LoopVectorize.cpp

Go to the documentation of this file.

//===- LoopVectorize.cpp - A Loop Vectorizer ------------------------------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
//
// This is the LLVM loop vectorizer. This pass modifies 'vectorizable' loops
// and generates target-independent LLVM-IR.
// The vectorizer uses the TargetTransformInfo analysis to estimate the costs
// of instructions in order to estimate the profitability of vectorization.
//
// The loop vectorizer combines consecutive loop iterations into a single
// 'wide' iteration. After this transformation the index is incremented
// by the SIMD vector width, and not by one.
//
// This pass has three parts:
// 1. The main loop pass that drives the different parts.
// 2. LoopVectorizationLegality - A unit that checks for the legality
//    of the vectorization.
// 3. InnerLoopVectorizer - A unit that performs the actual
//    widening of instructions.
// 4. LoopVectorizationCostModel - A unit that checks for the profitability
//    of vectorization. It decides on the optimal vector width, which
//    can be one, if vectorization is not profitable.
//
// There is a development effort going on to migrate loop vectorizer to the
// VPlan infrastructure and to introduce outer loop vectorization support (see
// docs/VectorizationPlan.rst and
// http://lists.llvm.org/pipermail/llvm-dev/2017-December/119523.html). For this
// purpose, we temporarily introduced the VPlan-native vectorization path: an
// alternative vectorization path that is natively implemented on top of the
// VPlan infrastructure. See EnableVPlanNativePath for enabling.
//
//===----------------------------------------------------------------------===//
//
// The reduction-variable vectorization is based on the paper:
//  D. Nuzman and R. Henderson. Multi-platform Auto-vectorization.
//
// 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.
//
//  S. Maleki, Y. Gao, M. Garzaran, T. Wong and D. Padua.  An Evaluation of
//  Vectorizing Compilers.
//
//===----------------------------------------------------------------------===//
 
#include "llvm/Transforms/Vectorize/LoopVectorize.h"
#include "LoopVectorizationPlanner.h"
#include "VPRecipeBuilder.h"
#include "VPlan.h"
#include "VPlanHCFGBuilder.h"
#include "VPlanTransforms.h"
#include "llvm/ADT/APInt.h"
#include "llvm/ADT/ArrayRef.h"
#include "llvm/ADT/DenseMap.h"
#include "llvm/ADT/DenseMapInfo.h"
#include "llvm/ADT/Hashing.h"
#include "llvm/ADT/MapVector.h"
#include "llvm/ADT/STLExtras.h"
#include "llvm/ADT/SmallPtrSet.h"
#include "llvm/ADT/SmallSet.h"
#include "llvm/ADT/SmallVector.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/ADT/StringRef.h"
#include "llvm/ADT/Twine.h"
#include "llvm/ADT/iterator_range.h"
#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/BasicAliasAnalysis.h"
#include "llvm/Analysis/BlockFrequencyInfo.h"
#include "llvm/Analysis/CFG.h"
#include "llvm/Analysis/CodeMetrics.h"
#include "llvm/Analysis/DemandedBits.h"
#include "llvm/Analysis/GlobalsModRef.h"
#include "llvm/Analysis/LoopAccessAnalysis.h"
#include "llvm/Analysis/LoopAnalysisManager.h"
#include "llvm/Analysis/LoopInfo.h"
#include "llvm/Analysis/LoopIterator.h"
#include "llvm/Analysis/OptimizationRemarkEmitter.h"
#include "llvm/Analysis/ProfileSummaryInfo.h"
#include "llvm/Analysis/ScalarEvolution.h"
#include "llvm/Analysis/ScalarEvolutionExpressions.h"
#include "llvm/Analysis/TargetLibraryInfo.h"
#include "llvm/Analysis/TargetTransformInfo.h"
#include "llvm/Analysis/ValueTracking.h"
#include "llvm/Analysis/VectorUtils.h"
#include "llvm/IR/Attributes.h"
#include "llvm/IR/BasicBlock.h"
#include "llvm/IR/CFG.h"
#include "llvm/IR/Constant.h"
#include "llvm/IR/Constants.h"
#include "llvm/IR/DataLayout.h"
#include "llvm/IR/DebugInfo.h"
#include "llvm/IR/DebugInfoMetadata.h"
#include "llvm/IR/DebugLoc.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/InstrTypes.h"
#include "llvm/IR/Instruction.h"
#include "llvm/IR/Instructions.h"
#include "llvm/IR/IntrinsicInst.h"
#include "llvm/IR/Intrinsics.h"
#include "llvm/IR/Metadata.h"
#include "llvm/IR/Module.h"
#include "llvm/IR/Operator.h"
#include "llvm/IR/PatternMatch.h"
#include "llvm/IR/Type.h"
#include "llvm/IR/Use.h"
#include "llvm/IR/User.h"
#include "llvm/IR/Value.h"
#include "llvm/IR/ValueHandle.h"
#include "llvm/IR/Verifier.h"
#include "llvm/Support/Casting.h"
#include "llvm/Support/CommandLine.h"
#include "llvm/Support/Compiler.h"
#include "llvm/Support/Debug.h"
#include "llvm/Support/ErrorHandling.h"
#include "llvm/Support/InstructionCost.h"
#include "llvm/Support/MathExtras.h"
#include "llvm/Support/raw_ostream.h"
#include "llvm/Transforms/Utils/BasicBlockUtils.h"
#include "llvm/Transforms/Utils/InjectTLIMappings.h"
#include "llvm/Transforms/Utils/LoopSimplify.h"
#include "llvm/Transforms/Utils/LoopUtils.h"
#include "llvm/Transforms/Utils/LoopVersioning.h"
#include "llvm/Transforms/Utils/ScalarEvolutionExpander.h"
#include "llvm/Transforms/Utils/SizeOpts.h"
#include "llvm/Transforms/Vectorize/LoopVectorizationLegality.h"
#include <algorithm>
#include <cassert>
#include <cmath>
#include <cstdint>
#include <functional>
#include <iterator>
#include <limits>
#include <map>
#include <memory>
#include <string>
#include <tuple>
#include <utility>
 
using namespace llvm;
 
#define LV_NAME "loop-vectorize"
#define DEBUG_TYPE LV_NAME
 
#ifndef NDEBUG
const char VerboseDebug[] = DEBUG_TYPE "-verbose";
#endif
 
/// @{
/// Metadata attribute names
const char LLVMLoopVectorizeFollowupAll[] = "llvm.loop.vectorize.followup_all";
const char LLVMLoopVectorizeFollowupVectorized[] =
    "llvm.loop.vectorize.followup_vectorized";
const char LLVMLoopVectorizeFollowupEpilogue[] =
    "llvm.loop.vectorize.followup_epilogue";
/// @}
 
STATISTIC(LoopsVectorized, "Number of loops vectorized");
STATISTIC(LoopsAnalyzed, "Number of loops analyzed for vectorization");
STATISTIC(LoopsEpilogueVectorized, "Number of epilogues vectorized");
 
static cl::opt<bool> EnableEpilogueVectorization(
    "enable-epilogue-vectorization", cl::init(true), cl::Hidden,
    cl::desc("Enable vectorization of epilogue loops."));
 
static cl::opt<unsigned> EpilogueVectorizationForceVF(
    "epilogue-vectorization-force-VF", cl::init(1), cl::Hidden,
    cl::desc("When epilogue vectorization is enabled, and a value greater than "
             "1 is specified, forces the given VF for all applicable epilogue "
             "loops."));
 
static cl::opt<unsigned> EpilogueVectorizationMinVF(
    "epilogue-vectorization-minimum-VF", cl::init(16), cl::Hidden,
    cl::desc("Only loops with vectorization factor equal to or larger than "
             "the specified value are considered for epilogue vectorization."));
 
/// Loops with a known constant trip count below this number are vectorized only
/// if no scalar iteration overheads are incurred.
static cl::opt<unsigned> TinyTripCountVectorThreshold(
    "vectorizer-min-trip-count", cl::init(16), cl::Hidden,
    cl::desc("Loops with a constant trip count that is smaller than this "
             "value are vectorized only if no scalar iteration overheads "
             "are incurred."));
 
static cl::opt<unsigned> VectorizeMemoryCheckThreshold(
    "vectorize-memory-check-threshold", cl::init(128), cl::Hidden,
    cl::desc("The maximum allowed number of runtime memory checks"));
 
// Option prefer-predicate-over-epilogue indicates that an epilogue is undesired,
// that predication is preferred, and this lists all options. I.e., the
// vectorizer will try to fold the tail-loop (epilogue) into the vector body
// and predicate the instructions accordingly. If tail-folding fails, there are
// different fallback strategies depending on these values:
namespace PreferPredicateTy {
  enum Option {
    ScalarEpilogue = 0,
    PredicateElseScalarEpilogue,
    PredicateOrDontVectorize
  };
} // namespace PreferPredicateTy
 
static cl::opt<PreferPredicateTy::Option> PreferPredicateOverEpilogue(
    "prefer-predicate-over-epilogue",
    cl::init(PreferPredicateTy::ScalarEpilogue),
    cl::Hidden,
    cl::desc("Tail-folding and predication preferences over creating a scalar "
             "epilogue loop."),
    cl::values(clEnumValN(PreferPredicateTy::ScalarEpilogue,
                         "scalar-epilogue",
                         "Don't tail-predicate loops, create scalar epilogue"),
              clEnumValN(PreferPredicateTy::PredicateElseScalarEpilogue,
                         "predicate-else-scalar-epilogue",
                         "prefer tail-folding, create scalar epilogue if tail "
                         "folding fails."),
              clEnumValN(PreferPredicateTy::PredicateOrDontVectorize,
                         "predicate-dont-vectorize",
                         "prefers tail-folding, don't attempt vectorization if "
                         "tail-folding fails.")));
 
static cl::opt<TailFoldingStyle> ForceTailFoldingStyle(
    "force-tail-folding-style", cl::desc("Force the tail folding style"),
    cl::init(TailFoldingStyle::None),
    cl::values(
        clEnumValN(TailFoldingStyle::None, "none", "Disable tail folding"),
        clEnumValN(
            TailFoldingStyle::Data, "data",
            "Create lane mask for data only, using active.lane.mask intrinsic"),
        clEnumValN(TailFoldingStyle::DataWithoutLaneMask,
                   "data-without-lane-mask",
                   "Create lane mask with compare/stepvector"),
        clEnumValN(TailFoldingStyle::DataAndControlFlow, "data-and-control",
                   "Create lane mask using active.lane.mask intrinsic, and use "
                   "it for both data and control flow"),
        clEnumValN(
            TailFoldingStyle::DataAndControlFlowWithoutRuntimeCheck,
            "data-and-control-without-rt-check",
            "Similar to data-and-control, but remove the runtime check")));
 
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."));
 
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"));
 
/// An interleave-group may need masking if it resides in a block that needs
/// predication, or in order to mask away gaps.
static cl::opt<bool> EnableMaskedInterleavedMemAccesses(
    "enable-masked-interleaved-mem-accesses", cl::init(false), cl::Hidden,
    cl::desc("Enable vectorization on masked interleaved memory accesses in a loop"));
 
static cl::opt<unsigned> TinyTripCountInterleaveThreshold(
    "tiny-trip-count-interleave-threshold", cl::init(128), cl::Hidden,
    cl::desc("We don't interleave loops with a estimated constant trip count "
             "below this number"));
 
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."));
 
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<bool> ForceTargetSupportsScalableVectors(
    "force-target-supports-scalable-vectors", cl::init(false), cl::Hidden,
    cl::desc(
        "Pretend that scalable vectors are supported, even if the target does "
        "not support them. This flag should only be used for 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(true), 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"));
 
/// Interleave small loops with scalar reductions.
static cl::opt<bool> InterleaveSmallLoopScalarReduction(
    "interleave-small-loop-scalar-reduction", cl::init(false), cl::Hidden,
    cl::desc("Enable interleaving for loops with small iteration counts that "
             "contain scalar reductions to expose ILP."));
 
/// 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(true), 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<bool>
    PreferInLoopReductions("prefer-inloop-reductions", cl::init(false),
                           cl::Hidden,
                           cl::desc("Prefer in-loop vector reductions, "
                                    "overriding the targets preference."));
 
static cl::opt<bool> ForceOrderedReductions(
    "force-ordered-reductions", cl::init(false), cl::Hidden,
    cl::desc("Enable the vectorisation of loops with in-order (strict) "
             "FP reductions"));
 
static cl::opt<bool> PreferPredicatedReductionSelect(
    "prefer-predicated-reduction-select", cl::init(false), cl::Hidden,
    cl::desc(
        "Prefer predicating a reduction operation over an after loop select."));
 
namespace llvm {
cl::opt<bool> EnableVPlanNativePath(
    "enable-vplan-native-path", cl::Hidden,
    cl::desc("Enable VPlan-native vectorization path with "
             "support for outer loop vectorization."));
}
 
// This flag enables the stress testing of the VPlan H-CFG construction in the
// VPlan-native vectorization path. It must be used in conjuction with
// -enable-vplan-native-path. -vplan-verify-hcfg can also be used to enable the
// verification of the H-CFGs built.
static cl::opt<bool> VPlanBuildStressTest(
    "vplan-build-stress-test", cl::init(false), cl::Hidden,
    cl::desc(
        "Build VPlan for every supported loop nest in the function and bail "
        "out right after the build (stress test the VPlan H-CFG construction "
        "in the VPlan-native vectorization path)."));
 
cl::opt<bool> llvm::EnableLoopInterleaving(
    "interleave-loops", cl::init(true), cl::Hidden,
    cl::desc("Enable loop interleaving in Loop vectorization passes"));
cl::opt<bool> llvm::EnableLoopVectorization(
    "vectorize-loops", cl::init(true), cl::Hidden,
    cl::desc("Run the Loop vectorization passes"));
 
static cl::opt<bool> PrintVPlansInDotFormat(
    "vplan-print-in-dot-format", cl::Hidden,
    cl::desc("Use dot format instead of plain text when dumping VPlans"));
 
static cl::opt<cl::boolOrDefault> ForceSafeDivisor(
    "force-widen-divrem-via-safe-divisor", cl::Hidden,
    cl::desc(
        "Override cost based safe divisor widening for div/rem instructions"));
 
/// A helper function that returns true if the given type is irregular. The
/// type is irregular if its allocated size doesn't equal the store size of an
/// element of the corresponding vector type.
static bool hasIrregularType(Type *Ty, const DataLayout &DL) {
  // Determine if an array of N elements of type Ty is "bitcast compatible"
  // with a <N x Ty> vector.
  // This is only true if there is no padding between the array elements.
  return DL.getTypeAllocSizeInBits(Ty) != DL.getTypeSizeInBits(Ty);
}
 
/// A helper function that returns the reciprocal of the block probability of
/// predicated blocks. If we return X, we are assuming the predicated block
/// will execute once for every X iterations of the loop header.
///
/// TODO: We should use actual block probability here, if available. Currently,
///       we always assume predicated blocks have a 50% chance of executing.
static unsigned getReciprocalPredBlockProb() { return 2; }
 
/// Returns "best known" trip count for the specified loop \p L as defined by
/// the following procedure:
///   1) Returns exact trip count if it is known.
///   2) Returns expected trip count according to profile data if any.
///   3) Returns upper bound estimate if it is known.
///   4) Returns std::nullopt if all of the above failed.
static std::optional<unsigned> getSmallBestKnownTC(ScalarEvolution &SE,
                                                   Loop *L) {
  // Check if exact trip count is known.
  if (unsigned ExpectedTC = SE.getSmallConstantTripCount(L))
    return ExpectedTC;
 
  // Check if there is an expected trip count available from profile data.
  if (LoopVectorizeWithBlockFrequency)
    if (auto EstimatedTC = getLoopEstimatedTripCount(L))
      return *EstimatedTC;
 
  // Check if upper bound estimate is known.
  if (unsigned ExpectedTC = SE.getSmallConstantMaxTripCount(L))
    return ExpectedTC;
 
  return std::nullopt;
}
 
/// Return a vector containing interleaved elements from multiple
/// smaller input vectors.
static Value *interleaveVectors(IRBuilderBase &Builder, ArrayRef<Value *> Vals,
                                const Twine &Name) {
  unsigned Factor = Vals.size();
  assert(Factor > 1 && "Tried to interleave invalid number of vectors");
 
  VectorType *VecTy = cast<VectorType>(Vals[0]->getType());
#ifndef NDEBUG
  for (Value *Val : Vals)
    assert(Val->getType() == VecTy && "Tried to interleave mismatched types");
#endif
 
  // Scalable vectors cannot use arbitrary shufflevectors (only splats), so
  // must use intrinsics to interleave.
  if (VecTy->isScalableTy()) {
    VectorType *WideVecTy = VectorType::getDoubleElementsVectorType(VecTy);
    return Builder.CreateIntrinsic(
        WideVecTy, Intrinsic::experimental_vector_interleave2, Vals,
        /*FMFSource=*/nullptr, Name);
  }
 
  // Fixed length. Start by concatenating all vectors into a wide vector.
  Value *WideVec = concatenateVectors(Builder, Vals);
 
  // Interleave the elements into the wide vector.
  const unsigned NumElts = VecTy->getElementCount().getFixedValue();
  return Builder.CreateShuffleVector(
      WideVec, createInterleaveMask(NumElts, Factor), Name);
}
 
namespace {
// Forward declare GeneratedRTChecks.
class GeneratedRTChecks;
 
using SCEV2ValueTy = DenseMap<const SCEV *, Value *>;
} // namespace
 
namespace llvm {
 
AnalysisKey ShouldRunExtraVectorPasses::Key;
 
/// InnerLoopVectorizer vectorizes loops which contain only one basic
/// block to a specified vectorization factor (VF).
/// This class performs the widening of scalars into vectors, or multiple
/// scalars. This class also implements the following features:
/// * It inserts an epilogue loop for handling loops that don't have iteration
///   counts that are known to be a multiple of the vectorization factor.
/// * It handles the code generation for reduction variables.
/// * Scalarization (implementation using scalars) of un-vectorizable
///   instructions.
/// InnerLoopVectorizer does not perform any vectorization-legality
/// checks, and relies on the caller to check for the different legality
/// aspects. The InnerLoopVectorizer relies on the
/// LoopVectorizationLegality class to provide information about the induction
/// and reduction variables that were found to a given vectorization factor.
class InnerLoopVectorizer {
public:
  InnerLoopVectorizer(Loop *OrigLoop, PredicatedScalarEvolution &PSE,
                      LoopInfo *LI, DominatorTree *DT,
                      const TargetLibraryInfo *TLI,
                      const TargetTransformInfo *TTI, AssumptionCache *AC,
                      OptimizationRemarkEmitter *ORE, ElementCount VecWidth,
                      ElementCount MinProfitableTripCount,
                      unsigned UnrollFactor, LoopVectorizationLegality *LVL,
                      LoopVectorizationCostModel *CM, BlockFrequencyInfo *BFI,
                      ProfileSummaryInfo *PSI, GeneratedRTChecks &RTChecks)
      : OrigLoop(OrigLoop), PSE(PSE), LI(LI), DT(DT), TLI(TLI), TTI(TTI),
        AC(AC), ORE(ORE), VF(VecWidth), UF(UnrollFactor),
        Builder(PSE.getSE()->getContext()), Legal(LVL), Cost(CM), BFI(BFI),
        PSI(PSI), RTChecks(RTChecks) {
    // Query this against the original loop and save it here because the profile
    // of the original loop header may change as the transformation happens.
    OptForSizeBasedOnProfile = llvm::shouldOptimizeForSize(
        OrigLoop->getHeader(), PSI, BFI, PGSOQueryType::IRPass);
 
    if (MinProfitableTripCount.isZero())
      this->MinProfitableTripCount = VecWidth;
    else
      this->MinProfitableTripCount = MinProfitableTripCount;
  }
 
  virtual ~InnerLoopVectorizer() = default;
 
  /// Create a new empty loop that will contain vectorized instructions later
  /// on, while the old loop will be used as the scalar remainder. Control flow
  /// is generated around the vectorized (and scalar epilogue) loops consisting
  /// of various checks and bypasses. Return the pre-header block of the new
  /// loop and the start value for the canonical induction, if it is != 0. The
  /// latter is the case when vectorizing the epilogue loop. In the case of
  /// epilogue vectorization, this function is overriden to handle the more
  /// complex control flow around the loops.  \p ExpandedSCEVs is used to
  /// look up SCEV expansions for expressions needed during skeleton creation.
  virtual std::pair<BasicBlock *, Value *>
  createVectorizedLoopSkeleton(const SCEV2ValueTy &ExpandedSCEVs);
 
  /// Fix the vectorized code, taking care of header phi's, live-outs, and more.
  void fixVectorizedLoop(VPTransformState &State, VPlan &Plan);
 
  // Return true if any runtime check is added.
  bool areSafetyChecksAdded() { return AddedSafetyChecks; }
 
  /// A type for vectorized values in the new loop. Each value from the
  /// original loop, when vectorized, is represented by UF vector values in the
  /// new unrolled loop, where UF is the unroll factor.
  using VectorParts = SmallVector<Value *, 2>;
 
  /// A helper function to scalarize a single Instruction in the innermost loop.
  /// Generates a sequence of scalar instances for each lane between \p MinLane
  /// and \p MaxLane, times each part between \p MinPart and \p MaxPart,
  /// inclusive. Uses the VPValue operands from \p RepRecipe instead of \p
  /// Instr's operands.
  void scalarizeInstruction(const Instruction *Instr,
                            VPReplicateRecipe *RepRecipe,
                            const VPIteration &Instance,
                            VPTransformState &State);
 
  /// Try to vectorize interleaved access group \p Group with the base address
  /// given in \p Addr, optionally masking the vector operations if \p
  /// BlockInMask is non-null. Use \p State to translate given VPValues to IR
  /// values in the vectorized loop.
  void vectorizeInterleaveGroup(const InterleaveGroup<Instruction> *Group,
                                ArrayRef<VPValue *> VPDefs,
                                VPTransformState &State, VPValue *Addr,
                                ArrayRef<VPValue *> StoredValues,
                                VPValue *BlockInMask, bool NeedsMaskForGaps);
 
  /// Fix the non-induction PHIs in \p Plan.
  void fixNonInductionPHIs(VPlan &Plan, VPTransformState &State);
 
  /// Returns true if the reordering of FP operations is not allowed, but we are
  /// able to vectorize with strict in-order reductions for the given RdxDesc.
  bool useOrderedReductions(const RecurrenceDescriptor &RdxDesc);
 
  // Returns the resume value (bc.merge.rdx) for a reduction as
  // generated by fixReduction.
  PHINode *getReductionResumeValue(const RecurrenceDescriptor &RdxDesc);
 
  /// Create a new phi node for the induction variable \p OrigPhi to resume
  /// iteration count in the scalar epilogue, from where the vectorized loop
  /// left off. \p Step is the SCEV-expanded induction step to use. In cases
  /// where the loop skeleton is more complicated (i.e., epilogue vectorization)
  /// and the resume values can come from an additional bypass block, the \p
  /// AdditionalBypass pair provides information about the bypass block and the
  /// end value on the edge from bypass to this loop.
  PHINode *createInductionResumeValue(
      PHINode *OrigPhi, const InductionDescriptor &ID, Value *Step,
      ArrayRef<BasicBlock *> BypassBlocks,
      std::pair<BasicBlock *, Value *> AdditionalBypass = {nullptr, nullptr});
 
  /// Returns the original loop trip count.
  Value *getTripCount() const { return TripCount; }
 
  /// Used to set the trip count after ILV's construction and after the
  /// preheader block has been executed. Note that this always holds the trip
  /// count of the original loop for both main loop and epilogue vectorization.
  void setTripCount(Value *TC) { TripCount = TC; }
 
protected:
  friend class LoopVectorizationPlanner;
 
  /// A small list of PHINodes.
  using PhiVector = SmallVector<PHINode *, 4>;
 
