Reverts wrong modification to MachineBlockPlacement & BranchFolding; uses a new strat...

[oota-llvm.git] / lib / CodeGen / CodeGenPrepare.cpp

//===- CodeGenPrepare.cpp - Prepare a function for code generation --------===//
//
//                     The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This pass munges the code in the input function to better prepare it for
// SelectionDAG-based code generation. This works around limitations in it's
// basic-block-at-a-time approach. It should eventually be removed.
//
//===----------------------------------------------------------------------===//

#include "llvm/CodeGen/Passes.h"
#include "llvm/ADT/DenseMap.h"
#include "llvm/ADT/SetVector.h"
#include "llvm/ADT/SmallPtrSet.h"
#include "llvm/ADT/SmallSet.h"
#include "llvm/ADT/SmallVector.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/Analysis/InstructionSimplify.h"
#include "llvm/Analysis/MemoryLocation.h"
#include "llvm/Analysis/TargetLibraryInfo.h"
#include "llvm/Analysis/TargetTransformInfo.h"
#include "llvm/Analysis/ValueTracking.h"
#include "llvm/IR/CallSite.h"
#include "llvm/IR/Constants.h"
#include "llvm/IR/DataLayout.h"
#include "llvm/IR/DerivedTypes.h"
#include "llvm/IR/Dominators.h"
#include "llvm/IR/Function.h"
#include "llvm/IR/GetElementPtrTypeIterator.h"
#include "llvm/IR/IRBuilder.h"
#include "llvm/IR/InlineAsm.h"
#include "llvm/IR/InstIterator.h"
#include "llvm/IR/InstrTypes.h"
#include "llvm/IR/Instructions.h"
#include "llvm/IR/IntrinsicInst.h"
#include "llvm/IR/MDBuilder.h"
#include "llvm/IR/NoFolder.h"
#include "llvm/IR/PatternMatch.h"
#include "llvm/IR/Statepoint.h"
#include "llvm/IR/ValueHandle.h"
#include "llvm/IR/ValueMap.h"
#include "llvm/Pass.h"
#include "llvm/Support/CommandLine.h"
#include "llvm/Support/Debug.h"
#include "llvm/Support/raw_ostream.h"
#include "llvm/Target/TargetLowering.h"
#include "llvm/Target/TargetSubtargetInfo.h"
#include "llvm/Transforms/Utils/BasicBlockUtils.h"
#include "llvm/Transforms/Utils/BuildLibCalls.h"
#include "llvm/Transforms/Utils/BypassSlowDivision.h"
#include "llvm/Transforms/Utils/Local.h"
#include "llvm/Transforms/Utils/SimplifyLibCalls.h"
using namespace llvm;
using namespace llvm::PatternMatch;

#define DEBUG_TYPE "codegenprepare"

STATISTIC(NumBlocksElim, "Number of blocks eliminated");
STATISTIC(NumPHIsElim,   "Number of trivial PHIs eliminated");
STATISTIC(NumGEPsElim,   "Number of GEPs converted to casts");
STATISTIC(NumCmpUses, "Number of uses of Cmp expressions replaced with uses of "
                      "sunken Cmps");
STATISTIC(NumCastUses, "Number of uses of Cast expressions replaced with uses "
                       "of sunken Casts");
STATISTIC(NumMemoryInsts, "Number of memory instructions whose address "
                          "computations were sunk");
STATISTIC(NumExtsMoved,  "Number of [s|z]ext instructions combined with loads");
STATISTIC(NumExtUses,    "Number of uses of [s|z]ext instructions optimized");
STATISTIC(NumAndsAdded,
          "Number of and mask instructions added to form ext loads");
STATISTIC(NumAndUses, "Number of uses of and mask instructions optimized");
STATISTIC(NumRetsDup,    "Number of return instructions duplicated");
STATISTIC(NumDbgValueMoved, "Number of debug value instructions moved");
STATISTIC(NumSelectsExpanded, "Number of selects turned into branches");
STATISTIC(NumAndCmpsMoved, "Number of and/cmp's pushed into branches");
STATISTIC(NumStoreExtractExposed, "Number of store(extractelement) exposed");

static cl::opt<bool> DisableBranchOpts(
  "disable-cgp-branch-opts", cl::Hidden, cl::init(false),
  cl::desc("Disable branch optimizations in CodeGenPrepare"));

static cl::opt<bool>
    DisableGCOpts("disable-cgp-gc-opts", cl::Hidden, cl::init(false),
                  cl::desc("Disable GC optimizations in CodeGenPrepare"));

