BasicTTIImpl.h

//===- BasicTTIImpl.h -------------------------------------------*- C++ -*-===//
//
// 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
//
//===----------------------------------------------------------------------===//
//
/// \file
/// This file provides a helper that implements much of the TTI interface in
/// terms of the target-independent code generator and TargetLowering
/// interfaces.
//
//===----------------------------------------------------------------------===//

#ifndef LLVM_CODEGEN_BASICTTIIMPL_H
#define LLVM_CODEGEN_BASICTTIIMPL_H

#include "llvm/ADT/APInt.h"
#include "llvm/ADT/ArrayRef.h"
#include "llvm/ADT/BitVector.h"
#include "llvm/ADT/SmallPtrSet.h"
#include "llvm/ADT/SmallVector.h"
#include "llvm/Analysis/LoopInfo.h"
#include "llvm/Analysis/TargetTransformInfo.h"
#include "llvm/Analysis/TargetTransformInfoImpl.h"
#include "llvm/CodeGen/ISDOpcodes.h"
#include "llvm/CodeGen/TargetLowering.h"
#include "llvm/CodeGen/TargetSubtargetInfo.h"
#include "llvm/CodeGen/ValueTypes.h"
#include "llvm/IR/BasicBlock.h"
#include "llvm/IR/Constant.h"
#include "llvm/IR/Constants.h"
#include "llvm/IR/DataLayout.h"
#include "llvm/IR/DerivedTypes.h"
#include "llvm/IR/InstrTypes.h"
#include "llvm/IR/Instruction.h"
#include "llvm/IR/Instructions.h"
#include "llvm/IR/Intrinsics.h"
#include "llvm/IR/Operator.h"
#include "llvm/IR/Type.h"
#include "llvm/IR/Value.h"
#include "llvm/MC/MCSchedule.h"
#include "llvm/Support/Casting.h"
#include "llvm/Support/CommandLine.h"
#include "llvm/Support/ErrorHandling.h"
#include "llvm/Support/MachineValueType.h"
#include "llvm/Support/MathExtras.h"
#include <algorithm>
#include <cassert>
#include <cstdint>
#include <limits>
#include <utility>

namespace llvm {

class Function;
class GlobalValue;
class LLVMContext;
class ScalarEvolution;
class SCEV;
class TargetMachine;

extern cl::opt<unsigned> PartialUnrollingThreshold;

/// Base class which can be used to help build a TTI implementation.
///
/// This class provides as much implementation of the TTI interface as is
/// possible using the target independent parts of the code generator.
///
/// In order to subclass it, your class must implement a getST() method to
/// return the subtarget, and a getTLI() method to return the target lowering.
/// We need these methods implemented in the derived class so that this class
/// doesn't have to duplicate storage for them.
template <typename T>
class BasicTTIImplBase : public TargetTransformInfoImplCRTPBase<T> {
private:
  using BaseT = TargetTransformInfoImplCRTPBase<T>;
  using TTI = TargetTransformInfo;

  /// Estimate a cost of Broadcast as an extract and sequence of insert
  /// operations.
  unsigned getBroadcastShuffleOverhead(FixedVectorType *VTy) {
    unsigned Cost = 0;
    // Broadcast cost is equal to the cost of extracting the zero'th element
    // plus the cost of inserting it into every element of the result vector.
    Cost += static_cast<T *>(this)->getVectorInstrCost(
        Instruction::ExtractElement, VTy, 0);

    for (int i = 0, e = VTy->getNumElements(); i < e; ++i) {
      Cost += static_cast<T *>(this)->getVectorInstrCost(
          Instruction::InsertElement, VTy, i);
    }
    return Cost;
  }

  /// Estimate a cost of shuffle as a sequence of extract and insert
  /// operations.
  unsigned getPermuteShuffleOverhead(FixedVectorType *VTy) {
    unsigned Cost = 0;
    // Shuffle cost is equal to the cost of extracting element from its argument
    // plus the cost of inserting them onto the result vector.

    // e.g. <4 x float> has a mask of <0,5,2,7> i.e we need to extract from
    // index 0 of first vector, index 1 of second vector,index 2 of first
    // vector and finally index 3 of second vector and insert them at index
    // <0,1,2,3> of result vector.
    for (int i = 0, e = VTy->getNumElements(); i < e; ++i) {
      Cost += static_cast<T *>(this)->getVectorInstrCost(
          Instruction::InsertElement, VTy, i);
      Cost += static_cast<T *>(this)->getVectorInstrCost(
          Instruction::ExtractElement, VTy, i);
    }
    return Cost;
  }

  /// Estimate a cost of subvector extraction as a sequence of extract and
  /// insert operations.
  unsigned getExtractSubvectorOverhead(FixedVectorType *VTy, int Index,
                                       FixedVectorType *SubVTy) {
    assert(VTy && SubVTy &&
           "Can only extract subvectors from vectors");
    int NumSubElts = SubVTy->getNumElements();
    assert((Index + NumSubElts) <= (int)VTy->getNumElements() &&
           "SK_ExtractSubvector index out of range");

    unsigned Cost = 0;
    // Subvector extraction cost is equal to the cost of extracting element from
    // the source type plus the cost of inserting them into the result vector
    // type.
    for (int i = 0; i != NumSubElts; ++i) {
      Cost += static_cast<T *>(this)->getVectorInstrCost(
          Instruction::ExtractElement, VTy, i + Index);
      Cost += static_cast<T *>(this)->getVectorInstrCost(
          Instruction::InsertElement, SubVTy, i);
    }
    return Cost;
  }