  /// A type for scalarized values in the new loop. Each value from the
  /// original loop, when scalarized, is represented by UF x VF scalar values
  /// in the new unrolled loop, where UF is the unroll factor and VF is the
  /// vectorization factor.
  using ScalarParts = SmallVector<SmallVector<Value *, 4>, 2>;
 
  /// Set up the values of the IVs correctly when exiting the vector loop.
  void fixupIVUsers(PHINode *OrigPhi, const InductionDescriptor &II,
                    Value *VectorTripCount, Value *EndValue,
                    BasicBlock *MiddleBlock, BasicBlock *VectorHeader,
                    VPlan &Plan, VPTransformState &State);
 
  /// Handle all cross-iteration phis in the header.
  void fixCrossIterationPHIs(VPTransformState &State);
 
  /// Create the exit value of first order recurrences in the middle block and
  /// update their users.
  void fixFixedOrderRecurrence(VPFirstOrderRecurrencePHIRecipe *PhiR,
                               VPTransformState &State);
 
  /// Create code for the loop exit value of the reduction.
  void fixReduction(VPReductionPHIRecipe *Phi, VPTransformState &State);
 
  /// Iteratively sink the scalarized operands of a predicated instruction into
  /// the block that was created for it.
  void sinkScalarOperands(Instruction *PredInst);
 
  /// Shrinks vector element sizes to the smallest bitwidth they can be legally
  /// represented as.
  void truncateToMinimalBitwidths(VPTransformState &State);
 
  /// Returns (and creates if needed) the trip count of the widened loop.
  Value *getOrCreateVectorTripCount(BasicBlock *InsertBlock);
 
  /// Returns a bitcasted value to the requested vector type.
  /// Also handles bitcasts of vector<float> <-> vector<pointer> types.
  Value *createBitOrPointerCast(Value *V, VectorType *DstVTy,
                                const DataLayout &DL);
 
  /// Emit a bypass check to see if the vector trip count is zero, including if
  /// it overflows.
  void emitIterationCountCheck(BasicBlock *Bypass);
 
  /// Emit a bypass check to see if all of the SCEV assumptions we've
  /// had to make are correct. Returns the block containing the checks or
  /// nullptr if no checks have been added.
  BasicBlock *emitSCEVChecks(BasicBlock *Bypass);
 
  /// Emit bypass checks to check any memory assumptions we may have made.
  /// Returns the block containing the checks or nullptr if no checks have been
  /// added.
  BasicBlock *emitMemRuntimeChecks(BasicBlock *Bypass);
 
  /// Emit basic blocks (prefixed with \p Prefix) for the iteration check,
  /// vector loop preheader, middle block and scalar preheader.
  void createVectorLoopSkeleton(StringRef Prefix);
 
  /// Create new phi nodes for the induction variables to resume iteration count
  /// in the scalar epilogue, from where the vectorized loop left off.
  /// In cases where the loop skeleton is more complicated (eg. epilogue
  /// vectorization) and the resume values can come from an additional bypass
  /// block, the \p AdditionalBypass pair provides information about the bypass
  /// block and the end value on the edge from bypass to this loop.
  void createInductionResumeValues(
      const SCEV2ValueTy &ExpandedSCEVs,
      std::pair<BasicBlock *, Value *> AdditionalBypass = {nullptr, nullptr});
 
  /// Complete the loop skeleton by adding debug MDs, creating appropriate
  /// conditional branches in the middle block, preparing the builder and
  /// running the verifier. Return the preheader of the completed vector loop.
  BasicBlock *completeLoopSkeleton();
 
  /// Collect poison-generating recipes that may generate a poison value that is
  /// used after vectorization, even when their operands are not poison. Those
  /// recipes meet the following conditions:
  ///  * Contribute to the address computation of a recipe generating a widen
  ///    memory load/store (VPWidenMemoryInstructionRecipe or
  ///    VPInterleaveRecipe).
  ///  * Such a widen memory load/store has at least one underlying Instruction
  ///    that is in a basic block that needs predication and after vectorization
  ///    the generated instruction won't be predicated.
  void collectPoisonGeneratingRecipes(VPTransformState &State);
 
  /// Allow subclasses to override and print debug traces before/after vplan
  /// execution, when trace information is requested.
  virtual void printDebugTracesAtStart(){};
  virtual void printDebugTracesAtEnd(){};
 
  /// The original loop.
  Loop *OrigLoop;
 
  /// 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;
 
  /// Target Library Info.
  const TargetLibraryInfo *TLI;
 
  /// Target Transform Info.
  const TargetTransformInfo *TTI;
 
  /// Assumption Cache.
  AssumptionCache *AC;
 
  /// Interface to emit optimization remarks.
  OptimizationRemarkEmitter *ORE;
 
  /// The vectorization SIMD factor to use. Each vector will have this many
  /// vector elements.
  ElementCount VF;
 
  ElementCount MinProfitableTripCount;
 
  /// The vectorization unroll factor to use. Each scalar is vectorized to this
  /// many different vector instructions.
  unsigned UF;
 
  /// The builder that we use
  IRBuilder<> Builder;
 
  // --- Vectorization state ---
 
  /// The vector-loop preheader.
  BasicBlock *LoopVectorPreHeader;
 
  /// The scalar-loop preheader.
  BasicBlock *LoopScalarPreHeader;
 
  /// Middle Block between the vector and the scalar.
  BasicBlock *LoopMiddleBlock;
 
  /// The unique ExitBlock of the scalar loop if one exists.  Note that
  /// there can be multiple exiting edges reaching this block.
  BasicBlock *LoopExitBlock;
 
  /// The scalar loop body.
  BasicBlock *LoopScalarBody;
 
  /// A list of all bypass blocks. The first block is the entry of the loop.
  SmallVector<BasicBlock *, 4> LoopBypassBlocks;
 
  /// Store instructions that were predicated.
  SmallVector<Instruction *, 4> PredicatedInstructions;
 
  /// Trip count of the original loop.
  Value *TripCount = nullptr;
 
  /// Trip count of the widened loop (TripCount - TripCount % (VF*UF))
  Value *VectorTripCount = nullptr;
 
  /// The legality analysis.
  LoopVectorizationLegality *Legal;
 
  /// The profitablity analysis.
  LoopVectorizationCostModel *Cost;
 
  // Record whether runtime checks are added.
  bool AddedSafetyChecks = false;
 
  // Holds the end values for each induction variable. We save the end values
  // so we can later fix-up the external users of the induction variables.
  DenseMap<PHINode *, Value *> IVEndValues;
 
  /// BFI and PSI are used to check for profile guided size optimizations.
  BlockFrequencyInfo *BFI;
  ProfileSummaryInfo *PSI;
 
  // Whether this loop should be optimized for size based on profile guided size
  // optimizatios.
  bool OptForSizeBasedOnProfile;
 
  /// Structure to hold information about generated runtime checks, responsible
  /// for cleaning the checks, if vectorization turns out unprofitable.
  GeneratedRTChecks &RTChecks;
 
  // Holds the resume values for reductions in the loops, used to set the
  // correct start value of reduction PHIs when vectorizing the epilogue.
  SmallMapVector<const RecurrenceDescriptor *, PHINode *, 4>
      ReductionResumeValues;
};
 
class InnerLoopUnroller : public InnerLoopVectorizer {
public:
  InnerLoopUnroller(Loop *OrigLoop, PredicatedScalarEvolution &PSE,
                    LoopInfo *LI, DominatorTree *DT,
                    const TargetLibraryInfo *TLI,
                    const TargetTransformInfo *TTI, AssumptionCache *AC,
                    OptimizationRemarkEmitter *ORE, unsigned UnrollFactor,
                    LoopVectorizationLegality *LVL,
                    LoopVectorizationCostModel *CM, BlockFrequencyInfo *BFI,
                    ProfileSummaryInfo *PSI, GeneratedRTChecks &Check)
      : InnerLoopVectorizer(OrigLoop, PSE, LI, DT, TLI, TTI, AC, ORE,
                            ElementCount::getFixed(1),
                            ElementCount::getFixed(1), UnrollFactor, LVL, CM,
                            BFI, PSI, Check) {}
};
 
/// Encapsulate information regarding vectorization of a loop and its epilogue.
/// This information is meant to be updated and used across two stages of
/// epilogue vectorization.
struct EpilogueLoopVectorizationInfo {
  ElementCount MainLoopVF = ElementCount::getFixed(0);
  unsigned MainLoopUF = 0;
  ElementCount EpilogueVF = ElementCount::getFixed(0);
  unsigned EpilogueUF = 0;
  BasicBlock *MainLoopIterationCountCheck = nullptr;
  BasicBlock *EpilogueIterationCountCheck = nullptr;
  BasicBlock *SCEVSafetyCheck = nullptr;
  BasicBlock *MemSafetyCheck = nullptr;
  Value *TripCount = nullptr;
  Value *VectorTripCount = nullptr;
 
  EpilogueLoopVectorizationInfo(ElementCount MVF, unsigned MUF,
                                ElementCount EVF, unsigned EUF)
      : MainLoopVF(MVF), MainLoopUF(MUF), EpilogueVF(EVF), EpilogueUF(EUF) {
    assert(EUF == 1 &&
           "A high UF for the epilogue loop is likely not beneficial.");
  }
};
 
/// An extension of the inner loop vectorizer that creates a skeleton for a
/// vectorized loop that has its epilogue (residual) also vectorized.
/// The idea is to run the vplan on a given loop twice, firstly to setup the
/// skeleton and vectorize the main loop, and secondly to complete the skeleton
/// from the first step and vectorize the epilogue.  This is achieved by
/// deriving two concrete strategy classes from this base class and invoking
/// them in succession from the loop vectorizer planner.
class InnerLoopAndEpilogueVectorizer : public InnerLoopVectorizer {
public:
  InnerLoopAndEpilogueVectorizer(
      Loop *OrigLoop, PredicatedScalarEvolution &PSE, LoopInfo *LI,
      DominatorTree *DT, const TargetLibraryInfo *TLI,
      const TargetTransformInfo *TTI, AssumptionCache *AC,
      OptimizationRemarkEmitter *ORE, EpilogueLoopVectorizationInfo &EPI,
      LoopVectorizationLegality *LVL, llvm::LoopVectorizationCostModel *CM,
      BlockFrequencyInfo *BFI, ProfileSummaryInfo *PSI,
      GeneratedRTChecks &Checks)
      : InnerLoopVectorizer(OrigLoop, PSE, LI, DT, TLI, TTI, AC, ORE,
                            EPI.MainLoopVF, EPI.MainLoopVF, EPI.MainLoopUF, LVL,
                            CM, BFI, PSI, Checks),
        EPI(EPI) {}
 
  // Override this function to handle the more complex control flow around the
  // three loops.
  std::pair<BasicBlock *, Value *> createVectorizedLoopSkeleton(
      const SCEV2ValueTy &ExpandedSCEVs) final {
    return createEpilogueVectorizedLoopSkeleton(ExpandedSCEVs);
  }
 
  /// The interface for creating a vectorized skeleton using one of two
  /// different strategies, each corresponding to one execution of the vplan
  /// as described above.
  virtual std::pair<BasicBlock *, Value *>
  createEpilogueVectorizedLoopSkeleton(const SCEV2ValueTy &ExpandedSCEVs) = 0;
 
  /// Holds and updates state information required to vectorize the main loop
  /// and its epilogue in two separate passes. This setup helps us avoid
  /// regenerating and recomputing runtime safety checks. It also helps us to
  /// shorten the iteration-count-check path length for the cases where the
  /// iteration count of the loop is so small that the main vector loop is
  /// completely skipped.
  EpilogueLoopVectorizationInfo &EPI;
};
 
/// A specialized derived class of inner loop vectorizer that performs
/// vectorization of *main* loops in the process of vectorizing loops and their
/// epilogues.
class EpilogueVectorizerMainLoop : public InnerLoopAndEpilogueVectorizer {
public:
  EpilogueVectorizerMainLoop(
      Loop *OrigLoop, PredicatedScalarEvolution &PSE, LoopInfo *LI,
      DominatorTree *DT, const TargetLibraryInfo *TLI,
      const TargetTransformInfo *TTI, AssumptionCache *AC,
      OptimizationRemarkEmitter *ORE, EpilogueLoopVectorizationInfo &EPI,
      LoopVectorizationLegality *LVL, llvm::LoopVectorizationCostModel *CM,
      BlockFrequencyInfo *BFI, ProfileSummaryInfo *PSI,
      GeneratedRTChecks &Check)
      : InnerLoopAndEpilogueVectorizer(OrigLoop, PSE, LI, DT, TLI, TTI, AC, ORE,
                                       EPI, LVL, CM, BFI, PSI, Check) {}
  /// Implements the interface for creating a vectorized skeleton using the
  /// *main loop* strategy (ie the first pass of vplan execution).
  std::pair<BasicBlock *, Value *>
  createEpilogueVectorizedLoopSkeleton(const SCEV2ValueTy &ExpandedSCEVs) final;
 
protected:
  /// Emits an iteration count bypass check once for the main loop (when \p
  /// ForEpilogue is false) and once for the epilogue loop (when \p
  /// ForEpilogue is true).
  BasicBlock *emitIterationCountCheck(BasicBlock *Bypass, bool ForEpilogue);
  void printDebugTracesAtStart() override;
  void printDebugTracesAtEnd() override;
};
 
// A specialized derived class of inner loop vectorizer that performs
// vectorization of *epilogue* loops in the process of vectorizing loops and
// their epilogues.
class EpilogueVectorizerEpilogueLoop : public InnerLoopAndEpilogueVectorizer {
public:
  EpilogueVectorizerEpilogueLoop(
      Loop *OrigLoop, PredicatedScalarEvolution &PSE, LoopInfo *LI,
      DominatorTree *DT, const TargetLibraryInfo *TLI,
      const TargetTransformInfo *TTI, AssumptionCache *AC,
      OptimizationRemarkEmitter *ORE, EpilogueLoopVectorizationInfo &EPI,
      LoopVectorizationLegality *LVL, llvm::LoopVectorizationCostModel *CM,
      BlockFrequencyInfo *BFI, ProfileSummaryInfo *PSI,
      GeneratedRTChecks &Checks)
      : InnerLoopAndEpilogueVectorizer(OrigLoop, PSE, LI, DT, TLI, TTI, AC, ORE,
                                       EPI, LVL, CM, BFI, PSI, Checks) {
    TripCount = EPI.TripCount;
  }
  /// Implements the interface for creating a vectorized skeleton using the
  /// *epilogue loop* strategy (ie the second pass of vplan execution).
  std::pair<BasicBlock *, Value *>
  createEpilogueVectorizedLoopSkeleton(const SCEV2ValueTy &ExpandedSCEVs) final;
 
protected:
  /// Emits an iteration count bypass check after the main vector loop has
  /// finished to see if there are any iterations left to execute by either
  /// the vector epilogue or the scalar epilogue.
  BasicBlock *emitMinimumVectorEpilogueIterCountCheck(
                                                      BasicBlock *Bypass,
                                                      BasicBlock *Insert);
  void printDebugTracesAtStart() override;
  void printDebugTracesAtEnd() override;
};
} // end namespace llvm
 
/// Look for a meaningful debug location on the instruction or it's
/// operands.
static DebugLoc getDebugLocFromInstOrOperands(Instruction *I) {
  if (!I)
    return DebugLoc();
 
  DebugLoc Empty;
  if (I->getDebugLoc() != Empty)
    return I->getDebugLoc();
 
  for (Use &Op : I->operands()) {
    if (Instruction *OpInst = dyn_cast<Instruction>(Op))
      if (OpInst->getDebugLoc() != Empty)
        return OpInst->getDebugLoc();
  }
 
  return I->getDebugLoc();
}
 
/// Write a \p DebugMsg about vectorization to the debug output stream. If \p I
/// is passed, the message relates to that particular instruction.
#ifndef NDEBUG
static void debugVectorizationMessage(const StringRef Prefix,
                                      const StringRef DebugMsg,
                                      Instruction *I) {
  dbgs() << "LV: " << Prefix << DebugMsg;
  if (I != nullptr)
    dbgs() << " " << *I;
  else
    dbgs() << '.';
  dbgs() << '\n';
}
#endif
 
/// Create an analysis remark that explains why vectorization failed
///
/// \p PassName is the name of the pass (e.g. can be AlwaysPrint).  \p
/// RemarkName is the identifier for the remark.  If \p I is passed it is an
/// instruction that prevents vectorization.  Otherwise \p TheLoop is used for
/// the location of the remark.  \return the remark object that can be
/// streamed to.
static OptimizationRemarkAnalysis createLVAnalysis(const char *PassName,
    StringRef RemarkName, Loop *TheLoop, Instruction *I) {
  Value *CodeRegion = TheLoop->getHeader();
  DebugLoc DL = TheLoop->getStartLoc();
 
  if (I) {
    CodeRegion = I->getParent();
    // If there is no debug location attached to the instruction, revert back to
    // using the loop's.
    if (I->getDebugLoc())
      DL = I->getDebugLoc();
  }
 
  return OptimizationRemarkAnalysis(PassName, RemarkName, DL, CodeRegion);
}
 
namespace llvm {
 
/// Return a value for Step multiplied by VF.
Value *createStepForVF(IRBuilderBase &B, Type *Ty, ElementCount VF,
                       int64_t Step) {
  assert(Ty->isIntegerTy() && "Expected an integer step");
  return B.CreateElementCount(Ty, VF.multiplyCoefficientBy(Step));
}
 
/// Return the runtime value for VF.
Value *getRuntimeVF(IRBuilderBase &B, Type *Ty, ElementCount VF) {
  return B.CreateElementCount(Ty, VF);
}
 
const SCEV *createTripCountSCEV(Type *IdxTy, PredicatedScalarEvolution &PSE,
                                Loop *OrigLoop) {
  const SCEV *BackedgeTakenCount = PSE.getBackedgeTakenCount();
  assert(!isa<SCEVCouldNotCompute>(BackedgeTakenCount) && "Invalid loop count");
 
  ScalarEvolution &SE = *PSE.getSE();
  return SE.getTripCountFromExitCount(BackedgeTakenCount, IdxTy, OrigLoop);
}
 
void reportVectorizationFailure(const StringRef DebugMsg,
                                const StringRef OREMsg, const StringRef ORETag,
                                OptimizationRemarkEmitter *ORE, Loop *TheLoop,
                                Instruction *I) {
  LLVM_DEBUG(debugVectorizationMessage("Not vectorizing: ", DebugMsg, I));
  LoopVectorizeHints Hints(TheLoop, true /* doesn't matter */, *ORE);
  ORE->emit(
      createLVAnalysis(Hints.vectorizeAnalysisPassName(), ORETag, TheLoop, I)
      << "loop not vectorized: " << OREMsg);
}
 
void reportVectorizationInfo(const StringRef Msg, const StringRef ORETag,
                             OptimizationRemarkEmitter *ORE, Loop *TheLoop,
                             Instruction *I) {
  LLVM_DEBUG(debugVectorizationMessage("", Msg, I));
  LoopVectorizeHints Hints(TheLoop, true /* doesn't matter */, *ORE);
  ORE->emit(
      createLVAnalysis(Hints.vectorizeAnalysisPassName(), ORETag, TheLoop, I)
      << Msg);
}
 
/// Report successful vectorization of the loop. In case an outer loop is
/// vectorized, prepend "outer" to the vectorization remark.
static void reportVectorization(OptimizationRemarkEmitter *ORE, Loop *TheLoop,
                                VectorizationFactor VF, unsigned IC) {
  LLVM_DEBUG(debugVectorizationMessage(
      "Vectorizing: ", TheLoop->isInnermost() ? "innermost loop" : "outer loop",
      nullptr));
  StringRef LoopType = TheLoop->isInnermost() ? "" : "outer ";
  ORE->emit([&]() {
    return OptimizationRemark(LV_NAME, "Vectorized", TheLoop->getStartLoc(),
                              TheLoop->getHeader())
           << "vectorized " << LoopType << "loop (vectorization width: "
           << ore::NV("VectorizationFactor", VF.Width)
           << ", interleaved count: " << ore::NV("InterleaveCount", IC) << ")";
  });
}
 
} // end namespace llvm
 
#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
 
void InnerLoopVectorizer::collectPoisonGeneratingRecipes(
    VPTransformState &State) {
 
  // Collect recipes in the backward slice of `Root` that may generate a poison
  // value that is used after vectorization.
  SmallPtrSet<VPRecipeBase *, 16> Visited;
  auto collectPoisonGeneratingInstrsInBackwardSlice([&](VPRecipeBase *Root) {
    SmallVector<VPRecipeBase *, 16> Worklist;
    Worklist.push_back(Root);
 
    // Traverse the backward slice of Root through its use-def chain.
    while (!Worklist.empty()) {
      VPRecipeBase *CurRec = Worklist.back();
      Worklist.pop_back();
 
      if (!Visited.insert(CurRec).second)
        continue;
 
      // Prune search if we find another recipe generating a widen memory
      // instruction. Widen memory instructions involved in address computation
      // will lead to gather/scatter instructions, which don't need to be
      // handled.
      if (isa<VPWidenMemoryInstructionRecipe>(CurRec) ||
          isa<VPInterleaveRecipe>(CurRec) ||
          isa<VPScalarIVStepsRecipe>(CurRec) ||
          isa<VPCanonicalIVPHIRecipe>(CurRec) ||
          isa<VPActiveLaneMaskPHIRecipe>(CurRec))
        continue;
 
      // This recipe contributes to the address computation of a widen
      // load/store. If the underlying instruction has poison-generating flags,
      // drop them directly.
      if (auto *RecWithFlags = dyn_cast<VPRecipeWithIRFlags>(CurRec)) {
        RecWithFlags->dropPoisonGeneratingFlags();
      } else {
        Instruction *Instr = CurRec->getUnderlyingInstr();
        (void)Instr;
        assert((!Instr || !Instr->hasPoisonGeneratingFlags()) &&
               "found instruction with poison generating flags not covered by "
               "VPRecipeWithIRFlags");
      }
 
      // Add new definitions to the worklist.
      for (VPValue *operand : CurRec->operands())
        if (VPRecipeBase *OpDef = operand->getDefiningRecipe())
          Worklist.push_back(OpDef);
    }
  });
 
  // Traverse all the recipes in the VPlan and collect the poison-generating
  // recipes in the backward slice starting at the address of a VPWidenRecipe or
  // VPInterleaveRecipe.
  auto Iter = vp_depth_first_deep(State.Plan->getEntry());
  for (VPBasicBlock *VPBB : VPBlockUtils::blocksOnly<VPBasicBlock>(Iter)) {
    for (VPRecipeBase &Recipe : *VPBB) {
      if (auto *WidenRec = dyn_cast<VPWidenMemoryInstructionRecipe>(&Recipe)) {
        Instruction &UnderlyingInstr = WidenRec->getIngredient();
        VPRecipeBase *AddrDef = WidenRec->getAddr()->getDefiningRecipe();
        if (AddrDef && WidenRec->isConsecutive() &&
            Legal->blockNeedsPredication(UnderlyingInstr.getParent()))
          collectPoisonGeneratingInstrsInBackwardSlice(AddrDef);
      } else if (auto *InterleaveRec = dyn_cast<VPInterleaveRecipe>(&Recipe)) {
        VPRecipeBase *AddrDef = InterleaveRec->getAddr()->getDefiningRecipe();
        if (AddrDef) {
          // Check if any member of the interleave group needs predication.
          const InterleaveGroup<Instruction> *InterGroup =
              InterleaveRec->getInterleaveGroup();
          bool NeedPredication = false;
          for (int I = 0, NumMembers = InterGroup->getNumMembers();
               I < NumMembers; ++I) {
            Instruction *Member = InterGroup->getMember(I);
            if (Member)
              NeedPredication |=
                  Legal->blockNeedsPredication(Member->getParent());
          }
 
          if (NeedPredication)
            collectPoisonGeneratingInstrsInBackwardSlice(AddrDef);
        }
      }
    }
  }
}
 
