatch();
-
- // Start with the conditional branch and walk up the block.
- Worklist.push_back(Latch->getTerminator()->getOperand(0));
-
- while (Worklist.size()) {
- Instruction *I = dyn_cast<Instruction>(Worklist.back());
- Worklist.pop_back();
-
- // Look at instructions inside this loop.
- // Stop when reaching PHI nodes.
- // TODO: we need to follow values all over the loop, not only in this block.
- if (!I || !TheLoop->contains(I) || isa<PHINode>(I))
- continue;
-
- // This is a known uniform.
- Uniforms.insert(I);
-
- // Insert all operands.
- Worklist.insert(Worklist.end(), I->op_begin(), I->op_end());
- }
-}
-
-namespace {
-/// \brief Analyses memory accesses in a loop.
-///
-/// Checks whether run time pointer checks are needed and builds sets for data
-/// dependence checking.
-class AccessAnalysis {
-public:
- /// \brief Read or write access location.
- typedef PointerIntPair<Value *, 1, bool> MemAccessInfo;
- typedef SmallPtrSet<MemAccessInfo, 8> MemAccessInfoSet;
-
- /// \brief Set of potential dependent memory accesses.
- typedef EquivalenceClasses<MemAccessInfo> DepCandidates;
-
- AccessAnalysis(DataLayout *Dl, DepCandidates &DA) :
- DL(Dl), DepCands(DA), AreAllWritesIdentified(true),
- AreAllReadsIdentified(true), IsRTCheckNeeded(false) {}
-
- /// \brief Register a load and whether it is only read from.
- void addLoad(Value *Ptr, bool IsReadOnly) {
- Accesses.insert(MemAccessInfo(Ptr, false));
- if (IsReadOnly)
- ReadOnlyPtr.insert(Ptr);
- }
-
- /// \brief Register a store.
- void addStore(Value *Ptr) {
- Accesses.insert(MemAccessInfo(Ptr, true));
- }
-
- /// \brief Check whether we can check the pointers at runtime for
- /// non-intersection.
- bool canCheckPtrAtRT(LoopVectorizationLegality::RuntimePointerCheck &RtCheck,
- unsigned &NumComparisons, ScalarEvolution *SE,
- Loop *TheLoop);
-
- /// \brief Goes over all memory accesses, checks whether a RT check is needed
- /// and builds sets of dependent accesses.
- void buildDependenceSets() {
- // Process read-write pointers first.
- processMemAccesses(false);
- // Next, process read pointers.
- processMemAccesses(true);
- }
-
- bool isRTCheckNeeded() { return IsRTCheckNeeded; }
-
- bool isDependencyCheckNeeded() { return !CheckDeps.empty(); }
-
- MemAccessInfoSet &getDependenciesToCheck() { return CheckDeps; }
-
-private:
- typedef SetVector<MemAccessInfo> PtrAccessSet;
- typedef DenseMap<Value*, MemAccessInfo> UnderlyingObjToAccessMap;
-
- /// \brief Go over all memory access or only the deferred ones if
- /// \p UseDeferred is true and check whether runtime pointer checks are needed
- /// and build sets of dependency check candidates.
- void processMemAccesses(bool UseDeferred);
-
- /// Set of all accesses.
- PtrAccessSet Accesses;
-
- /// Set of access to check after all writes have been processed.
- PtrAccessSet DeferredAccesses;
-
- /// Map of pointers to last access encountered.
- UnderlyingObjToAccessMap ObjToLastAccess;
-
- /// Set of accesses that need a further dependence check.
- MemAccessInfoSet CheckDeps;
-
- /// Set of pointers that are read only.
- SmallPtrSet<Value*, 16> ReadOnlyPtr;
-
- /// Set of underlying objects already written to.
- SmallPtrSet<Value*, 16> WriteObjects;
-
- DataLayout *DL;
-
- /// Sets of potentially dependent accesses - members of one set share an
- /// underlying pointer. The set "CheckDeps" identfies which sets really need a
- /// dependence check.
- DepCandidates &DepCands;
-
- bool AreAllWritesIdentified;
- bool AreAllReadsIdentified;
- bool IsRTCheckNeeded;
-};
-
-} // end anonymous namespace
-
-/// \brief Check whether a pointer can participate in a runtime bounds check.
-static bool hasComputableBounds(ScalarEvolution *SE, Value *Ptr) {
- const SCEV *PtrScev = SE->getSCEV(Ptr);
- const SCEVAddRecExpr *AR = dyn_cast<SCEVAddRecExpr>(PtrScev);
- if (!AR)
- return false;
-
- return AR->isAffine();
-}
-
-bool AccessAnalysis::canCheckPtrAtRT(
- LoopVectorizationLegality::RuntimePointerCheck &RtCheck,
- unsigned &NumComparisons, ScalarEvolution *SE,
- Loop *TheLoop) {
- // Find pointers with computable bounds. We are going to use this information
- // to place a runtime bound check.
- unsigned NumReadPtrChecks = 0;
- unsigned NumWritePtrChecks = 0;
- bool CanDoRT = true;
-
- bool IsDepCheckNeeded = isDependencyCheckNeeded();
- // We assign consecutive id to access from different dependence sets.
- // Accesses within the same set don't need a runtime check.
- unsigned RunningDepId = 1;
- DenseMap<Value *, unsigned> DepSetId;
-
- for (PtrAccessSet::iterator AI = Accesses.begin(), AE = Accesses.end();
- AI != AE; ++AI) {
- const MemAccessInfo &Access = *AI;
- Value *Ptr = Access.getPointer();
- bool IsWrite = Access.getInt();
-
- // Just add write checks if we have both.
- if (!IsWrite && Accesses.count(MemAccessInfo(Ptr, true)))
- continue;
-
- if (IsWrite)
- ++NumWritePtrChecks;
- else
- ++NumReadPtrChecks;
-
- if (hasComputableBounds(SE, Ptr)) {
- // The id of the dependence set.
- unsigned DepId;
-
- if (IsDepCheckNeeded) {
- Value *Leader = DepCands.getLeaderValue(Access).getPointer();
- unsigned &LeaderId = DepSetId[Leader];
- if (!LeaderId)
- LeaderId = RunningDepId++;
- DepId = LeaderId;
- } else
- // Each access has its own dependence set.
- DepId = RunningDepId++;
-
- RtCheck.insert(SE, TheLoop, Ptr, IsWrite, DepId);
-
- DEBUG(dbgs() << "LV: Found a runtime check ptr:" << *Ptr << '\n');
- } else {
- CanDoRT = false;
- }
- }
-
- if (IsDepCheckNeeded && CanDoRT && RunningDepId == 2)
- NumComparisons = 0; // Only one dependence set.
- else {
- NumComparisons = (NumWritePtrChecks * (NumReadPtrChecks +
- NumWritePtrChecks - 1));
- }
-
- // If the pointers that we would use for the bounds comparison have different
- // address spaces, assume the values aren't directly comparable, so we can't
- // use them for the runtime check. We also have to assume they could
- // overlap. In the future there should be metadata for whether address spaces
- // are disjoint.
- unsigned NumPointers = RtCheck.Pointers.size();
- for (unsigned i = 0; i < NumPointers; ++i) {
- for (unsigned j = i + 1; j < NumPointers; ++j) {
- // Only need to check pointers between two different dependency sets.
- if (RtCheck.DependencySetId[i] == RtCheck.DependencySetId[j])
- continue;
-
- Value *PtrI = RtCheck.Pointers[i];
- Value *PtrJ = RtCheck.Pointers[j];
-
- unsigned ASi = PtrI->getType()->getPointerAddressSpace();
- unsigned ASj = PtrJ->getType()->getPointerAddressSpace();
- if (ASi != ASj) {
- DEBUG(dbgs() << "LV: Runtime check would require comparison between"
- " different address spaces\n");
- return false;
- }
- }
- }
-
- return CanDoRT;
-}
-
-static bool isFunctionScopeIdentifiedObject(Value *Ptr) {
- return isNoAliasArgument(Ptr) || isNoAliasCall(Ptr) || isa<AllocaInst>(Ptr);
-}
-
-void AccessAnalysis::processMemAccesses(bool UseDeferred) {
- // We process the set twice: first we process read-write pointers, last we
- // process read-only pointers. This allows us to skip dependence tests for
- // read-only pointers.
-
- PtrAccessSet &S = UseDeferred ? DeferredAccesses : Accesses;
- for (PtrAccessSet::iterator AI = S.begin(), AE = S.end(); AI != AE; ++AI) {
- const MemAccessInfo &Access = *AI;
- Value *Ptr = Access.getPointer();
- bool IsWrite = Access.getInt();
-
- 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".
- bool IsReadOnlyPtr = ReadOnlyPtr.count(Ptr) && !IsWrite;
- if (!UseDeferred && IsReadOnlyPtr) {
- DeferredAccesses.insert(Access);
- continue;
- }
-
- bool NeedDepCheck = false;
- // Check whether there is the possiblity of dependency because of underlying
- // objects being the same.
- typedef SmallVector<Value*, 16> ValueVector;
- ValueVector TempObjects;
- GetUnderlyingObjects(Ptr, TempObjects, DL);
- for (ValueVector::iterator UI = TempObjects.begin(), UE = TempObjects.end();
- UI != UE; ++UI) {
- Value *UnderlyingObj = *UI;
-
- // If this is a write then it needs to be an identified object. If this a
- // read and all writes (so far) are identified function scope objects we
- // don't need an identified underlying object but only an Argument (the
- // next write is going to invalidate this assumption if it is
- // unidentified).
- // This is a micro-optimization for the case where all writes are
- // identified and we have one argument pointer.
- // Otherwise, we do need a runtime check.
- if ((IsWrite && !isFunctionScopeIdentifiedObject(UnderlyingObj)) ||
- (!IsWrite && (!AreAllWritesIdentified ||
- !isa<Argument>(UnderlyingObj)) &&
- !isIdentifiedObject(UnderlyingObj))) {
- DEBUG(dbgs() << "LV: Found an unidentified " <<
- (IsWrite ? "write" : "read" ) << " ptr: " << *UnderlyingObj <<
- "\n");
- IsRTCheckNeeded = (IsRTCheckNeeded ||
- !isIdentifiedObject(UnderlyingObj) ||
- !AreAllReadsIdentified);
-
- if (IsWrite)
- AreAllWritesIdentified = false;
- if (!IsWrite)
- AreAllReadsIdentified = false;
- }
-
- // 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) && WriteObjects.count(UnderlyingObj))
- NeedDepCheck = true;
-
- if (IsWrite)
- WriteObjects.insert(UnderlyingObj);
-
- // Create sets of pointers connected by shared underlying objects.
- UnderlyingObjToAccessMap::iterator Prev =
- ObjToLastAccess.find(UnderlyingObj);
- if (Prev != ObjToLastAccess.end())
- DepCands.unionSets(Access, Prev->second);
-
- ObjToLastAccess[UnderlyingObj] = Access;
- }
-
- if (NeedDepCheck)
- CheckDeps.insert(Access);
- }
-}
-
-namespace {
-/// \brief Checks memory dependences among accesses to the same underlying
-/// object to determine whether there vectorization is legal or not (and at
-/// which vectorization factor).
-///
-/// This class works under the assumption that we already checked that memory
-/// locations with different underlying pointers are "must-not alias".
-/// We use the ScalarEvolution framework to symbolically evalutate access
-/// functions pairs. Since we currently don't restructure the loop we can rely
-/// on the program order of memory accesses to determine their safety.
-/// At the moment we will only deem accesses as safe for:
-/// * A negative constant distance assuming program order.
-///
-/// Safe: tmp = a[i + 1]; OR a[i + 1] = x;
-/// a[i] = tmp; y = a[i];
-///
-/// The latter case is safe because later checks guarantuee that there can't
-/// be a cycle through a phi node (that is, we check that "x" and "y" is not
-/// the same variable: a header phi can only be an induction or a reduction, a
-/// reduction can't have a memory sink, an induction can't have a memory
-/// source). This is important and must not be violated (or we have to
-/// resort to checking for cycles through memory).
-///
-/// * A positive constant distance assuming program order that is bigger
-/// than the biggest memory access.
-///
-/// tmp = a[i] OR b[i] = x
-/// a[i+2] = tmp y = b[i+2];
-///
-/// Safe distance: 2 x sizeof(a[0]), and 2 x sizeof(b[0]), respectively.
-///
-/// * Zero distances and all accesses have the same size.
-///
-class MemoryDepChecker {
-public:
- typedef PointerIntPair<Value *, 1, bool> MemAccessInfo;
- typedef SmallPtrSet<MemAccessInfo, 8> MemAccessInfoSet;
-
- MemoryDepChecker(ScalarEvolution *Se, DataLayout *Dl, const Loop *L) :
- SE(Se), DL(Dl), InnermostLoop(L), AccessIdx(0) {}
-
- /// \brief Register the location (instructions are given increasing numbers)
- /// of a write access.