  /// Estimate a cost of subvector insertion as a sequence of extract and
  /// insert operations.
  unsigned getInsertSubvectorOverhead(FixedVectorType *VTy, int Index,
                                      FixedVectorType *SubVTy) {
    assert(VTy && SubVTy &&
           "Can only insert subvectors into vectors");
    int NumSubElts = SubVTy->getNumElements();
    assert((Index + NumSubElts) <= (int)VTy->getNumElements() &&
           "SK_InsertSubvector index out of range");

    unsigned Cost = 0;
    // Subvector insertion cost is equal to the cost of extracting element from
    // the source type plus the cost of inserting them into the result vector
    // type.
    for (int i = 0; i != NumSubElts; ++i) {
      Cost += static_cast<T *>(this)->getVectorInstrCost(
          Instruction::ExtractElement, SubVTy, i);
      Cost += static_cast<T *>(this)->getVectorInstrCost(
          Instruction::InsertElement, VTy, i + Index);
    }
    return Cost;
  }

  /// Local query method delegates up to T which *must* implement this!
  const TargetSubtargetInfo *getST() const {
    return static_cast<const T *>(this)->getST();
  }

  /// Local query method delegates up to T which *must* implement this!
  const TargetLoweringBase *getTLI() const {
    return static_cast<const T *>(this)->getTLI();
  }

  static ISD::MemIndexedMode getISDIndexedMode(TTI::MemIndexedMode M) {
    switch (M) {
      case TTI::MIM_Unindexed:
        return ISD::UNINDEXED;
      case TTI::MIM_PreInc:
        return ISD::PRE_INC;
      case TTI::MIM_PreDec:
        return ISD::PRE_DEC;
      case TTI::MIM_PostInc:
        return ISD::POST_INC;
      case TTI::MIM_PostDec:
        return ISD::POST_DEC;
    }
    llvm_unreachable("Unexpected MemIndexedMode");
  }

protected:
  explicit BasicTTIImplBase(const TargetMachine *TM, const DataLayout &DL)
      : BaseT(DL) {}
  virtual ~BasicTTIImplBase() = default;

  using TargetTransformInfoImplBase::DL;

public:
  /// \name Scalar TTI Implementations
  /// @{
  bool allowsMisalignedMemoryAccesses(LLVMContext &Context, unsigned BitWidth,
                                      unsigned AddressSpace, unsigned Alignment,
                                      bool *Fast) const {
    EVT E = EVT::getIntegerVT(Context, BitWidth);
    return getTLI()->allowsMisalignedMemoryAccesses(
        E, AddressSpace, Alignment, MachineMemOperand::MONone, Fast);
  }

  bool hasBranchDivergence() { return false; }

  bool useGPUDivergenceAnalysis() { return false; }

  bool isSourceOfDivergence(const Value *V) { return false; }

  bool isAlwaysUniform(const Value *V) { return false; }

  unsigned getFlatAddressSpace() {
    // Return an invalid address space.
    return -1;
  }

  bool collectFlatAddressOperands(SmallVectorImpl<int> &OpIndexes,
                                  Intrinsic::ID IID) const {
    return false;
  }

  bool isNoopAddrSpaceCast(unsigned FromAS, unsigned ToAS) const {
    return getTLI()->isNoopAddrSpaceCast(FromAS, ToAS);
  }

  Value *rewriteIntrinsicWithAddressSpace(IntrinsicInst *II, Value *OldV,
                                          Value *NewV) const {
    return nullptr;
  }

  bool isLegalAddImmediate(int64_t imm) {
    return getTLI()->isLegalAddImmediate(imm);
  }

  bool isLegalICmpImmediate(int64_t imm) {
    return getTLI()->isLegalICmpImmediate(imm);
  }

  bool isLegalAddressingMode(Type *Ty, GlobalValue *BaseGV, int64_t BaseOffset,
                             bool HasBaseReg, int64_t Scale,
                             unsigned AddrSpace, Instruction *I = nullptr) {
    TargetLoweringBase::AddrMode AM;
    AM.BaseGV = BaseGV;
    AM.BaseOffs = BaseOffset;
    AM.HasBaseReg = HasBaseReg;
    AM.Scale = Scale;
    return getTLI()->isLegalAddressingMode(DL, AM, Ty, AddrSpace, I);
  }

  bool isIndexedLoadLegal(TTI::MemIndexedMode M, Type *Ty,
                          const DataLayout &DL) const {
    EVT VT = getTLI()->getValueType(DL, Ty);
    return getTLI()->isIndexedLoadLegal(getISDIndexedMode(M), VT);
  }

  bool isIndexedStoreLegal(TTI::MemIndexedMode M, Type *Ty,
                           const DataLayout &DL) const {
    EVT VT = getTLI()->getValueType(DL, Ty);
    return getTLI()->isIndexedStoreLegal(getISDIndexedMode(M), VT);
  }

  bool isLSRCostLess(TTI::LSRCost C1, TTI::LSRCost C2) {
    return TargetTransformInfoImplBase::isLSRCostLess(C1, C2);
  }

  bool isProfitableLSRChainElement(Instruction *I) {
    return TargetTransformInfoImplBase::isProfitableLSRChainElement(I);
  }

  int getScalingFactorCost(Type *Ty, GlobalValue *BaseGV, int64_t BaseOffset,
                           bool HasBaseReg, int64_t Scale, unsigned AddrSpace) {
    TargetLoweringBase::AddrMode AM;
    AM.BaseGV = BaseGV;
    AM.BaseOffs = BaseOffset;
    AM.HasBaseReg = HasBaseReg;
    AM.Scale = Scale;
    return getTLI()->getScalingFactorCost(DL, AM, Ty, AddrSpace);
  }

  bool isTruncateFree(Type *Ty1, Type *Ty2) {
    return getTLI()->isTruncateFree(Ty1, Ty2);
  }

  bool isProfitableToHoist(Instruction *I) {
    return getTLI()->isProfitableToHoist(I);
  }

  bool useAA() const { return getST()->useAA(); }

  bool isTypeLegal(Type *Ty) {
    EVT VT = getTLI()->getValueType(DL, Ty);
    return getTLI()->isTypeLegal(VT);
  }

  int getGEPCost(Type *PointeeType, const Value *Ptr,
                 ArrayRef<const Value *> Operands) {
    return BaseT::getGEPCost(PointeeType, Ptr, Operands);
  }

  unsigned getEstimatedNumberOfCaseClusters(const SwitchInst &SI,
                                            unsigned &JumpTableSize,
                                            ProfileSummaryInfo *PSI,
                                            BlockFrequencyInfo *BFI) {
    /// Try to find the estimated number of clusters. Note that the number of
    /// clusters identified in this function could be different from the actual
    /// numbers found in lowering. This function ignore switches that are
    /// lowered with a mix of jump table / bit test / BTree. This function was
    /// initially intended to be used when estimating the cost of switch in
    /// inline cost heuristic, but it's a generic cost model to be used in other
    /// places (e.g., in loop unrolling).
    unsigned N = SI.getNumCases();
    const TargetLoweringBase *TLI = getTLI();
    const DataLayout &DL = this->getDataLayout();