PHINode *InnerLoopVectorizer::getReductionResumeValue(
    const RecurrenceDescriptor &RdxDesc) {
  auto It = ReductionResumeValues.find(&RdxDesc);
  assert(It != ReductionResumeValues.end() &&
         "Expected to find a resume value for the reduction.");
  return It->second;
}
 
namespace llvm {
 
// Loop vectorization cost-model hints how the scalar epilogue loop should be
// lowered.
enum ScalarEpilogueLowering {
 
  // The default: allowing scalar epilogues.
  CM_ScalarEpilogueAllowed,
 
  // Vectorization with OptForSize: don't allow epilogues.
  CM_ScalarEpilogueNotAllowedOptSize,
 
  // A special case of vectorisation with OptForSize: loops with a very small
  // trip count are considered for vectorization under OptForSize, thereby
  // making sure the cost of their loop body is dominant, free of runtime
  // guards and scalar iteration overheads.
  CM_ScalarEpilogueNotAllowedLowTripLoop,
 
  // Loop hint predicate indicating an epilogue is undesired.
  CM_ScalarEpilogueNotNeededUsePredicate,
 
  // Directive indicating we must either tail fold or not vectorize
  CM_ScalarEpilogueNotAllowedUsePredicate
};
 
using InstructionVFPair = std::pair<Instruction *, ElementCount>;
 
/// LoopVectorizationCostModel - estimates the expected speedups due to
/// vectorization.
/// In many cases vectorization is not profitable. This can happen because of
/// a number of reasons. In this class we mainly attempt to predict the
/// expected speedup/slowdowns due to the supported instruction set. We use the
/// TargetTransformInfo to query the different backends for the cost of
/// different operations.
class LoopVectorizationCostModel {
public:
  LoopVectorizationCostModel(ScalarEpilogueLowering SEL, Loop *L,
                             PredicatedScalarEvolution &PSE, LoopInfo *LI,
                             LoopVectorizationLegality *Legal,
                             const TargetTransformInfo &TTI,
                             const TargetLibraryInfo *TLI, DemandedBits *DB,
                             AssumptionCache *AC,
                             OptimizationRemarkEmitter *ORE, const Function *F,
                             const LoopVectorizeHints *Hints,
                             InterleavedAccessInfo &IAI)
      : ScalarEpilogueStatus(SEL), TheLoop(L), PSE(PSE), LI(LI), Legal(Legal),
        TTI(TTI), TLI(TLI), DB(DB), AC(AC), ORE(ORE), TheFunction(F),
        Hints(Hints), InterleaveInfo(IAI) {}
 
  /// \return An upper bound for the vectorization factors (both fixed and
  /// scalable). If the factors are 0, vectorization and interleaving should be
  /// avoided up front.
  FixedScalableVFPair computeMaxVF(ElementCount UserVF, unsigned UserIC);
 
  /// \return True if runtime checks are required for vectorization, and false
  /// otherwise.
  bool runtimeChecksRequired();
 
  /// Setup cost-based decisions for user vectorization factor.
  /// \return true if the UserVF is a feasible VF to be chosen.
  bool selectUserVectorizationFactor(ElementCount UserVF) {
    collectUniformsAndScalars(UserVF);
    collectInstsToScalarize(UserVF);
    return expectedCost(UserVF).first.isValid();
  }
 
  /// \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.
  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(ElementCount VF, InstructionCost LoopCost);
 
  /// Memory access instruction may be vectorized in more than one way.
  /// Form of instruction after vectorization depends on cost.
  /// This function takes cost-based decisions for Load/Store instructions
  /// and collects them in a map. This decisions map is used for building
  /// the lists of loop-uniform and loop-scalar instructions.
  /// The calculated cost is saved with widening decision in order to
  /// avoid redundant calculations.
  void setCostBasedWideningDecision(ElementCount VF);
 
  /// A struct that represents some properties of the register usage
  /// of a loop.
  struct RegisterUsage {
    /// Holds the number of loop invariant values that are used in the loop.
    /// The key is ClassID of target-provided register class.
    SmallMapVector<unsigned, unsigned, 4> LoopInvariantRegs;
    /// Holds the maximum number of concurrent live intervals in the loop.
    /// The key is ClassID of target-provided register class.
    SmallMapVector<unsigned, unsigned, 4> MaxLocalUsers;
  };
 
  /// \return Returns information about the register usages of the loop for the
  /// given vectorization factors.
  SmallVector<RegisterUsage, 8>
  calculateRegisterUsage(ArrayRef<ElementCount> VFs);
 
  /// Collect values we want to ignore in the cost model.
  void collectValuesToIgnore();
 
  /// Collect all element types in the loop for which widening is needed.
  void collectElementTypesForWidening();
 
  /// Split reductions into those that happen in the loop, and those that happen
  /// outside. In loop reductions are collected into InLoopReductions.
  void collectInLoopReductions();
 
  /// Returns true if we should use strict in-order reductions for the given
  /// RdxDesc. This is true if the -enable-strict-reductions flag is passed,
  /// the IsOrdered flag of RdxDesc is set and we do not allow reordering
  /// of FP operations.
  bool useOrderedReductions(const RecurrenceDescriptor &RdxDesc) const {
    return !Hints->allowReordering() && RdxDesc.isOrdered();
  }
 
  /// \returns The smallest bitwidth each instruction can be represented with.
  /// The vector equivalents of these instructions should be truncated to this
  /// type.
  const MapVector<Instruction *, uint64_t> &getMinimalBitwidths() const {
    return MinBWs;
  }
 
  /// \returns True if it is more profitable to scalarize instruction \p I for
  /// vectorization factor \p VF.
  bool isProfitableToScalarize(Instruction *I, ElementCount VF) const {
    assert(VF.isVector() &&
           "Profitable to scalarize relevant only for VF > 1.");
 
    // Cost model is not run in the VPlan-native path - return conservative
    // result until this changes.
    if (EnableVPlanNativePath)
      return false;
 
    auto Scalars = InstsToScalarize.find(VF);
    assert(Scalars != InstsToScalarize.end() &&
           "VF not yet analyzed for scalarization profitability");
    return Scalars->second.contains(I);
  }
 
  /// Returns true if \p I is known to be uniform after vectorization.
  bool isUniformAfterVectorization(Instruction *I, ElementCount VF) const {
    // Pseudo probe needs to be duplicated for each unrolled iteration and
    // vector lane so that profiled loop trip count can be accurately
    // accumulated instead of being under counted.
    if (isa<PseudoProbeInst>(I))
      return false;
 
    if (VF.isScalar())
      return true;
 
    // Cost model is not run in the VPlan-native path - return conservative
    // result until this changes.
    if (EnableVPlanNativePath)
      return false;
 
    auto UniformsPerVF = Uniforms.find(VF);
    assert(UniformsPerVF != Uniforms.end() &&
           "VF not yet analyzed for uniformity");
    return UniformsPerVF->second.count(I);
  }
 
  /// Returns true if \p I is known to be scalar after vectorization.
  bool isScalarAfterVectorization(Instruction *I, ElementCount VF) const {
    if (VF.isScalar())
      return true;
 
    // Cost model is not run in the VPlan-native path - return conservative
    // result until this changes.
    if (EnableVPlanNativePath)
      return false;
 
    auto ScalarsPerVF = Scalars.find(VF);
    assert(ScalarsPerVF != Scalars.end() &&
           "Scalar values are not calculated for VF");
    return ScalarsPerVF->second.count(I);
  }
 
  /// \returns True if instruction \p I can be truncated to a smaller bitwidth
  /// for vectorization factor \p VF.
  bool canTruncateToMinimalBitwidth(Instruction *I, ElementCount VF) const {
    return VF.isVector() && MinBWs.contains(I) &&
           !isProfitableToScalarize(I, VF) &&
           !isScalarAfterVectorization(I, VF);
  }
 
  /// Decision that was taken during cost calculation for memory instruction.
  enum InstWidening {
    CM_Unknown,
    CM_Widen,         // For consecutive accesses with stride +1.
    CM_Widen_Reverse, // For consecutive accesses with stride -1.
    CM_Interleave,
    CM_GatherScatter,
    CM_Scalarize
  };
 
  /// Save vectorization decision \p W and \p Cost taken by the cost model for
  /// instruction \p I and vector width \p VF.
  void setWideningDecision(Instruction *I, ElementCount VF, InstWidening W,
                           InstructionCost Cost) {
    assert(VF.isVector() && "Expected VF >=2");
    WideningDecisions[std::make_pair(I, VF)] = std::make_pair(W, Cost);
  }
 
  /// Save vectorization decision \p W and \p Cost taken by the cost model for
  /// interleaving group \p Grp and vector width \p VF.
  void setWideningDecision(const InterleaveGroup<Instruction> *Grp,
                           ElementCount VF, InstWidening W,
                           InstructionCost Cost) {
    assert(VF.isVector() && "Expected VF >=2");
    /// Broadcast this decicion to all instructions inside the group.
    /// But the cost will be assigned to one instruction only.
    for (unsigned i = 0; i < Grp->getFactor(); ++i) {
      if (auto *I = Grp->getMember(i)) {
        if (Grp->getInsertPos() == I)
          WideningDecisions[std::make_pair(I, VF)] = std::make_pair(W, Cost);
        else
          WideningDecisions[std::make_pair(I, VF)] = std::make_pair(W, 0);
      }
    }
  }
 
  /// Return the cost model decision for the given instruction \p I and vector
  /// width \p VF. Return CM_Unknown if this instruction did not pass
  /// through the cost modeling.
  InstWidening getWideningDecision(Instruction *I, ElementCount VF) const {
    assert(VF.isVector() && "Expected VF to be a vector VF");
    // Cost model is not run in the VPlan-native path - return conservative
    // result until this changes.
    if (EnableVPlanNativePath)
      return CM_GatherScatter;
 
    std::pair<Instruction *, ElementCount> InstOnVF = std::make_pair(I, VF);
    auto Itr = WideningDecisions.find(InstOnVF);
    if (Itr == WideningDecisions.end())
      return CM_Unknown;
    return Itr->second.first;
  }
 
  /// Return the vectorization cost for the given instruction \p I and vector
  /// width \p VF.
  InstructionCost getWideningCost(Instruction *I, ElementCount VF) {
    assert(VF.isVector() && "Expected VF >=2");
    std::pair<Instruction *, ElementCount> InstOnVF = std::make_pair(I, VF);
    assert(WideningDecisions.contains(InstOnVF) &&
           "The cost is not calculated");
    return WideningDecisions[InstOnVF].second;
  }
 
  /// Return True if instruction \p I is an optimizable truncate whose operand
  /// is an induction variable. Such a truncate will be removed by adding a new
  /// induction variable with the destination type.
  bool isOptimizableIVTruncate(Instruction *I, ElementCount VF) {
    // If the instruction is not a truncate, return false.
    auto *Trunc = dyn_cast<TruncInst>(I);
    if (!Trunc)
      return false;
 
    // Get the source and destination types of the truncate.
    Type *SrcTy = ToVectorTy(cast<CastInst>(I)->getSrcTy(), VF);
    Type *DestTy = ToVectorTy(cast<CastInst>(I)->getDestTy(), VF);
 
    // If the truncate is free for the given types, return false. Replacing a
    // free truncate with an induction variable would add an induction variable
    // update instruction to each iteration of the loop. We exclude from this
    // check the primary induction variable since it will need an update
    // instruction regardless.
    Value *Op = Trunc->getOperand(0);
    if (Op != Legal->getPrimaryInduction() && TTI.isTruncateFree(SrcTy, DestTy))
      return false;
 
    // If the truncated value is not an induction variable, return false.
    return Legal->isInductionPhi(Op);
  }
 
  /// Collects the instructions to scalarize for each predicated instruction in
  /// the loop.
  void collectInstsToScalarize(ElementCount VF);
 
  /// Collect Uniform and Scalar values for the given \p VF.
  /// The sets depend on CM decision for Load/Store instructions
  /// that may be vectorized as interleave, gather-scatter or scalarized.
  void collectUniformsAndScalars(ElementCount VF) {
    // Do the analysis once.
    if (VF.isScalar() || Uniforms.contains(VF))
      return;
    setCostBasedWideningDecision(VF);
    collectLoopUniforms(VF);
    collectLoopScalars(VF);
  }
 
  /// 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, Align Alignment) const {
    return Legal->isConsecutivePtr(DataType, Ptr) &&
           TTI.isLegalMaskedStore(DataType, Alignment);
  }
 
  /// 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, Align Alignment) const {
    return Legal->isConsecutivePtr(DataType, Ptr) &&
           TTI.isLegalMaskedLoad(DataType, Alignment);
  }
 
  /// Returns true if the target machine can represent \p V as a masked gather
  /// or scatter operation.
  bool isLegalGatherOrScatter(Value *V, ElementCount VF) {
    bool LI = isa<LoadInst>(V);
    bool SI = isa<StoreInst>(V);
    if (!LI && !SI)
      return false;
    auto *Ty = getLoadStoreType(V);
    Align Align = getLoadStoreAlignment(V);
    if (VF.isVector())
      Ty = VectorType::get(Ty, VF);
    return (LI && TTI.isLegalMaskedGather(Ty, Align)) ||
           (SI && TTI.isLegalMaskedScatter(Ty, Align));
  }
 
  /// Returns true if the target machine supports all of the reduction
  /// variables found for the given VF.
  bool canVectorizeReductions(ElementCount VF) const {
    return (all_of(Legal->getReductionVars(), [&](auto &Reduction) -> bool {
      const RecurrenceDescriptor &RdxDesc = Reduction.second;
      return TTI.isLegalToVectorizeReduction(RdxDesc, VF);
    }));
  }
 
  /// Given costs for both strategies, return true if the scalar predication
  /// lowering should be used for div/rem.  This incorporates an override
  /// option so it is not simply a cost comparison.
  bool isDivRemScalarWithPredication(InstructionCost ScalarCost,
                                     InstructionCost SafeDivisorCost) const {
    switch (ForceSafeDivisor) {
    case cl::BOU_UNSET:
      return ScalarCost < SafeDivisorCost;
    case cl::BOU_TRUE:
      return false;
    case cl::BOU_FALSE:
      return true;
    };
    llvm_unreachable("impossible case value");
  }
 
  /// Returns true if \p I is an instruction which requires predication and
  /// for which our chosen predication strategy is scalarization (i.e. we
  /// don't have an alternate strategy such as masking available).
  /// \p VF is the vectorization factor that will be used to vectorize \p I.
  bool isScalarWithPredication(Instruction *I, ElementCount VF) const;
 
  /// Returns true if \p I is an instruction that needs to be predicated
  /// at runtime.  The result is independent of the predication mechanism.
  /// Superset of instructions that return true for isScalarWithPredication.
  bool isPredicatedInst(Instruction *I) const;
 
  /// Return the costs for our two available strategies for lowering a
  /// div/rem operation which requires speculating at least one lane.
  /// First result is for scalarization (will be invalid for scalable
  /// vectors); second is for the safe-divisor strategy.
  std::pair<InstructionCost, InstructionCost>
  getDivRemSpeculationCost(Instruction *I,
                           ElementCount VF) const;
 
  /// Returns true if \p I is a memory instruction with consecutive memory
  /// access that can be widened.
  bool memoryInstructionCanBeWidened(Instruction *I, ElementCount VF);
 
  /// Returns true if \p I is a memory instruction in an interleaved-group
  /// of memory accesses that can be vectorized with wide vector loads/stores
  /// and shuffles.
  bool interleavedAccessCanBeWidened(Instruction *I, ElementCount VF);
 
  /// Check if \p Instr belongs to any interleaved access group.
  bool isAccessInterleaved(Instruction *Instr) {
    return InterleaveInfo.isInterleaved(Instr);
  }
 
  /// Get the interleaved access group that \p Instr belongs to.
  const InterleaveGroup<Instruction> *
  getInterleavedAccessGroup(Instruction *Instr) {
    return InterleaveInfo.getInterleaveGroup(Instr);
  }
 
  /// Returns true if we're required to use a scalar epilogue for at least
  /// the final iteration of the original loop.
  bool requiresScalarEpilogue(bool IsVectorizing) const {
    if (!isScalarEpilogueAllowed())
      return false;
    // If we might exit from anywhere but the latch, must run the exiting
    // iteration in scalar form.
    if (TheLoop->getExitingBlock() != TheLoop->getLoopLatch())
      return true;
    return IsVectorizing && InterleaveInfo.requiresScalarEpilogue();
  }
 
  /// Returns true if we're required to use a scalar epilogue for at least
  /// the final iteration of the original loop for all VFs in \p Range.
  /// A scalar epilogue must either be required for all VFs in \p Range or for
  /// none.
  bool requiresScalarEpilogue(VFRange Range) const {
    auto RequiresScalarEpilogue = [this](ElementCount VF) {
      return requiresScalarEpilogue(VF.isVector());
    };
    bool IsRequired = all_of(Range, RequiresScalarEpilogue);
    assert(
        (IsRequired || none_of(Range, RequiresScalarEpilogue)) &&
        "all VFs in range must agree on whether a scalar epilogue is required");
    return IsRequired;
  }
 
  /// Returns true if a scalar epilogue is not allowed due to optsize or a
  /// loop hint annotation.
  bool isScalarEpilogueAllowed() const {
    return ScalarEpilogueStatus == CM_ScalarEpilogueAllowed;
  }
 
  /// Returns the TailFoldingStyle that is best for the current loop.
  TailFoldingStyle
  getTailFoldingStyle(bool IVUpdateMayOverflow = true) const {
    if (!CanFoldTailByMasking)
      return TailFoldingStyle::None;
 
    if (ForceTailFoldingStyle.getNumOccurrences())
      return ForceTailFoldingStyle;
 
    return TTI.getPreferredTailFoldingStyle(IVUpdateMayOverflow);
  }
 
  /// Returns true if all loop blocks should be masked to fold tail loop.
  bool foldTailByMasking() const {
    return getTailFoldingStyle() != TailFoldingStyle::None;
  }
 
  /// Returns true if the instructions in this block requires predication
  /// for any reason, e.g. because tail folding now requires a predicate
  /// or because the block in the original loop was predicated.
  bool blockNeedsPredicationForAnyReason(BasicBlock *BB) const {
    return foldTailByMasking() || Legal->blockNeedsPredication(BB);
  }
 
  /// Returns true if the Phi is part of an inloop reduction.
  bool isInLoopReduction(PHINode *Phi) const {
    return InLoopReductions.contains(Phi);
  }
 
  /// 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.
  InstructionCost getVectorIntrinsicCost(CallInst *CI, ElementCount VF) const;
 
  /// 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.
  InstructionCost getVectorCallCost(CallInst *CI, ElementCount VF,
                                    Function **Variant,
                                    bool *NeedsMask = nullptr) const;
 
  /// Invalidates decisions already taken by the cost model.
  void invalidateCostModelingDecisions() {
    WideningDecisions.clear();
    Uniforms.clear();
    Scalars.clear();
  }
 
  /// The vectorization cost is a combination of the cost itself and a boolean
  /// indicating whether any of the contributing operations will actually
  /// operate on vector values after type legalization in the backend. If this
  /// latter value is false, then all operations will be scalarized (i.e. no
  /// vectorization has actually taken place).
  using VectorizationCostTy = std::pair<InstructionCost, bool>;
 
  /// Returns the expected execution cost. The unit of the cost does
  /// not matter because we use the 'cost' units to compare different
  /// vector widths. The cost that is returned is *not* normalized by
  /// the factor width. If \p Invalid is not nullptr, this function
  /// will add a pair(Instruction*, ElementCount) to \p Invalid for
  /// each instruction that has an Invalid cost for the given VF.
  VectorizationCostTy
  expectedCost(ElementCount VF,
               SmallVectorImpl<InstructionVFPair> *Invalid = nullptr);
 
  bool hasPredStores() const { return NumPredStores > 0; }
 
  /// Returns true if epilogue vectorization is considered profitable, and
  /// false otherwise.
  /// \p VF is the vectorization factor chosen for the original loop.
  bool isEpilogueVectorizationProfitable(const ElementCount VF) const;
 
private:
  unsigned NumPredStores = 0;
 
  /// \return An upper bound for the vectorization factors for both
  /// fixed and scalable vectorization, where the minimum-known number of
  /// elements is a power-of-2 larger than zero. If scalable vectorization is
  /// disabled or unsupported, then the scalable part will be equal to
  /// ElementCount::getScalable(0).
  FixedScalableVFPair computeFeasibleMaxVF(unsigned ConstTripCount,
                                           ElementCount UserVF,
                                           bool FoldTailByMasking);
 
  /// \return the maximized element count based on the targets vector
  /// registers and the loop trip-count, but limited to a maximum safe VF.
  /// This is a helper function of computeFeasibleMaxVF.
  ElementCount getMaximizedVFForTarget(unsigned ConstTripCount,
                                       unsigned SmallestType,
                                       unsigned WidestType,
                                       ElementCount MaxSafeVF,
                                       bool FoldTailByMasking);
 
  /// \return the maximum legal scalable VF, based on the safe max number
  /// of elements.
  ElementCount getMaxLegalScalableVF(unsigned MaxSafeElements);
 
  /// Returns the execution time cost of an instruction for a given vector
  /// width. Vector width of one means scalar.
  VectorizationCostTy getInstructionCost(Instruction *I, ElementCount VF);
 
  /// The cost-computation logic from getInstructionCost which provides
  /// the vector type as an output parameter.
  InstructionCost getInstructionCost(Instruction *I, ElementCount VF,
                                     Type *&VectorTy);
 
  /// Return the cost of instructions in an inloop reduction pattern, if I is
  /// part of that pattern.
  std::optional<InstructionCost>
  getReductionPatternCost(Instruction *I, ElementCount VF, Type *VectorTy,
                          TTI::TargetCostKind CostKind);
 
  /// Calculate vectorization cost of memory instruction \p I.
  InstructionCost getMemoryInstructionCost(Instruction *I, ElementCount VF);
 
  /// The cost computation for scalarized memory instruction.
  InstructionCost getMemInstScalarizationCost(Instruction *I, ElementCount VF);
 
  /// The cost computation for interleaving group of memory instructions.
  InstructionCost getInterleaveGroupCost(Instruction *I, ElementCount VF);
 
  /// The cost computation for Gather/Scatter instruction.
  InstructionCost getGatherScatterCost(Instruction *I, ElementCount VF);
 
  /// The cost computation for widening instruction \p I with consecutive
  /// memory access.
  InstructionCost getConsecutiveMemOpCost(Instruction *I, ElementCount VF);
 
  /// The cost calculation for Load/Store instruction \p I with uniform pointer -
  /// Load: scalar load + broadcast.
  /// Store: scalar store + (loop invariant value stored? 0 : extract of last
  /// element)
  InstructionCost getUniformMemOpCost(Instruction *I, ElementCount VF);
 
  /// Estimate the overhead of scalarizing an instruction. This is a
  /// convenience wrapper for the type-based getScalarizationOverhead API.
  InstructionCost getScalarizationOverhead(Instruction *I, ElementCount VF,
                                           TTI::TargetCostKind CostKind) const;
 
  /// Returns true if an artificially high cost for emulated masked memrefs
  /// should be used.
  bool useEmulatedMaskMemRefHack(Instruction *I, ElementCount VF);
 
  /// 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;
 
  /// A type representing the costs for instructions if they were to be
  /// scalarized rather than vectorized. The entries are Instruction-Cost
  /// pairs.
  using ScalarCostsTy = DenseMap<Instruction *, InstructionCost>;
 
  /// A set containing all BasicBlocks that are known to present after
  /// vectorization as a predicated block.
  DenseMap<ElementCount, SmallPtrSet<BasicBlock *, 4>>
      PredicatedBBsAfterVectorization;
 
  /// Records whether it is allowed to have the original scalar loop execute at
  /// least once. This may be needed as a fallback loop in case runtime
  /// aliasing/dependence checks fail, or to handle the tail/remainder
  /// iterations when the trip count is unknown or doesn't divide by the VF,
  /// or as a peel-loop to handle gaps in interleave-groups.
  /// Under optsize and when the trip count is very small we don't allow any
  /// iterations to execute in the scalar loop.
  ScalarEpilogueLowering ScalarEpilogueStatus = CM_ScalarEpilogueAllowed;
 