- void addAccess(StoreInst *SI) {
- Value *Ptr = SI->getPointerOperand();
- Accesses[MemAccessInfo(Ptr, true)].push_back(AccessIdx);
- InstMap.push_back(SI);
- ++AccessIdx;
- }
-
- /// \brief Register the location (instructions are given increasing numbers)
- /// of a write access.
- void addAccess(LoadInst *LI) {
- Value *Ptr = LI->getPointerOperand();
- Accesses[MemAccessInfo(Ptr, false)].push_back(AccessIdx);
- InstMap.push_back(LI);
- ++AccessIdx;
- }
-
- /// \brief Check whether the dependencies between the accesses are safe.
- ///
- /// Only checks sets with elements in \p CheckDeps.
- bool areDepsSafe(AccessAnalysis::DepCandidates &AccessSets,
- MemAccessInfoSet &CheckDeps);
-
- /// \brief The maximum number of bytes of a vector register we can vectorize
- /// the accesses safely with.
- unsigned getMaxSafeDepDistBytes() { return MaxSafeDepDistBytes; }
-
-private:
- ScalarEvolution *SE;
- DataLayout *DL;
- const Loop *InnermostLoop;
-
- /// \brief Maps access locations (ptr, read/write) to program order.
- DenseMap<MemAccessInfo, std::vector<unsigned> > Accesses;
-
- /// \brief Memory access instructions in program order.
- SmallVector<Instruction *, 16> InstMap;
-
- /// \brief The program order index to be used for the next instruction.
- unsigned AccessIdx;
-
- // We can access this many bytes in parallel safely.
- unsigned MaxSafeDepDistBytes;
-
- /// \brief Check whether there is a plausible dependence between the two
- /// accesses.
- ///
- /// Access \p A must happen before \p B in program order. The two indices
- /// identify the index into the program order map.
- ///
- /// This function checks whether there is a plausible dependence (or the
- /// absence of such can't be proved) between the two accesses. If there is a
- /// plausible dependence but the dependence distance is bigger than one
- /// element access it records this distance in \p MaxSafeDepDistBytes (if this
- /// distance is smaller than any other distance encountered so far).
- /// Otherwise, this function returns true signaling a possible dependence.
- bool isDependent(const MemAccessInfo &A, unsigned AIdx,
- const MemAccessInfo &B, unsigned BIdx);
-
- /// \brief Check whether the data dependence could prevent store-load
- /// forwarding.
- bool couldPreventStoreLoadForward(unsigned Distance, unsigned TypeByteSize);
-};
-
-} // end anonymous namespace
-
-static bool isInBoundsGep(Value *Ptr) {
- if (GetElementPtrInst *GEP = dyn_cast<GetElementPtrInst>(Ptr))
- return GEP->isInBounds();
- return false;
-}
-
-/// \brief Check whether the access through \p Ptr has a constant stride.
-static int isStridedPtr(ScalarEvolution *SE, DataLayout *DL, Value *Ptr,
- const Loop *Lp) {
- const Type *Ty = Ptr->getType();
- assert(Ty->isPointerTy() && "Unexpected non ptr");
-
- // Make sure that the pointer does not point to aggregate types.
- const PointerType *PtrTy = cast<PointerType>(Ty);
- if (PtrTy->getElementType()->isAggregateType()) {
- DEBUG(dbgs() << "LV: Bad stride - Not a pointer to a scalar type" << *Ptr <<
- "\n");
- return 0;
- }
-
- const SCEV *PtrScev = SE->getSCEV(Ptr);
- const SCEVAddRecExpr *AR = dyn_cast<SCEVAddRecExpr>(PtrScev);
- if (!AR) {
- DEBUG(dbgs() << "LV: Bad stride - Not an AddRecExpr pointer "
- << *Ptr << " SCEV: " << *PtrScev << "\n");
- return 0;
- }
-
- // The accesss function must stride over the innermost loop.
- if (Lp != AR->getLoop()) {
- DEBUG(dbgs() << "LV: Bad stride - Not striding over innermost loop " <<
- *Ptr << " SCEV: " << *PtrScev << "\n");
- }
-
- // The address calculation must not wrap. Otherwise, a dependence could be
- // inverted.
- // An inbounds getelementptr that is a AddRec with a unit stride
- // cannot wrap per definition. The unit stride requirement is checked later.
- // An getelementptr without an inbounds attribute and unit stride would have
- // to access the pointer value "0" which is undefined behavior in address
- // space 0, therefore we can also vectorize this case.
- bool IsInBoundsGEP = isInBoundsGep(Ptr);
- bool IsNoWrapAddRec = AR->getNoWrapFlags(SCEV::NoWrapMask);
- bool IsInAddressSpaceZero = PtrTy->getAddressSpace() == 0;
- if (!IsNoWrapAddRec && !IsInBoundsGEP && !IsInAddressSpaceZero) {
- DEBUG(dbgs() << "LV: Bad stride - Pointer may wrap in the address space "
- << *Ptr << " SCEV: " << *PtrScev << "\n");
- return 0;
- }
-
- // Check the step is constant.
- const SCEV *Step = AR->getStepRecurrence(*SE);
-
- // Calculate the pointer stride and check if it is consecutive.
- const SCEVConstant *C = dyn_cast<SCEVConstant>(Step);
- if (!C) {
- DEBUG(dbgs() << "LV: Bad stride - Not a constant strided " << *Ptr <<
- " SCEV: " << *PtrScev << "\n");
- return 0;
- }
-
- int64_t Size = DL->getTypeAllocSize(PtrTy->getElementType());
- const APInt &APStepVal = C->getValue()->getValue();
-
- // Huge step value - give up.
- if (APStepVal.getBitWidth() > 64)
- return 0;
-
- int64_t StepVal = APStepVal.getSExtValue();
-
- // Strided access.
- int64_t Stride = StepVal / Size;
- int64_t Rem = StepVal % Size;
- if (Rem)
- return 0;
-
- // If the SCEV could wrap but we have an inbounds gep with a unit stride we
- // know we can't "wrap around the address space". In case of address space
- // zero we know that this won't happen without triggering undefined behavior.
- if (!IsNoWrapAddRec && (IsInBoundsGEP || IsInAddressSpaceZero) &&
- Stride != 1 && Stride != -1)
- return 0;
-
- return Stride;
-}
-
-bool MemoryDepChecker::couldPreventStoreLoadForward(unsigned Distance,
- unsigned TypeByteSize) {
- // If loads occur at a distance that is not a multiple of a feasible vector
- // factor store-load forwarding does not take place.
- // Positive dependences might cause troubles because vectorizing them might
- // prevent store-load forwarding making vectorized code run a lot slower.
- // a[i] = a[i-3] ^ a[i-8];
- // The stores to a[i:i+1] don't align with the stores to a[i-3:i-2] and
- // hence on your typical architecture store-load forwarding does not take
- // place. Vectorizing in such cases does not make sense.
- // Store-load forwarding distance.
- const unsigned NumCyclesForStoreLoadThroughMemory = 8*TypeByteSize;
- // Maximum vector factor.
- unsigned MaxVFWithoutSLForwardIssues = MaxVectorWidth*TypeByteSize;
- if(MaxSafeDepDistBytes < MaxVFWithoutSLForwardIssues)
- MaxVFWithoutSLForwardIssues = MaxSafeDepDistBytes;
-
- for (unsigned vf = 2*TypeByteSize; vf <= MaxVFWithoutSLForwardIssues;
- vf *= 2) {
- if (Distance % vf && Distance / vf < NumCyclesForStoreLoadThroughMemory) {
- MaxVFWithoutSLForwardIssues = (vf >>=1);
- break;
- }
- }
-
- if (MaxVFWithoutSLForwardIssues< 2*TypeByteSize) {
- DEBUG(dbgs() << "LV: Distance " << Distance <<
- " that could cause a store-load forwarding conflict\n");
- return true;
- }
-
- if (MaxVFWithoutSLForwardIssues < MaxSafeDepDistBytes &&
- MaxVFWithoutSLForwardIssues != MaxVectorWidth*TypeByteSize)
- MaxSafeDepDistBytes = MaxVFWithoutSLForwardIssues;
- return false;
-}
-
-bool MemoryDepChecker::isDependent(const MemAccessInfo &A, unsigned AIdx,
- const MemAccessInfo &B, unsigned BIdx) {
- assert (AIdx < BIdx && "Must pass arguments in program order");
-
- Value *APtr = A.getPointer();
- Value *BPtr = B.getPointer();
- bool AIsWrite = A.getInt();
- bool BIsWrite = B.getInt();
-
- // Two reads are independent.
- if (!AIsWrite && !BIsWrite)
- return false;
-
- const SCEV *AScev = SE->getSCEV(APtr);
- const SCEV *BScev = SE->getSCEV(BPtr);
-
- int StrideAPtr = isStridedPtr(SE, DL, APtr, InnermostLoop);
- int StrideBPtr = isStridedPtr(SE, DL, BPtr, InnermostLoop);
-
- const SCEV *Src = AScev;
- const SCEV *Sink = BScev;
-
- // If the induction step is negative we have to invert source and sink of the
- // dependence.
- if (StrideAPtr < 0) {
- //Src = BScev;
- //Sink = AScev;
- std::swap(APtr, BPtr);
- std::swap(Src, Sink);
- std::swap(AIsWrite, BIsWrite);
- std::swap(AIdx, BIdx);
- std::swap(StrideAPtr, StrideBPtr);
- }
-
- const SCEV *Dist = SE->getMinusSCEV(Sink, Src);
-
- DEBUG(dbgs() << "LV: Src Scev: " << *Src << "Sink Scev: " << *Sink
- << "(Induction step: " << StrideAPtr << ")\n");
- DEBUG(dbgs() << "LV: Distance for " << *InstMap[AIdx] << " to "
- << *InstMap[BIdx] << ": " << *Dist << "\n");
-
- // Need consecutive accesses. We don't want to vectorize
- // "A[B[i]] += ..." and similar code or pointer arithmetic that could wrap in
- // the address space.
- if (!StrideAPtr || !StrideBPtr || StrideAPtr != StrideBPtr){
- DEBUG(dbgs() << "Non-consecutive pointer access\n");
- return true;
- }
-
- const SCEVConstant *C = dyn_cast<SCEVConstant>(Dist);
- if (!C) {
- DEBUG(dbgs() << "LV: Dependence because of non constant distance\n");
- return true;
- }
-
- Type *ATy = APtr->getType()->getPointerElementType();
- Type *BTy = BPtr->getType()->getPointerElementType();
- unsigned TypeByteSize = DL->getTypeAllocSize(ATy);
-
- // Negative distances are not plausible dependencies.
- const APInt &Val = C->getValue()->getValue();
- if (Val.isNegative()) {
- bool IsTrueDataDependence = (AIsWrite && !BIsWrite);
- if (IsTrueDataDependence &&
- (couldPreventStoreLoadForward(Val.abs().getZExtValue(), TypeByteSize) ||
- ATy != BTy))
- return true;
-
- DEBUG(dbgs() << "LV: Dependence is negative: NoDep\n");
- return false;
- }
-
- // Write to the same location with the same size.
- // Could be improved to assert type sizes are the same (i32 == float, etc).
- if (Val == 0) {
- if (ATy == BTy)
- return false;
- DEBUG(dbgs() << "LV: Zero dependence difference but different types\n");
- return true;
- }
-
- assert(Val.isStrictlyPositive() && "Expect a positive value");
-
- // Positive distance bigger than max vectorization factor.
- if (ATy != BTy) {
- DEBUG(dbgs() <<
- "LV: ReadWrite-Write positive dependency with different types\n");
- return false;
- }
-
- unsigned Distance = (unsigned) Val.getZExtValue();
-
- // Bail out early if passed-in parameters make vectorization not feasible.
- unsigned ForcedFactor = VectorizationFactor ? VectorizationFactor : 1;
- unsigned ForcedUnroll = VectorizationUnroll ? VectorizationUnroll : 1;
-
- // The distance must be bigger than the size needed for a vectorized version
- // of the operation and the size of the vectorized operation must not be
- // bigger than the currrent maximum size.
- if (Distance < 2*TypeByteSize ||
- 2*TypeByteSize > MaxSafeDepDistBytes ||
- Distance < TypeByteSize * ForcedUnroll * ForcedFactor) {
- DEBUG(dbgs() << "LV: Failure because of Positive distance "
- << Val.getSExtValue() << '\n');
- return true;
- }
-
- MaxSafeDepDistBytes = Distance < MaxSafeDepDistBytes ?