static cl::opt<bool> DisableSelectToBranch(
  "disable-cgp-select2branch", cl::Hidden, cl::init(false),
  cl::desc("Disable select to branch conversion."));

static cl::opt<bool> AddrSinkUsingGEPs(
  "addr-sink-using-gep", cl::Hidden, cl::init(false),
  cl::desc("Address sinking in CGP using GEPs."));

static cl::opt<bool> EnableAndCmpSinking(
   "enable-andcmp-sinking", cl::Hidden, cl::init(true),
   cl::desc("Enable sinkinig and/cmp into branches."));

static cl::opt<bool> DisableStoreExtract(
    "disable-cgp-store-extract", cl::Hidden, cl::init(false),
    cl::desc("Disable store(extract) optimizations in CodeGenPrepare"));

static cl::opt<bool> StressStoreExtract(
    "stress-cgp-store-extract", cl::Hidden, cl::init(false),
    cl::desc("Stress test store(extract) optimizations in CodeGenPrepare"));

static cl::opt<bool> DisableExtLdPromotion(
    "disable-cgp-ext-ld-promotion", cl::Hidden, cl::init(false),
    cl::desc("Disable ext(promotable(ld)) -> promoted(ext(ld)) optimization in "
             "CodeGenPrepare"));

static cl::opt<bool> StressExtLdPromotion(
    "stress-cgp-ext-ld-promotion", cl::Hidden, cl::init(false),
    cl::desc("Stress test ext(promotable(ld)) -> promoted(ext(ld)) "
             "optimization in CodeGenPrepare"));

namespace {
typedef SmallPtrSet<Instruction *, 16> SetOfInstrs;
typedef PointerIntPair<Type *, 1, bool> TypeIsSExt;
typedef DenseMap<Instruction *, TypeIsSExt> InstrToOrigTy;
class TypePromotionTransaction;

  class CodeGenPrepare : public FunctionPass {
    const TargetMachine *TM;
    const TargetLowering *TLI;
    const TargetTransformInfo *TTI;
    const TargetLibraryInfo *TLInfo;

    /// As we scan instructions optimizing them, this is the next instruction
    /// to optimize. Transforms that can invalidate this should update it.
    BasicBlock::iterator CurInstIterator;

    /// Keeps track of non-local addresses that have been sunk into a block.
    /// This allows us to avoid inserting duplicate code for blocks with
    /// multiple load/stores of the same address.
    ValueMap<Value*, Value*> SunkAddrs;

    /// Keeps track of all instructions inserted for the current function.
    SetOfInstrs InsertedInsts;
    /// Keeps track of the type of the related instruction before their
    /// promotion for the current function.
    InstrToOrigTy PromotedInsts;

    /// True if CFG is modified in any way.
    bool ModifiedDT;

    /// True if optimizing for size.
    bool OptSize;

    /// DataLayout for the Function being processed.
    const DataLayout *DL;

  public:
    static char ID; // Pass identification, replacement for typeid
    explicit CodeGenPrepare(const TargetMachine *TM = nullptr)
        : FunctionPass(ID), TM(TM), TLI(nullptr), TTI(nullptr), DL(nullptr) {
        initializeCodeGenPreparePass(*PassRegistry::getPassRegistry());
      }
    bool runOnFunction(Function &F) override;

    const char *getPassName() const override { return "CodeGen Prepare"; }

    void getAnalysisUsage(AnalysisUsage &AU) const override {
      AU.addPreserved<DominatorTreeWrapperPass>();
      AU.addRequired<TargetLibraryInfoWrapperPass>();
      AU.addRequired<TargetTransformInfoWrapperPass>();
    }