    JumpTableSize = 0;
    bool IsJTAllowed = TLI->areJTsAllowed(SI.getParent()->getParent());

    // Early exit if both a jump table and bit test are not allowed.
    if (N < 1 || (!IsJTAllowed && DL.getIndexSizeInBits(0u) < N))
      return N;

    APInt MaxCaseVal = SI.case_begin()->getCaseValue()->getValue();
    APInt MinCaseVal = MaxCaseVal;
    for (auto CI : SI.cases()) {
      const APInt &CaseVal = CI.getCaseValue()->getValue();
      if (CaseVal.sgt(MaxCaseVal))
        MaxCaseVal = CaseVal;
      if (CaseVal.slt(MinCaseVal))
        MinCaseVal = CaseVal;
    }

    // Check if suitable for a bit test
    if (N <= DL.getIndexSizeInBits(0u)) {
      SmallPtrSet<const BasicBlock *, 4> Dests;
      for (auto I : SI.cases())
        Dests.insert(I.getCaseSuccessor());

      if (TLI->isSuitableForBitTests(Dests.size(), N, MinCaseVal, MaxCaseVal,
                                     DL))
        return 1;
    }

    // Check if suitable for a jump table.
    if (IsJTAllowed) {
      if (N < 2 || N < TLI->getMinimumJumpTableEntries())
        return N;
      uint64_t Range =
          (MaxCaseVal - MinCaseVal)
              .getLimitedValue(std::numeric_limits<uint64_t>::max() - 1) + 1;
      // Check whether a range of clusters is dense enough for a jump table
      if (TLI->isSuitableForJumpTable(&SI, N, Range, PSI, BFI)) {
        JumpTableSize = Range;
        return 1;
      }
    }
    return N;
  }

  bool shouldBuildLookupTables() {
    const TargetLoweringBase *TLI = getTLI();
    return TLI->isOperationLegalOrCustom(ISD::BR_JT, MVT::Other) ||
           TLI->isOperationLegalOrCustom(ISD::BRIND, MVT::Other);
  }

  bool haveFastSqrt(Type *Ty) {
    const TargetLoweringBase *TLI = getTLI();
    EVT VT = TLI->getValueType(DL, Ty);
    return TLI->isTypeLegal(VT) &&
           TLI->isOperationLegalOrCustom(ISD::FSQRT, VT);
  }

  bool isFCmpOrdCheaperThanFCmpZero(Type *Ty) {
    return true;
  }

  unsigned getFPOpCost(Type *Ty) {
    // Check whether FADD is available, as a proxy for floating-point in
    // general.
    const TargetLoweringBase *TLI = getTLI();
    EVT VT = TLI->getValueType(DL, Ty);
    if (TLI->isOperationLegalOrCustomOrPromote(ISD::FADD, VT))
      return TargetTransformInfo::TCC_Basic;
    return TargetTransformInfo::TCC_Expensive;
  }

  unsigned getInliningThresholdMultiplier() { return 1; }

  int getInlinerVectorBonusPercent() { return 150; }

  void getUnrollingPreferences(Loop *L, ScalarEvolution &SE,
                               TTI::UnrollingPreferences &UP) {
    // This unrolling functionality is target independent, but to provide some
    // motivation for its intended use, for x86:

    // According to the Intel 64 and IA-32 Architectures Optimization Reference
    // Manual, Intel Core models and later have a loop stream detector (and
    // associated uop queue) that can benefit from partial unrolling.
    // The relevant requirements are:
    //  - The loop must have no more than 4 (8 for Nehalem and later) branches
    //    taken, and none of them may be calls.
    //  - The loop can have no more than 18 (28 for Nehalem and later) uops.

    // According to the Software Optimization Guide for AMD Family 15h
    // Processors, models 30h-4fh (Steamroller and later) have a loop predictor
    // and loop buffer which can benefit from partial unrolling.
    // The relevant requirements are:
    //  - The loop must have fewer than 16 branches
    //  - The loop must have less than 40 uops in all executed loop branches

    // The number of taken branches in a loop is hard to estimate here, and
    // benchmarking has revealed that it is better not to be conservative when
    // estimating the branch count. As a result, we'll ignore the branch limits
    // until someone finds a case where it matters in practice.

    unsigned MaxOps;
    const TargetSubtargetInfo *ST = getST();
    if (PartialUnrollingThreshold.getNumOccurrences() > 0)
      MaxOps = PartialUnrollingThreshold;
    else if (ST->getSchedModel().LoopMicroOpBufferSize > 0)
      MaxOps = ST->getSchedModel().LoopMicroOpBufferSize;
    else
      return;

    // Scan the loop: don't unroll loops with calls.
    for (BasicBlock *BB : L->blocks()) {
      for (Instruction &I : *BB) {
        if (isa<CallInst>(I) || isa<InvokeInst>(I)) {
          if (const Function *F = cast<CallBase>(I).getCalledFunction()) {
            if (!static_cast<T *>(this)->isLoweredToCall(F))
              continue;
          }

          return;
        }
      }
    }

    // Enable runtime and partial unrolling up to the specified size.
    // Enable using trip count upper bound to unroll loops.
    UP.Partial = UP.Runtime = UP.UpperBound = true;
    UP.PartialThreshold = MaxOps;