- Distance : MaxSafeDepDistBytes;
-
- bool IsTrueDataDependence = (!AIsWrite && BIsWrite);
- if (IsTrueDataDependence &&
- couldPreventStoreLoadForward(Distance, TypeByteSize))
- return true;
-
- DEBUG(dbgs() << "LV: Positive distance " << Val.getSExtValue() <<
- " with max VF = " << MaxSafeDepDistBytes / TypeByteSize << '\n');
-
- return false;
-}
-
-bool
-MemoryDepChecker::areDepsSafe(AccessAnalysis::DepCandidates &AccessSets,
- MemAccessInfoSet &CheckDeps) {
-
- MaxSafeDepDistBytes = -1U;
- while (!CheckDeps.empty()) {
- MemAccessInfo CurAccess = *CheckDeps.begin();
-
- // Get the relevant memory access set.
- EquivalenceClasses<MemAccessInfo>::iterator I =
- AccessSets.findValue(AccessSets.getLeaderValue(CurAccess));
-
- // Check accesses within this set.
- EquivalenceClasses<MemAccessInfo>::member_iterator AI, AE;
- AI = AccessSets.member_begin(I), AE = AccessSets.member_end();
-
- // Check every access pair.
- while (AI != AE) {
- CheckDeps.erase(*AI);
- EquivalenceClasses<MemAccessInfo>::member_iterator OI = llvm::next(AI);
- while (OI != AE) {
- // Check every accessing instruction pair in program order.
- for (std::vector<unsigned>::iterator I1 = Accesses[*AI].begin(),
- I1E = Accesses[*AI].end(); I1 != I1E; ++I1)
- for (std::vector<unsigned>::iterator I2 = Accesses[*OI].begin(),
- I2E = Accesses[*OI].end(); I2 != I2E; ++I2) {
- if (*I1 < *I2 && isDependent(*AI, *I1, *OI, *I2))
- return false;
- if (*I2 < *I1 && isDependent(*OI, *I2, *AI, *I1))
- return false;
- }
- ++OI;
- }
- AI++;
- }
- }
- return true;
-}
-
-bool LoopVectorizationLegality::canVectorizeMemory() {
-
- typedef SmallVector<Value*, 16> ValueVector;
- typedef SmallPtrSet<Value*, 16> ValueSet;
-
- // Holds the Load and Store *instructions*.
- ValueVector Loads;
- ValueVector Stores;
-
- // Holds all the different accesses in the loop.
- unsigned NumReads = 0;
- unsigned NumReadWrites = 0;
-
- PtrRtCheck.Pointers.clear();
- PtrRtCheck.Need = false;
-
- const bool IsAnnotatedParallel = TheLoop->isAnnotatedParallel();
- MemoryDepChecker DepChecker(SE, DL, TheLoop);
-
- // For each block.
- for (Loop::block_iterator bb = TheLoop->block_begin(),
- be = TheLoop->block_end(); bb != be; ++bb) {
-
- // Scan the BB and collect legal loads and stores.
- for (BasicBlock::iterator it = (*bb)->begin(), e = (*bb)->end(); it != e;
- ++it) {
-
- // If this is a load, save it. If this instruction can read from memory
- // but is not a load, then we quit. Notice that we don't handle function
- // calls that read or write.
- if (it->mayReadFromMemory()) {
- // Many math library functions read the rounding mode. We will only
- // vectorize a loop if it contains known function calls that don't set
- // the flag. Therefore, it is safe to ignore this read from memory.
- CallInst *Call = dyn_cast<CallInst>(it);
- if (Call && getIntrinsicIDForCall(Call, TLI))
- continue;
-
- LoadInst *Ld = dyn_cast<LoadInst>(it);
- if (!Ld) return false;
- if (!Ld->isSimple() && !IsAnnotatedParallel) {
- DEBUG(dbgs() << "LV: Found a non-simple load.\n");
- return false;
- }
- Loads.push_back(Ld);
- DepChecker.addAccess(Ld);
- continue;
- }
-
- // Save 'store' instructions. Abort if other instructions write to memory.
- if (it->mayWriteToMemory()) {
- StoreInst *St = dyn_cast<StoreInst>(it);
- if (!St) return false;
- if (!St->isSimple() && !IsAnnotatedParallel) {
- DEBUG(dbgs() << "LV: Found a non-simple store.\n");
- return false;
- }
- Stores.push_back(St);
- DepChecker.addAccess(St);
- }
- } // Next instr.
- } // Next block.
-
- // Now we have two lists that hold the loads and the stores.
- // Next, we find the pointers that they use.
-
- // Check if we see any stores. If there are no stores, then we don't
- // care if the pointers are *restrict*.
- if (!Stores.size()) {
- DEBUG(dbgs() << "LV: Found a read-only loop!\n");
- return true;
- }
-
- AccessAnalysis::DepCandidates DependentAccesses;
- AccessAnalysis Accesses(DL, DependentAccesses);
-
- // Holds the analyzed pointers. We don't want to call GetUnderlyingObjects
- // multiple times on the same object. If the ptr is accessed twice, once
- // for read and once for write, it will only appear once (on the write
- // list). This is okay, since we are going to check for conflicts between
- // writes and between reads and writes, but not between reads and reads.
- ValueSet Seen;
-
- ValueVector::iterator I, IE;
- for (I = Stores.begin(), IE = Stores.end(); I != IE; ++I) {
- StoreInst *ST = cast<StoreInst>(*I);
- Value* Ptr = ST->getPointerOperand();
-
- if (isUniform(Ptr)) {
- DEBUG(dbgs() << "LV: We don't allow storing to uniform addresses\n");
- return false;
- }
-
- // 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)) {
- ++NumReadWrites;
- Accesses.addStore(Ptr);
- }
- }
-
- if (IsAnnotatedParallel) {
- DEBUG(dbgs()
- << "LV: A loop annotated parallel, ignore memory dependency "
- << "checks.\n");
- return true;
- }
-
- for (I = Loads.begin(), IE = Loads.end(); I != IE; ++I) {
- LoadInst *LD = cast<LoadInst>(*I);
- Value* Ptr = LD->getPointerOperand();
- // If we did *not* see this pointer before, insert it to the
- // read list. If we *did* see it before, then it is already in
- // the read-write list. This allows us to vectorize expressions
- // such as A[i] += x; Because the address of A[i] is a read-write
- // pointer. This only works if the index of A[i] is consecutive.
- // If the address of i is unknown (for example A[B[i]]) then we may
- // 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)) {
- ++NumReads;
- IsReadOnlyPtr = true;
- }
- Accesses.addLoad(Ptr, IsReadOnlyPtr);
- }
-
- // If we write (or read-write) to a single destination and there are no
- // other reads in this loop then is it safe to vectorize.
- if (NumReadWrites == 1 && NumReads == 0) {
- DEBUG(dbgs() << "LV: Found a write-only loop!\n");
- return true;
- }
-
- // Build dependence sets and check whether we need a runtime pointer bounds
- // check.
- Accesses.buildDependenceSets();
- bool NeedRTCheck = Accesses.isRTCheckNeeded();
-
- // Find pointers with computable bounds. We are going to use this information
- // to place a runtime bound check.
- unsigned NumComparisons = 0;
- bool CanDoRT = false;
- if (NeedRTCheck)
- CanDoRT = Accesses.canCheckPtrAtRT(PtrRtCheck, NumComparisons, SE, TheLoop);
-
-
- DEBUG(dbgs() << "LV: We need to do " << NumComparisons <<
- " pointer comparisons.\n");
-
- // If we only have one set of dependences to check pointers among we don't
- // need a runtime check.
- if (NumComparisons == 0 && NeedRTCheck)
- NeedRTCheck = false;
-
- // Check that we did not collect too many pointers or found an unsizeable
- // pointer.
- if (!CanDoRT || NumComparisons > RuntimeMemoryCheckThreshold) {
- PtrRtCheck.reset();
- CanDoRT = false;
- }
-
- if (CanDoRT) {
- DEBUG(dbgs() << "LV: We can perform a memory runtime check if needed.\n");
- }
-
- if (NeedRTCheck && !CanDoRT) {
- DEBUG(dbgs() << "LV: We can't vectorize because we can't find " <<
- "the array bounds.\n");
- PtrRtCheck.reset();
- return false;
- }
-
- PtrRtCheck.Need = NeedRTCheck;
-
- bool CanVecMem = true;
- if (Accesses.isDependencyCheckNeeded()) {
- DEBUG(dbgs() << "LV: Checking memory dependencies\n");
- CanVecMem = DepChecker.areDepsSafe(DependentAccesses,
- Accesses.getDependenciesToCheck());
- MaxSafeDepDistBytes = DepChecker.getMaxSafeDepDistBytes();
- }
-
- DEBUG(dbgs() << "LV: We" << (NeedRTCheck ? "" : " don't") <<
- " need a runtime memory check.\n");
-
- return CanVecMem;
-}
-
-static bool hasMultipleUsesOf(Instruction *I,
- SmallPtrSet<Instruction *, 8> &Insts) {
- unsigned NumUses = 0;
- for(User::op_iterator Use = I->op_begin(), E = I->op_end(); Use != E; ++Use) {
- if (Insts.count(dyn_cast<Instruction>(*Use)))
- ++NumUses;
- if (NumUses > 1)
- return true;
- }
-
- return false;
-}
-
-static bool areAllUsesIn(Instruction *I, SmallPtrSet<Instruction *, 8> &Set) {
- for(User::op_iterator Use = I->op_begin(), E = I->op_end(); Use != E; ++Use)
- if (!Set.count(dyn_cast<Instruction>(*Use)))
- return false;
- return true;
-}
-
-bool LoopVectorizationLegality::AddReductionVar(PHINode *Phi,
- ReductionKind Kind) {
- if (Phi->getNumIncomingValues() != 2)
- return false;
-
- // Reduction variables are only found in the loop header block.
- if (Phi->getParent() != TheLoop->getHeader())
- return false;
-
- // Obtain the reduction start value from the value that comes from the loop
- // preheader.
- Value *RdxStart = Phi->getIncomingValueForBlock(TheLoop->getLoopPreheader());
-
- // ExitInstruction is the single value which is used outside the loop.
- // We only allow for a single reduction value to be used outside the loop.
- // This includes users of the reduction, variables (which form a cycle
- // which ends in the phi node).
- Instruction *ExitInstruction = 0;
- // Indicates that we found a reduction operation in our scan.
- bool FoundReduxOp = false;
-
- // We start with the PHI node and scan for all of the users of this
- // instruction. All users must be instructions that can be used as reduction
- // variables (such as ADD). We must have a single out-of-block user. The cycle
- // must include the original PHI.
- bool FoundStartPHI = false;
-
- // To recognize min/max patterns formed by a icmp select sequence, we store
- // the number of instruction we saw from the recognized min/max pattern,
- // to make sure we only see exactly the two instructions.
- unsigned NumCmpSelectPatternInst = 0;
- ReductionInstDesc ReduxDesc(false, 0);
-
- SmallPtrSet<Instruction *, 8> VisitedInsts;
- SmallVector<Instruction *, 8> Worklist;
- Worklist.push_back(Phi);
- VisitedInsts.insert(Phi);
-
- // A value in the reduction can be used:
- // - By the reduction:
- // - Reduction operation:
- // - One use of reduction value (safe).
- // - Multiple use of reduction value (not safe).
- // - PHI:
- // - All uses of the PHI must be the reduction (safe).
- // - Otherwise, not safe.
- // - By one instruction outside of the loop (safe).
- // - By further instructions outside of the loop (not safe).
- // - By an instruction that is not part of the reduction (not safe).
- // This is either:
- // * An instruction type other than PHI or the reduction operation.
- // * A PHI in the header other than the initial PHI.
- while (!Worklist.empty()) {
- Instruction *Cur = Worklist.back();
- Worklist.pop_back();
-
- // No Users.
- // If the instruction has no users then this is a broken chain and can't be
- // a reduction variable.
- if (Cur->use_empty())
- return false;
-
- bool IsAPhi = isa<PHINode>(Cur);
+ DEBUG(dbgs() << "LV: Found an induction variable.\n");
- // A header PHI use other than the original PHI.
- if (Cur != Phi && IsAPhi && Cur->getParent() == Phi->getParent())
- return false;
+ // Until we explicitly handle the case of an induction variable with
+ // an outside loop user we have to give up vectorizing this loop.
+ if (hasOutsideLoopUser(TheLoop, &*it, AllowedExit)) {
+ emitAnalysis(VectorizationReport(&*it) <<
+ "use of induction value outside of the "
+ "loop is not handled by vectorizer");
+ return false;
+ }
- // Reductions of instructions such as Div, and Sub is only possible if the
- // LHS is the reduction variable.
- if (!Cur->isCommutative() && !IsAPhi && !isa<SelectInst>(Cur) &&
- !isa<ICmpInst>(Cur) && !isa<FCmpInst>(Cur) &&
- !VisitedInsts.count(dyn_cast<Instruction>(Cur->getOperand(0))))
- return false;
+ continue;
+ }
- // Any reduction instruction must be of one of the allowed kinds.