  /// All blocks of loop are to be masked to fold tail of scalar iterations.
  bool CanFoldTailByMasking = false;
 
  /// A map holding scalar costs for different vectorization factors. The
  /// presence of a cost for an instruction in the mapping indicates that the
  /// instruction will be scalarized when vectorizing with the associated
  /// vectorization factor. The entries are VF-ScalarCostTy pairs.
  DenseMap<ElementCount, ScalarCostsTy> InstsToScalarize;
 
  /// Holds the instructions known to be uniform after vectorization.
  /// The data is collected per VF.
  DenseMap<ElementCount, SmallPtrSet<Instruction *, 4>> Uniforms;
 
  /// Holds the instructions known to be scalar after vectorization.
  /// The data is collected per VF.
  DenseMap<ElementCount, SmallPtrSet<Instruction *, 4>> Scalars;
 
  /// Holds the instructions (address computations) that are forced to be
  /// scalarized.
  DenseMap<ElementCount, SmallPtrSet<Instruction *, 4>> ForcedScalars;
 
  /// PHINodes of the reductions that should be expanded in-loop.
  SmallPtrSet<PHINode *, 4> InLoopReductions;
 
  /// A Map of inloop reduction operations and their immediate chain operand.
  /// FIXME: This can be removed once reductions can be costed correctly in
  /// VPlan. This was added to allow quick lookup of the inloop operations.
  DenseMap<Instruction *, Instruction *> InLoopReductionImmediateChains;
 
  /// Returns the expected difference in cost from scalarizing the expression
  /// feeding a predicated instruction \p PredInst. The instructions to
  /// scalarize and their scalar costs are collected in \p ScalarCosts. A
  /// non-negative return value implies the expression will be scalarized.
  /// Currently, only single-use chains are considered for scalarization.
  InstructionCost computePredInstDiscount(Instruction *PredInst,
                                          ScalarCostsTy &ScalarCosts,
                                          ElementCount VF);
 
  /// Collect the instructions that are uniform after vectorization. An
  /// instruction is uniform if we represent it with a single scalar value in
  /// the vectorized loop corresponding to each vector iteration. Examples of
  /// uniform instructions include pointer operands of consecutive or
  /// interleaved memory accesses. Note that although uniformity implies an
  /// instruction will be scalar, the reverse is not true. In general, a
  /// scalarized instruction will be represented by VF scalar values in the
  /// vectorized loop, each corresponding to an iteration of the original
  /// scalar loop.
  void collectLoopUniforms(ElementCount VF);
 
  /// Collect the instructions that are scalar after vectorization. An
  /// instruction is scalar if it is known to be uniform or will be scalarized
  /// during vectorization. collectLoopScalars should only add non-uniform nodes
  /// to the list if they are used by a load/store instruction that is marked as
  /// CM_Scalarize. Non-uniform scalarized instructions will be represented by
  /// VF values in the vectorized loop, each corresponding to an iteration of
  /// the original scalar loop.
  void collectLoopScalars(ElementCount VF);
 
  /// Keeps cost model vectorization decision and cost for instructions.
  /// Right now it is used for memory instructions only.
  using DecisionList = DenseMap<std::pair<Instruction *, ElementCount>,
                                std::pair<InstWidening, InstructionCost>>;
 
  DecisionList WideningDecisions;
 
  /// Returns true if \p V is expected to be vectorized and it needs to be
  /// extracted.
  bool needsExtract(Value *V, ElementCount VF) const {
    Instruction *I = dyn_cast<Instruction>(V);
    if (VF.isScalar() || !I || !TheLoop->contains(I) ||
        TheLoop->isLoopInvariant(I))
      return false;
 
    // Assume we can vectorize V (and hence we need extraction) if the
    // scalars are not computed yet. This can happen, because it is called
    // via getScalarizationOverhead from setCostBasedWideningDecision, before
    // the scalars are collected. That should be a safe assumption in most
    // cases, because we check if the operands have vectorizable types
    // beforehand in LoopVectorizationLegality.
    return !Scalars.contains(VF) || !isScalarAfterVectorization(I, VF);
  };
 
  /// Returns a range containing only operands needing to be extracted.
  SmallVector<Value *, 4> filterExtractingOperands(Instruction::op_range Ops,
                                                   ElementCount VF) const {
    return SmallVector<Value *, 4>(make_filter_range(
        Ops, [this, VF](Value *V) { return this->needsExtract(V, VF); }));
  }
 
public:
  /// The loop that we evaluate.
  Loop *TheLoop;
 
  /// Predicated scalar evolution analysis.
  PredicatedScalarEvolution &PSE;
 
  /// Loop Info analysis.
  LoopInfo *LI;
 
  /// Vectorization legality.
  LoopVectorizationLegality *Legal;
 
  /// Vector target information.
  const TargetTransformInfo &TTI;
 
  /// Target Library Info.
  const TargetLibraryInfo *TLI;
 
  /// Demanded bits analysis.
  DemandedBits *DB;
 
  /// Assumption cache.
  AssumptionCache *AC;
 
  /// Interface to emit optimization remarks.
  OptimizationRemarkEmitter *ORE;
 
  const Function *TheFunction;
 
  /// Loop Vectorize Hint.
  const LoopVectorizeHints *Hints;
 
  /// The interleave access information contains groups of interleaved accesses
  /// with the same stride and close to each other.
  InterleavedAccessInfo &InterleaveInfo;
 
  /// Values to ignore in the cost model.
  SmallPtrSet<const Value *, 16> ValuesToIgnore;
 
  /// Values to ignore in the cost model when VF > 1.
  SmallPtrSet<const Value *, 16> VecValuesToIgnore;
 
  /// All element types found in the loop.
  SmallPtrSet<Type *, 16> ElementTypesInLoop;
};
} // end namespace llvm
 
namespace {
/// Helper struct to manage generating runtime checks for vectorization.
///
/// The runtime checks are created up-front in temporary blocks to allow better
/// estimating the cost and un-linked from the existing IR. After deciding to
/// vectorize, the checks are moved back. If deciding not to vectorize, the
/// temporary blocks are completely removed.
class GeneratedRTChecks {
  /// Basic block which contains the generated SCEV checks, if any.
  BasicBlock *SCEVCheckBlock = nullptr;
 
  /// The value representing the result of the generated SCEV checks. If it is
  /// nullptr, either no SCEV checks have been generated or they have been used.
  Value *SCEVCheckCond = nullptr;
 
  /// Basic block which contains the generated memory runtime checks, if any.
  BasicBlock *MemCheckBlock = nullptr;
 
  /// The value representing the result of the generated memory runtime checks.
  /// If it is nullptr, either no memory runtime checks have been generated or
  /// they have been used.
  Value *MemRuntimeCheckCond = nullptr;
 
  DominatorTree *DT;
  LoopInfo *LI;
  TargetTransformInfo *TTI;
 
  SCEVExpander SCEVExp;
  SCEVExpander MemCheckExp;
 
  bool CostTooHigh = false;
 
public:
  GeneratedRTChecks(ScalarEvolution &SE, DominatorTree *DT, LoopInfo *LI,
                    TargetTransformInfo *TTI, const DataLayout &DL)
      : DT(DT), LI(LI), TTI(TTI), SCEVExp(SE, DL, "scev.check"),
        MemCheckExp(SE, DL, "scev.check") {}
 
  /// Generate runtime checks in SCEVCheckBlock and MemCheckBlock, so we can
  /// accurately estimate the cost of the runtime checks. The blocks are
  /// un-linked from the IR and is added back during vector code generation. If
  /// there is no vector code generation, the check blocks are removed
  /// completely.
  void Create(Loop *L, const LoopAccessInfo &LAI,
              const SCEVPredicate &UnionPred, ElementCount VF, unsigned IC) {
 
    // Hard cutoff to limit compile-time increase in case a very large number of
    // runtime checks needs to be generated.
    // TODO: Skip cutoff if the loop is guaranteed to execute, e.g. due to
    // profile info.
    CostTooHigh =
        LAI.getNumRuntimePointerChecks() > VectorizeMemoryCheckThreshold;
    if (CostTooHigh)
      return;
 
    BasicBlock *LoopHeader = L->getHeader();
    BasicBlock *Preheader = L->getLoopPreheader();
 
    // Use SplitBlock to create blocks for SCEV & memory runtime checks to
    // ensure the blocks are properly added to LoopInfo & DominatorTree. Those
    // may be used by SCEVExpander. The blocks will be un-linked from their
    // predecessors and removed from LI & DT at the end of the function.
    if (!UnionPred.isAlwaysTrue()) {
      SCEVCheckBlock = SplitBlock(Preheader, Preheader->getTerminator(), DT, LI,
                                  nullptr, "vector.scevcheck");
 
      SCEVCheckCond = SCEVExp.expandCodeForPredicate(
          &UnionPred, SCEVCheckBlock->getTerminator());
    }
 
    const auto &RtPtrChecking = *LAI.getRuntimePointerChecking();
    if (RtPtrChecking.Need) {
      auto *Pred = SCEVCheckBlock ? SCEVCheckBlock : Preheader;
      MemCheckBlock = SplitBlock(Pred, Pred->getTerminator(), DT, LI, nullptr,
                                 "vector.memcheck");
 
      auto DiffChecks = RtPtrChecking.getDiffChecks();
      if (DiffChecks) {
        Value *RuntimeVF = nullptr;
        MemRuntimeCheckCond = addDiffRuntimeChecks(
            MemCheckBlock->getTerminator(), *DiffChecks, MemCheckExp,
            [VF, &RuntimeVF](IRBuilderBase &B, unsigned Bits) {
              if (!RuntimeVF)
                RuntimeVF = getRuntimeVF(B, B.getIntNTy(Bits), VF);
              return RuntimeVF;
            },
            IC);
      } else {
        MemRuntimeCheckCond = addRuntimeChecks(
            MemCheckBlock->getTerminator(), L, RtPtrChecking.getChecks(),
            MemCheckExp, VectorizerParams::HoistRuntimeChecks);
      }
      assert(MemRuntimeCheckCond &&
             "no RT checks generated although RtPtrChecking "
             "claimed checks are required");
    }
 
    if (!MemCheckBlock && !SCEVCheckBlock)
      return;
 
    // Unhook the temporary block with the checks, update various places
    // accordingly.
    if (SCEVCheckBlock)
      SCEVCheckBlock->replaceAllUsesWith(Preheader);
    if (MemCheckBlock)
      MemCheckBlock->replaceAllUsesWith(Preheader);
 
    if (SCEVCheckBlock) {
      SCEVCheckBlock->getTerminator()->moveBefore(Preheader->getTerminator());
      new UnreachableInst(Preheader->getContext(), SCEVCheckBlock);
      Preheader->getTerminator()->eraseFromParent();
    }
    if (MemCheckBlock) {
      MemCheckBlock->getTerminator()->moveBefore(Preheader->getTerminator());
      new UnreachableInst(Preheader->getContext(), MemCheckBlock);
      Preheader->getTerminator()->eraseFromParent();
    }
 
    DT->changeImmediateDominator(LoopHeader, Preheader);
    if (MemCheckBlock) {
      DT->eraseNode(MemCheckBlock);
      LI->removeBlock(MemCheckBlock);
    }
    if (SCEVCheckBlock) {
      DT->eraseNode(SCEVCheckBlock);
      LI->removeBlock(SCEVCheckBlock);
    }
  }
 
  InstructionCost getCost() {
    if (SCEVCheckBlock || MemCheckBlock)
      LLVM_DEBUG(dbgs() << "Calculating cost of runtime checks:\n");
 
    if (CostTooHigh) {
      InstructionCost Cost;
      Cost.setInvalid();
      LLVM_DEBUG(dbgs() << "  number of checks exceeded threshold\n");
      return Cost;
    }
 
    InstructionCost RTCheckCost = 0;
    if (SCEVCheckBlock)
      for (Instruction &I : *SCEVCheckBlock) {
        if (SCEVCheckBlock->getTerminator() == &I)
          continue;
        InstructionCost C =
            TTI->getInstructionCost(&I, TTI::TCK_RecipThroughput);
        LLVM_DEBUG(dbgs() << "  " << C << "  for " << I << "\n");
        RTCheckCost += C;
      }
    if (MemCheckBlock)
      for (Instruction &I : *MemCheckBlock) {
        if (MemCheckBlock->getTerminator() == &I)
          continue;
        InstructionCost C =
            TTI->getInstructionCost(&I, TTI::TCK_RecipThroughput);
        LLVM_DEBUG(dbgs() << "  " << C << "  for " << I << "\n");
        RTCheckCost += C;
      }
 
    if (SCEVCheckBlock || MemCheckBlock)
      LLVM_DEBUG(dbgs() << "Total cost of runtime checks: " << RTCheckCost
                        << "\n");
 
    return RTCheckCost;
  }
 
  /// Remove the created SCEV & memory runtime check blocks & instructions, if
  /// unused.
  ~GeneratedRTChecks() {
    SCEVExpanderCleaner SCEVCleaner(SCEVExp);
    SCEVExpanderCleaner MemCheckCleaner(MemCheckExp);
    if (!SCEVCheckCond)
      SCEVCleaner.markResultUsed();
 
    if (!MemRuntimeCheckCond)
      MemCheckCleaner.markResultUsed();
 
    if (MemRuntimeCheckCond) {
      auto &SE = *MemCheckExp.getSE();
      // Memory runtime check generation creates compares that use expanded
      // values. Remove them before running the SCEVExpanderCleaners.
      for (auto &I : make_early_inc_range(reverse(*MemCheckBlock))) {
        if (MemCheckExp.isInsertedInstruction(&I))
          continue;
        SE.forgetValue(&I);
        I.eraseFromParent();
      }
    }
    MemCheckCleaner.cleanup();
    SCEVCleaner.cleanup();
 
    if (SCEVCheckCond)
      SCEVCheckBlock->eraseFromParent();
    if (MemRuntimeCheckCond)
      MemCheckBlock->eraseFromParent();
  }
 
  /// Adds the generated SCEVCheckBlock before \p LoopVectorPreHeader and
  /// adjusts the branches to branch to the vector preheader or \p Bypass,
  /// depending on the generated condition.
  BasicBlock *emitSCEVChecks(BasicBlock *Bypass,
                             BasicBlock *LoopVectorPreHeader,
                             BasicBlock *LoopExitBlock) {
    if (!SCEVCheckCond)
      return nullptr;
 
    Value *Cond = SCEVCheckCond;
    // Mark the check as used, to prevent it from being removed during cleanup.
    SCEVCheckCond = nullptr;
    if (auto *C = dyn_cast<ConstantInt>(Cond))
      if (C->isZero())
        return nullptr;
 
    auto *Pred = LoopVectorPreHeader->getSinglePredecessor();
 
    BranchInst::Create(LoopVectorPreHeader, SCEVCheckBlock);
    // Create new preheader for vector loop.
    if (auto *PL = LI->getLoopFor(LoopVectorPreHeader))
      PL->addBasicBlockToLoop(SCEVCheckBlock, *LI);
 
    SCEVCheckBlock->getTerminator()->eraseFromParent();
    SCEVCheckBlock->moveBefore(LoopVectorPreHeader);
    Pred->getTerminator()->replaceSuccessorWith(LoopVectorPreHeader,
                                                SCEVCheckBlock);
 
    DT->addNewBlock(SCEVCheckBlock, Pred);
    DT->changeImmediateDominator(LoopVectorPreHeader, SCEVCheckBlock);
 
    ReplaceInstWithInst(SCEVCheckBlock->getTerminator(),
                        BranchInst::Create(Bypass, LoopVectorPreHeader, Cond));
    return SCEVCheckBlock;
  }
 
  /// Adds the generated MemCheckBlock before \p LoopVectorPreHeader and adjusts
  /// the branches to branch to the vector preheader or \p Bypass, depending on
  /// the generated condition.
  BasicBlock *emitMemRuntimeChecks(BasicBlock *Bypass,
                                   BasicBlock *LoopVectorPreHeader) {
    // Check if we generated code that checks in runtime if arrays overlap.
    if (!MemRuntimeCheckCond)
      return nullptr;
 
    auto *Pred = LoopVectorPreHeader->getSinglePredecessor();
    Pred->getTerminator()->replaceSuccessorWith(LoopVectorPreHeader,
                                                MemCheckBlock);
 
    DT->addNewBlock(MemCheckBlock, Pred);
    DT->changeImmediateDominator(LoopVectorPreHeader, MemCheckBlock);
    MemCheckBlock->moveBefore(LoopVectorPreHeader);
 
    if (auto *PL = LI->getLoopFor(LoopVectorPreHeader))
      PL->addBasicBlockToLoop(MemCheckBlock, *LI);
 
    ReplaceInstWithInst(
        MemCheckBlock->getTerminator(),
        BranchInst::Create(Bypass, LoopVectorPreHeader, MemRuntimeCheckCond));
    MemCheckBlock->getTerminator()->setDebugLoc(
        Pred->getTerminator()->getDebugLoc());
 
    // Mark the check as used, to prevent it from being removed during cleanup.
    MemRuntimeCheckCond = nullptr;
    return MemCheckBlock;
  }
};
} // namespace
 
static bool useActiveLaneMask(TailFoldingStyle Style) {
  return Style == TailFoldingStyle::Data ||
         Style == TailFoldingStyle::DataAndControlFlow ||
         Style == TailFoldingStyle::DataAndControlFlowWithoutRuntimeCheck;
}
 
static bool useActiveLaneMaskForControlFlow(TailFoldingStyle Style) {
  return Style == TailFoldingStyle::DataAndControlFlow ||
         Style == TailFoldingStyle::DataAndControlFlowWithoutRuntimeCheck;
}
 
// Return true if \p OuterLp is an outer loop annotated with hints for explicit
// vectorization. The loop needs to be annotated with #pragma omp simd
// simdlen(#) or #pragma clang vectorize(enable) vectorize_width(#). If the
// vector length information is not provided, vectorization is not considered
// explicit. Interleave hints are not allowed either. These limitations will be
// relaxed in the future.
// Please, note that we are currently forced to abuse the pragma 'clang
// vectorize' semantics. This pragma provides *auto-vectorization hints*
// (i.e., LV must check that vectorization is legal) whereas pragma 'omp simd'
// provides *explicit vectorization hints* (LV can bypass legal checks and
// assume that vectorization is legal). However, both hints are implemented
// using the same metadata (llvm.loop.vectorize, processed by
// LoopVectorizeHints). This will be fixed in the future when the native IR
// representation for pragma 'omp simd' is introduced.
static bool isExplicitVecOuterLoop(Loop *OuterLp,
                                   OptimizationRemarkEmitter *ORE) {
  assert(!OuterLp->isInnermost() && "This is not an outer loop");
  LoopVectorizeHints Hints(OuterLp, true /*DisableInterleaving*/, *ORE);
 
  // Only outer loops with an explicit vectorization hint are supported.
  // Unannotated outer loops are ignored.
  if (Hints.getForce() == LoopVectorizeHints::FK_Undefined)
    return false;
 
  Function *Fn = OuterLp->getHeader()->getParent();
  if (!Hints.allowVectorization(Fn, OuterLp,
                                true /*VectorizeOnlyWhenForced*/)) {
    LLVM_DEBUG(dbgs() << "LV: Loop hints prevent outer loop vectorization.\n");
    return false;
  }
 
  if (Hints.getInterleave() > 1) {
    // TODO: Interleave support is future work.
    LLVM_DEBUG(dbgs() << "LV: Not vectorizing: Interleave is not supported for "
                         "outer loops.\n");
    Hints.emitRemarkWithHints();
    return false;
  }
 
  return true;
}
 
static void collectSupportedLoops(Loop &L, LoopInfo *LI,
                                  OptimizationRemarkEmitter *ORE,
                                  SmallVectorImpl<Loop *> &V) {
  // Collect inner loops and outer loops without irreducible control flow. For
  // now, only collect outer loops that have explicit vectorization hints. If we
  // are stress testing the VPlan H-CFG construction, we collect the outermost
  // loop of every loop nest.
  if (L.isInnermost() || VPlanBuildStressTest ||
      (EnableVPlanNativePath && isExplicitVecOuterLoop(&L, ORE))) {
    LoopBlocksRPO RPOT(&L);
    RPOT.perform(LI);
    if (!containsIrreducibleCFG<const BasicBlock *>(RPOT, *LI)) {
      V.push_back(&L);
      // TODO: Collect inner loops inside marked outer loops in case
      // vectorization fails for the outer loop. Do not invoke
      // 'containsIrreducibleCFG' again for inner loops when the outer loop is
      // already known to be reducible. We can use an inherited attribute for
      // that.
      return;
    }
  }
  for (Loop *InnerL : L)
    collectSupportedLoops(*InnerL, LI, ORE, V);
}
 
//===----------------------------------------------------------------------===//
// Implementation of LoopVectorizationLegality, InnerLoopVectorizer and
// LoopVectorizationCostModel and LoopVectorizationPlanner.
//===----------------------------------------------------------------------===//
 
/// Compute the transformed value of Index at offset StartValue using step
/// StepValue.
/// For integer induction, returns StartValue + Index * StepValue.
/// For pointer induction, returns StartValue[Index * StepValue].
/// FIXME: The newly created binary instructions should contain nsw/nuw
/// flags, which can be found from the original scalar operations.
static Value *
emitTransformedIndex(IRBuilderBase &B, Value *Index, Value *StartValue,
                     Value *Step,
                     InductionDescriptor::InductionKind InductionKind,
                     const BinaryOperator *InductionBinOp) {
  Type *StepTy = Step->getType();
  Value *CastedIndex = StepTy->isIntegerTy()
                           ? B.CreateSExtOrTrunc(Index, StepTy)
                           : B.CreateCast(Instruction::SIToFP, Index, StepTy);
  if (CastedIndex != Index) {
    CastedIndex->setName(CastedIndex->getName() + ".cast");
    Index = CastedIndex;
  }
 
  // Note: the IR at this point is broken. We cannot use SE to create any new
  // SCEV and then expand it, hoping that SCEV's simplification will give us
  // a more optimal code. Unfortunately, attempt of doing so on invalid IR may
  // lead to various SCEV crashes. So all we can do is to use builder and rely
  // on InstCombine for future simplifications. Here we handle some trivial
  // cases only.
  auto CreateAdd = [&B](Value *X, Value *Y) {
    assert(X->getType() == Y->getType() && "Types don't match!");
    if (auto *CX = dyn_cast<ConstantInt>(X))
      if (CX->isZero())
        return Y;
    if (auto *CY = dyn_cast<ConstantInt>(Y))
      if (CY->isZero())
        return X;
    return B.CreateAdd(X, Y);
  };
 
  // We allow X to be a vector type, in which case Y will potentially be
  // splatted into a vector with the same element count.
  auto CreateMul = [&B](Value *X, Value *Y) {
    assert(X->getType()->getScalarType() == Y->getType() &&
           "Types don't match!");
    if (auto *CX = dyn_cast<ConstantInt>(X))
      if (CX->isOne())
        return Y;
    if (auto *CY = dyn_cast<ConstantInt>(Y))
      if (CY->isOne())
        return X;
    VectorType *XVTy = dyn_cast<VectorType>(X->getType());
    if (XVTy && !isa<VectorType>(Y->getType()))
      Y = B.CreateVectorSplat(XVTy->getElementCount(), Y);
    return B.CreateMul(X, Y);
  };
 
  switch (InductionKind) {
  case InductionDescriptor::IK_IntInduction: {
    assert(!isa<VectorType>(Index->getType()) &&
           "Vector indices not supported for integer inductions yet");
    assert(Index->getType() == StartValue->getType() &&
           "Index type does not match StartValue type");
    if (isa<ConstantInt>(Step) && cast<ConstantInt>(Step)->isMinusOne())
      return B.CreateSub(StartValue, Index);
    auto *Offset = CreateMul(Index, Step);
    return CreateAdd(StartValue, Offset);
  }
  case InductionDescriptor::IK_PtrInduction: {
    return B.CreateGEP(B.getInt8Ty(), StartValue, CreateMul(Index, Step));
  }
  case InductionDescriptor::IK_FpInduction: {
    assert(!isa<VectorType>(Index->getType()) &&
           "Vector indices not supported for FP inductions yet");
    assert(Step->getType()->isFloatingPointTy() && "Expected FP Step value");
    assert(InductionBinOp &&
           (InductionBinOp->getOpcode() == Instruction::FAdd ||
            InductionBinOp->getOpcode() == Instruction::FSub) &&
           "Original bin op should be defined for FP induction");
 