  private:
    bool eliminateFallThrough(Function &F);
    bool eliminateMostlyEmptyBlocks(Function &F);
    bool canMergeBlocks(const BasicBlock *BB, const BasicBlock *DestBB) const;
    void eliminateMostlyEmptyBlock(BasicBlock *BB);
    bool optimizeBlock(BasicBlock &BB, bool& ModifiedDT);
    bool optimizeInst(Instruction *I, bool& ModifiedDT);
    bool optimizeMemoryInst(Instruction *I, Value *Addr,
                            Type *AccessTy, unsigned AS);
    bool optimizeInlineAsmInst(CallInst *CS);
    bool optimizeCallInst(CallInst *CI, bool& ModifiedDT);
    bool moveExtToFormExtLoad(Instruction *&I);
    bool optimizeExtUses(Instruction *I);
    bool optimizeLoadExt(LoadInst *I);
    bool optimizeSelectInst(SelectInst *SI);
    bool optimizeShuffleVectorInst(ShuffleVectorInst *SI);
    bool optimizeSwitchInst(SwitchInst *CI);
    bool optimizeExtractElementInst(Instruction *Inst);
    bool dupRetToEnableTailCallOpts(BasicBlock *BB);
    bool placeDbgValues(Function &F);
    bool sinkAndCmp(Function &F);
    bool extLdPromotion(TypePromotionTransaction &TPT, LoadInst *&LI,
                        Instruction *&Inst,
                        const SmallVectorImpl<Instruction *> &Exts,
                        unsigned CreatedInstCost);
    bool splitBranchCondition(Function &F);
    bool simplifyOffsetableRelocate(Instruction &I);
    void stripInvariantGroupMetadata(Instruction &I);
  };
}

char CodeGenPrepare::ID = 0;
INITIALIZE_TM_PASS(CodeGenPrepare, "codegenprepare",
                   "Optimize for code generation", false, false)

FunctionPass *llvm::createCodeGenPreparePass(const TargetMachine *TM) {
  return new CodeGenPrepare(TM);
}

namespace {

bool StoreAddressDependOnValue(StoreInst* SI, Value* DepVal);
Value* GetUntaintedAddress(Value* CurrentAddress);

// The depth we trace down a variable to look for its dependence set.
const unsigned kDependenceDepth = 4;

// Recursively looks for variables that 'Val' depends on at the given depth
// 'Depth', and adds them in 'DepSet'. If 'InsertOnlyLeafNodes' is true, only
// inserts the leaf node values; otherwise, all visited nodes are included in
// 'DepSet'. Note that constants will be ignored.
template <typename SetType>
void recursivelyFindDependence(SetType* DepSet, Value* Val,
                               bool InsertOnlyLeafNodes = false,
                               unsigned Depth = kDependenceDepth) {
  if (Val == nullptr) {
    return;
  }
  if (!InsertOnlyLeafNodes && !isa<Constant>(Val)) {
    DepSet->insert(Val);
  }
  if (Depth == 0) {
    // Cannot go deeper. Insert the leaf nodes.
    if (InsertOnlyLeafNodes && !isa<Constant>(Val)) {
      DepSet->insert(Val);
    }
    return;
  }

  // Go one step further to explore the dependence of the operands.
  Instruction* I = nullptr;
  if ((I = dyn_cast<Instruction>(Val))) {
    if (isa<LoadInst>(I)) {
      // A load is considerd the leaf load of the dependence tree. Done.
      DepSet->insert(Val);
      return;
    } else if (I->isBinaryOp()) {
      BinaryOperator* I = dyn_cast<BinaryOperator>(Val);
      Value *Op0 = I->getOperand(0), *Op1 = I->getOperand(1);
      recursivelyFindDependence(DepSet, Op0, Depth - 1);
      recursivelyFindDependence(DepSet, Op1, Depth - 1);
    } else if (I->isCast()) {
      Value* Op0 = I->getOperand(0);
      recursivelyFindDependence(DepSet, Op0, Depth - 1);
    } else if (I->getOpcode() == Instruction::Select) {
      Value* Op0 = I->getOperand(0);
      Value* Op1 = I->getOperand(1);
      Value* Op2 = I->getOperand(2);
      recursivelyFindDependence(DepSet, Op0, Depth - 1);
      recursivelyFindDependence(DepSet, Op1, Depth - 1);
      recursivelyFindDependence(DepSet, Op2, Depth - 1);
    } else if (I->getOpcode() == Instruction::GetElementPtr) {
      for (unsigned i = 0; i < I->getNumOperands(); i++) {
        recursivelyFindDependence(DepSet, I->getOperand(i), Depth - 1);
      }
    } else if (I->getOpcode() == Instruction::Store) {
      auto* SI = dyn_cast<StoreInst>(Val);
      recursivelyFindDependence(DepSet, SI->getPointerOperand(), Depth - 1);
      recursivelyFindDependence(DepSet, SI->getValueOperand(), Depth - 1);
    } else {
      Value* Op0 = nullptr;
      Value* Op1 = nullptr;
      switch (I->getOpcode()) {
        case Instruction::ICmp:
        case Instruction::FCmp: {
          Op0 = I->getOperand(0);
          Op1 = I->getOperand(1);
          recursivelyFindDependence(DepSet, Op0, Depth - 1);
          recursivelyFindDependence(DepSet, Op1, Depth - 1);
          break;
        }
        default: {
          // Be conservative. Add it and be done with it.
          DepSet->insert(Val);
          return;
        }
      }
    }
  } else if (isa<Constant>(Val)) {
    // Not interested in constant values. Done.
    return;
  } else {
    // Be conservative. Add it and be done with it.
    DepSet->insert(Val);
    return;
  }
}