    // Avoid unrolling when optimizing for size.
    UP.OptSizeThreshold = 0;
    UP.PartialOptSizeThreshold = 0;

    // Set number of instructions optimized when "back edge"
    // becomes "fall through" to default value of 2.
    UP.BEInsns = 2;
  }

  bool isHardwareLoopProfitable(Loop *L, ScalarEvolution &SE,
                                AssumptionCache &AC,
                                TargetLibraryInfo *LibInfo,
                                HardwareLoopInfo &HWLoopInfo) {
    return BaseT::isHardwareLoopProfitable(L, SE, AC, LibInfo, HWLoopInfo);
  }

  bool preferPredicateOverEpilogue(Loop *L, LoopInfo *LI, ScalarEvolution &SE,
                                   AssumptionCache &AC, TargetLibraryInfo *TLI,
                                   DominatorTree *DT,
                                   const LoopAccessInfo *LAI) {
    return BaseT::preferPredicateOverEpilogue(L, LI, SE, AC, TLI, DT, LAI);
  }

  bool emitGetActiveLaneMask() {
    return BaseT::emitGetActiveLaneMask();
  }

  int getInstructionLatency(const Instruction *I) {
    if (isa<LoadInst>(I))
      return getST()->getSchedModel().DefaultLoadLatency;

    return BaseT::getInstructionLatency(I);
  }

  virtual Optional<unsigned>
  getCacheSize(TargetTransformInfo::CacheLevel Level) const {
    return Optional<unsigned>(
      getST()->getCacheSize(static_cast<unsigned>(Level)));
  }

  virtual Optional<unsigned>
  getCacheAssociativity(TargetTransformInfo::CacheLevel Level) const {
    Optional<unsigned> TargetResult =
        getST()->getCacheAssociativity(static_cast<unsigned>(Level));

    if (TargetResult)
      return TargetResult;

    return BaseT::getCacheAssociativity(Level);
  }

  virtual unsigned getCacheLineSize() const {
    return getST()->getCacheLineSize();
  }

  virtual unsigned getPrefetchDistance() const {
    return getST()->getPrefetchDistance();
  }

  virtual unsigned getMinPrefetchStride(unsigned NumMemAccesses,
                                        unsigned NumStridedMemAccesses,
                                        unsigned NumPrefetches,
                                        bool HasCall) const {
    return getST()->getMinPrefetchStride(NumMemAccesses, NumStridedMemAccesses,
                                         NumPrefetches, HasCall);
  }

  virtual unsigned getMaxPrefetchIterationsAhead() const {
    return getST()->getMaxPrefetchIterationsAhead();
  }

  virtual bool enableWritePrefetching() const {
    return getST()->enableWritePrefetching();
  }

  /// @}

  /// \name Vector TTI Implementations
  /// @{

  unsigned getRegisterBitWidth(bool Vector) const { return 32; }

  /// Estimate the overhead of scalarizing an instruction. Insert and Extract
  /// are set if the demanded result elements need to be inserted and/or
  /// extracted from vectors.
  unsigned getScalarizationOverhead(VectorType *InTy, const APInt &DemandedElts,
                                    bool Insert, bool Extract) {
    /// FIXME: a bitfield is not a reasonable abstraction for talking about
    /// which elements are needed from a scalable vector
    auto *Ty = cast<FixedVectorType>(InTy);

    assert(DemandedElts.getBitWidth() == Ty->getNumElements() &&
           "Vector size mismatch");

    unsigned Cost = 0;

    for (int i = 0, e = Ty->getNumElements(); i < e; ++i) {
      if (!DemandedElts[i])
        continue;
      if (Insert)
        Cost += static_cast<T *>(this)->getVectorInstrCost(
            Instruction::InsertElement, Ty, i);
      if (Extract)
        Cost += static_cast<T *>(this)->getVectorInstrCost(
            Instruction::ExtractElement, Ty, i);
    }

    return Cost;
  }

  /// Helper wrapper for the DemandedElts variant of getScalarizationOverhead.
  unsigned getScalarizationOverhead(VectorType *InTy, bool Insert,
                                    bool Extract) {
    auto *Ty = cast<FixedVectorType>(InTy);

    APInt DemandedElts = APInt::getAllOnesValue(Ty->getNumElements());
    return static_cast<T *>(this)->getScalarizationOverhead(Ty, DemandedElts,
                                                            Insert, Extract);
  }

  /// Estimate the overhead of scalarizing an instructions unique
  /// non-constant operands. The types of the arguments are ordinarily
  /// scalar, in which case the costs are multiplied with VF.
  unsigned getOperandsScalarizationOverhead(ArrayRef<const Value *> Args,
                                            unsigned VF) {
    unsigned Cost = 0;
    SmallPtrSet<const Value*, 4> UniqueOperands;
    for (const Value *A : Args) {
      if (!isa<Constant>(A) && UniqueOperands.insert(A).second) {
        auto *VecTy = dyn_cast<VectorType>(A->getType());
        if (VecTy) {
          // If A is a vector operand, VF should be 1 or correspond to A.
          assert((VF == 1 ||
                  VF == cast<FixedVectorType>(VecTy)->getNumElements()) &&
                 "Vector argument does not match VF");
        }
        else
          VecTy = FixedVectorType::get(A->getType(), VF);

        Cost += getScalarizationOverhead(VecTy, false, true);
      }
    }

    return Cost;
  }

  unsigned getScalarizationOverhead(VectorType *InTy,
                                    ArrayRef<const Value *> Args) {
    auto *Ty = cast<FixedVectorType>(InTy);

    unsigned Cost = 0;