    Value *MulExp = B.CreateFMul(Step, Index);
    return B.CreateBinOp(InductionBinOp->getOpcode(), StartValue, MulExp,
                         "induction");
  }
  case InductionDescriptor::IK_NoInduction:
    return nullptr;
  }
  llvm_unreachable("invalid enum");
}
 
std::optional<unsigned> getMaxVScale(const Function &F,
                                     const TargetTransformInfo &TTI) {
  if (std::optional<unsigned> MaxVScale = TTI.getMaxVScale())
    return MaxVScale;
 
  if (F.hasFnAttribute(Attribute::VScaleRange))
    return F.getFnAttribute(Attribute::VScaleRange).getVScaleRangeMax();
 
  return std::nullopt;
}
 
/// For the given VF and UF and maximum trip count computed for the loop, return
/// whether the induction variable might overflow in the vectorized loop. If not,
/// then we know a runtime overflow check always evaluates to false and can be
/// removed.
static bool isIndvarOverflowCheckKnownFalse(
    const LoopVectorizationCostModel *Cost,
    ElementCount VF, std::optional<unsigned> UF = std::nullopt) {
  // Always be conservative if we don't know the exact unroll factor.
  unsigned MaxUF = UF ? *UF : Cost->TTI.getMaxInterleaveFactor(VF);
 
  Type *IdxTy = Cost->Legal->getWidestInductionType();
  APInt MaxUIntTripCount = cast<IntegerType>(IdxTy)->getMask();
 
  // We know the runtime overflow check is known false iff the (max) trip-count
  // is known and (max) trip-count + (VF * UF) does not overflow in the type of
  // the vector loop induction variable.
  if (unsigned TC =
          Cost->PSE.getSE()->getSmallConstantMaxTripCount(Cost->TheLoop)) {
    uint64_t MaxVF = VF.getKnownMinValue();
    if (VF.isScalable()) {
      std::optional<unsigned> MaxVScale =
          getMaxVScale(*Cost->TheFunction, Cost->TTI);
      if (!MaxVScale)
        return false;
      MaxVF *= *MaxVScale;
    }
 
    return (MaxUIntTripCount - TC).ugt(MaxVF * MaxUF);
  }
 
  return false;
}
 
// Return whether we allow using masked interleave-groups (for dealing with
// strided loads/stores that reside in predicated blocks, or for dealing
// with gaps).
static bool useMaskedInterleavedAccesses(const TargetTransformInfo &TTI) {
  // If an override option has been passed in for interleaved accesses, use it.
  if (EnableMaskedInterleavedMemAccesses.getNumOccurrences() > 0)
    return EnableMaskedInterleavedMemAccesses;
 
  return TTI.enableMaskedInterleavedAccessVectorization();
}
 
// 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, poison, <0, 3, 6, 9>   ; R elements
//   %G.vec = shuffle %wide.vec, poison, <1, 4, 7, 10>  ; G elements
//   %B.vec = shuffle %wide.vec, poison, <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, poison, <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(
    const InterleaveGroup<Instruction> *Group, ArrayRef<VPValue *> VPDefs,
    VPTransformState &State, VPValue *Addr, ArrayRef<VPValue *> StoredValues,
    VPValue *BlockInMask, bool NeedsMaskForGaps) {
  Instruction *Instr = Group->getInsertPos();
  const DataLayout &DL = Instr->getModule()->getDataLayout();
 
  // Prepare for the vector type of the interleaved load/store.
  Type *ScalarTy = getLoadStoreType(Instr);
  unsigned InterleaveFactor = Group->getFactor();
  auto *VecTy = VectorType::get(ScalarTy, VF * InterleaveFactor);
 
  // Prepare for the new pointers.
  SmallVector<Value *, 2> AddrParts;
  unsigned Index = Group->getIndex(Instr);
 
  // TODO: extend the masked interleaved-group support to reversed access.
  assert((!BlockInMask || !Group->isReverse()) &&
         "Reversed masked interleave-group not supported.");
 
  Value *Idx;
  // If the group is reverse, adjust the index to refer to the last vector lane
  // instead of the first. We adjust the index from the first vector lane,
  // rather than directly getting the pointer for lane VF - 1, because the
  // pointer operand of the interleaved access is supposed to be uniform. For
  // uniform instructions, we're only required to generate a value for the
  // first vector lane in each unroll iteration.
  if (Group->isReverse()) {
    Value *RuntimeVF = getRuntimeVF(Builder, Builder.getInt32Ty(), VF);
    Idx = Builder.CreateSub(RuntimeVF, Builder.getInt32(1));
    Idx = Builder.CreateMul(Idx, Builder.getInt32(Group->getFactor()));
    Idx = Builder.CreateAdd(Idx, Builder.getInt32(Index));
    Idx = Builder.CreateNeg(Idx);
  } else
    Idx = Builder.getInt32(-Index);
 
  for (unsigned Part = 0; Part < UF; Part++) {
    Value *AddrPart = State.get(Addr, VPIteration(Part, 0));
    if (auto *I = dyn_cast<Instruction>(AddrPart))
      State.setDebugLocFrom(I->getDebugLoc());
 
    // 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].
 
    bool InBounds = false;
    if (auto *gep = dyn_cast<GetElementPtrInst>(AddrPart->stripPointerCasts()))
      InBounds = gep->isInBounds();
    AddrPart = Builder.CreateGEP(ScalarTy, AddrPart, Idx, "", InBounds);
    AddrParts.push_back(AddrPart);
  }
 
  State.setDebugLocFrom(Instr->getDebugLoc());
  Value *PoisonVec = PoisonValue::get(VecTy);
 
  auto CreateGroupMask = [this, &BlockInMask, &State, &InterleaveFactor](
                             unsigned Part, Value *MaskForGaps) -> Value * {
    if (VF.isScalable()) {
      assert(!MaskForGaps && "Interleaved groups with gaps are not supported.");
      assert(InterleaveFactor == 2 &&
             "Unsupported deinterleave factor for scalable vectors");
      auto *BlockInMaskPart = State.get(BlockInMask, Part);
      SmallVector<Value *, 2> Ops = {BlockInMaskPart, BlockInMaskPart};
      auto *MaskTy =
          VectorType::get(Builder.getInt1Ty(), VF.getKnownMinValue() * 2, true);
      return Builder.CreateIntrinsic(
          MaskTy, Intrinsic::experimental_vector_interleave2, Ops,
          /*FMFSource=*/nullptr, "interleaved.mask");
    }
 
    if (!BlockInMask)
      return MaskForGaps;
 
    Value *BlockInMaskPart = State.get(BlockInMask, Part);
    Value *ShuffledMask = Builder.CreateShuffleVector(
        BlockInMaskPart,
        createReplicatedMask(InterleaveFactor, VF.getKnownMinValue()),
        "interleaved.mask");
    return MaskForGaps ? Builder.CreateBinOp(Instruction::And, ShuffledMask,
                                             MaskForGaps)
                       : ShuffledMask;
  };
 
  // Vectorize the interleaved load group.
  if (isa<LoadInst>(Instr)) {
    Value *MaskForGaps = nullptr;
    if (NeedsMaskForGaps) {
      MaskForGaps =
          createBitMaskForGaps(Builder, VF.getKnownMinValue(), *Group);
      assert(MaskForGaps && "Mask for Gaps is required but it is null");
    }
 
    // For each unroll part, create a wide load for the group.
    SmallVector<Value *, 2> NewLoads;
    for (unsigned Part = 0; Part < UF; Part++) {
      Instruction *NewLoad;
      if (BlockInMask || MaskForGaps) {
        assert(useMaskedInterleavedAccesses(*TTI) &&
               "masked interleaved groups are not allowed.");
        Value *GroupMask = CreateGroupMask(Part, MaskForGaps);
        NewLoad =
            Builder.CreateMaskedLoad(VecTy, AddrParts[Part], Group->getAlign(),
                                     GroupMask, PoisonVec, "wide.masked.vec");
      }
      else
        NewLoad = Builder.CreateAlignedLoad(VecTy, AddrParts[Part],
                                            Group->getAlign(), "wide.vec");
      Group->addMetadata(NewLoad);
      NewLoads.push_back(NewLoad);
    }
 
    if (VecTy->isScalableTy()) {
      assert(InterleaveFactor == 2 &&
             "Unsupported deinterleave factor for scalable vectors");
 
      for (unsigned Part = 0; Part < UF; ++Part) {
        // Scalable vectors cannot use arbitrary shufflevectors (only splats),
        // so must use intrinsics to deinterleave.
        Value *DI = Builder.CreateIntrinsic(
            Intrinsic::experimental_vector_deinterleave2, VecTy, NewLoads[Part],
            /*FMFSource=*/nullptr, "strided.vec");
        unsigned J = 0;
        for (unsigned I = 0; I < InterleaveFactor; ++I) {
          Instruction *Member = Group->getMember(I);
 
          if (!Member)
            continue;
 
          Value *StridedVec = Builder.CreateExtractValue(DI, I);
          // If this member has different type, cast the result type.
          if (Member->getType() != ScalarTy) {
            VectorType *OtherVTy = VectorType::get(Member->getType(), VF);
            StridedVec = createBitOrPointerCast(StridedVec, OtherVTy, DL);
          }
 
          if (Group->isReverse())
            StridedVec = Builder.CreateVectorReverse(StridedVec, "reverse");
 
          State.set(VPDefs[J], StridedVec, Part);
          ++J;
        }
      }
 
      return;
    }
 
    // For each member in the group, shuffle out the appropriate data from the
    // wide loads.
    unsigned J = 0;
    for (unsigned I = 0; I < InterleaveFactor; ++I) {
      Instruction *Member = Group->getMember(I);
 
      // Skip the gaps in the group.
      if (!Member)
        continue;
 
      auto StrideMask =
          createStrideMask(I, InterleaveFactor, VF.getKnownMinValue());
      for (unsigned Part = 0; Part < UF; Part++) {
        Value *StridedVec = Builder.CreateShuffleVector(
            NewLoads[Part], StrideMask, "strided.vec");
 
        // If this member has different type, cast the result type.
        if (Member->getType() != ScalarTy) {
          assert(!VF.isScalable() && "VF is assumed to be non scalable.");
          VectorType *OtherVTy = VectorType::get(Member->getType(), VF);
          StridedVec = createBitOrPointerCast(StridedVec, OtherVTy, DL);
        }
 
        if (Group->isReverse())
          StridedVec = Builder.CreateVectorReverse(StridedVec, "reverse");
 
        State.set(VPDefs[J], StridedVec, Part);
      }
      ++J;
    }
    return;
  }
 
  // The sub vector type for current instruction.
  auto *SubVT = VectorType::get(ScalarTy, VF);
 
  // Vectorize the interleaved store group.
  Value *MaskForGaps =
      createBitMaskForGaps(Builder, VF.getKnownMinValue(), *Group);
  assert((!MaskForGaps || useMaskedInterleavedAccesses(*TTI)) &&
         "masked interleaved groups are not allowed.");
  assert((!MaskForGaps || !VF.isScalable()) &&
         "masking gaps for scalable vectors is not yet supported.");
  for (unsigned Part = 0; Part < UF; Part++) {
    // Collect the stored vector from each member.
    SmallVector<Value *, 4> StoredVecs;
    unsigned StoredIdx = 0;
    for (unsigned i = 0; i < InterleaveFactor; i++) {
      assert((Group->getMember(i) || MaskForGaps) &&
             "Fail to get a member from an interleaved store group");
      Instruction *Member = Group->getMember(i);
 
      // Skip the gaps in the group.
      if (!Member) {
        Value *Undef = PoisonValue::get(SubVT);
        StoredVecs.push_back(Undef);
        continue;
      }
 
      Value *StoredVec = State.get(StoredValues[StoredIdx], Part);
      ++StoredIdx;
 
      if (Group->isReverse())
        StoredVec = Builder.CreateVectorReverse(StoredVec, "reverse");
 
      // If this member has different type, cast it to a unified type.
 
      if (StoredVec->getType() != SubVT)
        StoredVec = createBitOrPointerCast(StoredVec, SubVT, DL);
 
      StoredVecs.push_back(StoredVec);
    }
 
    // Interleave all the smaller vectors into one wider vector.
    Value *IVec = interleaveVectors(Builder, StoredVecs, "interleaved.vec");
    Instruction *NewStoreInstr;
    if (BlockInMask || MaskForGaps) {
      Value *GroupMask = CreateGroupMask(Part, MaskForGaps);
      NewStoreInstr = Builder.CreateMaskedStore(IVec, AddrParts[Part],
                                                Group->getAlign(), GroupMask);
    } else
      NewStoreInstr =
          Builder.CreateAlignedStore(IVec, AddrParts[Part], Group->getAlign());
 
    Group->addMetadata(NewStoreInstr);
  }
}
 
void InnerLoopVectorizer::scalarizeInstruction(const Instruction *Instr,
                                               VPReplicateRecipe *RepRecipe,
                                               const VPIteration &Instance,
                                               VPTransformState &State) {
  assert(!Instr->getType()->isAggregateType() && "Can't handle vectors");
 
  // llvm.experimental.noalias.scope.decl intrinsics must only be duplicated for
  // the first lane and part.
  if (isa<NoAliasScopeDeclInst>(Instr))
    if (!Instance.isFirstIteration())
      return;
 
  // Does this instruction return a value ?
  bool IsVoidRetTy = Instr->getType()->isVoidTy();
 
  Instruction *Cloned = Instr->clone();
  if (!IsVoidRetTy)
    Cloned->setName(Instr->getName() + ".cloned");
 
  RepRecipe->setFlags(Cloned);
 
  if (auto DL = Instr->getDebugLoc())
    State.setDebugLocFrom(DL);
 
  // Replace the operands of the cloned instructions with their scalar
  // equivalents in the new loop.
  for (const auto &I : enumerate(RepRecipe->operands())) {
    auto InputInstance = Instance;
    VPValue *Operand = I.value();
    if (vputils::isUniformAfterVectorization(Operand))
      InputInstance.Lane = VPLane::getFirstLane();
    Cloned->setOperand(I.index(), State.get(Operand, InputInstance));
  }
  State.addNewMetadata(Cloned, Instr);
 
  // Place the cloned scalar in the new loop.
  State.Builder.Insert(Cloned);
 
  State.set(RepRecipe, Cloned, Instance);
 
  // If we just cloned a new assumption, add it the assumption cache.
  if (auto *II = dyn_cast<AssumeInst>(Cloned))
    AC->registerAssumption(II);
 
  // End if-block.
  bool IfPredicateInstr = RepRecipe->getParent()->getParent()->isReplicator();
  if (IfPredicateInstr)
    PredicatedInstructions.push_back(Cloned);
}
 
Value *
InnerLoopVectorizer::getOrCreateVectorTripCount(BasicBlock *InsertBlock) {
  if (VectorTripCount)
    return VectorTripCount;
 
  Value *TC = getTripCount();
  IRBuilder<> Builder(InsertBlock->getTerminator());
 
  Type *Ty = TC->getType();
  // This is where we can make the step a runtime constant.
  Value *Step = createStepForVF(Builder, Ty, VF, UF);
 
  // If the tail is to be folded by masking, round the number of iterations N
  // up to a multiple of Step instead of rounding down. This is done by first
  // adding Step-1 and then rounding down. Note that it's ok if this addition
  // overflows: the vector induction variable will eventually wrap to zero given
  // that it starts at zero and its Step is a power of two; the loop will then
  // exit, with the last early-exit vector comparison also producing all-true.
  // For scalable vectors the VF is not guaranteed to be a power of 2, but this
  // is accounted for in emitIterationCountCheck that adds an overflow check.
  if (Cost->foldTailByMasking()) {
    assert(isPowerOf2_32(VF.getKnownMinValue() * UF) &&
           "VF*UF must be a power of 2 when folding tail by masking");
    Value *NumLanes = getRuntimeVF(Builder, Ty, VF * UF);
    TC = Builder.CreateAdd(
        TC, Builder.CreateSub(NumLanes, ConstantInt::get(Ty, 1)), "n.rnd.up");
  }
 
  // Now we need to generate the expression for the part of the loop that the
  // vectorized body will execute. This is equal to N - (N % Step) if scalar
  // iterations are not required for correctness, or N - Step, otherwise. Step
  // is equal to the vectorization factor (number of SIMD elements) times the
  // unroll factor (number of SIMD instructions).
  Value *R = Builder.CreateURem(TC, Step, "n.mod.vf");
 
  // There are cases where we *must* run at least one iteration in the remainder
  // loop.  See the cost model for when this can happen.  If the step evenly
  // divides the trip count, we set the remainder to be equal to the step. If
  // the step does not evenly divide the trip count, no adjustment is necessary
  // since there will already be scalar iterations. Note that the minimum
  // iterations check ensures that N >= Step.
  if (Cost->requiresScalarEpilogue(VF.isVector())) {
    auto *IsZero = Builder.CreateICmpEQ(R, ConstantInt::get(R->getType(), 0));
    R = Builder.CreateSelect(IsZero, Step, R);
  }
 
  VectorTripCount = Builder.CreateSub(TC, R, "n.vec");
 
  return VectorTripCount;
}
 
Value *InnerLoopVectorizer::createBitOrPointerCast(Value *V, VectorType *DstVTy,
                                                   const DataLayout &DL) {
  // Verify that V is a vector type with same number of elements as DstVTy.
  auto *DstFVTy = cast<VectorType>(DstVTy);
  auto VF = DstFVTy->getElementCount();
  auto *SrcVecTy = cast<VectorType>(V->getType());
  assert(VF == SrcVecTy->getElementCount() && "Vector dimensions do not match");
  Type *SrcElemTy = SrcVecTy->getElementType();
  Type *DstElemTy = DstFVTy->getElementType();
  assert((DL.getTypeSizeInBits(SrcElemTy) == DL.getTypeSizeInBits(DstElemTy)) &&
         "Vector elements must have same size");
 
  // Do a direct cast if element types are castable.
  if (CastInst::isBitOrNoopPointerCastable(SrcElemTy, DstElemTy, DL)) {
    return Builder.CreateBitOrPointerCast(V, DstFVTy);
  }
  // V cannot be directly casted to desired vector type.
  // May happen when V is a floating point vector but DstVTy is a vector of
  // pointers or vice-versa. Handle this using a two-step bitcast using an
  // intermediate Integer type for the bitcast i.e. Ptr <-> Int <-> Float.
  assert((DstElemTy->isPointerTy() != SrcElemTy->isPointerTy()) &&
         "Only one type should be a pointer type");
  assert((DstElemTy->isFloatingPointTy() != SrcElemTy->isFloatingPointTy()) &&
         "Only one type should be a floating point type");
  Type *IntTy =
      IntegerType::getIntNTy(V->getContext(), DL.getTypeSizeInBits(SrcElemTy));
  auto *VecIntTy = VectorType::get(IntTy, VF);
  Value *CastVal = Builder.CreateBitOrPointerCast(V, VecIntTy);
  return Builder.CreateBitOrPointerCast(CastVal, DstFVTy);
}
 
void InnerLoopVectorizer::emitIterationCountCheck(BasicBlock *Bypass) {
  Value *Count = getTripCount();
  // Reuse existing vector loop preheader for TC checks.
  // Note that new preheader block is generated for vector loop.
  BasicBlock *const TCCheckBlock = LoopVectorPreHeader;
  IRBuilder<> Builder(TCCheckBlock->getTerminator());
 
  // Generate code to check if the loop's trip count is less than VF * UF, or
  // equal to it in case a scalar epilogue is required; this implies that the
  // vector trip count is zero. This check also covers the case where adding one
  // to the backedge-taken count overflowed leading to an incorrect trip count
  // of zero. In this case we will also jump to the scalar loop.
  auto P = Cost->requiresScalarEpilogue(VF.isVector()) ? ICmpInst::ICMP_ULE
                                                       : ICmpInst::ICMP_ULT;
 
  // If tail is to be folded, vector loop takes care of all iterations.
  Type *CountTy = Count->getType();
  Value *CheckMinIters = Builder.getFalse();
  auto CreateStep = [&]() -> Value * {
    // Create step with max(MinProTripCount, UF * VF).
    if (UF * VF.getKnownMinValue() >= MinProfitableTripCount.getKnownMinValue())
      return createStepForVF(Builder, CountTy, VF, UF);
 
    Value *MinProfTC =
        createStepForVF(Builder, CountTy, MinProfitableTripCount, 1);
    if (!VF.isScalable())
      return MinProfTC;
    return Builder.CreateBinaryIntrinsic(
        Intrinsic::umax, MinProfTC, createStepForVF(Builder, CountTy, VF, UF));
  };
 
  TailFoldingStyle Style = Cost->getTailFoldingStyle();
  if (Style == TailFoldingStyle::None)
    CheckMinIters =
        Builder.CreateICmp(P, Count, CreateStep(), "min.iters.check");
  else if (VF.isScalable() &&
           !isIndvarOverflowCheckKnownFalse(Cost, VF, UF) &&
           Style != TailFoldingStyle::DataAndControlFlowWithoutRuntimeCheck) {
    // vscale is not necessarily a power-of-2, which means we cannot guarantee
    // an overflow to zero when updating induction variables and so an
    // additional overflow check is required before entering the vector loop.
 