// Helper function to create a Cast instruction.
Value* createCast(IRBuilder<true, NoFolder>& Builder, Value* DepVal,
                  Type* TargetIntegerType) {
  Instruction::CastOps CastOp = Instruction::BitCast;
  switch (DepVal->getType()->getTypeID()) {
    case Type::IntegerTyID: {
      CastOp = Instruction::SExt;
      break;
    }
    case Type::FloatTyID:
    case Type::DoubleTyID: {
      CastOp = Instruction::FPToSI;
      break;
    }
    case Type::PointerTyID: {
      CastOp = Instruction::PtrToInt;
      break;
    }
    default: { break; }
  }

  return Builder.CreateCast(CastOp, DepVal, TargetIntegerType);
}

// Given a value, if it's a tainted address, this function returns the
// instruction that ORs the "dependence value" with the "original address".
// Otherwise, returns nullptr.  This instruction is the first OR instruction
// where one of its operand is an AND instruction with an operand being 0.
//
// E.g., it returns '%4 = or i32 %3, %2' given 'CurrentAddress' is '%5'.
// %0 = load i32, i32* @y, align 4, !tbaa !1
// %cmp = icmp ne i32 %0, 42  // <== this is like the condition
// %1 = sext i1 %cmp to i32
// %2 = ptrtoint i32* @x to i32
// %3 = and i32 %1, 0
// %4 = or i32 %3, %2
// %5 = inttoptr i32 %4 to i32*
// store i32 1, i32* %5, align 4
Instruction* getOrAddress(Value* CurrentAddress) {
  // Is it a cast from integer to pointer type.
  Instruction* OrAddress = nullptr;
  Instruction* AndDep = nullptr;
  Instruction* CastToInt = nullptr;
  Value* ActualAddress = nullptr;
  Constant* ZeroConst = nullptr;

  const Instruction* CastToPtr = dyn_cast<Instruction>(CurrentAddress);
  if (CastToPtr && CastToPtr->getOpcode() == Instruction::IntToPtr) {
    // Is it an OR instruction: %1 = or %and, %actualAddress.
    if ((OrAddress = dyn_cast<Instruction>(CastToPtr->getOperand(0))) &&
        OrAddress->getOpcode() == Instruction::Or) {
      // The first operand should be and AND instruction.
      AndDep = dyn_cast<Instruction>(OrAddress->getOperand(0));
      if (AndDep && AndDep->getOpcode() == Instruction::And) {
        // Also make sure its first operand of the "AND" is 0, or the "AND" is
        // marked explicitly by "NoInstCombine".
        if ((ZeroConst = dyn_cast<Constant>(AndDep->getOperand(1))) &&
            ZeroConst->isNullValue()) {
          return OrAddress;
        }
      }
    }
  }
  // Looks like it's not been tainted.
  return nullptr;
}

// Given a value, if it's a tainted address, this function returns the
// instruction that taints the "dependence value". Otherwise, returns nullptr.
// This instruction is the last AND instruction where one of its operand is 0.
// E.g., it returns '%3' given 'CurrentAddress' is '%5'.
// %0 = load i32, i32* @y, align 4, !tbaa !1
// %cmp = icmp ne i32 %0, 42  // <== this is like the condition
// %1 = sext i1 %cmp to i32
// %2 = ptrtoint i32* @x to i32
// %3 = and i32 %1, 0
// %4 = or i32 %3, %2
// %5 = inttoptr i32 %4 to i32*
// store i32 1, i32* %5, align 4
Instruction* getAndDependence(Value* CurrentAddress) {
  // If 'CurrentAddress' is tainted, get the OR instruction.
  auto* OrAddress = getOrAddress(CurrentAddress);
  if (OrAddress == nullptr) {
    return nullptr;
  }

  // No need to check the operands.
  auto* AndDepInst = dyn_cast<Instruction>(OrAddress->getOperand(0));
  assert(AndDepInst);
  return AndDepInst;
}