    Cost += getScalarizationOverhead(Ty, true, false);
    if (!Args.empty())
      Cost += getOperandsScalarizationOverhead(Args, Ty->getNumElements());
    else
      // When no information on arguments is provided, we add the cost
      // associated with one argument as a heuristic.
      Cost += getScalarizationOverhead(Ty, false, true);

    return Cost;
  }

  unsigned getMaxInterleaveFactor(unsigned VF) { return 1; }

  unsigned getArithmeticInstrCost(
      unsigned Opcode, Type *Ty,
      TTI::TargetCostKind CostKind = TTI::TCK_RecipThroughput,
      TTI::OperandValueKind Opd1Info = TTI::OK_AnyValue,
      TTI::OperandValueKind Opd2Info = TTI::OK_AnyValue,
      TTI::OperandValueProperties Opd1PropInfo = TTI::OP_None,
      TTI::OperandValueProperties Opd2PropInfo = TTI::OP_None,
      ArrayRef<const Value *> Args = ArrayRef<const Value *>(),
      const Instruction *CxtI = nullptr) {
    // Check if any of the operands are vector operands.
    const TargetLoweringBase *TLI = getTLI();
    int ISD = TLI->InstructionOpcodeToISD(Opcode);
    assert(ISD && "Invalid opcode");

    // TODO: Handle more cost kinds.
    if (CostKind != TTI::TCK_RecipThroughput)
      return BaseT::getArithmeticInstrCost(Opcode, Ty, CostKind,
                                           Opd1Info, Opd2Info,
                                           Opd1PropInfo, Opd2PropInfo,
                                           Args, CxtI);

    std::pair<unsigned, MVT> LT = TLI->getTypeLegalizationCost(DL, Ty);

    bool IsFloat = Ty->isFPOrFPVectorTy();
    // Assume that floating point arithmetic operations cost twice as much as
    // integer operations.
    unsigned OpCost = (IsFloat ? 2 : 1);

    if (TLI->isOperationLegalOrPromote(ISD, LT.second)) {
      // The operation is legal. Assume it costs 1.
      // TODO: Once we have extract/insert subvector cost we need to use them.
      return LT.first * OpCost;
    }

    if (!TLI->isOperationExpand(ISD, LT.second)) {
      // If the operation is custom lowered, then assume that the code is twice
      // as expensive.
      return LT.first * 2 * OpCost;
    }

    // Else, assume that we need to scalarize this op.
    // TODO: If one of the types get legalized by splitting, handle this
    // similarly to what getCastInstrCost() does.
    if (auto *VTy = dyn_cast<VectorType>(Ty)) {
      unsigned Num = cast<FixedVectorType>(VTy)->getNumElements();
      unsigned Cost = static_cast<T *>(this)->getArithmeticInstrCost(
          Opcode, VTy->getScalarType(), CostKind);
      // Return the cost of multiple scalar invocation plus the cost of
      // inserting and extracting the values.
      return getScalarizationOverhead(VTy, Args) + Num * Cost;
    }

    // We don't know anything about this scalar instruction.
    return OpCost;
  }

  unsigned getShuffleCost(TTI::ShuffleKind Kind, VectorType *Tp, int Index,
                          VectorType *SubTp) {

    switch (Kind) {
    case TTI::SK_Broadcast:
      return getBroadcastShuffleOverhead(cast<FixedVectorType>(Tp));
    case TTI::SK_Select:
    case TTI::SK_Reverse:
    case TTI::SK_Transpose:
    case TTI::SK_PermuteSingleSrc:
    case TTI::SK_PermuteTwoSrc:
      return getPermuteShuffleOverhead(cast<FixedVectorType>(Tp));
    case TTI::SK_ExtractSubvector:
      return getExtractSubvectorOverhead(cast<FixedVectorType>(Tp), Index,
                                         cast<FixedVectorType>(SubTp));
    case TTI::SK_InsertSubvector:
      return getInsertSubvectorOverhead(cast<FixedVectorType>(Tp), Index,
                                        cast<FixedVectorType>(SubTp));
    }
    llvm_unreachable("Unknown TTI::ShuffleKind");
  }

  unsigned getCastInstrCost(unsigned Opcode, Type *Dst, Type *Src,
                            TTI::TargetCostKind CostKind,
                            const Instruction *I = nullptr) {
    if (BaseT::getCastInstrCost(Opcode, Dst, Src, CostKind, I) == 0)
      return 0;

    const TargetLoweringBase *TLI = getTLI();
    int ISD = TLI->InstructionOpcodeToISD(Opcode);
    assert(ISD && "Invalid opcode");
    std::pair<unsigned, MVT> SrcLT = TLI->getTypeLegalizationCost(DL, Src);
    std::pair<unsigned, MVT> DstLT = TLI->getTypeLegalizationCost(DL, Dst);

    unsigned SrcSize = SrcLT.second.getSizeInBits();
    unsigned DstSize = DstLT.second.getSizeInBits();
    bool IntOrPtrSrc = Src->isIntegerTy() || Src->isPointerTy();
    bool IntOrPtrDst = Dst->isIntegerTy() || Dst->isPointerTy();

    switch (Opcode) {
    default:
      break;
    case Instruction::Trunc:
      // Check for NOOP conversions.
      if (TLI->isTruncateFree(SrcLT.second, DstLT.second))
        return 0;
      LLVM_FALLTHROUGH;
    case Instruction::BitCast:
      // Bitcast between types that are legalized to the same type are free and
      // assume int to/from ptr of the same size is also free.
      if (SrcLT.first == DstLT.first && IntOrPtrSrc == IntOrPtrDst &&
          SrcSize == DstSize)
        return 0;
      break;
    case Instruction::FPExt:
      if (I && getTLI()->isExtFree(I))
        return 0;
      break;
    case Instruction::ZExt:
      if (TLI->isZExtFree(SrcLT.second, DstLT.second))
        return 0;
      LLVM_FALLTHROUGH;
    case Instruction::SExt:
      if (!I)
        break;

      if (getTLI()->isExtFree(I))
        return 0;