    // Get the maximum unsigned value for the type.
    Value *MaxUIntTripCount =
        ConstantInt::get(CountTy, cast<IntegerType>(CountTy)->getMask());
    Value *LHS = Builder.CreateSub(MaxUIntTripCount, Count);
 
    // Don't execute the vector loop if (UMax - n) < (VF * UF).
    CheckMinIters = Builder.CreateICmp(ICmpInst::ICMP_ULT, LHS, CreateStep());
  }
 
  // Create new preheader for vector loop.
  LoopVectorPreHeader =
      SplitBlock(TCCheckBlock, TCCheckBlock->getTerminator(), DT, LI, nullptr,
                 "vector.ph");
 
  assert(DT->properlyDominates(DT->getNode(TCCheckBlock),
                               DT->getNode(Bypass)->getIDom()) &&
         "TC check is expected to dominate Bypass");
 
  // Update dominator for Bypass & LoopExit (if needed).
  DT->changeImmediateDominator(Bypass, TCCheckBlock);
  if (!Cost->requiresScalarEpilogue(VF.isVector()))
    // If there is an epilogue which must run, there's no edge from the
    // middle block to exit blocks  and thus no need to update the immediate
    // dominator of the exit blocks.
    DT->changeImmediateDominator(LoopExitBlock, TCCheckBlock);
 
  ReplaceInstWithInst(
      TCCheckBlock->getTerminator(),
      BranchInst::Create(Bypass, LoopVectorPreHeader, CheckMinIters));
  LoopBypassBlocks.push_back(TCCheckBlock);
}
 
BasicBlock *InnerLoopVectorizer::emitSCEVChecks(BasicBlock *Bypass) {
  BasicBlock *const SCEVCheckBlock =
      RTChecks.emitSCEVChecks(Bypass, LoopVectorPreHeader, LoopExitBlock);
  if (!SCEVCheckBlock)
    return nullptr;
 
  assert(!(SCEVCheckBlock->getParent()->hasOptSize() ||
           (OptForSizeBasedOnProfile &&
            Cost->Hints->getForce() != LoopVectorizeHints::FK_Enabled)) &&
         "Cannot SCEV check stride or overflow when optimizing for size");
 
 
  // Update dominator only if this is first RT check.
  if (LoopBypassBlocks.empty()) {
    DT->changeImmediateDominator(Bypass, SCEVCheckBlock);
    if (!Cost->requiresScalarEpilogue(VF.isVector()))
      // If there is an epilogue which must run, there's no edge from the
      // middle block to exit blocks  and thus no need to update the immediate
      // dominator of the exit blocks.
      DT->changeImmediateDominator(LoopExitBlock, SCEVCheckBlock);
  }
 
  LoopBypassBlocks.push_back(SCEVCheckBlock);
  AddedSafetyChecks = true;
  return SCEVCheckBlock;
}
 
BasicBlock *InnerLoopVectorizer::emitMemRuntimeChecks(BasicBlock *Bypass) {
  // VPlan-native path does not do any analysis for runtime checks currently.
  if (EnableVPlanNativePath)
    return nullptr;
 
  BasicBlock *const MemCheckBlock =
      RTChecks.emitMemRuntimeChecks(Bypass, LoopVectorPreHeader);
 
  // Check if we generated 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.
  if (!MemCheckBlock)
    return nullptr;
 
  if (MemCheckBlock->getParent()->hasOptSize() || OptForSizeBasedOnProfile) {
    assert(Cost->Hints->getForce() == LoopVectorizeHints::FK_Enabled &&
           "Cannot emit memory checks when optimizing for size, unless forced "
           "to vectorize.");
    ORE->emit([&]() {
      return OptimizationRemarkAnalysis(DEBUG_TYPE, "VectorizationCodeSize",
                                        OrigLoop->getStartLoc(),
                                        OrigLoop->getHeader())
             << "Code-size may be reduced by not forcing "
                "vectorization, or by source-code modifications "
                "eliminating the need for runtime checks "
                "(e.g., adding 'restrict').";
    });
  }
 
  LoopBypassBlocks.push_back(MemCheckBlock);
 
  AddedSafetyChecks = true;
 
  return MemCheckBlock;
}
 
void InnerLoopVectorizer::createVectorLoopSkeleton(StringRef Prefix) {
  LoopScalarBody = OrigLoop->getHeader();
  LoopVectorPreHeader = OrigLoop->getLoopPreheader();
  assert(LoopVectorPreHeader && "Invalid loop structure");
  LoopExitBlock = OrigLoop->getUniqueExitBlock(); // may be nullptr
  assert((LoopExitBlock || Cost->requiresScalarEpilogue(VF.isVector())) &&
         "multiple exit loop without required epilogue?");
 
  LoopMiddleBlock =
      SplitBlock(LoopVectorPreHeader, LoopVectorPreHeader->getTerminator(), DT,
                 LI, nullptr, Twine(Prefix) + "middle.block");
  LoopScalarPreHeader =
      SplitBlock(LoopMiddleBlock, LoopMiddleBlock->getTerminator(), DT, LI,
                 nullptr, Twine(Prefix) + "scalar.ph");
 
  auto *ScalarLatchTerm = OrigLoop->getLoopLatch()->getTerminator();
 
  // Set up the middle block terminator.  Two cases:
  // 1) If we know that we must execute the scalar epilogue, emit an
  //    unconditional branch.
  // 2) Otherwise, we must have a single unique exit block (due to how we
  //    implement the multiple exit case).  In this case, set up a conditional
  //    branch from the middle block to the loop scalar preheader, and the
  //    exit block.  completeLoopSkeleton will update the condition to use an
  //    iteration check, if required to decide whether to execute the remainder.
  BranchInst *BrInst =
      Cost->requiresScalarEpilogue(VF.isVector())
          ? BranchInst::Create(LoopScalarPreHeader)
          : BranchInst::Create(LoopExitBlock, LoopScalarPreHeader,
                               Builder.getTrue());
  BrInst->setDebugLoc(ScalarLatchTerm->getDebugLoc());
  ReplaceInstWithInst(LoopMiddleBlock->getTerminator(), BrInst);
 
  // Update dominator for loop exit. During skeleton creation, only the vector
  // pre-header and the middle block are created. The vector loop is entirely
  // created during VPlan exection.
  if (!Cost->requiresScalarEpilogue(VF.isVector()))
    // If there is an epilogue which must run, there's no edge from the
    // middle block to exit blocks  and thus no need to update the immediate
    // dominator of the exit blocks.
    DT->changeImmediateDominator(LoopExitBlock, LoopMiddleBlock);
}
 
PHINode *InnerLoopVectorizer::createInductionResumeValue(
    PHINode *OrigPhi, const InductionDescriptor &II, Value *Step,
    ArrayRef<BasicBlock *> BypassBlocks,
    std::pair<BasicBlock *, Value *> AdditionalBypass) {
  Value *VectorTripCount = getOrCreateVectorTripCount(LoopVectorPreHeader);
  assert(VectorTripCount && "Expected valid arguments");
 
  Instruction *OldInduction = Legal->getPrimaryInduction();
  Value *&EndValue = IVEndValues[OrigPhi];
  Value *EndValueFromAdditionalBypass = AdditionalBypass.second;
  if (OrigPhi == OldInduction) {
    // We know what the end value is.
    EndValue = VectorTripCount;
  } else {
    IRBuilder<> B(LoopVectorPreHeader->getTerminator());
 
    // Fast-math-flags propagate from the original induction instruction.
    if (II.getInductionBinOp() && isa<FPMathOperator>(II.getInductionBinOp()))
      B.setFastMathFlags(II.getInductionBinOp()->getFastMathFlags());
 
    EndValue = emitTransformedIndex(B, VectorTripCount, II.getStartValue(),
                                    Step, II.getKind(), II.getInductionBinOp());
    EndValue->setName("ind.end");
 
    // Compute the end value for the additional bypass (if applicable).
    if (AdditionalBypass.first) {
      B.SetInsertPoint(AdditionalBypass.first,
                       AdditionalBypass.first->getFirstInsertionPt());
      EndValueFromAdditionalBypass =
          emitTransformedIndex(B, AdditionalBypass.second, II.getStartValue(),
                               Step, II.getKind(), II.getInductionBinOp());
      EndValueFromAdditionalBypass->setName("ind.end");
    }
  }
 
  // Create phi nodes to merge from the  backedge-taken check block.
  PHINode *BCResumeVal = PHINode::Create(OrigPhi->getType(), 3, "bc.resume.val",
                                         LoopScalarPreHeader->getTerminator());
  // Copy original phi DL over to the new one.
  BCResumeVal->setDebugLoc(OrigPhi->getDebugLoc());
 
  // The new PHI merges the original incoming value, in case of a bypass,
  // or the value at the end of the vectorized loop.
  BCResumeVal->addIncoming(EndValue, LoopMiddleBlock);
 
  // Fix the scalar body counter (PHI node).
  // The old induction's phi node in the scalar body needs the truncated
  // value.
  for (BasicBlock *BB : BypassBlocks)
    BCResumeVal->addIncoming(II.getStartValue(), BB);
 
  if (AdditionalBypass.first)
    BCResumeVal->setIncomingValueForBlock(AdditionalBypass.first,
                                          EndValueFromAdditionalBypass);
  return BCResumeVal;
}
 
/// Return the expanded step for \p ID using \p ExpandedSCEVs to look up SCEV
/// expansion results.
static Value *getExpandedStep(const InductionDescriptor &ID,
                              const SCEV2ValueTy &ExpandedSCEVs) {
  const SCEV *Step = ID.getStep();
  if (auto *C = dyn_cast<SCEVConstant>(Step))
    return C->getValue();
  if (auto *U = dyn_cast<SCEVUnknown>(Step))
    return U->getValue();
  auto I = ExpandedSCEVs.find(Step);
  assert(I != ExpandedSCEVs.end() && "SCEV must be expanded at this point");
  return I->second;
}
 
void InnerLoopVectorizer::createInductionResumeValues(
    const SCEV2ValueTy &ExpandedSCEVs,
    std::pair<BasicBlock *, Value *> AdditionalBypass) {
  assert(((AdditionalBypass.first && AdditionalBypass.second) ||
          (!AdditionalBypass.first && !AdditionalBypass.second)) &&
         "Inconsistent information about additional bypass.");
  // We are going to resume the execution of the scalar loop.
  // Go over all of the induction variables that we found and fix the
  // PHIs that are left in the scalar version of the loop.
  // The starting values of PHI nodes depend on the counter of the last
  // iteration in the vectorized loop.
  // If we come from a bypass edge then we need to start from the original
  // start value.
  for (const auto &InductionEntry : Legal->getInductionVars()) {
    PHINode *OrigPhi = InductionEntry.first;
    const InductionDescriptor &II = InductionEntry.second;
    PHINode *BCResumeVal = createInductionResumeValue(
        OrigPhi, II, getExpandedStep(II, ExpandedSCEVs), LoopBypassBlocks,
        AdditionalBypass);
    OrigPhi->setIncomingValueForBlock(LoopScalarPreHeader, BCResumeVal);
  }
}
 
BasicBlock *InnerLoopVectorizer::completeLoopSkeleton() {
  // The trip counts should be cached by now.
  Value *Count = getTripCount();
  Value *VectorTripCount = getOrCreateVectorTripCount(LoopVectorPreHeader);
 
  auto *ScalarLatchTerm = OrigLoop->getLoopLatch()->getTerminator();
 
  // Add a check in the middle block to see if we have completed
  // all of the iterations in the first vector loop.  Three cases:
  // 1) If we require a scalar epilogue, there is no conditional branch as
  //    we unconditionally branch to the scalar preheader.  Do nothing.
  // 2) If (N - N%VF) == N, then we *don't* need to run the remainder.
  //    Thus if tail is to be folded, we know we don't need to run the
  //    remainder and we can use the previous value for the condition (true).
  // 3) Otherwise, construct a runtime check.
  if (!Cost->requiresScalarEpilogue(VF.isVector()) &&
      !Cost->foldTailByMasking()) {
    // Here we use the same DebugLoc as the scalar loop latch terminator instead
    // of the corresponding compare because they may have ended up with
    // different line numbers and we want to avoid awkward line stepping while
    // debugging. Eg. if the compare has got a line number inside the loop.
    // TODO: At the moment, CreateICmpEQ will simplify conditions with constant
    // operands. Perform simplification directly on VPlan once the branch is
    // modeled there.
    IRBuilder<> B(LoopMiddleBlock->getTerminator());
    B.SetCurrentDebugLocation(ScalarLatchTerm->getDebugLoc());
    Value *CmpN = B.CreateICmpEQ(Count, VectorTripCount, "cmp.n");
    cast<BranchInst>(LoopMiddleBlock->getTerminator())->setCondition(CmpN);
  }
 
#ifdef EXPENSIVE_CHECKS
  assert(DT->verify(DominatorTree::VerificationLevel::Fast));
#endif
 
  return LoopVectorPreHeader;
}
 
std::pair<BasicBlock *, Value *>
InnerLoopVectorizer::createVectorizedLoopSkeleton(
    const SCEV2ValueTy &ExpandedSCEVs) {
  /*
   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.
 
       [ ] <-- old preheader - loop iteration number check and SCEVs in Plan's
     /  |      preheader are expanded here. Eventually all required SCEV
    /   |      expansion should happen here.
   /    v
  |    [ ] <-- vector loop bypass (may consist of multiple blocks).
  |  /  |
  | /   v
  ||   [ ]     <-- vector pre header.
  |/    |
  |     v
  |    [  ] \
  |    [  ]_|   <-- vector loop (created during VPlan execution).
  |     |
  |     v
  \   -[ ]   <--- middle-block.
   \/   |
   /\   v
   | ->[ ]     <--- new preheader.
   |    |
 (opt)  v      <-- edge from middle to exit iff epilogue is not required.
   |   [ ] \
   |   [ ]_|   <-- old scalar loop to handle remainder (scalar epilogue).
    \   |
     \  v
      >[ ]     <-- exit block(s).
   ...
   */
 
  // Create an empty vector loop, and prepare basic blocks for the runtime
  // checks.
  createVectorLoopSkeleton("");
 
  // Now, compare the new count to zero. If it is zero skip the vector loop and
  // jump to the scalar loop. This check also covers the case where the
  // backedge-taken count is uint##_max: adding one to it will overflow leading
  // to an incorrect trip count of zero. In this (rare) case we will also jump
  // to the scalar loop.
  emitIterationCountCheck(LoopScalarPreHeader);
 
  // Generate the code to check any assumptions that we've made for SCEV
  // expressions.
  emitSCEVChecks(LoopScalarPreHeader);
 
  // 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(LoopScalarPreHeader);
 
  // Emit phis for the new starting index of the scalar loop.
  createInductionResumeValues(ExpandedSCEVs);
 
  return {completeLoopSkeleton(), nullptr};
}
 
// Fix up external users of the induction variable. At this point, we are
// in LCSSA form, with all external PHIs that use the IV having one input value,
// coming from the remainder loop. We need those PHIs to also have a correct
// value for the IV when arriving directly from the middle block.
void InnerLoopVectorizer::fixupIVUsers(PHINode *OrigPhi,
                                       const InductionDescriptor &II,
                                       Value *VectorTripCount, Value *EndValue,
                                       BasicBlock *MiddleBlock,
                                       BasicBlock *VectorHeader, VPlan &Plan,
                                       VPTransformState &State) {
  // There are two kinds of external IV usages - those that use the value
  // computed in the last iteration (the PHI) and those that use the penultimate
  // value (the value that feeds into the phi from the loop latch).
  // We allow both, but they, obviously, have different values.
 
  assert(OrigLoop->getUniqueExitBlock() && "Expected a single exit block");
 
  DenseMap<Value *, Value *> MissingVals;
 
  // An external user of the last iteration's value should see the value that
  // the remainder loop uses to initialize its own IV.
  Value *PostInc = OrigPhi->getIncomingValueForBlock(OrigLoop->getLoopLatch());
  for (User *U : PostInc->users()) {
    Instruction *UI = cast<Instruction>(U);
    if (!OrigLoop->contains(UI)) {
      assert(isa<PHINode>(UI) && "Expected LCSSA form");
      MissingVals[UI] = EndValue;
    }
  }
 
  // An external user of the penultimate value need to see EndValue - Step.
  // The simplest way to get this is to recompute it from the constituent SCEVs,
  // that is Start + (Step * (CRD - 1)).
  for (User *U : OrigPhi->users()) {
    auto *UI = cast<Instruction>(U);
    if (!OrigLoop->contains(UI)) {
      assert(isa<PHINode>(UI) && "Expected LCSSA form");
      IRBuilder<> B(MiddleBlock->getTerminator());
 
      // Fast-math-flags propagate from the original induction instruction.
      if (II.getInductionBinOp() && isa<FPMathOperator>(II.getInductionBinOp()))
        B.setFastMathFlags(II.getInductionBinOp()->getFastMathFlags());
 
      Value *CountMinusOne = B.CreateSub(
          VectorTripCount, ConstantInt::get(VectorTripCount->getType(), 1));
      CountMinusOne->setName("cmo");
 
      VPValue *StepVPV = Plan.getSCEVExpansion(II.getStep());
      assert(StepVPV && "step must have been expanded during VPlan execution");
      Value *Step = StepVPV->isLiveIn() ? StepVPV->getLiveInIRValue()
                                        : State.get(StepVPV, {0, 0});
      Value *Escape =
          emitTransformedIndex(B, CountMinusOne, II.getStartValue(), Step,
                               II.getKind(), II.getInductionBinOp());
      Escape->setName("ind.escape");
      MissingVals[UI] = Escape;
    }
  }
 
  for (auto &I : MissingVals) {
    PHINode *PHI = cast<PHINode>(I.first);
    // One corner case we have to handle is two IVs "chasing" each-other,
    // that is %IV2 = phi [...], [ %IV1, %latch ]
    // In this case, if IV1 has an external use, we need to avoid adding both
    // "last value of IV1" and "penultimate value of IV2". So, verify that we
    // don't already have an incoming value for the middle block.
    if (PHI->getBasicBlockIndex(MiddleBlock) == -1) {
      PHI->addIncoming(I.second, MiddleBlock);
      Plan.removeLiveOut(PHI);
    }
  }
}
 
namespace {
 
struct CSEDenseMapInfo {
  static bool canHandle(const 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(const 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(const Instruction *LHS, const Instruction *RHS) {
    if (LHS == getEmptyKey() || RHS == getEmptyKey() ||
        LHS == getTombstoneKey() || RHS == getTombstoneKey())
      return LHS == RHS;
    return LHS->isIdenticalTo(RHS);
  }
};
 
} // end anonymous namespace
 
///Perform cse of induction variable instructions.
static void cse(BasicBlock *BB) {
  // Perform simple cse.
  SmallDenseMap<Instruction *, Instruction *, 4, CSEDenseMapInfo> CSEMap;
  for (Instruction &In : llvm::make_early_inc_range(*BB)) {
    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;
    }
 
    CSEMap[&In] = &In;
  }
}
 
InstructionCost LoopVectorizationCostModel::getVectorCallCost(
    CallInst *CI, ElementCount VF, Function **Variant, bool *NeedsMask) const {
  Function *F = CI->getCalledFunction();
  Type *ScalarRetTy = CI->getType();
  SmallVector<Type *, 4> Tys, ScalarTys;
  bool MaskRequired = Legal->isMaskRequired(CI);
  for (auto &ArgOp : CI->args())
    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.
  TTI::TargetCostKind CostKind = TTI::TCK_RecipThroughput;
  InstructionCost ScalarCallCost =
      TTI.getCallInstrCost(F, ScalarRetTy, ScalarTys, CostKind);
  if (VF.isScalar())
    return ScalarCallCost;
 
  // Compute corresponding vector type for return value and arguments.
  Type *RetTy = ToVectorTy(ScalarRetTy, VF);
  for (Type *ScalarTy : ScalarTys)
    Tys.push_back(ToVectorTy(ScalarTy, VF));
 
  // Compute costs of unpacking argument values for the scalar calls and
  // packing the return values to a vector.
  InstructionCost ScalarizationCost =
      getScalarizationOverhead(CI, VF, CostKind);
 
  InstructionCost Cost =
      ScalarCallCost * VF.getKnownMinValue() + ScalarizationCost;
 
  // If we can't emit a vector call for this function, then the currently found
  // cost is the cost we need to return.
  InstructionCost MaskCost = 0;
  VFShape Shape = VFShape::get(*CI, VF, MaskRequired);
  if (NeedsMask)
    *NeedsMask = MaskRequired;
  Function *VecFunc = VFDatabase(*CI).getVectorizedFunction(Shape);
  // If we want an unmasked vector function but can't find one matching the VF,
  // maybe we can find vector function that does use a mask and synthesize
  // an all-true mask.
  if (!VecFunc && !MaskRequired) {
    Shape = VFShape::get(*CI, VF, /*HasGlobalPred=*/true);
    VecFunc = VFDatabase(*CI).getVectorizedFunction(Shape);
    // If we found one, add in the cost of creating a mask
    if (VecFunc) {
      if (NeedsMask)
        *NeedsMask = true;
      MaskCost = TTI.getShuffleCost(
          TargetTransformInfo::SK_Broadcast,
          VectorType::get(
              IntegerType::getInt1Ty(VecFunc->getFunctionType()->getContext()),
              VF));
    }
  }
 
  // We don't support masked function calls yet, but we can scalarize a
  // masked call with branches (unless VF is scalable).
  if (!TLI || CI->isNoBuiltin() || !VecFunc)
    return VF.isScalable() ? InstructionCost::getInvalid() : Cost;
 
  // If the corresponding vector cost is cheaper, return its cost.
  InstructionCost VectorCallCost =
      TTI.getCallInstrCost(nullptr, RetTy, Tys, CostKind) + MaskCost;
  if (VectorCallCost < Cost) {
    *Variant = VecFunc;
    Cost = VectorCallCost;
  }
  return Cost;
}
 
static Type *MaybeVectorizeType(Type *Elt, ElementCount VF) {
  if (VF.isScalar() || (!Elt->isIntOrPtrTy() && !Elt->isFloatingPointTy()))
    return Elt;
  return VectorType::get(Elt, VF);
}
 
InstructionCost
LoopVectorizationCostModel::getVectorIntrinsicCost(CallInst *CI,
                                                   ElementCount VF) const {
  Intrinsic::ID ID = getVectorIntrinsicIDForCall(CI, TLI);
  assert(ID && "Expected intrinsic call!");
  Type *RetTy = MaybeVectorizeType(CI->getType(), VF);
  FastMathFlags FMF;
  if (auto *FPMO = dyn_cast<FPMathOperator>(CI))
    FMF = FPMO->getFastMathFlags();
 
  SmallVector<const Value *> Arguments(CI->args());
  FunctionType *FTy = CI->getCalledFunction()->getFunctionType();
  SmallVector<Type *> ParamTys;
  std::transform(FTy->param_begin(), FTy->param_end(),
                 std::back_inserter(ParamTys),
                 [&](Type *Ty) { return MaybeVectorizeType(Ty, VF); });
 
  IntrinsicCostAttributes CostAttrs(ID, RetTy, Arguments, ParamTys, FMF,
                                    dyn_cast<IntrinsicInst>(CI));
  return TTI.getIntrinsicInstrCost(CostAttrs,
                                   TargetTransformInfo::TCK_RecipThroughput);
}
 
static Type *smallestIntegerVectorType(Type *T1, Type *T2) {
  auto *I1 = cast<IntegerType>(cast<VectorType>(T1)->getElementType());
  auto *I2 = cast<IntegerType>(cast<VectorType>(T2)->getElementType());
  return I1->getBitWidth() < I2->getBitWidth() ? T1 : T2;
}
 
static Type *largestIntegerVectorType(Type *T1, Type *T2) {
  auto *I1 = cast<IntegerType>(cast<VectorType>(T1)->getElementType());
  auto *I2 = cast<IntegerType>(cast<VectorType>(T2)->getElementType());
  return I1->getBitWidth() > I2->getBitWidth() ? T1 : T2;
}
 
void InnerLoopVectorizer::truncateToMinimalBitwidths(VPTransformState &State) {
  // 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.
  SmallPtrSet<Value *, 4> Erased;
  for (const auto &KV : Cost->getMinimalBitwidths()) {
    // If the value wasn't vectorized, we must maintain the original scalar
    // type. The absence of the value from State indicates that it
    // wasn't vectorized.
    // FIXME: Should not rely on getVPValue at this point.
    VPValue *Def = State.Plan->getVPValue(KV.first, true);
    if (!State.hasAnyVectorValue(Def))
      continue;
    // If the instruction is defined outside the loop, only update the first
    // part; the first part will be re-used for all other parts.
    unsigned UFToUse = OrigLoop->contains(KV.first) ? UF : 1;
    for (unsigned Part = 0; Part < UFToUse; ++Part) {
      Value *I = State.get(Def, Part);
      if (Erased.count(I) || I->use_empty() || !isa<Instruction>(I))
        continue;
      Type *OriginalTy = I->getType();
      Type *ScalarTruncatedTy =
          IntegerType::get(OriginalTy->getContext(), KV.second);
      auto *TruncatedTy = VectorType::get(
          ScalarTruncatedTy, cast<VectorType>(OriginalTy)->getElementCount());
      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 (auto *BO = dyn_cast<BinaryOperator>(I)) {
        Value *Op0 = ShrinkOperand(BO->getOperand(0));
        Value *Op1 = ShrinkOperand(BO->getOperand(1));
        NewI = B.CreateBinOp(BO->getOpcode(), Op0, Op1);
 