// Given a value, if it's a tainted address, this function returns
// the "dependence value", which is the first operand in the AND instruction.
// E.g., it returns '%1' given 'CurrentAddress' is '%5'.
// %0 = load i32, i32* @y, align 4, !tbaa !1
// %cmp = icmp ne i32 %0, 42  // <== this is like the condition
// %1 = sext i1 %cmp to i32
// %2 = ptrtoint i32* @x to i32
// %3 = and i32 %1, 0
// %4 = or i32 %3, %2
// %5 = inttoptr i32 %4 to i32*
// store i32 1, i32* %5, align 4
Value* getDependence(Value* CurrentAddress) {
  auto* AndInst = getAndDependence(CurrentAddress);
  if (AndInst == nullptr) {
    return nullptr;
  }
  return AndInst->getOperand(0);
}

// Given an address that has been tainted, returns the only condition it depends
// on, if any; otherwise, returns nullptr.
Value* getConditionDependence(Value* Address) {
  auto* Dep = getDependence(Address);
  if (Dep == nullptr) {
    // 'Address' has not been dependence-tainted.
    return nullptr;
  }

  Value* Operand = Dep;
  while (true) {
    auto* Inst = dyn_cast<Instruction>(Operand);
    if (Inst == nullptr) {
      // Non-instruction type does not have condition dependence.
      return nullptr;
    }
    if (Inst->getOpcode() == Instruction::ICmp) {
      return Inst;
    } else {
      if (Inst->getNumOperands() != 1) {
        return nullptr;
      } else {
        Operand = Inst->getOperand(0);
      }
    }
  }
}

// Conservatively decides whether the dependence set of 'Val1' includes the
// dependence set of 'Val2'. If 'ExpandSecondValue' is false, we do not expand
// 'Val2' and use that single value as its dependence set.
// If it returns true, it means the dependence set of 'Val1' includes that of
// 'Val2'; otherwise, it only means we cannot conclusively decide it.
bool dependenceSetInclusion(Value* Val1, Value* Val2,
                            int Val1ExpandLevel = 2 * kDependenceDepth,
                            int Val2ExpandLevel = kDependenceDepth) {
  typedef SmallSet<Value*, 8> IncludingSet;
  typedef SmallSet<Value*, 4> IncludedSet;

  IncludingSet DepSet1;
  IncludedSet DepSet2;
  // Look for more depths for the including set.
  recursivelyFindDependence(&DepSet1, Val1, false /*Insert all visited nodes*/,
                            Val1ExpandLevel);
  recursivelyFindDependence(&DepSet2, Val2, true /*Only insert leaf nodes*/,
                            Val2ExpandLevel);

  auto set_inclusion = [](IncludingSet FullSet, IncludedSet Subset) {
    for (auto* Dep : Subset) {
      if (0 == FullSet.count(Dep)) {
        return false;
      }
    }
    return true;
  };
  bool inclusion = set_inclusion(DepSet1, DepSet2);
  DEBUG(dbgs() << "[dependenceSetInclusion]: " << inclusion << "\n");
  DEBUG(dbgs() << "Including set for: " << *Val1 << "\n");
  DEBUG(for (const auto* Dep : DepSet1) { dbgs() << "\t\t" << *Dep << "\n"; });
  DEBUG(dbgs() << "Included set for: " << *Val2 << "\n");
  DEBUG(for (const auto* Dep : DepSet2) { dbgs() << "\t\t" << *Dep << "\n"; });

  return inclusion;
}

// Recursively iterates through the operands spawned from 'DepVal'. If there
// exists a single value that 'DepVal' only depends on, we call that value the
// root dependence of 'DepVal' and return it. Otherwise, return 'DepVal'.
Value* getRootDependence(Value* DepVal) {
  SmallSet<Value*, 8> DepSet;
  for (unsigned depth = kDependenceDepth; depth > 0; --depth) {
    recursivelyFindDependence(&DepSet, DepVal, true /*Only insert leaf nodes*/,
                              depth);
    if (DepSet.size() == 1) {
      return *DepSet.begin();
    }
    DepSet.clear();
  }
  return DepVal;
}