- ReduxDesc = isReductionInstr(Cur, Kind, ReduxDesc);
- if (!ReduxDesc.IsReduction)
- return false;
+ RecurrenceDescriptor RedDes;
+ if (RecurrenceDescriptor::isReductionPHI(Phi, TheLoop, RedDes)) {
+ if (RedDes.hasUnsafeAlgebra())
+ Requirements->addUnsafeAlgebraInst(RedDes.getUnsafeAlgebraInst());
+ AllowedExit.insert(RedDes.getLoopExitInstr());
+ Reductions[Phi] = RedDes;
+ continue;
+ }
- // A reduction operation must only have one use of the reduction value.
- if (!IsAPhi && Kind != RK_IntegerMinMax && Kind != RK_FloatMinMax &&
- hasMultipleUsesOf(Cur, VisitedInsts))
- return false;
+ emitAnalysis(VectorizationReport(&*it) <<
+ "value that could not be identified as "
+ "reduction is used outside the loop");
+ DEBUG(dbgs() << "LV: Found an unidentified PHI."<< *Phi <<"\n");
+ return false;
+ }// end of PHI handling
- // All inputs to a PHI node must be a reduction value.
- if(IsAPhi && Cur != Phi && !areAllUsesIn(Cur, VisitedInsts))
- return false;
+ // We handle calls that:
+ // * Are debug info intrinsics.
+ // * Have a mapping to an IR intrinsic.
+ // * Have a vector version available.
+ CallInst *CI = dyn_cast<CallInst>(it);
+ if (CI && !getIntrinsicIDForCall(CI, TLI) && !isa<DbgInfoIntrinsic>(CI) &&
+ !(CI->getCalledFunction() && TLI &&
+ TLI->isFunctionVectorizable(CI->getCalledFunction()->getName()))) {
+ emitAnalysis(VectorizationReport(&*it)
+ << "call instruction cannot be vectorized");
+ DEBUG(dbgs() << "LV: Found a non-intrinsic, non-libfunc callsite.\n");
+ return false;
+ }
- if (Kind == RK_IntegerMinMax && (isa<ICmpInst>(Cur) ||
- isa<SelectInst>(Cur)))
- ++NumCmpSelectPatternInst;
- if (Kind == RK_FloatMinMax && (isa<FCmpInst>(Cur) ||
- isa<SelectInst>(Cur)))
- ++NumCmpSelectPatternInst;
-
- // Check whether we found a reduction operator.
- FoundReduxOp |= !IsAPhi;
-
- // Process users of current instruction. Push non PHI nodes after PHI nodes
- // onto the stack. This way we are going to have seen all inputs to PHI
- // nodes once we get to them.
- SmallVector<Instruction *, 8> NonPHIs;
- SmallVector<Instruction *, 8> PHIs;
- for (Value::use_iterator UI = Cur->use_begin(), E = Cur->use_end(); UI != E;
- ++UI) {
- Instruction *Usr = cast<Instruction>(*UI);
-
- // Check if we found the exit user.
- BasicBlock *Parent = Usr->getParent();
- if (!TheLoop->contains(Parent)) {
- // Exit if you find multiple outside users or if the header phi node is
- // being used. In this case the user uses the value of the previous
- // iteration, in which case we would loose "VF-1" iterations of the
- // reduction operation if we vectorize.
- if (ExitInstruction != 0 || Cur == Phi)
+ // Intrinsics such as powi,cttz and ctlz are legal to vectorize if the
+ // second argument is the same (i.e. loop invariant)
+ if (CI &&
+ hasVectorInstrinsicScalarOpd(getIntrinsicIDForCall(CI, TLI), 1)) {
+ auto *SE = PSE.getSE();
+ if (!SE->isLoopInvariant(PSE.getSCEV(CI->getOperand(1)), TheLoop)) {
+ emitAnalysis(VectorizationReport(&*it)
+ << "intrinsic instruction cannot be vectorized");
+ DEBUG(dbgs() << "LV: Found unvectorizable intrinsic " << *CI << "\n");
return false;
+ }
+ }
- // The instruction used by an outside user must be the last instruction
- // before we feed back to the reduction phi. Otherwise, we loose VF-1
- // operations on the value.
- if (std::find(Phi->op_begin(), Phi->op_end(), Cur) == Phi->op_end())
- return false;
-
- ExitInstruction = Cur;
- continue;
+ // Check that the instruction return type is vectorizable.
+ // Also, we can't vectorize extractelement instructions.
+ if ((!VectorType::isValidElementType(it->getType()) &&
+ !it->getType()->isVoidTy()) || isa<ExtractElementInst>(it)) {
+ emitAnalysis(VectorizationReport(&*it)
+ << "instruction return type cannot be vectorized");
+ DEBUG(dbgs() << "LV: Found unvectorizable type.\n");
+ return false;
}
- // Process instructions only once (termination).
- if (VisitedInsts.insert(Usr)) {
- if (isa<PHINode>(Usr))
- PHIs.push_back(Usr);
- else
- NonPHIs.push_back(Usr);
+ // Check that the stored type is vectorizable.
+ if (StoreInst *ST = dyn_cast<StoreInst>(it)) {
+ Type *T = ST->getValueOperand()->getType();
+ if (!VectorType::isValidElementType(T)) {
+ emitAnalysis(VectorizationReport(ST) <<
+ "store instruction cannot be vectorized");
+ return false;
+ }
+ if (EnableMemAccessVersioning)
+ collectStridedAccess(ST);
}
- // Remember that we completed the cycle.
- if (Usr == Phi)
- FoundStartPHI = true;
- }
- Worklist.append(PHIs.begin(), PHIs.end());
- Worklist.append(NonPHIs.begin(), NonPHIs.end());
- }
- // This means we have seen one but not the other instruction of the
- // pattern or more than just a select and cmp.
- if ((Kind == RK_IntegerMinMax || Kind == RK_FloatMinMax) &&
- NumCmpSelectPatternInst != 2)
- return false;
+ if (EnableMemAccessVersioning)
+ if (LoadInst *LI = dyn_cast<LoadInst>(it))
+ collectStridedAccess(LI);
- if (!FoundStartPHI || !FoundReduxOp || !ExitInstruction)
- return false;
+ // Reduction instructions are allowed to have exit users.
+ // All other instructions must not have external users.
+ if (hasOutsideLoopUser(TheLoop, &*it, AllowedExit)) {
+ emitAnalysis(VectorizationReport(&*it) <<
+ "value cannot be used outside the loop");
+ return false;
+ }
+
+ } // next instr.
- // We found a reduction var if we have reached the original phi node and we
- // only have a single instruction with out-of-loop users.
+ }
- // This instruction is allowed to have out-of-loop users.
- AllowedExit.insert(ExitInstruction);
+ if (!Induction) {
+ DEBUG(dbgs() << "LV: Did not find one integer induction var.\n");
+ if (Inductions.empty()) {
+ emitAnalysis(VectorizationReport()
+ << "loop induction variable could not be identified");
+ return false;
+ }
+ }
- // Save the description of this reduction variable.
- ReductionDescriptor RD(RdxStart, ExitInstruction, Kind,
- ReduxDesc.MinMaxKind);
- Reductions[Phi] = RD;
- // We've ended the cycle. This is a reduction variable if we have an
- // outside user and it has a binary op.
+ // Now we know the widest induction type, check if our found induction
+ // is the same size. If it's not, unset it here and InnerLoopVectorizer
+ // will create another.
+ if (Induction && WidestIndTy != Induction->getType())
+ Induction = nullptr;
return true;
}
-/// Returns true if the instruction is a Select(ICmp(X, Y), X, Y) instruction
-/// pattern corresponding to a min(X, Y) or max(X, Y).
-LoopVectorizationLegality::ReductionInstDesc
-LoopVectorizationLegality::isMinMaxSelectCmpPattern(Instruction *I,
- ReductionInstDesc &Prev) {
-
- assert((isa<ICmpInst>(I) || isa<FCmpInst>(I) || isa<SelectInst>(I)) &&
- "Expect a select instruction");
- Instruction *Cmp = 0;
- SelectInst *Select = 0;
-
- // We must handle the select(cmp()) as a single instruction. Advance to the
- // select.
- if ((Cmp = dyn_cast<ICmpInst>(I)) || (Cmp = dyn_cast<FCmpInst>(I))) {
- if (!Cmp->hasOneUse() || !(Select = dyn_cast<SelectInst>(*I->use_begin())))
- return ReductionInstDesc(false, I);
- return ReductionInstDesc(Select, Prev.MinMaxKind);
- }
-
- // Only handle single use cases for now.
- if (!(Select = dyn_cast<SelectInst>(I)))
- return ReductionInstDesc(false, I);
- if (!(Cmp = dyn_cast<ICmpInst>(I->getOperand(0))) &&
- !(Cmp = dyn_cast<FCmpInst>(I->getOperand(0))))
- return ReductionInstDesc(false, I);
- if (!Cmp->hasOneUse())
- return ReductionInstDesc(false, I);
-
- Value *CmpLeft;
- Value *CmpRight;
-
- // Look for a min/max pattern.
- if (m_UMin(m_Value(CmpLeft), m_Value(CmpRight)).match(Select))
- return ReductionInstDesc(Select, MRK_UIntMin);
- else if (m_UMax(m_Value(CmpLeft), m_Value(CmpRight)).match(Select))
- return ReductionInstDesc(Select, MRK_UIntMax);
- else if (m_SMax(m_Value(CmpLeft), m_Value(CmpRight)).match(Select))
- return ReductionInstDesc(Select, MRK_SIntMax);
- else if (m_SMin(m_Value(CmpLeft), m_Value(CmpRight)).match(Select))
- return ReductionInstDesc(Select, MRK_SIntMin);
- else if (m_OrdFMin(m_Value(CmpLeft), m_Value(CmpRight)).match(Select))
- return ReductionInstDesc(Select, MRK_FloatMin);
- else if (m_OrdFMax(m_Value(CmpLeft), m_Value(CmpRight)).match(Select))
- return ReductionInstDesc(Select, MRK_FloatMax);
- else if (m_UnordFMin(m_Value(CmpLeft), m_Value(CmpRight)).match(Select))
- return ReductionInstDesc(Select, MRK_FloatMin);
- else if (m_UnordFMax(m_Value(CmpLeft), m_Value(CmpRight)).match(Select))
- return ReductionInstDesc(Select, MRK_FloatMax);
-
- return ReductionInstDesc(false, I);
-}
+void LoopVectorizationLegality::collectStridedAccess(Value *MemAccess) {
+ Value *Ptr = nullptr;
+ if (LoadInst *LI = dyn_cast<LoadInst>(MemAccess))
+ Ptr = LI->getPointerOperand();
+ else if (StoreInst *SI = dyn_cast<StoreInst>(MemAccess))
+ Ptr = SI->getPointerOperand();
+ else
+ return;
-LoopVectorizationLegality::ReductionInstDesc
-LoopVectorizationLegality::isReductionInstr(Instruction *I,
- ReductionKind Kind,
- ReductionInstDesc &Prev) {
- bool FP = I->getType()->isFloatingPointTy();
- bool FastMath = (FP && I->isCommutative() && I->isAssociative());
- switch (I->getOpcode()) {
- default:
- return ReductionInstDesc(false, I);
- case Instruction::PHI:
- if (FP && (Kind != RK_FloatMult && Kind != RK_FloatAdd &&
- Kind != RK_FloatMinMax))
- return ReductionInstDesc(false, I);
- return ReductionInstDesc(I, Prev.MinMaxKind);
- case Instruction::Sub:
- case Instruction::Add:
- return ReductionInstDesc(Kind == RK_IntegerAdd, I);
- case Instruction::Mul:
- return ReductionInstDesc(Kind == RK_IntegerMult, I);
- case Instruction::And:
- return ReductionInstDesc(Kind == RK_IntegerAnd, I);
- case Instruction::Or:
- return ReductionInstDesc(Kind == RK_IntegerOr, I);
- case Instruction::Xor:
- return ReductionInstDesc(Kind == RK_IntegerXor, I);
- case Instruction::FMul:
- return ReductionInstDesc(Kind == RK_FloatMult && FastMath, I);
- case Instruction::FAdd:
- return ReductionInstDesc(Kind == RK_FloatAdd && FastMath, I);
- case Instruction::FCmp:
- case Instruction::ICmp:
- case Instruction::Select:
- if (Kind != RK_IntegerMinMax &&
- (!HasFunNoNaNAttr || Kind != RK_FloatMinMax))
- return ReductionInstDesc(false, I);
- return isMinMaxSelectCmpPattern(I, Prev);
- }
+ Value *Stride = getStrideFromPointer(Ptr, PSE.getSE(), TheLoop);
+ if (!Stride)
+ return;
+
+ DEBUG(dbgs() << "LV: Found a strided access that we can version");
+ DEBUG(dbgs() << " Ptr: " << *Ptr << " Stride: " << *Stride << "\n");
+ Strides[Ptr] = Stride;
+ StrideSet.insert(Stride);
}
-LoopVectorizationLegality::InductionKind
-LoopVectorizationLegality::isInductionVariable(PHINode *Phi) {
- Type *PhiTy = Phi->getType();
- // We only handle integer and pointer inductions variables.