      // If this is a zext/sext of a load, return 0 if the corresponding
      // extending load exists on target.
      if (I && isa<LoadInst>(I->getOperand(0))) {
        EVT ExtVT = EVT::getEVT(Dst);
        EVT LoadVT = EVT::getEVT(Src);
        unsigned LType =
          ((Opcode == Instruction::ZExt) ? ISD::ZEXTLOAD : ISD::SEXTLOAD);
        if (TLI->isLoadExtLegal(LType, ExtVT, LoadVT))
          return 0;
      }
      break;
    case Instruction::AddrSpaceCast:
      if (TLI->isFreeAddrSpaceCast(Src->getPointerAddressSpace(),
                                   Dst->getPointerAddressSpace()))
        return 0;
      break;
    }

    auto *SrcVTy = dyn_cast<VectorType>(Src);
    auto *DstVTy = dyn_cast<VectorType>(Dst);

    // If the cast is marked as legal (or promote) then assume low cost.
    if (SrcLT.first == DstLT.first &&
        TLI->isOperationLegalOrPromote(ISD, DstLT.second))
      return SrcLT.first;

    // Handle scalar conversions.
    if (!SrcVTy && !DstVTy) {
      // Just check the op cost. If the operation is legal then assume it costs
      // 1.
      if (!TLI->isOperationExpand(ISD, DstLT.second))
        return 1;

      // Assume that illegal scalar instruction are expensive.
      return 4;
    }

    // Check vector-to-vector casts.
    if (DstVTy && SrcVTy) {
      // If the cast is between same-sized registers, then the check is simple.
      if (SrcLT.first == DstLT.first &&
          SrcLT.second.getSizeInBits() == DstLT.second.getSizeInBits()) {

        // Assume that Zext is done using AND.
        if (Opcode == Instruction::ZExt)
          return SrcLT.first;

        // Assume that sext is done using SHL and SRA.
        if (Opcode == Instruction::SExt)
          return SrcLT.first * 2;

        // Just check the op cost. If the operation is legal then assume it
        // costs
        // 1 and multiply by the type-legalization overhead.
        if (!TLI->isOperationExpand(ISD, DstLT.second))
          return SrcLT.first * 1;
      }

      // If we are legalizing by splitting, query the concrete TTI for the cost
      // of casting the original vector twice. We also need to factor in the
      // cost of the split itself. Count that as 1, to be consistent with
      // TLI->getTypeLegalizationCost().
      bool SplitSrc =
          TLI->getTypeAction(Src->getContext(), TLI->getValueType(DL, Src)) ==
          TargetLowering::TypeSplitVector;
      bool SplitDst =
          TLI->getTypeAction(Dst->getContext(), TLI->getValueType(DL, Dst)) ==
          TargetLowering::TypeSplitVector;
      if ((SplitSrc || SplitDst) &&
          cast<FixedVectorType>(SrcVTy)->getNumElements() > 1 &&
          cast<FixedVectorType>(DstVTy)->getNumElements() > 1) {
        Type *SplitDstTy = VectorType::getHalfElementsVectorType(DstVTy);
        Type *SplitSrcTy = VectorType::getHalfElementsVectorType(SrcVTy);
        T *TTI = static_cast<T *>(this);
        // If both types need to be split then the split is free.
        unsigned SplitCost =
            (!SplitSrc || !SplitDst) ? TTI->getVectorSplitCost() : 0;
        return SplitCost +
               (2 * TTI->getCastInstrCost(Opcode, SplitDstTy, SplitSrcTy,
                                          CostKind, I));
      }

      // In other cases where the source or destination are illegal, assume
      // the operation will get scalarized.
      unsigned Num = cast<FixedVectorType>(DstVTy)->getNumElements();
      unsigned Cost = static_cast<T *>(this)->getCastInstrCost(
          Opcode, Dst->getScalarType(), Src->getScalarType(),
          CostKind, I);

      // Return the cost of multiple scalar invocation plus the cost of
      // inserting and extracting the values.
      return getScalarizationOverhead(DstVTy, true, true) + Num * Cost;
    }

    // We already handled vector-to-vector and scalar-to-scalar conversions.
    // This
    // is where we handle bitcast between vectors and scalars. We need to assume
    //  that the conversion is scalarized in one way or another.
    if (Opcode == Instruction::BitCast) {
      // Illegal bitcasts are done by storing and loading from a stack slot.
      return (SrcVTy ? getScalarizationOverhead(SrcVTy, false, true) : 0) +
             (DstVTy ? getScalarizationOverhead(DstVTy, true, false) : 0);
    }

    llvm_unreachable("Unhandled cast");
  }

  unsigned getExtractWithExtendCost(unsigned Opcode, Type *Dst,
                                    VectorType *VecTy, unsigned Index) {
    return static_cast<T *>(this)->getVectorInstrCost(
               Instruction::ExtractElement, VecTy, Index) +
           static_cast<T *>(this)->getCastInstrCost(Opcode, Dst,
                                                    VecTy->getElementType(),
                                                    TTI::TCK_RecipThroughput);
  }

  unsigned getCFInstrCost(unsigned Opcode, TTI::TargetCostKind CostKind) {
    return BaseT::getCFInstrCost(Opcode, CostKind);
  }

  unsigned getCmpSelInstrCost(unsigned Opcode, Type *ValTy, Type *CondTy,
                              TTI::TargetCostKind CostKind,
                              const Instruction *I = nullptr) {
    const TargetLoweringBase *TLI = getTLI();
    int ISD = TLI->InstructionOpcodeToISD(Opcode);
    assert(ISD && "Invalid opcode");

    // TODO: Handle other cost kinds.
    if (CostKind != TTI::TCK_RecipThroughput)
      return BaseT::getCmpSelInstrCost(Opcode, ValTy, CondTy, CostKind, I);