        // Any wrapping introduced by shrinking this operation shouldn't be
        // considered undefined behavior. So, we can't unconditionally copy
        // arithmetic wrapping flags to NewI.
        cast<BinaryOperator>(NewI)->copyIRFlags(I, /*IncludeWrapFlags=*/false);
      } else if (auto *CI = dyn_cast<ICmpInst>(I)) {
        Value *Op0 = ShrinkOperand(BO->getOperand(0));
        Value *Op1 = ShrinkOperand(BO->getOperand(1));
        NewI = B.CreateICmp(CI->getPredicate(), Op0, Op1);
      } else if (auto *SI = dyn_cast<SelectInst>(I)) {
        Value *TV = ShrinkOperand(SI->getTrueValue());
        Value *FV = ShrinkOperand(SI->getFalseValue());
        NewI = B.CreateSelect(SI->getCondition(), TV, FV);
      } else if (auto *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 (auto *SI = dyn_cast<ShuffleVectorInst>(I)) {
        auto Elements0 =
            cast<VectorType>(SI->getOperand(0)->getType())->getElementCount();
        auto *O0 = B.CreateZExtOrTrunc(
            SI->getOperand(0), VectorType::get(ScalarTruncatedTy, Elements0));
        auto Elements1 =
            cast<VectorType>(SI->getOperand(1)->getType())->getElementCount();
        auto *O1 = B.CreateZExtOrTrunc(
            SI->getOperand(1), VectorType::get(ScalarTruncatedTy, Elements1));
 
        NewI = B.CreateShuffleVector(O0, O1, SI->getShuffleMask());
      } else if (isa<LoadInst>(I) || isa<PHINode>(I)) {
        // Don't do anything with the operands, just extend the result.
        continue;
      } else if (auto *IE = dyn_cast<InsertElementInst>(I)) {
        auto Elements =
            cast<VectorType>(IE->getOperand(0)->getType())->getElementCount();
        auto *O0 = B.CreateZExtOrTrunc(
            IE->getOperand(0), VectorType::get(ScalarTruncatedTy, Elements));
        auto *O1 = B.CreateZExtOrTrunc(IE->getOperand(1), ScalarTruncatedTy);
        NewI = B.CreateInsertElement(O0, O1, IE->getOperand(2));
      } else if (auto *EE = dyn_cast<ExtractElementInst>(I)) {
        auto Elements =
            cast<VectorType>(EE->getOperand(0)->getType())->getElementCount();
        auto *O0 = B.CreateZExtOrTrunc(
            EE->getOperand(0), VectorType::get(ScalarTruncatedTy, Elements));
        NewI = B.CreateExtractElement(O0, EE->getOperand(2));
      } else {
        // If we don't know what to do, be conservative and don't do anything.
        continue;
      }
 
      // Lastly, extend the result.
      NewI->takeName(cast<Instruction>(I));
      Value *Res = B.CreateZExtOrTrunc(NewI, OriginalTy);
      I->replaceAllUsesWith(Res);
      cast<Instruction>(I)->eraseFromParent();
      Erased.insert(I);
      State.reset(Def, Res, Part);
    }
  }
 
  // We'll have created a bunch of ZExts that are now parentless. Clean up.
  for (const auto &KV : Cost->getMinimalBitwidths()) {
    // If the value wasn't vectorized, we must maintain the original scalar
    // type. The absence of the value from State indicates that it
    // wasn't vectorized.
    // FIXME: Should not rely on getVPValue at this point.
    VPValue *Def = State.Plan->getVPValue(KV.first, true);
    if (!State.hasAnyVectorValue(Def))
      continue;
    unsigned UFToUse = OrigLoop->contains(KV.first) ? UF : 1;
    for (unsigned Part = 0; Part < UFToUse; ++Part) {
      Value *I = State.get(Def, Part);
      ZExtInst *Inst = dyn_cast<ZExtInst>(I);
      if (Inst && Inst->use_empty()) {
        Value *NewI = Inst->getOperand(0);
        Inst->eraseFromParent();
        State.reset(Def, NewI, Part);
      }
    }
  }
}
 
void InnerLoopVectorizer::fixVectorizedLoop(VPTransformState &State,
                                            VPlan &Plan) {
  // Insert truncates and extends for any truncated instructions as hints to
  // InstCombine.
  if (VF.isVector())
    truncateToMinimalBitwidths(State);
 
  // Fix widened non-induction PHIs by setting up the PHI operands.
  if (EnableVPlanNativePath)
    fixNonInductionPHIs(Plan, State);
 
  // At this point every instruction in the original loop is widened to a
  // vector form. Now we need to fix the recurrences in the loop. These PHI
  // nodes are currently empty because we did not want to introduce cycles.
  // This is the second stage of vectorizing recurrences.
  fixCrossIterationPHIs(State);
 
  // Forget the original basic block.
  PSE.getSE()->forgetLoop(OrigLoop);
 
  // After vectorization, the exit blocks of the original loop will have
  // additional predecessors. Invalidate SCEVs for the exit phis in case SE
  // looked through single-entry phis.
  SmallVector<BasicBlock *> ExitBlocks;
  OrigLoop->getExitBlocks(ExitBlocks);
  for (BasicBlock *Exit : ExitBlocks)
    for (PHINode &PN : Exit->phis())
      PSE.getSE()->forgetValue(&PN);
 
  VPBasicBlock *LatchVPBB = Plan.getVectorLoopRegion()->getExitingBasicBlock();
  Loop *VectorLoop = LI->getLoopFor(State.CFG.VPBB2IRBB[LatchVPBB]);
  if (Cost->requiresScalarEpilogue(VF.isVector())) {
    // No edge from the middle block to the unique exit block has been inserted
    // and there is nothing to fix from vector loop; phis should have incoming
    // from scalar loop only.
  } else {
    // TODO: Check VPLiveOuts to see if IV users need fixing instead of checking
    // the cost model.
 
    // If we inserted an edge from the middle block to the unique exit block,
    // update uses outside the loop (phis) to account for the newly inserted
    // edge.
 
    // Fix-up external users of the induction variables.
    for (const auto &Entry : Legal->getInductionVars())
      fixupIVUsers(Entry.first, Entry.second,
                   getOrCreateVectorTripCount(VectorLoop->getLoopPreheader()),
                   IVEndValues[Entry.first], LoopMiddleBlock,
                   VectorLoop->getHeader(), Plan, State);
  }
 
  // Fix LCSSA phis not already fixed earlier. Extracts may need to be generated
  // in the exit block, so update the builder.
  State.Builder.SetInsertPoint(State.CFG.ExitBB,
                               State.CFG.ExitBB->getFirstNonPHIIt());
  for (const auto &KV : Plan.getLiveOuts())
    KV.second->fixPhi(Plan, State);
 
  for (Instruction *PI : PredicatedInstructions)
    sinkScalarOperands(&*PI);
 
  // Remove redundant induction instructions.
  cse(VectorLoop->getHeader());
 
  // Set/update profile weights for the vector and remainder loops as original
  // loop iterations are now distributed among them. Note that original loop
  // represented by LoopScalarBody becomes remainder loop after vectorization.
  //
  // For cases like foldTailByMasking() and requiresScalarEpiloque() we may
  // end up getting slightly roughened result but that should be OK since
  // profile is not inherently precise anyway. Note also possible bypass of
  // vector code caused by legality checks is ignored, assigning all the weight
  // to the vector loop, optimistically.
  //
  // For scalable vectorization we can't know at compile time how many iterations
  // of the loop are handled in one vector iteration, so instead assume a pessimistic
  // vscale of '1'.
  setProfileInfoAfterUnrolling(LI->getLoopFor(LoopScalarBody), VectorLoop,
                               LI->getLoopFor(LoopScalarBody),
                               VF.getKnownMinValue() * UF);
}
 
void InnerLoopVectorizer::fixCrossIterationPHIs(VPTransformState &State) {
  // In order to support recurrences we need to be able to vectorize Phi nodes.
  // Phi nodes have cycles, so we need to vectorize them in two stages. This is
  // stage #2: We now need to fix the recurrences by adding incoming edges to
  // the currently empty PHI nodes. At this point every instruction in the
  // original loop is widened to a vector form so we can use them to construct
  // the incoming edges.
  VPBasicBlock *Header =
      State.Plan->getVectorLoopRegion()->getEntryBasicBlock();
 
  // Gather all VPReductionPHIRecipe and sort them so that Intermediate stores
  // sank outside of the loop would keep the same order as they had in the
  // original loop.
  SmallVector<VPReductionPHIRecipe *> ReductionPHIList;
  for (VPRecipeBase &R : Header->phis()) {
    if (auto *ReductionPhi = dyn_cast<VPReductionPHIRecipe>(&R))
      ReductionPHIList.emplace_back(ReductionPhi);
  }
  stable_sort(ReductionPHIList, [this](const VPReductionPHIRecipe *R1,
                                       const VPReductionPHIRecipe *R2) {
    auto *IS1 = R1->getRecurrenceDescriptor().IntermediateStore;
    auto *IS2 = R2->getRecurrenceDescriptor().IntermediateStore;
 
    // If neither of the recipes has an intermediate store, keep the order the
    // same.
    if (!IS1 && !IS2)
      return false;
 
    // If only one of the recipes has an intermediate store, then move it
    // towards the beginning of the list.
    if (IS1 && !IS2)
      return true;
 
    if (!IS1 && IS2)
      return false;
 
    // If both recipes have an intermediate store, then the recipe with the
    // later store should be processed earlier. So it should go to the beginning
    // of the list.
    return DT->dominates(IS2, IS1);
  });
 
  for (VPReductionPHIRecipe *ReductionPhi : ReductionPHIList)
    fixReduction(ReductionPhi, State);
 
  for (VPRecipeBase &R : Header->phis()) {
    if (auto *FOR = dyn_cast<VPFirstOrderRecurrencePHIRecipe>(&R))
      fixFixedOrderRecurrence(FOR, State);
  }
}
 
void InnerLoopVectorizer::fixFixedOrderRecurrence(
    VPFirstOrderRecurrencePHIRecipe *PhiR, VPTransformState &State) {
  // This is the second phase of vectorizing first-order recurrences. An
  // overview of the transformation is described below. Suppose we have the
  // following loop.
  //
  //   for (int i = 0; i < n; ++i)
  //     b[i] = a[i] - a[i - 1];
  //
  // There is a first-order recurrence on "a". For this loop, the shorthand
  // scalar IR looks like:
  //
  //   scalar.ph:
  //     s_init = a[-1]
  //     br scalar.body
  //
  //   scalar.body:
  //     i = phi [0, scalar.ph], [i+1, scalar.body]
  //     s1 = phi [s_init, scalar.ph], [s2, scalar.body]
  //     s2 = a[i]
  //     b[i] = s2 - s1
  //     br cond, scalar.body, ...
  //
  // In this example, s1 is a recurrence because it's value depends on the
  // previous iteration. In the first phase of vectorization, we created a
  // vector phi v1 for s1. We now complete the vectorization and produce the
  // shorthand vector IR shown below (for VF = 4, UF = 1).
  //
  //   vector.ph:
  //     v_init = vector(..., ..., ..., a[-1])
  //     br vector.body
  //
  //   vector.body
  //     i = phi [0, vector.ph], [i+4, vector.body]
  //     v1 = phi [v_init, vector.ph], [v2, vector.body]
  //     v2 = a[i, i+1, i+2, i+3];
  //     v3 = vector(v1(3), v2(0, 1, 2))
  //     b[i, i+1, i+2, i+3] = v2 - v3
  //     br cond, vector.body, middle.block
  //
  //   middle.block:
  //     x = v2(3)
  //     br scalar.ph
  //
  //   scalar.ph:
  //     s_init = phi [x, middle.block], [a[-1], otherwise]
  //     br scalar.body
  //
  // After execution completes the vector loop, we extract the next value of
  // the recurrence (x) to use as the initial value in the scalar loop.
 
  // Extract the last vector element in the middle block. This will be the
  // initial value for the recurrence when jumping to the scalar loop.
  VPValue *PreviousDef = PhiR->getBackedgeValue();
  Value *Incoming = State.get(PreviousDef, UF - 1);
  auto *ExtractForScalar = Incoming;
  auto *IdxTy = Builder.getInt32Ty();
  Value *RuntimeVF = nullptr;
  if (VF.isVector()) {
    auto *One = ConstantInt::get(IdxTy, 1);
    Builder.SetInsertPoint(LoopMiddleBlock->getTerminator());
    RuntimeVF = getRuntimeVF(Builder, IdxTy, VF);
    auto *LastIdx = Builder.CreateSub(RuntimeVF, One);
    ExtractForScalar =
        Builder.CreateExtractElement(Incoming, LastIdx, "vector.recur.extract");
  }
 
  auto RecurSplice = cast<VPInstruction>(*PhiR->user_begin());
  assert(PhiR->getNumUsers() == 1 &&
         RecurSplice->getOpcode() ==
             VPInstruction::FirstOrderRecurrenceSplice &&
         "recurrence phi must have a single user: FirstOrderRecurrenceSplice");
  SmallVector<VPLiveOut *> LiveOuts;
  for (VPUser *U : RecurSplice->users())
    if (auto *LiveOut = dyn_cast<VPLiveOut>(U))
      LiveOuts.push_back(LiveOut);
 
  if (!LiveOuts.empty()) {
    // Extract the second last element in the middle block if the
    // Phi is used outside the loop. We need to extract the phi itself
    // and not the last element (the phi update in the current iteration). This
    // will be the value when jumping to the exit block from the
    // LoopMiddleBlock, when the scalar loop is not run at all.
    Value *ExtractForPhiUsedOutsideLoop = nullptr;
    if (VF.isVector()) {
      auto *Idx = Builder.CreateSub(RuntimeVF, ConstantInt::get(IdxTy, 2));
      ExtractForPhiUsedOutsideLoop = Builder.CreateExtractElement(
          Incoming, Idx, "vector.recur.extract.for.phi");
    } else {
      assert(UF > 1 && "VF and UF cannot both be 1");
      // When loop is unrolled without vectorizing, initialize
      // ExtractForPhiUsedOutsideLoop with the value just prior to unrolled
      // value of `Incoming`. This is analogous to the vectorized case above:
      // extracting the second last element when VF > 1.
      ExtractForPhiUsedOutsideLoop = State.get(PreviousDef, UF - 2);
    }
 
    for (VPLiveOut *LiveOut : LiveOuts) {
      assert(!Cost->requiresScalarEpilogue(VF.isVector()));
      PHINode *LCSSAPhi = LiveOut->getPhi();
      LCSSAPhi->addIncoming(ExtractForPhiUsedOutsideLoop, LoopMiddleBlock);
      State.Plan->removeLiveOut(LCSSAPhi);
    }
  }
 
  // Fix the initial value of the original recurrence in the scalar loop.
  Builder.SetInsertPoint(LoopScalarPreHeader, LoopScalarPreHeader->begin());
  PHINode *Phi = cast<PHINode>(PhiR->getUnderlyingValue());
  auto *Start = Builder.CreatePHI(Phi->getType(), 2, "scalar.recur.init");
  auto *ScalarInit = PhiR->getStartValue()->getLiveInIRValue();
  for (auto *BB : predecessors(LoopScalarPreHeader)) {
    auto *Incoming = BB == LoopMiddleBlock ? ExtractForScalar : ScalarInit;
    Start->addIncoming(Incoming, BB);
  }
 
  Phi->setIncomingValueForBlock(LoopScalarPreHeader, Start);
  Phi->setName("scalar.recur");
}
 
void InnerLoopVectorizer::fixReduction(VPReductionPHIRecipe *PhiR,
                                       VPTransformState &State) {
  PHINode *OrigPhi = cast<PHINode>(PhiR->getUnderlyingValue());
  // Get it's reduction variable descriptor.
  assert(Legal->isReductionVariable(OrigPhi) &&
         "Unable to find the reduction variable");
  const RecurrenceDescriptor &RdxDesc = PhiR->getRecurrenceDescriptor();
 
  RecurKind RK = RdxDesc.getRecurrenceKind();
  TrackingVH<Value> ReductionStartValue = RdxDesc.getRecurrenceStartValue();
  Instruction *LoopExitInst = RdxDesc.getLoopExitInstr();
  if (auto *I = dyn_cast<Instruction>(&*ReductionStartValue))
    State.setDebugLocFrom(I->getDebugLoc());
 
  VPValue *LoopExitInstDef = PhiR->getBackedgeValue();
  // This is the vector-clone of the value that leaves the loop.
  Type *VecTy = State.get(LoopExitInstDef, 0)->getType();
 
  // 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,
                         LoopMiddleBlock->getFirstInsertionPt());
 
  State.setDebugLocFrom(LoopExitInst->getDebugLoc());
 
  Type *PhiTy = OrigPhi->getType();
 
  VPBasicBlock *LatchVPBB =
      PhiR->getParent()->getEnclosingLoopRegion()->getExitingBasicBlock();
  BasicBlock *VectorLoopLatch = State.CFG.VPBB2IRBB[LatchVPBB];
  // If tail is folded by masking, the vector value to leave the loop should be
  // a Select choosing between the vectorized LoopExitInst and vectorized Phi,
  // instead of the former. For an inloop reduction the reduction will already
  // be predicated, and does not need to be handled here.
  if (Cost->foldTailByMasking() && !PhiR->isInLoop()) {
    for (unsigned Part = 0; Part < UF; ++Part) {
      Value *VecLoopExitInst = State.get(LoopExitInstDef, Part);
      SelectInst *Sel = nullptr;
      for (User *U : VecLoopExitInst->users()) {
        if (isa<SelectInst>(U)) {
          assert((!Sel || U == Sel) &&
                 "Reduction exit feeding two different selects");
          Sel = cast<SelectInst>(U);
        } else
          assert(isa<PHINode>(U) && "Reduction exit must feed Phi's or select");
      }
      assert(Sel && "Reduction exit feeds no select");
      State.reset(LoopExitInstDef, Sel, Part);
 
      // If the target can create a predicated operator for the reduction at no
      // extra cost in the loop (for example a predicated vadd), it can be
      // cheaper for the select to remain in the loop than be sunk out of it,
      // and so use the select value for the phi instead of the old
      // LoopExitValue.
      if (PreferPredicatedReductionSelect ||
          TTI->preferPredicatedReductionSelect(
              RdxDesc.getOpcode(), PhiTy,
              TargetTransformInfo::ReductionFlags())) {
        auto *VecRdxPhi =
            cast<PHINode>(State.get(PhiR, Part));
        VecRdxPhi->setIncomingValueForBlock(VectorLoopLatch, Sel);
      }
    }
  }
 
  // 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.isVector() && PhiTy != RdxDesc.getRecurrenceType()) {
    assert(!PhiR->isInLoop() && "Unexpected truncated inloop reduction!");
    Type *RdxVecTy = VectorType::get(RdxDesc.getRecurrenceType(), VF);
    Builder.SetInsertPoint(VectorLoopLatch->getTerminator());
    VectorParts RdxParts(UF);
    for (unsigned Part = 0; Part < UF; ++Part) {
      RdxParts[Part] = State.get(LoopExitInstDef, Part);
      Value *Trunc = Builder.CreateTrunc(RdxParts[Part], RdxVecTy);
      Value *Extnd = RdxDesc.isSigned() ? Builder.CreateSExt(Trunc, VecTy)
                                        : Builder.CreateZExt(Trunc, VecTy);
      for (User *U : llvm::make_early_inc_range(RdxParts[Part]->users()))
        if (U != Trunc) {
          U->replaceUsesOfWith(RdxParts[Part], Extnd);
          RdxParts[Part] = Extnd;
        }
    }
    Builder.SetInsertPoint(LoopMiddleBlock,
                           LoopMiddleBlock->getFirstInsertionPt());
    for (unsigned Part = 0; Part < UF; ++Part) {
      RdxParts[Part] = Builder.CreateTrunc(RdxParts[Part], RdxVecTy);
      State.reset(LoopExitInstDef, RdxParts[Part], Part);
    }
  }
 
  // Reduce all of the unrolled parts into a single vector.
  Value *ReducedPartRdx = State.get(LoopExitInstDef, 0);
  unsigned Op = RecurrenceDescriptor::getOpcode(RK);
 
  // The middle block terminator has already been assigned a DebugLoc here (the
  // OrigLoop's single latch terminator). We want the whole middle block to
  // appear to execute on this line because: (a) it is all compiler generated,
  // (b) these instructions are always executed after evaluating the latch
  // conditional branch, and (c) other passes may add new predecessors which
  // terminate on this line. This is the easiest way to ensure we don't
  // accidentally cause an extra step back into the loop while debugging.
  State.setDebugLocFrom(LoopMiddleBlock->getTerminator()->getDebugLoc());
  if (PhiR->isOrdered())
    ReducedPartRdx = State.get(LoopExitInstDef, UF - 1);
  else {
    // Floating-point operations should have some FMF to enable the reduction.
    IRBuilderBase::FastMathFlagGuard FMFG(Builder);
    Builder.setFastMathFlags(RdxDesc.getFastMathFlags());
    for (unsigned Part = 1; Part < UF; ++Part) {
      Value *RdxPart = State.get(LoopExitInstDef, Part);
      if (Op != Instruction::ICmp && Op != Instruction::FCmp)
        ReducedPartRdx = Builder.CreateBinOp(
            (Instruction::BinaryOps)Op, RdxPart, ReducedPartRdx, "bin.rdx");
      else if (RecurrenceDescriptor::isAnyOfRecurrenceKind(RK))
        ReducedPartRdx = createAnyOfOp(Builder, ReductionStartValue, RK,
                                       ReducedPartRdx, RdxPart);
      else
        ReducedPartRdx = createMinMaxOp(Builder, RK, ReducedPartRdx, RdxPart);
    }
  }
 
  // Create the reduction after the loop. Note that inloop reductions create the
  // target reduction in the loop using a Reduction recipe.
  if (VF.isVector() && !PhiR->isInLoop()) {
    ReducedPartRdx =
        createTargetReduction(Builder, RdxDesc, ReducedPartRdx, OrigPhi);
    // 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 (PhiTy != RdxDesc.getRecurrenceType())
      ReducedPartRdx = RdxDesc.isSigned()
                           ? Builder.CreateSExt(ReducedPartRdx, PhiTy)
                           : Builder.CreateZExt(ReducedPartRdx, PhiTy);
  }
 
  PHINode *ResumePhi =
      dyn_cast<PHINode>(PhiR->getStartValue()->getUnderlyingValue());
 
  // Create a phi node that merges control-flow from the backedge-taken check
  // block and the middle block.
  PHINode *BCBlockPhi = PHINode::Create(PhiTy, 2, "bc.merge.rdx",
                                        LoopScalarPreHeader->getTerminator());
 
  // If we are fixing reductions in the epilogue loop then we should already
  // have created a bc.merge.rdx Phi after the main vector body. Ensure that
  // we carry over the incoming values correctly.
  for (auto *Incoming : predecessors(LoopScalarPreHeader)) {
    if (Incoming == LoopMiddleBlock)
      BCBlockPhi->addIncoming(ReducedPartRdx, Incoming);
    else if (ResumePhi && llvm::is_contained(ResumePhi->blocks(), Incoming))
      BCBlockPhi->addIncoming(ResumePhi->getIncomingValueForBlock(Incoming),
                              Incoming);
    else
      BCBlockPhi->addIncoming(ReductionStartValue, Incoming);
  }
 
  // Set the resume value for this reduction
  ReductionResumeValues.insert({&RdxDesc, BCBlockPhi});
 
  // If there were stores of the reduction value to a uniform memory address
  // inside the loop, create the final store here.
  if (StoreInst *SI = RdxDesc.IntermediateStore) {
    StoreInst *NewSI =
        Builder.CreateAlignedStore(ReducedPartRdx, SI->getPointerOperand(),
                                   SI->getAlign());
    propagateMetadata(NewSI, SI);
 
    // If the reduction value is used in other places,
    // then let the code below create PHI's for that.
  }
 
  // 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 PHI nodes
  // in the exit blocks.  See comment on analogous loop in
  // fixFixedOrderRecurrence for a more complete explaination of the logic.
  if (!Cost->requiresScalarEpilogue(VF.isVector()))
    for (PHINode &LCSSAPhi : LoopExitBlock->phis())
      if (llvm::is_contained(LCSSAPhi.incoming_values(), LoopExitInst)) {
        LCSSAPhi.addIncoming(ReducedPartRdx, LoopMiddleBlock);
        State.Plan->removeLiveOut(&LCSSAPhi);
      }
 