- if (!PhiTy->isIntegerTy() && !PhiTy->isPointerTy())
- return IK_NoInduction;
+void LoopVectorizationLegality::collectLoopUniforms() {
+ // We now know that the loop is vectorizable!
+ // Collect variables that will remain uniform after vectorization.
+ std::vector<Value*> Worklist;
+ BasicBlock *Latch = TheLoop->getLoopLatch();
- // Check that the PHI is consecutive.
- const SCEV *PhiScev = SE->getSCEV(Phi);
- const SCEVAddRecExpr *AR = dyn_cast<SCEVAddRecExpr>(PhiScev);
- if (!AR) {
- DEBUG(dbgs() << "LV: PHI is not a poly recurrence.\n");
- return IK_NoInduction;
- }
- const SCEV *Step = AR->getStepRecurrence(*SE);
+ // Start with the conditional branch and walk up the block.
+ Worklist.push_back(Latch->getTerminator()->getOperand(0));
- // Integer inductions need to have a stride of one.
- if (PhiTy->isIntegerTy()) {
- if (Step->isOne())
- return IK_IntInduction;
- if (Step->isAllOnesValue())
- return IK_ReverseIntInduction;
- return IK_NoInduction;
+ // Also add all consecutive pointer values; these values will be uniform
+ // after vectorization (and subsequent cleanup) and, until revectorization is
+ // supported, all dependencies must also be uniform.
+ for (Loop::block_iterator B = TheLoop->block_begin(),
+ BE = TheLoop->block_end(); B != BE; ++B)
+ for (BasicBlock::iterator I = (*B)->begin(), IE = (*B)->end();
+ I != IE; ++I)
+ if (I->getType()->isPointerTy() && isConsecutivePtr(&*I))
+ Worklist.insert(Worklist.end(), I->op_begin(), I->op_end());
+
+ while (!Worklist.empty()) {
+ Instruction *I = dyn_cast<Instruction>(Worklist.back());
+ Worklist.pop_back();
+
+ // Look at instructions inside this loop.
+ // Stop when reaching PHI nodes.
+ // TODO: we need to follow values all over the loop, not only in this block.
+ if (!I || !TheLoop->contains(I) || isa<PHINode>(I))
+ continue;
+
+ // This is a known uniform.
+ Uniforms.insert(I);
+
+ // Insert all operands.
+ Worklist.insert(Worklist.end(), I->op_begin(), I->op_end());
}
+}
- // Calculate the pointer stride and check if it is consecutive.
- const SCEVConstant *C = dyn_cast<SCEVConstant>(Step);
- if (!C)
- return IK_NoInduction;
+bool LoopVectorizationLegality::canVectorizeMemory() {
+ LAI = &LAA->getInfo(TheLoop, Strides);
+ auto &OptionalReport = LAI->getReport();
+ if (OptionalReport)
+ emitAnalysis(VectorizationReport(*OptionalReport));
+ if (!LAI->canVectorizeMemory())
+ return false;
+
+ if (LAI->hasStoreToLoopInvariantAddress()) {
+ emitAnalysis(
+ VectorizationReport()
+ << "write to a loop invariant address could not be vectorized");
+ DEBUG(dbgs() << "LV: We don't allow storing to uniform addresses\n");
+ return false;
+ }
- assert(PhiTy->isPointerTy() && "The PHI must be a pointer");
- uint64_t Size = DL->getTypeAllocSize(PhiTy->getPointerElementType());
- if (C->getValue()->equalsInt(Size))
- return IK_PtrInduction;
- else if (C->getValue()->equalsInt(0 - Size))
- return IK_ReversePtrInduction;
+ Requirements->addRuntimePointerChecks(LAI->getNumRuntimePointerChecks());
+ PSE.addPredicate(LAI->PSE.getUnionPredicate());
- return IK_NoInduction;
+ return true;
}
bool LoopVectorizationLegality::isInductionVariable(const Value *V) {
}
bool LoopVectorizationLegality::blockNeedsPredication(BasicBlock *BB) {
- assert(TheLoop->contains(BB) && "Unknown block used");
-
- // Blocks that do not dominate the latch need predication.
- BasicBlock* Latch = TheLoop->getLoopLatch();
- return !DT->dominates(BB, Latch);
+ return LoopAccessInfo::blockNeedsPredication(BB, TheLoop, DT);
}
bool LoopVectorizationLegality::blockCanBePredicated(BasicBlock *BB,
- SmallPtrSet<Value *, 8>& SafePtrs) {
+ 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.
- if (it->mayWriteToMemory() || it->mayThrow())
+ if (it->mayWriteToMemory()) {
+ StoreInst *SI = dyn_cast<StoreInst>(it);
+ // We only support predication of stores in basic blocks with one
+ // predecessor.
+ 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;
// The instructions below can trap.
case Instruction::SDiv:
case Instruction::URem:
case Instruction::SRem:
- return false;
+ return false;
}
}
return true;
}
+void InterleavedAccessInfo::collectConstStridedAccesses(
+ MapVector<Instruction *, StrideDescriptor> &StrideAccesses,
+ const ValueToValueMap &Strides) {
+ // Holds load/store instructions in program order.
+ SmallVector<Instruction *, 16> AccessList;
+
+ for (auto *BB : TheLoop->getBlocks()) {
+ bool IsPred = LoopAccessInfo::blockNeedsPredication(BB, TheLoop, DT);
+
+ for (auto &I : *BB) {
+ if (!isa<LoadInst>(&I) && !isa<StoreInst>(&I))
+ continue;
+ // FIXME: Currently we can't handle mixed accesses and predicated accesses
+ if (IsPred)
+ return;
+
+ AccessList.push_back(&I);
+ }
+ }
+
+ if (AccessList.empty())
+ return;
+
+ auto &DL = TheLoop->getHeader()->getModule()->getDataLayout();
+ for (auto I : AccessList) {
+ LoadInst *LI = dyn_cast<LoadInst>(I);
+ StoreInst *SI = dyn_cast<StoreInst>(I);
+
+ Value *Ptr = LI ? LI->getPointerOperand() : SI->getPointerOperand();
+ int Stride = isStridedPtr(PSE, Ptr, TheLoop, Strides);
+
+ // The factor of the corresponding interleave group.
+ unsigned Factor = std::abs(Stride);
+
+ // Ignore the access if the factor is too small or too large.
+ if (Factor < 2 || Factor > MaxInterleaveGroupFactor)
+ continue;
+
+ const SCEV *Scev = replaceSymbolicStrideSCEV(PSE, Strides, Ptr);
+ PointerType *PtrTy = dyn_cast<PointerType>(Ptr->getType());
+ unsigned Size = DL.getTypeAllocSize(PtrTy->getElementType());
+
+ // An alignment of 0 means target ABI alignment.
+ unsigned Align = LI ? LI->getAlignment() : SI->getAlignment();
+ if (!Align)
+ Align = DL.getABITypeAlignment(PtrTy->getElementType());
+
+ StrideAccesses[I] = StrideDescriptor(Stride, Scev, Size, Align);
+ }
+}
+
+// Analyze interleaved accesses and collect them into interleave groups.
+//
+// Notice that the vectorization on interleaved groups will change instruction
+// orders and may break dependences. But the memory dependence check guarantees
+// that there is no overlap between two pointers of different strides, element
+// sizes or underlying bases.
+//
+// For pointers sharing the same stride, element size and underlying base, no
+// need to worry about Read-After-Write dependences and Write-After-Read
+// dependences.
+//
+// E.g. The RAW dependence: A[i] = a;
+// b = A[i];
+// This won't exist as it is a store-load forwarding conflict, which has
+// already been checked and forbidden in the dependence check.
+//
+// E.g. The WAR dependence: a = A[i]; // (1)
+// A[i] = b; // (2)
+// The store group of (2) is always inserted at or below (2), and the load group
+// of (1) is always inserted at or above (1). The dependence is safe.
+void InterleavedAccessInfo::analyzeInterleaving(
+ const ValueToValueMap &Strides) {
+ DEBUG(dbgs() << "LV: Analyzing interleaved accesses...\n");
+
+ // Holds all the stride accesses.
+ MapVector<Instruction *, StrideDescriptor> StrideAccesses;
+ collectConstStridedAccesses(StrideAccesses, Strides);
+
+ if (StrideAccesses.empty())
+ return;
+
+ // Holds all interleaved store groups temporarily.
+ SmallSetVector<InterleaveGroup *, 4> StoreGroups;
+ // Holds all interleaved load groups temporarily.
+ SmallSetVector<InterleaveGroup *, 4> LoadGroups;
+
+ // Search the load-load/write-write pair B-A in bottom-up order and try to
+ // insert B into the interleave group of A according to 3 rules:
+ // 1. A and B have the same stride.
+ // 2. A and B have the same memory object size.
+ // 3. B belongs to the group according to the distance.
+ //
+ // The bottom-up order can avoid breaking the Write-After-Write dependences
+ // between two pointers of the same base.
+ // E.g. A[i] = a; (1)
+ // A[i] = b; (2)
+ // A[i+1] = c (3)
+ // We form the group (2)+(3) in front, so (1) has to form groups with accesses
+ // above (1), which guarantees that (1) is always above (2).
+ for (auto I = StrideAccesses.rbegin(), E = StrideAccesses.rend(); I != E;
+ ++I) {
+ Instruction *A = I->first;
+ StrideDescriptor DesA = I->second;
+
+ InterleaveGroup *Group = getInterleaveGroup(A);
+ if (!Group) {
+ DEBUG(dbgs() << "LV: Creating an interleave group with:" << *A << '\n');
+ Group = createInterleaveGroup(A, DesA.Stride, DesA.Align);
+ }
+
+ if (A->mayWriteToMemory())
+ StoreGroups.insert(Group);
+ else
+ LoadGroups.insert(Group);
+
+ for (auto II = std::next(I); II != E; ++II) {
+ Instruction *B = II->first;
+ StrideDescriptor DesB = II->second;
+
+ // Ignore if B is already in a group or B is a different memory operation.
+ if (isInterleaved(B) || A->mayReadFromMemory() != B->mayReadFromMemory())
+ continue;
+
+ // Check the rule 1 and 2.
+ if (DesB.Stride != DesA.Stride || DesB.Size != DesA.Size)
+ continue;
+
+ // Calculate the distance and prepare for the rule 3.
+ const SCEVConstant *DistToA = dyn_cast<SCEVConstant>(
+ PSE.getSE()->getMinusSCEV(DesB.Scev, DesA.Scev));
+ if (!DistToA)
+ continue;
+
+ int DistanceToA = DistToA->getAPInt().getSExtValue();
+
+ // Skip if the distance is not multiple of size as they are not in the
+ // same group.
+ if (DistanceToA % static_cast<int>(DesA.Size))
+ continue;
+
+ // The index of B is the index of A plus the related index to A.
+ int IndexB =
+ Group->getIndex(A) + DistanceToA / static_cast<int>(DesA.Size);
+
+ // Try to insert B into the group.
+ if (Group->insertMember(B, IndexB, DesB.Align)) {
+ DEBUG(dbgs() << "LV: Inserted:" << *B << '\n'
+ << " into the interleave group with" << *A << '\n');
+ InterleaveGroupMap[B] = Group;
+
+ // Set the first load in program order as the insert position.
+ if (B->mayReadFromMemory())
+ Group->setInsertPos(B);
+ }
+ } // Iteration on instruction B
+ } // Iteration on instruction A
+
+ // Remove interleaved store groups with gaps.
+ for (InterleaveGroup *Group : StoreGroups)
+ if (Group->getNumMembers() != Group->getFactor())
+ releaseGroup(Group);
+
+ // Remove interleaved load groups that don't have the first and last member.
+ // This guarantees that we won't do speculative out of bounds loads.
+ for (InterleaveGroup *Group : LoadGroups)
+ if (!Group->getMember(0) || !Group->getMember(Group->getFactor() - 1))
+ releaseGroup(Group);
+}
+
LoopVectorizationCostModel::VectorizationFactor
-LoopVectorizationCostModel::selectVectorizationFactor(bool OptForSize,
- unsigned UserVF) {
+LoopVectorizationCostModel::selectVectorizationFactor(bool OptForSize) {
// Width 1 means no vectorize
VectorizationFactor Factor = { 1U, 0U };
- if (OptForSize && Legal->getRuntimePointerCheck()->Need) {
- DEBUG(dbgs() << "LV: Aborting. Runtime ptr check is required in Os.\n");
+ if (OptForSize && Legal->getRuntimePointerChecking()->Need) {
+ emitAnalysis(VectorizationReport() <<
+ "runtime pointer checks needed. Enable vectorization of this "
+ "loop with '#pragma clang loop vectorize(enable)' when "
+ "compiling with -Os/-Oz");
+ DEBUG(dbgs() <<
+ "LV: Aborting. Runtime ptr check is required with -Os/-Oz.\n");
+ return Factor;
+ }
+
+ if (!EnableCondStoresVectorization && Legal->getNumPredStores()) {
+ emitAnalysis(VectorizationReport() <<
+ "store that is conditionally executed prevents vectorization");
+ DEBUG(dbgs() << "LV: No vectorization. There are conditional stores.\n");
return Factor;
}
// Find the trip count.