    // Selects on vectors are actually vector selects.
    if (ISD == ISD::SELECT) {
      assert(CondTy && "CondTy must exist");
      if (CondTy->isVectorTy())
        ISD = ISD::VSELECT;
    }
    std::pair<unsigned, MVT> LT = TLI->getTypeLegalizationCost(DL, ValTy);

    if (!(ValTy->isVectorTy() && !LT.second.isVector()) &&
        !TLI->isOperationExpand(ISD, LT.second)) {
      // The operation is legal. Assume it costs 1. Multiply
      // by the type-legalization overhead.
      return LT.first * 1;
    }

    // Otherwise, assume that the cast is scalarized.
    // TODO: If one of the types get legalized by splitting, handle this
    // similarly to what getCastInstrCost() does.
    if (auto *ValVTy = dyn_cast<VectorType>(ValTy)) {
      unsigned Num = cast<FixedVectorType>(ValVTy)->getNumElements();
      if (CondTy)
        CondTy = CondTy->getScalarType();
      unsigned Cost = static_cast<T *>(this)->getCmpSelInstrCost(
          Opcode, ValVTy->getScalarType(), CondTy, CostKind, I);

      // Return the cost of multiple scalar invocation plus the cost of
      // inserting and extracting the values.
      return getScalarizationOverhead(ValVTy, true, false) + Num * Cost;
    }

    // Unknown scalar opcode.
    return 1;
  }

  unsigned getVectorInstrCost(unsigned Opcode, Type *Val, unsigned Index) {
    std::pair<unsigned, MVT> LT =
        getTLI()->getTypeLegalizationCost(DL, Val->getScalarType());

    return LT.first;
  }

  unsigned getMemoryOpCost(unsigned Opcode, Type *Src, MaybeAlign Alignment,
                           unsigned AddressSpace,
                           TTI::TargetCostKind CostKind,
                           const Instruction *I = nullptr) {
    assert(!Src->isVoidTy() && "Invalid type");
    // Assume types, such as structs, are expensive.
    if (getTLI()->getValueType(DL, Src,  true) == MVT::Other)
      return 4;
    std::pair<unsigned, MVT> LT = getTLI()->getTypeLegalizationCost(DL, Src);

    // Assuming that all loads of legal types cost 1.
    unsigned Cost = LT.first;
    if (CostKind != TTI::TCK_RecipThroughput)
      return Cost;

    if (Src->isVectorTy() &&
        Src->getPrimitiveSizeInBits() < LT.second.getSizeInBits()) {
      // This is a vector load that legalizes to a larger type than the vector
      // itself. Unless the corresponding extending load or truncating store is
      // legal, then this will scalarize.
      TargetLowering::LegalizeAction LA = TargetLowering::Expand;
      EVT MemVT = getTLI()->getValueType(DL, Src);
      if (Opcode == Instruction::Store)
        LA = getTLI()->getTruncStoreAction(LT.second, MemVT);
      else
        LA = getTLI()->getLoadExtAction(ISD::EXTLOAD, LT.second, MemVT);

      if (LA != TargetLowering::Legal && LA != TargetLowering::Custom) {
        // This is a vector load/store for some illegal type that is scalarized.
        // We must account for the cost of building or decomposing the vector.
        Cost += getScalarizationOverhead(cast<VectorType>(Src),
                                         Opcode != Instruction::Store,
                                         Opcode == Instruction::Store);
      }
    }

    return Cost;
  }

  unsigned getInterleavedMemoryOpCost(unsigned Opcode, Type *VecTy,
                                      unsigned Factor,
                                      ArrayRef<unsigned> Indices,
                                      unsigned Alignment, unsigned AddressSpace,
                                      TTI::TargetCostKind CostKind,
                                      bool UseMaskForCond = false,
                                      bool UseMaskForGaps = false) {
    auto *VT = cast<FixedVectorType>(VecTy);

    unsigned NumElts = VT->getNumElements();
    assert(Factor > 1 && NumElts % Factor == 0 && "Invalid interleave factor");

    unsigned NumSubElts = NumElts / Factor;
    auto *SubVT = FixedVectorType::get(VT->getElementType(), NumSubElts);

    // Firstly, the cost of load/store operation.
    unsigned Cost;
    if (UseMaskForCond || UseMaskForGaps)
      Cost = static_cast<T *>(this)->getMaskedMemoryOpCost(
          Opcode, VecTy, Alignment, AddressSpace, CostKind);
    else
      Cost = static_cast<T *>(this)->getMemoryOpCost(
          Opcode, VecTy, MaybeAlign(Alignment), AddressSpace, CostKind);

    // Legalize the vector type, and get the legalized and unlegalized type
    // sizes.
    MVT VecTyLT = getTLI()->getTypeLegalizationCost(DL, VecTy).second;
    unsigned VecTySize =
        static_cast<T *>(this)->getDataLayout().getTypeStoreSize(VecTy);
    unsigned VecTyLTSize = VecTyLT.getStoreSize();

    // Return the ceiling of dividing A by B.
    auto ceil = [](unsigned A, unsigned B) { return (A + B - 1) / B; };

    // Scale the cost of the memory operation by the fraction of legalized
    // instructions that will actually be used. We shouldn't account for the
    // cost of dead instructions since they will be removed.
    //
    // E.g., An interleaved load of factor 8:
    //       %vec = load <16 x i64>, <16 x i64>* %ptr
    //       %v0 = shufflevector %vec, undef, <0, 8>
    //
    // If <16 x i64> is legalized to 8 v2i64 loads, only 2 of the loads will be
    // used (those corresponding to elements [0:1] and [8:9] of the unlegalized
    // type). The other loads are unused.
    //
    // We only scale the cost of loads since interleaved store groups aren't
    // allowed to have gaps.
    if (Opcode == Instruction::Load && VecTySize > VecTyLTSize) {
      // The number of loads of a legal type it will take to represent a load
      // of the unlegalized vector type.
      unsigned NumLegalInsts = ceil(VecTySize, VecTyLTSize);

      // The number of elements of the unlegalized type that correspond to a
      // single legal instruction.
      unsigned NumEltsPerLegalInst = ceil(NumElts, NumLegalInsts);