// This function actually taints 'DepVal' to the address to 'SI'. If the
// address
// of 'SI' already depends on whatever 'DepVal' depends on, this function
// doesn't do anything and returns false. Otherwise, returns true.
//
// This effect forces the store and any stores that comes later to depend on
// 'DepVal'. For example, we have a condition "cond", and a store instruction
// "s: STORE addr, val". If we want "s" (and any later store) to depend on
// "cond", we do the following:
// %conv = sext i1 %cond to i32
// %addrVal = ptrtoint i32* %addr to i32
// %andCond = and i32 conv, 0;
// %orAddr = or i32 %andCond, %addrVal;
// %NewAddr = inttoptr i32 %orAddr to i32*;
//
// This is a more concrete example:
// ------
// %0 = load i32, i32* @y, align 4, !tbaa !1
// %cmp = icmp ne i32 %0, 42  // <== this is like the condition
// %1 = sext i1 %cmp to i32
// %2 = ptrtoint i32* @x to i32
// %3 = and i32 %1, 0
// %4 = or i32 %3, %2
// %5 = inttoptr i32 %4 to i32*
// store i32 1, i32* %5, align 4
bool taintStoreAddress(StoreInst* SI, Value* DepVal,
                       const char* calling_func = __builtin_FUNCTION()) {
  DEBUG(dbgs() << "Called from " << calling_func << '\n');
  IRBuilder<true, NoFolder> Builder(SI);
  BasicBlock* BB = SI->getParent();
  Value* Address = SI->getPointerOperand();
  Type* TargetIntegerType =
      IntegerType::get(Address->getContext(),
                       BB->getModule()->getDataLayout().getPointerSizeInBits());

  // Does SI's address already depends on whatever 'DepVal' depends on?
  if (StoreAddressDependOnValue(SI, DepVal)) {
    return false;
  }

  // Figure out if there's a root variable 'DepVal' depends on. For example, we
  // can extract "getelementptr inbounds %struct, %struct* %0, i64 0, i32 123"
  // to be "%struct* %0" since all other operands are constant.
  DepVal = getRootDependence(DepVal);

  // Is this already a dependence-tainted store?
  Value* OldDep = getDependence(Address);
  if (OldDep) {
    // The address of 'SI' has already been tainted.  Just need to absorb the
    // DepVal to the existing dependence in the address of SI.
    Instruction* AndDep = getAndDependence(Address);
    IRBuilder<true, NoFolder> Builder(AndDep);
    Value* NewDep = nullptr;
    if (DepVal->getType() == AndDep->getType()) {
      NewDep = Builder.CreateAnd(OldDep, DepVal);
    } else {
      NewDep = Builder.CreateAnd(
          OldDep, createCast(Builder, DepVal, TargetIntegerType));
    }

    auto* NewDepInst = dyn_cast<Instruction>(NewDep);

    // Use the new AND instruction as the dependence
    AndDep->setOperand(0, NewDep);
    return true;
  }

  // SI's address has not been tainted. Now taint it with 'DepVal'.
  Value* CastDepToInt = createCast(Builder, DepVal, TargetIntegerType);
  Value* PtrToIntCast = Builder.CreatePtrToInt(Address, TargetIntegerType);
  Value* AndDepVal =
      Builder.CreateAnd(CastDepToInt, ConstantInt::get(TargetIntegerType, 0));
  auto AndInst = dyn_cast<Instruction>(AndDepVal);
  // XXX-comment: The original IR InstCombiner would change our and instruction
  // to a select and then the back end optimize the condition out.  We attach a
  // flag to instructions and set it here to inform the InstCombiner to not to
  // touch this and instruction at all.
  Value* OrAddr = Builder.CreateOr(AndDepVal, PtrToIntCast);
  Value* NewAddr = Builder.CreateIntToPtr(OrAddr, Address->getType());

  DEBUG(dbgs() << "[taintStoreAddress]\n"
               << "Original store: " << *SI << '\n');
  SI->setOperand(1, NewAddr);

  // Debug output.
  DEBUG(dbgs() << "\tTargetIntegerType: " << *TargetIntegerType << '\n'
               << "\tCast dependence value to integer: " << *CastDepToInt
               << '\n'
               << "\tCast address to integer: " << *PtrToIntCast << '\n'
               << "\tAnd dependence value: " << *AndDepVal << '\n'
               << "\tOr address: " << *OrAddr << '\n'
               << "\tCast or instruction to address: " << *NewAddr << "\n\n");

  return true;
}

// Looks for the previous store in the if block --- 'BrBB', which makes the
// speculative store 'StoreToHoist' safe.
Value* getSpeculativeStoreInPrevBB(StoreInst* StoreToHoist, BasicBlock* BrBB) {
  assert(StoreToHoist && "StoreToHoist must be a real store");

  Value* StorePtr = StoreToHoist->getPointerOperand();