- unsigned TC = SE->getSmallConstantTripCount(TheLoop, TheLoop->getLoopLatch());
+ unsigned TC = SE->getSmallConstantTripCount(TheLoop);
DEBUG(dbgs() << "LV: Found trip count: " << TC << '\n');
- unsigned WidestType = getWidestType();
+ MinBWs = computeMinimumValueSizes(TheLoop->getBlocks(), *DB, &TTI);
+ unsigned SmallestType, WidestType;
+ std::tie(SmallestType, WidestType) = getSmallestAndWidestTypes();
unsigned WidestRegister = TTI.getRegisterBitWidth(true);
unsigned MaxSafeDepDist = -1U;
if (Legal->getMaxSafeDepDistBytes() != -1U)
WidestRegister = ((WidestRegister < MaxSafeDepDist) ?
WidestRegister : MaxSafeDepDist);
unsigned MaxVectorSize = WidestRegister / WidestType;
- DEBUG(dbgs() << "LV: The Widest type: " << WidestType << " bits.\n");
+
+ DEBUG(dbgs() << "LV: The Smallest and Widest types: " << SmallestType << " / "
+ << WidestType << " bits.\n");
DEBUG(dbgs() << "LV: The Widest register is: "
<< WidestRegister << " bits.\n");
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;
+ if (MaximizeBandwidth && !OptForSize) {
+ // Collect all viable vectorization factors.
+ SmallVector<unsigned, 8> VFs;
+ unsigned NewMaxVectorSize = WidestRegister / SmallestType;
+ for (unsigned VS = MaxVectorSize; VS <= NewMaxVectorSize; VS *= 2)
+ VFs.push_back(VS);
+
+ // For each VF calculate its register usage.
+ auto RUs = calculateRegisterUsage(VFs);
+
+ // Select the largest VF which doesn't require more registers than existing
+ // ones.
+ unsigned TargetNumRegisters = TTI.getNumberOfRegisters(true);
+ for (int i = RUs.size() - 1; i >= 0; --i) {
+ if (RUs[i].MaxLocalUsers <= TargetNumRegisters) {
+ VF = VFs[i];
+ break;
+ }
+ }
+ }
// If we optimize the program for size, avoid creating the tail loop.
if (OptForSize) {
// If we are unable to calculate the trip count then don't try to vectorize.
if (TC < 2) {
- DEBUG(dbgs() << "LV: Aborting. A tail loop is required in Os.\n");
+ emitAnalysis
+ (VectorizationReport() <<
+ "unable to calculate the loop count due to complex control flow");
+ DEBUG(dbgs() << "LV: Aborting. A tail loop is required with -Os/-Oz.\n");
return Factor;
}
if (VF == 0)
VF = MaxVectorSize;
-
- // If the trip count that we found modulo the vectorization factor is not
- // zero then we require a tail.
- if (VF < 2) {
- DEBUG(dbgs() << "LV: Aborting. A tail loop is required in Os.\n");
+ else {
+ // If the trip count that we found modulo the vectorization factor is not
+ // zero then we require a tail.
+ emitAnalysis(VectorizationReport() <<
+ "cannot optimize for size and vectorize at the "
+ "same time. Enable vectorization of this loop "
+ "with '#pragma clang loop vectorize(enable)' "
+ "when compiling with -Os/-Oz");
+ DEBUG(dbgs() << "LV: Aborting. A tail loop is required with -Os/-Oz.\n");
return Factor;
}
}
+ int UserVF = Hints->getWidth();
if (UserVF != 0) {
assert(isPowerOf2_32(UserVF) && "VF needs to be a power of two");
DEBUG(dbgs() << "LV: Using user VF " << UserVF << ".\n");
}
float Cost = expectedCost(1);
+#ifndef NDEBUG
+ const float ScalarCost = Cost;
+#endif /* NDEBUG */
unsigned Width = 1;
- DEBUG(dbgs() << "LV: Scalar loop costs: " << (int)Cost << ".\n");
+ DEBUG(dbgs() << "LV: Scalar loop costs: " << (int)ScalarCost << ".\n");
+
+ bool ForceVectorization = Hints->getForce() == LoopVectorizeHints::FK_Enabled;
+ // Ignore scalar width, because the user explicitly wants vectorization.
+ if (ForceVectorization && VF > 1) {
+ Width = 2;
+ Cost = expectedCost(Width) / (float)Width;
+ }
+
for (unsigned i=2; i <= VF; i*=2) {
// Notice that the vector loop needs to be executed less times, so
// we need to divide the cost of the vector loops by the width of
}
}
- DEBUG(dbgs() << "LV: Selecting VF = : "<< Width << ".\n");
+ DEBUG(if (ForceVectorization && Width > 1 && Cost >= ScalarCost) dbgs()
+ << "LV: Vectorization seems to be not beneficial, "
+ << "but was forced by a user.\n");
+ DEBUG(dbgs() << "LV: Selecting VF: "<< Width << ".\n");
Factor.Width = Width;
Factor.Cost = Width * Cost;
return Factor;
}
-unsigned LoopVectorizationCostModel::getWidestType() {
+std::pair<unsigned, unsigned>
+LoopVectorizationCostModel::getSmallestAndWidestTypes() {
+ unsigned MinWidth = -1U;
unsigned MaxWidth = 8;
+ const DataLayout &DL = TheFunction->getParent()->getDataLayout();
// For each block.
for (Loop::block_iterator bb = TheLoop->block_begin(),
for (BasicBlock::iterator it = BB->begin(), e = BB->end(); it != e; ++it) {
Type *T = it->getType();
+ // Skip ignored values.
+ if (ValuesToIgnore.count(&*it))
+ continue;
+
// Only examine Loads, Stores and PHINodes.
if (!isa<LoadInst>(it) && !isa<StoreInst>(it) && !isa<PHINode>(it))
continue;
- // Examine PHI nodes that are reduction variables.
- if (PHINode *PN = dyn_cast<PHINode>(it))
- if (!Legal->getReductionVars()->count(PN))
+ // Examine PHI nodes that are reduction variables. Update the type to
+ // account for the recurrence type.
+ if (PHINode *PN = dyn_cast<PHINode>(it)) {
+ if (!Legal->isReductionVariable(PN))
continue;
+ RecurrenceDescriptor RdxDesc = (*Legal->getReductionVars())[PN];
+ T = RdxDesc.getRecurrenceType();
+ }
// Examine the stored values.
if (StoreInst *ST = dyn_cast<StoreInst>(it))
// Ignore loaded pointer types and stored pointer types that are not
// consecutive. However, we do want to take consecutive stores/loads of
// pointer vectors into account.
- if (T->isPointerTy() && !isConsecutiveLoadOrStore(it))
+ if (T->isPointerTy() && !isConsecutiveLoadOrStore(&*it))
continue;
+ MinWidth = std::min(MinWidth,
+ (unsigned)DL.getTypeSizeInBits(T->getScalarType()));
MaxWidth = std::max(MaxWidth,
- (unsigned)DL->getTypeSizeInBits(T->getScalarType()));
+ (unsigned)DL.getTypeSizeInBits(T->getScalarType()));
}
}
- return MaxWidth;
+ return {MinWidth, MaxWidth};
}
-unsigned
-LoopVectorizationCostModel::selectUnrollFactor(bool OptForSize,
- unsigned UserUF,
- unsigned VF,
- unsigned LoopCost) {
+unsigned LoopVectorizationCostModel::selectInterleaveCount(bool OptForSize,
+ unsigned VF,
+ unsigned LoopCost) {
- // -- The unroll heuristics --
- // We unroll the loop in order to expose ILP and reduce the loop overhead.
+ // -- The interleave heuristics --
+ // We interleave the loop in order to expose ILP and reduce the loop overhead.
// There are many micro-architectural considerations that we can't predict
- // at this level. For example frontend pressure (on decode or fetch) due to
+ // at this level. For example, frontend pressure (on decode or fetch) due to
// code size, or the number and capabilities of the execution ports.
//
- // We use the following heuristics to select the unroll factor:
- // 1. If the code has reductions the we unroll in order to break the cross
+ // We use the following heuristics to select the interleave count:
+ // 1. If the code has reductions, then we interleave to break the cross
// iteration dependency.
- // 2. If the loop is really small then we unroll in order to reduce the loop
+ // 2. If the loop is really small, then we interleave to reduce the loop
// overhead.
- // 3. We don't unroll if we think that we will spill registers to memory due
- // to the increased register pressure.
+ // 3. We don't interleave if we think that we will spill registers to memory
+ // due to the increased register pressure.
- // Use the user preference, unless 'auto' is selected.
- if (UserUF != 0)
- return UserUF;
-
- // When we optimize for size we don't unroll.
+ // When we optimize for size, we don't interleave.
if (OptForSize)
return 1;
- // We used the distance for the unroll factor.
+ // We used the distance for the interleave count.
if (Legal->getMaxSafeDepDistBytes() != -1U)
return 1;
- // Do not unroll loops with a relatively small trip count.
- unsigned TC = SE->getSmallConstantTripCount(TheLoop,
- TheLoop->getLoopLatch());
- if (TC > 1 && TC < TinyTripCountUnrollThreshold)
+ // Do not interleave loops with a relatively small trip count.
+ unsigned TC = SE->getSmallConstantTripCount(TheLoop);
+ if (TC > 1 && TC < TinyTripCountInterleaveThreshold)
return 1;
- unsigned TargetVectorRegisters = TTI.getNumberOfRegisters(true);
- DEBUG(dbgs() << "LV: The target has " << TargetVectorRegisters <<
- " vector registers\n");
+ unsigned TargetNumRegisters = TTI.getNumberOfRegisters(VF > 1);
+ DEBUG(dbgs() << "LV: The target has " << TargetNumRegisters <<
+ " registers\n");
+
+ if (VF == 1) {
+ if (ForceTargetNumScalarRegs.getNumOccurrences() > 0)
+ TargetNumRegisters = ForceTargetNumScalarRegs;
+ } else {
+ if (ForceTargetNumVectorRegs.getNumOccurrences() > 0)
+ TargetNumRegisters = ForceTargetNumVectorRegs;
+ }
- LoopVectorizationCostModel::RegisterUsage R = calculateRegisterUsage();
+ RegisterUsage R = calculateRegisterUsage({VF})[0];
// We divide by these constants so assume that we have at least one
// instruction that uses at least one register.
R.MaxLocalUsers = std::max(R.MaxLocalUsers, 1U);
R.NumInstructions = std::max(R.NumInstructions, 1U);
- // We calculate the unroll factor using the following formula.
+ // We calculate the interleave count using the following formula.
// Subtract the number of loop invariants from the number of available
- // registers. These registers are used by all of the unrolled instances.
+ // registers. These registers are used by all of the interleaved instances.
// Next, divide the remaining registers by the number of registers that is
// required by the loop, in order to estimate how many parallel instances
- // fit without causing spills.
- unsigned UF = (TargetVectorRegisters - R.LoopInvariantRegs) / R.MaxLocalUsers;
+ // fit without causing spills. All of this is rounded down if necessary to be
+ // a power of two. We want power of two interleave count to simplify any
+ // addressing operations or alignment considerations.
+ unsigned IC = PowerOf2Floor((TargetNumRegisters - R.LoopInvariantRegs) /
+ R.MaxLocalUsers);
+
+ // Don't count the induction variable as interleaved.
+ if (EnableIndVarRegisterHeur)
+ IC = PowerOf2Floor((TargetNumRegisters - R.LoopInvariantRegs - 1) /
+ std::max(1U, (R.MaxLocalUsers - 1)));
- // Clamp the unroll factor ranges to reasonable factors.
- unsigned MaxUnrollSize = TTI.getMaximumUnrollFactor();
+ // Clamp the interleave ranges to reasonable counts.
+ unsigned MaxInterleaveCount = TTI.getMaxInterleaveFactor(VF);
+
+ // Check if the user has overridden the max.
+ if (VF == 1) {
+ if (ForceTargetMaxScalarInterleaveFactor.getNumOccurrences() > 0)
+ MaxInterleaveCount = ForceTargetMaxScalarInterleaveFactor;
+ } else {
+ if (ForceTargetMaxVectorInterleaveFactor.getNumOccurrences() > 0)
+ MaxInterleaveCount = ForceTargetMaxVectorInterleaveFactor;
+ }
// If we did not calculate the cost for VF (because the user selected the VF)
// then we calculate the cost of VF here.
if (LoopCost == 0)
LoopCost = expectedCost(VF);
- // Clamp the calculated UF to be between the 1 and the max unroll factor
+ // Clamp the calculated IC to be between the 1 and the max interleave count
// that the target allows.