  // Fix the scalar loop reduction variable with the incoming reduction sum
  // from the vector body and from the backedge value.
  int IncomingEdgeBlockIdx =
      OrigPhi->getBasicBlockIndex(OrigLoop->getLoopLatch());
  assert(IncomingEdgeBlockIdx >= 0 && "Invalid block index");
  // Pick the other block.
  int SelfEdgeBlockIdx = (IncomingEdgeBlockIdx ? 0 : 1);
  OrigPhi->setIncomingValue(SelfEdgeBlockIdx, BCBlockPhi);
  OrigPhi->setIncomingValue(IncomingEdgeBlockIdx, LoopExitInst);
}
 
void InnerLoopVectorizer::sinkScalarOperands(Instruction *PredInst) {
  // The basic block and loop containing the predicated instruction.
  auto *PredBB = PredInst->getParent();
  auto *VectorLoop = LI->getLoopFor(PredBB);
 
  // Initialize a worklist with the operands of the predicated instruction.
  SetVector<Value *> Worklist(PredInst->op_begin(), PredInst->op_end());
 
  // Holds instructions that we need to analyze again. An instruction may be
  // reanalyzed if we don't yet know if we can sink it or not.
  SmallVector<Instruction *, 8> InstsToReanalyze;
 
  // Returns true if a given use occurs in the predicated block. Phi nodes use
  // their operands in their corresponding predecessor blocks.
  auto isBlockOfUsePredicated = [&](Use &U) -> bool {
    auto *I = cast<Instruction>(U.getUser());
    BasicBlock *BB = I->getParent();
    if (auto *Phi = dyn_cast<PHINode>(I))
      BB = Phi->getIncomingBlock(
          PHINode::getIncomingValueNumForOperand(U.getOperandNo()));
    return BB == PredBB;
  };
 
  // Iteratively sink the scalarized operands of the predicated instruction
  // into the block we created for it. When an instruction is sunk, it's
  // operands are then added to the worklist. The algorithm ends after one pass
  // through the worklist doesn't sink a single instruction.
  bool Changed;
  do {
    // Add the instructions that need to be reanalyzed to the worklist, and
    // reset the changed indicator.
    Worklist.insert(InstsToReanalyze.begin(), InstsToReanalyze.end());
    InstsToReanalyze.clear();
    Changed = false;
 
    while (!Worklist.empty()) {
      auto *I = dyn_cast<Instruction>(Worklist.pop_back_val());
 
      // We can't sink an instruction if it is a phi node, is not in the loop,
      // may have side effects or may read from memory.
      // TODO Could dor more granular checking to allow sinking a load past non-store instructions.
      if (!I || isa<PHINode>(I) || !VectorLoop->contains(I) ||
          I->mayHaveSideEffects() || I->mayReadFromMemory())
          continue;
 
      // If the instruction is already in PredBB, check if we can sink its
      // operands. In that case, VPlan's sinkScalarOperands() succeeded in
      // sinking the scalar instruction I, hence it appears in PredBB; but it
      // may have failed to sink I's operands (recursively), which we try
      // (again) here.
      if (I->getParent() == PredBB) {
        Worklist.insert(I->op_begin(), I->op_end());
        continue;
      }
 
      // It's legal to sink the instruction if all its uses occur in the
      // predicated block. Otherwise, there's nothing to do yet, and we may
      // need to reanalyze the instruction.
      if (!llvm::all_of(I->uses(), isBlockOfUsePredicated)) {
        InstsToReanalyze.push_back(I);
        continue;
      }
 
      // Move the instruction to the beginning of the predicated block, and add
      // it's operands to the worklist.
      I->moveBefore(&*PredBB->getFirstInsertionPt());
      Worklist.insert(I->op_begin(), I->op_end());
 
      // The sinking may have enabled other instructions to be sunk, so we will
      // need to iterate.
      Changed = true;
    }
  } while (Changed);
}
 
void InnerLoopVectorizer::fixNonInductionPHIs(VPlan &Plan,
                                              VPTransformState &State) {
  auto Iter = vp_depth_first_deep(Plan.getEntry());
  for (VPBasicBlock *VPBB : VPBlockUtils::blocksOnly<VPBasicBlock>(Iter)) {
    for (VPRecipeBase &P : VPBB->phis()) {
      VPWidenPHIRecipe *VPPhi = dyn_cast<VPWidenPHIRecipe>(&P);
      if (!VPPhi)
        continue;
      PHINode *NewPhi = cast<PHINode>(State.get(VPPhi, 0));
      // Make sure the builder has a valid insert point.
      Builder.SetInsertPoint(NewPhi);
      for (unsigned i = 0; i < VPPhi->getNumOperands(); ++i) {
        VPValue *Inc = VPPhi->getIncomingValue(i);
        VPBasicBlock *VPBB = VPPhi->getIncomingBlock(i);
        NewPhi->addIncoming(State.get(Inc, 0), State.CFG.VPBB2IRBB[VPBB]);
      }
    }
  }
}
 
bool InnerLoopVectorizer::useOrderedReductions(
    const RecurrenceDescriptor &RdxDesc) {
  return Cost->useOrderedReductions(RdxDesc);
}
 
void LoopVectorizationCostModel::collectLoopScalars(ElementCount VF) {
  // We should not collect Scalars more than once per VF. Right now, this
  // function is called from collectUniformsAndScalars(), which already does
  // this check. Collecting Scalars for VF=1 does not make any sense.
  assert(VF.isVector() && !Scalars.contains(VF) &&
         "This function should not be visited twice for the same VF");
 
  // This avoids any chances of creating a REPLICATE recipe during planning
  // since that would result in generation of scalarized code during execution,
  // which is not supported for scalable vectors.
  if (VF.isScalable()) {
    Scalars[VF].insert(Uniforms[VF].begin(), Uniforms[VF].end());
    return;
  }
 
  SmallSetVector<Instruction *, 8> Worklist;
 
  // These sets are used to seed the analysis with pointers used by memory
  // accesses that will remain scalar.
  SmallSetVector<Instruction *, 8> ScalarPtrs;
  SmallPtrSet<Instruction *, 8> PossibleNonScalarPtrs;
  auto *Latch = TheLoop->getLoopLatch();
 
  // A helper that returns true if the use of Ptr by MemAccess will be scalar.
  // The pointer operands of loads and stores will be scalar as long as the
  // memory access is not a gather or scatter operation. The value operand of a
  // store will remain scalar if the store is scalarized.
  auto isScalarUse = [&](Instruction *MemAccess, Value *Ptr) {
    InstWidening WideningDecision = getWideningDecision(MemAccess, VF);
    assert(WideningDecision != CM_Unknown &&
           "Widening decision should be ready at this moment");
    if (auto *Store = dyn_cast<StoreInst>(MemAccess))
      if (Ptr == Store->getValueOperand())
        return WideningDecision == CM_Scalarize;
    assert(Ptr == getLoadStorePointerOperand(MemAccess) &&
           "Ptr is neither a value or pointer operand");
    return WideningDecision != CM_GatherScatter;
  };
 
  // A helper that returns true if the given value is a bitcast or
  // getelementptr instruction contained in the loop.
  auto isLoopVaryingBitCastOrGEP = [&](Value *V) {
    return ((isa<BitCastInst>(V) && V->getType()->isPointerTy()) ||
            isa<GetElementPtrInst>(V)) &&
           !TheLoop->isLoopInvariant(V);
  };
 
  // A helper that evaluates a memory access's use of a pointer. If the use will
  // be a scalar use and the pointer is only used by memory accesses, we place
  // the pointer in ScalarPtrs. Otherwise, the pointer is placed in
  // PossibleNonScalarPtrs.
  auto evaluatePtrUse = [&](Instruction *MemAccess, Value *Ptr) {
    // We only care about bitcast and getelementptr instructions contained in
    // the loop.
    if (!isLoopVaryingBitCastOrGEP(Ptr))
      return;
 
    // If the pointer has already been identified as scalar (e.g., if it was
    // also identified as uniform), there's nothing to do.
    auto *I = cast<Instruction>(Ptr);
    if (Worklist.count(I))
      return;
 
    // If the use of the pointer will be a scalar use, and all users of the
    // pointer are memory accesses, place the pointer in ScalarPtrs. Otherwise,
    // place the pointer in PossibleNonScalarPtrs.
    if (isScalarUse(MemAccess, Ptr) && llvm::all_of(I->users(), [&](User *U) {
          return isa<LoadInst>(U) || isa<StoreInst>(U);
        }))
      ScalarPtrs.insert(I);
    else
      PossibleNonScalarPtrs.insert(I);
  };
 
  // We seed the scalars analysis with three classes of instructions: (1)
  // instructions marked uniform-after-vectorization and (2) bitcast,
  // getelementptr and (pointer) phi instructions used by memory accesses
  // requiring a scalar use.
  //
  // (1) Add to the worklist all instructions that have been identified as
  // uniform-after-vectorization.
  Worklist.insert(Uniforms[VF].begin(), Uniforms[VF].end());
 
  // (2) Add to the worklist all bitcast and getelementptr instructions used by
  // memory accesses requiring a scalar use. The pointer operands of loads and
  // stores will be scalar as long as the memory accesses is not a gather or
  // scatter operation. The value operand of a store will remain scalar if the
  // store is scalarized.
  for (auto *BB : TheLoop->blocks())
    for (auto &I : *BB) {
      if (auto *Load = dyn_cast<LoadInst>(&I)) {
        evaluatePtrUse(Load, Load->getPointerOperand());
      } else if (auto *Store = dyn_cast<StoreInst>(&I)) {
        evaluatePtrUse(Store, Store->getPointerOperand());
        evaluatePtrUse(Store, Store->getValueOperand());
      }
    }
  for (auto *I : ScalarPtrs)
    if (!PossibleNonScalarPtrs.count(I)) {
      LLVM_DEBUG(dbgs() << "LV: Found scalar instruction: " << *I << "\n");
      Worklist.insert(I);
    }
 
  // Insert the forced scalars.
  // FIXME: Currently VPWidenPHIRecipe() often creates a dead vector
  // induction variable when the PHI user is scalarized.
  auto ForcedScalar = ForcedScalars.find(VF);
  if (ForcedScalar != ForcedScalars.end())
    for (auto *I : ForcedScalar->second) {
      LLVM_DEBUG(dbgs() << "LV: Found (forced) scalar instruction: " << *I << "\n");
      Worklist.insert(I);
    }
 
  // Expand the worklist by looking through any bitcasts and getelementptr
  // instructions we've already identified as scalar. This is similar to the
  // expansion step in collectLoopUniforms(); however, here we're only
  // expanding to include additional bitcasts and getelementptr instructions.
  unsigned Idx = 0;
  while (Idx != Worklist.size()) {
    Instruction *Dst = Worklist[Idx++];
    if (!isLoopVaryingBitCastOrGEP(Dst->getOperand(0)))
      continue;
    auto *Src = cast<Instruction>(Dst->getOperand(0));
    if (llvm::all_of(Src->users(), [&](User *U) -> bool {
          auto *J = cast<Instruction>(U);
          return !TheLoop->contains(J) || Worklist.count(J) ||
                 ((isa<LoadInst>(J) || isa<StoreInst>(J)) &&
                  isScalarUse(J, Src));
        })) {
      Worklist.insert(Src);
      LLVM_DEBUG(dbgs() << "LV: Found scalar instruction: " << *Src << "\n");
    }
  }
 
  // An induction variable will remain scalar if all users of the induction
  // variable and induction variable update remain scalar.
  for (const auto &Induction : Legal->getInductionVars()) {
    auto *Ind = Induction.first;
    auto *IndUpdate = cast<Instruction>(Ind->getIncomingValueForBlock(Latch));
 
    // If tail-folding is applied, the primary induction variable will be used
    // to feed a vector compare.
    if (Ind == Legal->getPrimaryInduction() && foldTailByMasking())
      continue;
 
    // Returns true if \p Indvar is a pointer induction that is used directly by
    // load/store instruction \p I.
    auto IsDirectLoadStoreFromPtrIndvar = [&](Instruction *Indvar,
                                              Instruction *I) {
      return Induction.second.getKind() ==
                 InductionDescriptor::IK_PtrInduction &&
             (isa<LoadInst>(I) || isa<StoreInst>(I)) &&
             Indvar == getLoadStorePointerOperand(I) && isScalarUse(I, Indvar);
    };
 
    // Determine if all users of the induction variable are scalar after
    // vectorization.
    auto ScalarInd = llvm::all_of(Ind->users(), [&](User *U) -> bool {
      auto *I = cast<Instruction>(U);
      return I == IndUpdate || !TheLoop->contains(I) || Worklist.count(I) ||
             IsDirectLoadStoreFromPtrIndvar(Ind, I);
    });
    if (!ScalarInd)
      continue;
 
    // Determine if all users of the induction variable update instruction are
    // scalar after vectorization.
    auto ScalarIndUpdate =
        llvm::all_of(IndUpdate->users(), [&](User *U) -> bool {
          auto *I = cast<Instruction>(U);
          return I == Ind || !TheLoop->contains(I) || Worklist.count(I) ||
                 IsDirectLoadStoreFromPtrIndvar(IndUpdate, I);
        });
    if (!ScalarIndUpdate)
      continue;
 
    // The induction variable and its update instruction will remain scalar.
    Worklist.insert(Ind);
    Worklist.insert(IndUpdate);
    LLVM_DEBUG(dbgs() << "LV: Found scalar instruction: " << *Ind << "\n");
    LLVM_DEBUG(dbgs() << "LV: Found scalar instruction: " << *IndUpdate
                      << "\n");
  }
 
  Scalars[VF].insert(Worklist.begin(), Worklist.end());
}
 
bool LoopVectorizationCostModel::isScalarWithPredication(
    Instruction *I, ElementCount VF) const {
  if (!isPredicatedInst(I))
    return false;
 
  // Do we have a non-scalar lowering for this predicated
  // instruction? No - it is scalar with predication.
  switch(I->getOpcode()) {
  default:
    return true;
  case Instruction::Call:
    return !VFDatabase::hasMaskedVariant(*(cast<CallInst>(I)), VF);
  case Instruction::Load:
  case Instruction::Store: {
    auto *Ptr = getLoadStorePointerOperand(I);
    auto *Ty = getLoadStoreType(I);
    Type *VTy = Ty;
    if (VF.isVector())
      VTy = VectorType::get(Ty, VF);
    const Align Alignment = getLoadStoreAlignment(I);
    return isa<LoadInst>(I) ? !(isLegalMaskedLoad(Ty, Ptr, Alignment) ||
                                TTI.isLegalMaskedGather(VTy, Alignment))
                            : !(isLegalMaskedStore(Ty, Ptr, Alignment) ||
                                TTI.isLegalMaskedScatter(VTy, Alignment));
  }
  case Instruction::UDiv:
  case Instruction::SDiv:
  case Instruction::SRem:
  case Instruction::URem: {
    // We have the option to use the safe-divisor idiom to avoid predication.
    // The cost based decision here will always select safe-divisor for
    // scalable vectors as scalarization isn't legal.
    const auto [ScalarCost, SafeDivisorCost] = getDivRemSpeculationCost(I, VF);
    return isDivRemScalarWithPredication(ScalarCost, SafeDivisorCost);
  }
  }
}
 
bool LoopVectorizationCostModel::isPredicatedInst(Instruction *I) const {
  if (!blockNeedsPredicationForAnyReason(I->getParent()))
    return false;
 
  // Can we prove this instruction is safe to unconditionally execute?
  // If not, we must use some form of predication.
  switch(I->getOpcode()) {
  default:
    return false;
  case Instruction::Load:
  case Instruction::Store: {
    if (!Legal->isMaskRequired(I))
      return false;
    // When we know the load's address is loop invariant and the instruction
    // in the original scalar loop was unconditionally executed then we
    // don't need to mark it as a predicated instruction. Tail folding may
    // introduce additional predication, but we're guaranteed to always have
    // at least one active lane.  We call Legal->blockNeedsPredication here
    // because it doesn't query tail-folding.  For stores, we need to prove
    // both speculation safety (which follows from the same argument as loads),
    // but also must prove the value being stored is correct.  The easiest
    // form of the later is to require that all values stored are the same.
    if (Legal->isInvariant(getLoadStorePointerOperand(I)) &&
        (isa<LoadInst>(I) ||
         (isa<StoreInst>(I) &&
          TheLoop->isLoopInvariant(cast<StoreInst>(I)->getValueOperand()))) &&
        !Legal->blockNeedsPredication(I->getParent()))
      return false;
    return true;
  }
  case Instruction::UDiv:
  case Instruction::SDiv:
  case Instruction::SRem:
  case Instruction::URem:
    // TODO: We can use the loop-preheader as context point here and get
    // context sensitive reasoning
    return !isSafeToSpeculativelyExecute(I);
  case Instruction::Call:
    return Legal->isMaskRequired(I);
  }
}
 
std::pair<InstructionCost, InstructionCost>
LoopVectorizationCostModel::getDivRemSpeculationCost(Instruction *I,
                                                    ElementCount VF) const {
  assert(I->getOpcode() == Instruction::UDiv ||
         I->getOpcode() == Instruction::SDiv ||
         I->getOpcode() == Instruction::SRem ||
         I->getOpcode() == Instruction::URem);
  assert(!isSafeToSpeculativelyExecute(I));
 
  const TTI::TargetCostKind CostKind = TTI::TCK_RecipThroughput;
 
  // Scalarization isn't legal for scalable vector types
  InstructionCost ScalarizationCost = InstructionCost::getInvalid();
  if (!VF.isScalable()) {
    // Get the scalarization cost and scale this amount by the probability of
    // executing the predicated block. If the instruction is not predicated,
    // we fall through to the next case.
    ScalarizationCost = 0;
 
    // These instructions have a non-void type, so account for the phi nodes
    // that we will create. This cost is likely to be zero. The phi node
    // cost, if any, should be scaled by the block probability because it
    // models a copy at the end of each predicated block.
    ScalarizationCost += VF.getKnownMinValue() *
      TTI.getCFInstrCost(Instruction::PHI, CostKind);
 
    // The cost of the non-predicated instruction.
    ScalarizationCost += VF.getKnownMinValue() *
      TTI.getArithmeticInstrCost(I->getOpcode(), I->getType(), CostKind);
 
    // The cost of insertelement and extractelement instructions needed for
    // scalarization.
    ScalarizationCost += getScalarizationOverhead(I, VF, CostKind);
 
    // Scale the cost by the probability of executing the predicated blocks.
    // This assumes the predicated block for each vector lane is equally
    // likely.
    ScalarizationCost = ScalarizationCost / getReciprocalPredBlockProb();
  }
  InstructionCost SafeDivisorCost = 0;
 
  auto *VecTy = ToVectorTy(I->getType(), VF);
 
  // The cost of the select guard to ensure all lanes are well defined
  // after we speculate above any internal control flow.
  SafeDivisorCost += TTI.getCmpSelInstrCost(
    Instruction::Select, VecTy,
    ToVectorTy(Type::getInt1Ty(I->getContext()), VF),
    CmpInst::BAD_ICMP_PREDICATE, CostKind);
 
  // Certain instructions can be cheaper to vectorize if they have a constant
  // second vector operand. One example of this are shifts on x86.
  Value *Op2 = I->getOperand(1);
  auto Op2Info = TTI.getOperandInfo(Op2);
  if (Op2Info.Kind == TargetTransformInfo::OK_AnyValue &&
      Legal->isInvariant(Op2))
    Op2Info.Kind = TargetTransformInfo::OK_UniformValue;
 
  SmallVector<const Value *, 4> Operands(I->operand_values());
  SafeDivisorCost += TTI.getArithmeticInstrCost(
    I->getOpcode(), VecTy, CostKind,
    {TargetTransformInfo::OK_AnyValue, TargetTransformInfo::OP_None},
    Op2Info, Operands, I);
  return {ScalarizationCost, SafeDivisorCost};
}
 
bool LoopVectorizationCostModel::interleavedAccessCanBeWidened(
    Instruction *I, ElementCount VF) {
  assert(isAccessInterleaved(I) && "Expecting interleaved access.");
  assert(getWideningDecision(I, VF) == CM_Unknown &&
         "Decision should not be set yet.");
  auto *Group = getInterleavedAccessGroup(I);
  assert(Group && "Must have a group.");
 
  // If the instruction's allocated size doesn't equal it's type size, it
  // requires padding and will be scalarized.
  auto &DL = I->getModule()->getDataLayout();
  auto *ScalarTy = getLoadStoreType(I);
  if (hasIrregularType(ScalarTy, DL))
    return false;
 
  // If the group involves a non-integral pointer, we may not be able to
  // losslessly cast all values to a common type.
  unsigned InterleaveFactor = Group->getFactor();
  bool ScalarNI = DL.isNonIntegralPointerType(ScalarTy);
  for (unsigned i = 0; i < InterleaveFactor; i++) {
    Instruction *Member = Group->getMember(i);
    if (!Member)
      continue;
    auto *MemberTy = getLoadStoreType(Member);
    bool MemberNI = DL.isNonIntegralPointerType(MemberTy);
    // Don't coerce non-integral pointers to integers or vice versa.
    if (MemberNI != ScalarNI) {
      // TODO: Consider adding special nullptr value case here
      return false;
    } else if (MemberNI && ScalarNI &&
               ScalarTy->getPointerAddressSpace() !=
               MemberTy->getPointerAddressSpace()) {
      return false;
    }
  }
 
  // Check if masking is required.
  // A Group may need masking for one of two reasons: it resides in a block that
  // needs predication, or it was decided to use masking to deal with gaps
  // (either a gap at the end of a load-access that may result in a speculative
  // load, or any gaps in a store-access).
  bool PredicatedAccessRequiresMasking =
      blockNeedsPredicationForAnyReason(I->getParent()) &&
      Legal->isMaskRequired(I);
  bool LoadAccessWithGapsRequiresEpilogMasking =
      isa<LoadInst>(I) && Group->requiresScalarEpilogue() &&
      !isScalarEpilogueAllowed();
  bool StoreAccessWithGapsRequiresMasking =
      isa<StoreInst>(I) && (Group->getNumMembers() < Group->getFactor());
  if (!PredicatedAccessRequiresMasking &&
      !LoadAccessWithGapsRequiresEpilogMasking &&
      !StoreAccessWithGapsRequiresMasking)
    return true;
 
  // If masked interleaving is required, we expect that the user/target had
  // enabled it, because otherwise it either wouldn't have been created or
  // it should have been invalidated by the CostModel.
  assert(useMaskedInterleavedAccesses(TTI) &&
         "Masked interleave-groups for predicated accesses are not enabled.");
 
  if (Group->isReverse())
    return false;
 
  auto *Ty = getLoadStoreType(I);
  const Align Alignment = getLoadStoreAlignment(I);
  return isa<LoadInst>(I) ? TTI.isLegalMaskedLoad(Ty, Alignment)
                          : TTI.isLegalMaskedStore(Ty, Alignment);
}
 
bool LoopVectorizationCostModel::memoryInstructionCanBeWidened(
    Instruction *I, ElementCount VF) {
  // Get and ensure we have a valid memory instruction.
  assert((isa<LoadInst, StoreInst>(I)) && "Invalid memory instruction");
 
  auto *Ptr = getLoadStorePointerOperand(I);
  auto *ScalarTy = getLoadStoreType(I);
 
  // In order to be widened, the pointer should be consecutive, first of all.
  if (!Legal->isConsecutivePtr(ScalarTy, Ptr))
    return false;
 
  // If the instruction is a store located in a predicated block, it will be
  // scalarized.
  if (isScalarWithPredication(I, VF))
    return false;
 
  // If the instruction's allocated size doesn't equal it's type size, it
  // requires padding and will be scalarized.
  auto &DL = I->getModule()->getDataLayout();
  if (hasIrregularType(ScalarTy, DL))
    return false;
 
  return true;
}
 
void LoopVectorizationCostModel::collectLoopUniforms(ElementCount VF) {
  // We should not collect Uniforms more than once per VF. Right now,
  // this function is called from collectUniformsAndScalars(), which
  // already does this check. Collecting Uniforms for VF=1 does not make any</