  // Look for a store to the same pointer in BrBB.
  for (BasicBlock::reverse_iterator RI = BrBB->rbegin(), RE = BrBB->rend();
       RI != RE; ++RI) {
    Instruction* CurI = &*RI;

    StoreInst* SI = dyn_cast<StoreInst>(CurI);
    // Found the previous store make sure it stores to the same location.
    // XXX-update: If the previous store's original untainted address are the
    // same as 'StorePtr', we are also good to hoist the store.
    if (SI && (SI->getPointerOperand() == StorePtr ||
               GetUntaintedAddress(SI->getPointerOperand()) == StorePtr)) {
      // Found the previous store, return its value operand.
      return SI;
    }
  }

  assert(false &&
         "We should not reach here since this store is safe to speculate");
}

// XXX-comment: Returns true if it changes the code, false otherwise (the branch
// condition already depends on 'DepVal'.
bool taintConditionalBranch(BranchInst* BI, Value* DepVal) {
  assert(BI->isConditional());
  auto* Cond = BI->getOperand(0);
  if (dependenceSetInclusion(Cond, DepVal)) {
    // The dependence/ordering is self-evident.
    return false;
  }

  IRBuilder<true, NoFolder> Builder(BI);
  auto* AndDep =
      Builder.CreateAnd(DepVal, ConstantInt::get(DepVal->getType(), 0));
  auto* TruncAndDep =
      Builder.CreateTrunc(AndDep, IntegerType::get(DepVal->getContext(), 1));
  auto* OrCond = Builder.CreateOr(TruncAndDep, Cond);
  BI->setOperand(0, OrCond);

  // Debug output.
  DEBUG(dbgs() << "\tTainted branch condition:\n" << *BI->getParent());

  return true;
}

bool ConditionalBranchDependsOnValue(BranchInst* BI, Value* DepVal) {
  assert(BI->isConditional());
  auto* Cond = BI->getOperand(0);
  return dependenceSetInclusion(Cond, DepVal);
}

// XXX-update: For a relaxed load 'LI', find the first immediate atomic store or
// the first conditional branch. Returns nullptr if there's no such immediately
// following store/branch instructions, which we can only enforce the load with
// 'acquire'.
Instruction* findFirstStoreCondBranchInst(LoadInst* LI) {
  // In some situations, relaxed loads can be left as is:
  // 1. The relaxed load is used to calculate the address of the immediate
  // following store;
  // 2. The relaxed load is used as a condition in the immediate following
  // condition, and there are no stores in between. This is actually quite
  // common. E.g.,
  // int r1 = x.load(relaxed);
  // if (r1 != 0) {
  //   y.store(1, relaxed);
  // }
  // However, in this function, we don't deal with them directly. Instead, we
  // just find the immediate following store/condition branch and return it.

  auto* BB = LI->getParent();
  auto BE = BB->end();
  auto BBI = BasicBlock::iterator(LI);
  BBI++;
  while (true) {
    for (; BBI != BE; BBI++) {
      auto* Inst = dyn_cast<Instruction>(&*BBI);
      if (Inst == nullptr) {
        continue;
      }
      if (Inst->getOpcode() == Instruction::Store) {
        return Inst;
      } else if (Inst->getOpcode() == Instruction::Br) {
        auto* BrInst = dyn_cast<BranchInst>(Inst);
        if (BrInst->isConditional()) {
          return Inst;
        } else {
          // Reinitialize iterators with the destination of the unconditional
          // branch.
          BB = BrInst->getSuccessor(0);
          BBI = BB->begin();
          BE = BB->end();
          break;
        }
      }
    }
    if (BBI == BE) {
      return nullptr;
    }
  }
}

// XXX-comment: Returns whether the code has been changed.
bool taintMonotonicLoads(const SmallVector<LoadInst*, 1>& MonotonicLoadInsts) {
  bool Changed = false;
  for (auto* LI : MonotonicLoadInsts) {
    auto* FirstInst = findFirstStoreCondBranchInst(LI);
    if (FirstInst == nullptr) {
      // We don't seem to be able to taint a following store/conditional branch
      // instruction. Simply make it acquire.
      DEBUG(dbgs() << "[RelaxedLoad]: Transformed to acquire load\n"
                   << *LI << "\n");
      LI->setOrdering(Acquire);
      Changed = true;
      continue;
    }
    // Taint 'FirstInst', which could be a store or a condition branch
    // instruction.
    if (FirstInst->getOpcode() == Instruction::Store) {
      Changed |= taintStoreAddress(dyn_cast<StoreInst>(FirstInst), LI);
    } else if (FirstInst->getOpcode() == Instruction::Br) {
      Changed |= taintConditionalBranch(dyn_cast<BranchInst>(FirstInst), LI);
    } else {
      assert(false && "findFirstStoreCondBranchInst() should return a "
                    "store/condition branch instruction");
    }
  }
  return Changed;
}