- if (UF > MaxUnrollSize)
- UF = MaxUnrollSize;
- else if (UF < 1)
- UF = 1;
+ if (IC > MaxInterleaveCount)
+ IC = MaxInterleaveCount;
+ else if (IC < 1)
+ IC = 1;
+
+ // Interleave if we vectorized this loop and there is a reduction that could
+ // benefit from interleaving.
+ if (VF > 1 && Legal->getReductionVars()->size()) {
+ DEBUG(dbgs() << "LV: Interleaving because of reductions.\n");
+ return IC;
+ }
- bool HasReductions = Legal->getReductionVars()->size();
+ // Note that if we've already vectorized the loop we will have done the
+ // runtime check and so interleaving won't require further checks.
+ bool InterleavingRequiresRuntimePointerCheck =
+ (VF == 1 && Legal->getRuntimePointerChecking()->Need);
- // Decide if we want to unroll if we decided that it is legal to vectorize
- // but not profitable.
- if (VF == 1) {
- if (TheLoop->getNumBlocks() > 1 || !HasReductions ||
- LoopCost > SmallLoopCost)
- return 1;
+ // We want to interleave small loops in order to reduce the loop overhead and
+ // potentially expose ILP opportunities.
+ DEBUG(dbgs() << "LV: Loop cost is " << LoopCost << '\n');
+ if (!InterleavingRequiresRuntimePointerCheck && LoopCost < SmallLoopCost) {
+ // We assume that the cost overhead is 1 and we use the cost model
+ // to estimate the cost of the loop and interleave until the cost of the
+ // loop overhead is about 5% of the cost of the loop.
+ unsigned SmallIC =
+ std::min(IC, (unsigned)PowerOf2Floor(SmallLoopCost / LoopCost));
+
+ // Interleave until store/load ports (estimated by max interleave count) are
+ // saturated.
+ unsigned NumStores = Legal->getNumStores();
+ unsigned NumLoads = Legal->getNumLoads();
+ unsigned StoresIC = IC / (NumStores ? NumStores : 1);
+ unsigned LoadsIC = IC / (NumLoads ? NumLoads : 1);
+
+ // If we have a scalar reduction (vector reductions are already dealt with
+ // by this point), we can increase the critical path length if the loop
+ // we're interleaving is inside another loop. Limit, by default to 2, so the
+ // critical path only gets increased by one reduction operation.
+ if (Legal->getReductionVars()->size() &&
+ TheLoop->getLoopDepth() > 1) {
+ unsigned F = static_cast<unsigned>(MaxNestedScalarReductionIC);
+ SmallIC = std::min(SmallIC, F);
+ StoresIC = std::min(StoresIC, F);
+ LoadsIC = std::min(LoadsIC, F);
+ }
- return UF;
- }
+ if (EnableLoadStoreRuntimeInterleave &&
+ std::max(StoresIC, LoadsIC) > SmallIC) {
+ DEBUG(dbgs() << "LV: Interleaving to saturate store or load ports.\n");
+ return std::max(StoresIC, LoadsIC);
+ }
- if (HasReductions) {
- DEBUG(dbgs() << "LV: Unrolling because of reductions.\n");
- return UF;
+ DEBUG(dbgs() << "LV: Interleaving to reduce branch cost.\n");
+ return SmallIC;
}
- // We want to unroll tiny loops in order to reduce the loop overhead.
- // We assume that the cost overhead is 1 and we use the cost model
- // to estimate the cost of the loop and unroll until the cost of the
- // loop overhead is about 5% of the cost of the loop.
- DEBUG(dbgs() << "LV: Loop cost is " << LoopCost << '\n');
- if (LoopCost < SmallLoopCost) {
- DEBUG(dbgs() << "LV: Unrolling to reduce branch cost.\n");
- unsigned NewUF = SmallLoopCost / (LoopCost + 1);
- return std::min(NewUF, UF);
+ // Interleave if this is a large loop (small loops are already dealt with by
+ // this point) that could benefit from interleaving.
+ bool HasReductions = (Legal->getReductionVars()->size() > 0);
+ if (TTI.enableAggressiveInterleaving(HasReductions)) {
+ DEBUG(dbgs() << "LV: Interleaving to expose ILP.\n");
+ return IC;
}
- DEBUG(dbgs() << "LV: Not Unrolling.\n");
+ DEBUG(dbgs() << "LV: Not Interleaving.\n");
return 1;
}
-LoopVectorizationCostModel::RegisterUsage
-LoopVectorizationCostModel::calculateRegisterUsage() {
+SmallVector<LoopVectorizationCostModel::RegisterUsage, 8>
+LoopVectorizationCostModel::calculateRegisterUsage(
+ const SmallVector<unsigned, 8> &VFs) {
// This function calculates the register usage by measuring the highest number
// of values that are alive at a single location. Obviously, this is a very
// rough estimation. We scan the loop in a topological order in order and
LoopBlocksDFS DFS(TheLoop);
DFS.perform(LI);
- RegisterUsage R;
- R.NumInstructions = 0;
+ RegisterUsage RU;
+ RU.NumInstructions = 0;
// Each 'key' in the map opens a new interval. The values
// of the map are the index of the 'last seen' usage of the
unsigned Index = 0;
for (LoopBlocksDFS::RPOIterator bb = DFS.beginRPO(),
be = DFS.endRPO(); bb != be; ++bb) {
- R.NumInstructions += (*bb)->size();
- for (BasicBlock::iterator it = (*bb)->begin(), e = (*bb)->end(); it != e;
- ++it) {
- Instruction *I = it;
- IdxToInstr[Index++] = I;
+ RU.NumInstructions += (*bb)->size();
+ for (Instruction &I : **bb) {
+ IdxToInstr[Index++] = &I;
// Save the end location of each USE.
- for (unsigned i = 0; i < I->getNumOperands(); ++i) {
- Value *U = I->getOperand(i);
+ for (unsigned i = 0; i < I.getNumOperands(); ++i) {
+ Value *U = I.getOperand(i);
Instruction *Instr = dyn_cast<Instruction>(U);
// Ignore non-instruction values such as arguments, constants, etc.
TransposeEnds[it->second].push_back(it->first);
SmallSet<Instruction*, 8> OpenIntervals;
- unsigned MaxUsage = 0;
+ // Get the size of the widest register.
+ unsigned MaxSafeDepDist = -1U;
+ if (Legal->getMaxSafeDepDistBytes() != -1U)
+ MaxSafeDepDist = Legal->getMaxSafeDepDistBytes() * 8;
+ unsigned WidestRegister =
+ std::min(TTI.getRegisterBitWidth(true), MaxSafeDepDist);
+ const DataLayout &DL = TheFunction->getParent()->getDataLayout();
+
+ SmallVector<RegisterUsage, 8> RUs(VFs.size());
+ SmallVector<unsigned, 8> MaxUsages(VFs.size(), 0);
DEBUG(dbgs() << "LV(REG): Calculating max register usage:\n");
+
+ // A lambda that gets the register usage for the given type and VF.
+ auto GetRegUsage = [&DL, WidestRegister](Type *Ty, unsigned VF) {
+ unsigned TypeSize = DL.getTypeSizeInBits(Ty->getScalarType());
+ return std::max<unsigned>(1, VF * TypeSize / WidestRegister);
+ };
+
for (unsigned int i = 0; i < Index; ++i) {
Instruction *I = IdxToInstr[i];
// Ignore instructions that are never used within the loop.
if (!Ends.count(I)) continue;
+ // Skip ignored values.
+ if (ValuesToIgnore.count(I))
+ continue;
+
// Remove all of the instructions that end at this location.
InstrList &List = TransposeEnds[i];
- for (unsigned int j=0, e = List.size(); j < e; ++j)
+ for (unsigned int j = 0, e = List.size(); j < e; ++j)
OpenIntervals.erase(List[j]);
- // Count the number of live interals.
- MaxUsage = std::max(MaxUsage, OpenIntervals.size());
+ // For each VF find the maximum usage of registers.
+ for (unsigned j = 0, e = VFs.size(); j < e; ++j) {
+ if (VFs[j] == 1) {
+ MaxUsages[j] = std::max(MaxUsages[j], OpenIntervals.size());
+ continue;
+ }
+
+ // Count the number of live intervals.
+ unsigned RegUsage = 0;
+ for (auto Inst : OpenIntervals)
+ RegUsage += GetRegUsage(Inst->getType(), VFs[j]);
+ MaxUsages[j] = std::max(MaxUsages[j], RegUsage);
+ }
- DEBUG(dbgs() << "LV(REG): At #" << i << " Interval # " <<
- OpenIntervals.size() << '\n');
+ DEBUG(dbgs() << "LV(REG): At #" << i << " Interval # "
+ << OpenIntervals.size() << '\n');
// Add the current instruction to the list of open intervals.
OpenIntervals.insert(I);
}
- unsigned Invariant = LoopInvariants.size();
- DEBUG(dbgs() << "LV(REG): Found max usage: " << MaxUsage << '\n');
- DEBUG(dbgs() << "LV(REG): Found invariant usage: " << Invariant << '\n');
- DEBUG(dbgs() << "LV(REG): LoopSize: " << R.NumInstructions << '\n');
+ for (unsigned i = 0, e = VFs.size(); i < e; ++i) {
+ unsigned Invariant = 0;
+ if (VFs[i] == 1)
+ Invariant = LoopInvariants.size();
+ else {
+ for (auto Inst : LoopInvariants)
+ Invariant += GetRegUsage(Inst->getType(), VFs[i]);
+ }
+
+ DEBUG(dbgs() << "LV(REG): VF = " << VFs[i] << '\n');
+ DEBUG(dbgs() << "LV(REG): Found max usage: " << MaxUsages[i] << '\n');
+ DEBUG(dbgs() << "LV(REG): Found invariant usage: " << Invariant << '\n');
+ DEBUG(dbgs() << "LV(REG): LoopSize: " << RU.NumInstructions << '\n');
+
+ RU.LoopInvariantRegs = Invariant;
+ RU.MaxLocalUsers = MaxUsages[i];
+ RUs[i] = RU;
+ }
- R.LoopInvariantRegs = Invariant;
- R.MaxLocalUsers = MaxUsage;
- return R;
+ return RUs;
}
unsigned LoopVectorizationCostModel::expectedCost(unsigned VF) {
if (isa<DbgInfoIntrinsic>(it))
continue;
- unsigned C = getInstructionCost(it, VF);
+ // Skip ignored values.
+ if (ValuesToIgnore.count(&*it))
+ continue;
+
+ unsigned C = getInstructionCost(&*it, VF);
+
+ // Check if we should override the cost.
+ if (ForceTargetInstructionCost.getNumOccurrences() > 0)
+ C = ForceTargetInstructionCost;
+
BlockCost += C;
DEBUG(dbgs() << "LV: Found an estimated cost of " << C << " for VF " <<
VF << " For instruction: " << *it << '\n');
if (!C)
return true;
- const APInt &APStepVal = C->getValue()->getValue();
+ const APInt &APStepVal = C->getAPInt();
// Huge step value - give up.
if (APStepVal.getBitWidth() > 64)
return StepVal > MaxMergeDistance;
}
+static bool isStrideMul(Instruction *I, LoopVectorizationLegality *Legal) {
+ return Legal->hasStride(I->getOperand(0)) ||
+ Legal->hasStride(I->getOperand(1));
+}
+
unsigned
LoopVectorizationCostModel::getInstructionCost(Instruction *I, unsigned VF) {
// If we know that this instruction will remain uniform, check the cost of
VF = 1;
Type *RetTy = I->getType();
+ if (VF > 1 && MinBWs.count(I))
+ RetTy = IntegerType::get(RetTy->getContext(), MinBWs[I]);
Type *VectorTy = ToVectorTy(RetTy, VF);
// TODO: We need to estimate the cost of intrinsic calls.
case Instruction::And:
case Instruction::Or:
case Instruction::Xor: {
+ // Since we will replace the stride by 1 the multiplication should go away.
+ if (I->getOpcode() == Instruction::Mul && isStrideMul(I, Legal))
+ return 0;
// Certain instructions can be cheaper to vectorize if they have a constant
// second vector operand. One example of this are shifts on x86.
TargetTransformInfo::OperandValueKind Op1VK =
TargetTransformInfo::OK_AnyValue;
TargetTransformInfo::OperandValueKind Op2VK =
TargetTransformInfo::OK_AnyValue;
-
- if (isa<ConstantInt>(I->getOperand(1)))
+ TargetTransformInfo::OperandValueProperties Op1VP =
+ TargetTransformInfo::OP_None;
+ TargetTransformInfo::OperandValueProperties Op2VP =
+ TargetTransformInfo::OP_None;
+ Value *Op2 = I->getOperand(1);
+
+ // Check for a splat of a constant or for a non uniform vector of constants.