      // Determine which legal instructions will be used.
      BitVector UsedInsts(NumLegalInsts, false);
      for (unsigned Index : Indices)
        for (unsigned Elt = 0; Elt < NumSubElts; ++Elt)
          UsedInsts.set((Index + Elt * Factor) / NumEltsPerLegalInst);

      // Scale the cost of the load by the fraction of legal instructions that
      // will be used.
      Cost *= UsedInsts.count() / NumLegalInsts;
    }

    // Then plus the cost of interleave operation.
    if (Opcode == Instruction::Load) {
      // The interleave cost is similar to extract sub vectors' elements
      // from the wide vector, and insert them into sub vectors.
      //
      // E.g. An interleaved load of factor 2 (with one member of index 0):
      //      %vec = load <8 x i32>, <8 x i32>* %ptr
      //      %v0 = shuffle %vec, undef, <0, 2, 4, 6>         ; Index 0
      // The cost is estimated as extract elements at 0, 2, 4, 6 from the
      // <8 x i32> vector and insert them into a <4 x i32> vector.

      assert(Indices.size() <= Factor &&
             "Interleaved memory op has too many members");

      for (unsigned Index : Indices) {
        assert(Index < Factor && "Invalid index for interleaved memory op");

        // Extract elements from loaded vector for each sub vector.
        for (unsigned i = 0; i < NumSubElts; i++)
          Cost += static_cast<T *>(this)->getVectorInstrCost(
              Instruction::ExtractElement, VT, Index + i * Factor);
      }

      unsigned InsSubCost = 0;
      for (unsigned i = 0; i < NumSubElts; i++)
        InsSubCost += static_cast<T *>(this)->getVectorInstrCost(
            Instruction::InsertElement, SubVT, i);

      Cost += Indices.size() * InsSubCost;
    } else {
      // The interleave cost is extract all elements from sub vectors, and
      // insert them into the wide vector.
      //
      // E.g. An interleaved store of factor 2:
      //      %v0_v1 = shuffle %v0, %v1, <0, 4, 1, 5, 2, 6, 3, 7>
      //      store <8 x i32> %interleaved.vec, <8 x i32>* %ptr
      // The cost is estimated as extract all elements from both <4 x i32>
      // vectors and insert into the <8 x i32> vector.

      unsigned ExtSubCost = 0;
      for (unsigned i = 0; i < NumSubElts; i++)
        ExtSubCost += static_cast<T *>(this)->getVectorInstrCost(
            Instruction::ExtractElement, SubVT, i);
      Cost += ExtSubCost * Factor;

      for (unsigned i = 0; i < NumElts; i++)
        Cost += static_cast<T *>(this)
                    ->getVectorInstrCost(Instruction::InsertElement, VT, i);
    }

    if (!UseMaskForCond)
      return Cost;

    Type *I8Type = Type::getInt8Ty(VT->getContext());
    auto *MaskVT = FixedVectorType::get(I8Type, NumElts);
    SubVT = FixedVectorType::get(I8Type, NumSubElts);

    // The Mask shuffling cost is extract all the elements of the Mask
    // and insert each of them Factor times into the wide vector:
    //
    // E.g. an interleaved group with factor 3:
    //    %mask = icmp ult <8 x i32> %vec1, %vec2
    //    %interleaved.mask = shufflevector <8 x i1> %mask, <8 x i1> undef,
    //        <24 x i32> <0,0,0,1,1,1,2,2,2,3,3,3,4,4,4,5,5,5,6,6,6,7,7,7>
    // The cost is estimated as extract all mask elements from the <8xi1> mask
    // vector and insert them factor times into the <24xi1> shuffled mask
    // vector.
    for (unsigned i = 0; i < NumSubElts; i++)
      Cost += static_cast<T *>(this)->getVectorInstrCost(
          Instruction::ExtractElement, SubVT, i);

    for (unsigned i = 0; i < NumElts; i++)
      Cost += static_cast<T *>(this)->getVectorInstrCost(
          Instruction::InsertElement, MaskVT, i);

    // The Gaps mask is invariant and created outside the loop, therefore the
    // cost of creating it is not accounted for here. However if we have both
    // a MaskForGaps and some other mask that guards the execution of the
    // memory access, we need to account for the cost of And-ing the two masks
    // inside the loop.
    if (UseMaskForGaps)
      Cost += static_cast<T *>(this)->getArithmeticInstrCost(
          BinaryOperator::And, MaskVT, CostKind);

    return Cost;
  }

  /// Get intrinsic cost based on arguments.
  unsigned getIntrinsicInstrCost(const IntrinsicCostAttributes &ICA,
                                 TTI::TargetCostKind CostKind) {
    Intrinsic::ID IID = ICA.getID();
    auto *ConcreteTTI = static_cast<T *>(this);

    // Special case some scalar intrinsics.
    if (CostKind != TTI::TCK_RecipThroughput) {
      switch (IID) {
      default:
        break;
      case Intrinsic::cttz:
        if (getTLI()->isCheapToSpeculateCttz())
          return TargetTransformInfo::TCC_Basic;
        break;
      case Intrinsic::ctlz:
        if (getTLI()->isCheapToSpeculateCtlz())
          return TargetTransformInfo::TCC_Basic;
        break;
      case Intrinsic::memcpy:
        return ConcreteTTI->getMemcpyCost(ICA.getInst());
      // TODO: other libc intrinsics.
      }
      return BaseT::getIntrinsicInstrCost(ICA, CostKind);
    }

    if (BaseT::getIntrinsicInstrCost(ICA, CostKind) == 0)
      return 0;

    // TODO: Combine these two logic paths.
    if (ICA.isTypeBasedOnly())
      return getTypeBasedIntrinsicInstrCost(ICA, CostKind);

    Type *RetTy = ICA.getReturnType();
    unsigned VF = ICA.getVectorFactor();
    unsigned RetVF =
        (RetTy->isVectorTy() ? cast<FixedVectorType>(RetTy)->getNumElements()
                             : 1);