+ if (isa<ConstantInt>(Op2)) {
+ ConstantInt *CInt = cast<ConstantInt>(Op2);
+ if (CInt && CInt->getValue().isPowerOf2())
+ Op2VP = TargetTransformInfo::OP_PowerOf2;
Op2VK = TargetTransformInfo::OK_UniformConstantValue;
+ } else if (isa<ConstantVector>(Op2) || isa<ConstantDataVector>(Op2)) {
+ Op2VK = TargetTransformInfo::OK_NonUniformConstantValue;
+ Constant *SplatValue = cast<Constant>(Op2)->getSplatValue();
+ if (SplatValue) {
+ ConstantInt *CInt = dyn_cast<ConstantInt>(SplatValue);
+ if (CInt && CInt->getValue().isPowerOf2())
+ Op2VP = TargetTransformInfo::OP_PowerOf2;
+ Op2VK = TargetTransformInfo::OK_UniformConstantValue;
+ }
+ }
- return TTI.getArithmeticInstrCost(I->getOpcode(), VectorTy, Op1VK, Op2VK);
+ return TTI.getArithmeticInstrCost(I->getOpcode(), VectorTy, Op1VK, Op2VK,
+ Op1VP, Op2VP);
}
case Instruction::Select: {
SelectInst *SI = cast<SelectInst>(I);
case Instruction::ICmp:
case Instruction::FCmp: {
Type *ValTy = I->getOperand(0)->getType();
+ Instruction *Op0AsInstruction = dyn_cast<Instruction>(I->getOperand(0));
+ auto It = MinBWs.find(Op0AsInstruction);
+ if (VF > 1 && It != MinBWs.end())
+ ValTy = IntegerType::get(ValTy->getContext(), It->second);
VectorTy = ToVectorTy(ValTy, VF);
return TTI.getCmpSelInstrCost(I->getOpcode(), VectorTy);
}
return TTI.getAddressComputationCost(VectorTy) +
TTI.getMemoryOpCost(I->getOpcode(), VectorTy, Alignment, AS);
+ // For an interleaved access, calculate the total cost of the whole
+ // interleave group.
+ if (Legal->isAccessInterleaved(I)) {
+ auto Group = Legal->getInterleavedAccessGroup(I);
+ assert(Group && "Fail to get an interleaved access group.");
+
+ // Only calculate the cost once at the insert position.
+ if (Group->getInsertPos() != I)
+ return 0;
+
+ unsigned InterleaveFactor = Group->getFactor();
+ Type *WideVecTy =
+ VectorType::get(VectorTy->getVectorElementType(),
+ VectorTy->getVectorNumElements() * InterleaveFactor);
+
+ // Holds the indices of existing members in an interleaved load group.
+ // An interleaved store group doesn't need this as it dones't allow gaps.
+ SmallVector<unsigned, 4> Indices;
+ if (LI) {
+ for (unsigned i = 0; i < InterleaveFactor; i++)
+ if (Group->getMember(i))
+ Indices.push_back(i);
+ }
+
+ // Calculate the cost of the whole interleaved group.
+ unsigned Cost = TTI.getInterleavedMemoryOpCost(
+ I->getOpcode(), WideVecTy, Group->getFactor(), Indices,
+ Group->getAlignment(), AS);
+
+ if (Group->isReverse())
+ Cost +=
+ Group->getNumMembers() *
+ TTI.getShuffleCost(TargetTransformInfo::SK_Reverse, VectorTy, 0);
+
+ // FIXME: The interleaved load group with a huge gap could be even more
+ // expensive than scalar operations. Then we could ignore such group and
+ // use scalar operations instead.
+ return Cost;
+ }
+
// Scalarized loads/stores.
int ConsecutiveStride = Legal->isConsecutivePtr(Ptr);
bool Reverse = ConsecutiveStride < 0;
- unsigned ScalarAllocatedSize = DL->getTypeAllocSize(ValTy);
- unsigned VectorElementSize = DL->getTypeStoreSize(VectorTy)/VF;
+ const DataLayout &DL = I->getModule()->getDataLayout();
+ unsigned ScalarAllocatedSize = DL.getTypeAllocSize(ValTy);
+ unsigned VectorElementSize = DL.getTypeStoreSize(VectorTy) / VF;
if (!ConsecutiveStride || ScalarAllocatedSize != VectorElementSize) {
bool IsComplexComputation =
isLikelyComplexAddressComputation(Ptr, Legal, SE, TheLoop);
// Wide load/stores.
unsigned Cost = TTI.getAddressComputationCost(VectorTy);
- Cost += TTI.getMemoryOpCost(I->getOpcode(), VectorTy, Alignment, AS);
+ if (Legal->isMaskRequired(I))
+ Cost += TTI.getMaskedMemoryOpCost(I->getOpcode(), VectorTy, Alignment,
+ AS);
+ else
+ Cost += TTI.getMemoryOpCost(I->getOpcode(), VectorTy, Alignment, AS);
if (Reverse)
Cost += TTI.getShuffleCost(TargetTransformInfo::SK_Reverse,
Legal->isInductionVariable(I->getOperand(0)))
return TTI.getCastInstrCost(I->getOpcode(), I->getType(),
I->getOperand(0)->getType());
-
- Type *SrcVecTy = ToVectorTy(I->getOperand(0)->getType(), VF);
+
+ Type *SrcScalarTy = I->getOperand(0)->getType();
+ Type *SrcVecTy = ToVectorTy(SrcScalarTy, VF);
+ if (VF > 1 && MinBWs.count(I)) {
+ // This cast is going to be shrunk. This may remove the cast or it might
+ // turn it into slightly different cast. For example, if MinBW == 16,
+ // "zext i8 %1 to i32" becomes "zext i8 %1 to i16".
+ //
+ // Calculate the modified src and dest types.
+ Type *MinVecTy = VectorTy;
+ if (I->getOpcode() == Instruction::Trunc) {
+ SrcVecTy = smallestIntegerVectorType(SrcVecTy, MinVecTy);
+ VectorTy = largestIntegerVectorType(ToVectorTy(I->getType(), VF),
+ MinVecTy);
+ } else if (I->getOpcode() == Instruction::ZExt ||
+ I->getOpcode() == Instruction::SExt) {
+ SrcVecTy = largestIntegerVectorType(SrcVecTy, MinVecTy);
+ VectorTy = smallestIntegerVectorType(ToVectorTy(I->getType(), VF),
+ MinVecTy);
+ }
+ }
+
return TTI.getCastInstrCost(I->getOpcode(), VectorTy, SrcVecTy);
}
case Instruction::Call: {
+ bool NeedToScalarize;
CallInst *CI = cast<CallInst>(I);
- Intrinsic::ID ID = getIntrinsicIDForCall(CI, TLI);
- assert(ID && "Not an intrinsic call!");
- Type *RetTy = ToVectorTy(CI->getType(), VF);
- SmallVector<Type*, 4> Tys;
- for (unsigned i = 0, ie = CI->getNumArgOperands(); i != ie; ++i)
- Tys.push_back(ToVectorTy(CI->getArgOperand(i)->getType(), VF));
- return TTI.getIntrinsicInstrCost(ID, RetTy, Tys);
+ unsigned CallCost = getVectorCallCost(CI, VF, TTI, TLI, NeedToScalarize);
+ if (getIntrinsicIDForCall(CI, TLI))
+ return std::min(CallCost, getVectorIntrinsicCost(CI, VF, TTI, TLI));
+ return CallCost;
}
default: {
// We are scalarizing the instruction. Return the cost of the scalar
}// end of switch.
}
-Type* LoopVectorizationCostModel::ToVectorTy(Type *Scalar, unsigned VF) {
- if (Scalar->isVoidTy() || VF == 1)
- return Scalar;
- return VectorType::get(Scalar, VF);
-}
-
char LoopVectorize::ID = 0;
static const char lv_name[] = "Loop Vectorization";
INITIALIZE_PASS_BEGIN(LoopVectorize, LV_NAME, lv_name, false, false)
-INITIALIZE_AG_DEPENDENCY(TargetTransformInfo)
-INITIALIZE_PASS_DEPENDENCY(DominatorTree)
-INITIALIZE_PASS_DEPENDENCY(ScalarEvolution)
+INITIALIZE_PASS_DEPENDENCY(TargetTransformInfoWrapperPass)
+INITIALIZE_PASS_DEPENDENCY(BasicAAWrapperPass)
+INITIALIZE_PASS_DEPENDENCY(AAResultsWrapperPass)
+INITIALIZE_PASS_DEPENDENCY(GlobalsAAWrapperPass)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
+INITIALIZE_PASS_DEPENDENCY(BlockFrequencyInfoWrapperPass)
+INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
+INITIALIZE_PASS_DEPENDENCY(ScalarEvolutionWrapperPass)
INITIALIZE_PASS_DEPENDENCY(LCSSA)
-INITIALIZE_PASS_DEPENDENCY(LoopInfo)
+INITIALIZE_PASS_DEPENDENCY(LoopInfoWrapperPass)
INITIALIZE_PASS_DEPENDENCY(LoopSimplify)
+INITIALIZE_PASS_DEPENDENCY(LoopAccessAnalysis)
+INITIALIZE_PASS_DEPENDENCY(DemandedBits)
INITIALIZE_PASS_END(LoopVectorize, LV_NAME, lv_name, false, false)
namespace llvm {
- Pass *createLoopVectorizePass(bool NoUnrolling) {
- return new LoopVectorize(NoUnrolling);
+ Pass *createLoopVectorizePass(bool NoUnrolling, bool AlwaysVectorize) {
+ return new LoopVectorize(NoUnrolling, AlwaysVectorize);
}
}
}
-void InnerLoopUnroller::scalarizeInstruction(Instruction *Instr) {
+void InnerLoopUnroller::scalarizeInstruction(Instruction *Instr,
+ bool IfPredicateStore) {
assert(!Instr->getType()->isAggregateType() && "Can't handle vectors");
// Holds vector parameters or scalars, in case of uniform vals.
SmallVector<VectorParts, 4> Params;
// Does this instruction return a value ?
bool IsVoidRetTy = Instr->getType()->isVoidTy();
- Value *UndefVec = IsVoidRetTy ? 0 :
+ Value *UndefVec = IsVoidRetTy ? nullptr :
UndefValue::get(Instr->getType());
// Create a new entry in the WidenMap and initialize it to Undef or Null.
VectorParts &VecResults = WidenMap.splat(Instr, UndefVec);
+ VectorParts Cond;
+ if (IfPredicateStore) {
+ assert(Instr->getParent()->getSinglePredecessor() &&
+ "Only support single predecessor blocks");
+ Cond = createEdgeMask(Instr->getParent()->getSinglePredecessor(),
+ Instr->getParent());
+ }
+
// For each vector unroll 'part':
for (unsigned Part = 0; Part < UF; ++Part) {
// For each scalar that we create:
+ // Start an "if (pred) a[i] = ..." block.
+ Value *Cmp = nullptr;
+ if (IfPredicateStore) {
+ if (Cond[Part]->getType()->isVectorTy())
+ Cond[Part] =
+ Builder.CreateExtractElement(Cond[Part], Builder.getInt32(0));
+ Cmp = Builder.CreateICmp(ICmpInst::ICMP_EQ, Cond[Part],
+ ConstantInt::get(Cond[Part]->getType(), 1));
+ }
+
Instruction *Cloned = Instr->clone();
if (!IsVoidRetTy)
Cloned->setName(Instr->getName() + ".cloned");
// so that future users will be able to use it.
if (!IsVoidRetTy)
VecResults[Part] = Cloned;
+
+ // End if-block.
+ if (IfPredicateStore)
+ PredicatedStores.push_back(std::make_pair(cast<StoreInst>(Cloned),
+ Cmp));
}
}
-void
-InnerLoopUnroller::vectorizeMemoryInstruction(Instruction *Instr,
- LoopVectorizationLegality*) {
- return scalarizeInstruction(Instr);
+void InnerLoopUnroller::vectorizeMemoryInstruction(Instruction *Instr) {
+ StoreInst *SI = dyn_cast<StoreInst>(Instr);
+ bool IfPredicateStore = (SI && Legal->blockNeedsPredication(SI->getParent()));
+
+ return scalarizeInstruction(Instr, IfPredicateStore);
}
Value *InnerLoopUnroller::reverseVector(Value *Vec) {
return V;
}
-Value *InnerLoopUnroller::getConsecutiveVector(Value* Val, int StartIdx,
- bool Negate) {
+Value *InnerLoopUnroller::getStepVector(Value *Val, int StartIdx, Value *Step) {
// When unrolling and the VF is 1, we only need to add a simple scalar.
Type *ITy = Val->getType();
assert(!ITy->isVectorTy() && "Val must be a scalar");
- Constant *C = ConstantInt::get(ITy, StartIdx, Negate);
- return Builder.CreateAdd(Val, C, "induction");
+ Constant *C = ConstantInt::get(ITy, StartIdx);
+ return Builder.CreateAdd(Val, Builder.CreateMul(C, Step), "induction");
}
-
projects / oota-llvm.git / blobdiff