// Inserts a fake conditional branch right after the instruction 'SplitInst',
// and the branch condition is 'Condition'. 'SplitInst' will be placed in the
// newly created block.
void AddFakeConditionalBranch(Instruction* SplitInst, Value* Condition) {
  auto* BB = SplitInst->getParent();
  TerminatorInst* ThenTerm = nullptr;
  TerminatorInst* ElseTerm = nullptr;
  SplitBlockAndInsertIfThenElse(Condition, SplitInst, &ThenTerm, &ElseTerm);
  assert(ThenTerm && ElseTerm &&
         "Then/Else terminators cannot be empty after basic block spliting");
  auto* ThenBB = ThenTerm->getParent();
  auto* ElseBB = ElseTerm->getParent();
  auto* TailBB = ThenBB->getSingleSuccessor();
  assert(TailBB && "Tail block cannot be empty after basic block spliting");

  ThenBB->disableCanEliminateBlock();
  ThenBB->disableCanEliminateBlock();
  TailBB->disableCanEliminateBlock();
  ThenBB->setName(BB->getName() + "Then.Fake");
  ElseBB->setName(BB->getName() + "Else.Fake");
  DEBUG(dbgs() << "Add fake conditional branch:\n"
               << "Then Block:\n"
               << *ThenBB << "Else Block:\n"
               << *ElseBB << "\n");
}

// Returns true if the code is changed, and false otherwise.
void TaintRelaxedLoads(LoadInst* LI) {
  IRBuilder<true, NoFolder> Builder(LI->getNextNode());
  auto* FakeCondition = dyn_cast<Instruction>(Builder.CreateICmp(
      CmpInst::ICMP_EQ, LI, Constant::getNullValue(LI->getType())));
  AddFakeConditionalBranch(FakeCondition->getNextNode(), FakeCondition);
}

// XXX-comment: Returns whether the code has been changed.
bool AddFakeConditionalBranchAfterMonotonicLoads(
    const SmallVector<LoadInst*, 1>& MonotonicLoadInsts) {
  bool Changed = false;
  for (auto* LI : MonotonicLoadInsts) {
    auto* FirstInst = findFirstStoreCondBranchInst(LI);
    if (FirstInst != nullptr) {
      if (FirstInst->getOpcode() == Instruction::Store) {
        if (StoreAddressDependOnValue(dyn_cast<StoreInst>(FirstInst), LI)) {
          continue;
        }
      } else if (FirstInst->getOpcode() == Instruction::Br) {
        if (ConditionalBranchDependsOnValue(dyn_cast<BranchInst>(FirstInst),
                                            LI)) {
          continue;
        }
      } else {
        dbgs() << "FirstInst=" << *FirstInst << "\n";
        assert(false && "findFirstStoreCondBranchInst() should return a "
                        "store/condition branch instruction");
      }
    }

    // We really need to process the relaxed load now.
    TaintRelaxedLoads(LI);
    Changed = true;
  }
  return Changed;
}

/**** Implementations of public methods for dependence tainting ****/
Value* GetUntaintedAddress(Value* CurrentAddress) {
  auto* OrAddress = getOrAddress(CurrentAddress);
  if (OrAddress == nullptr) {
    // Is it tainted by a select instruction?
    auto* Inst = dyn_cast<Instruction>(CurrentAddress);
    if (nullptr != Inst && Inst->getOpcode() == Instruction::Select) {
      // A selection instruction.
      if (Inst->getOperand(1) == Inst->getOperand(2)) {
        return Inst->getOperand(1);
      }
    }

    return CurrentAddress;
  }
  Value* ActualAddress = nullptr;

  auto* CastToInt = dyn_cast<Instruction>(OrAddress->getOperand(1));
  if (CastToInt && CastToInt->getOpcode() == Instruction::PtrToInt) {
    return CastToInt->getOperand(0);
  } else {
    // This should be a IntToPtr constant expression.
    ConstantExpr* PtrToIntExpr =
        dyn_cast<ConstantExpr>(OrAddress->getOperand(1));
    if (PtrToIntExpr && PtrToIntExpr->getOpcode() == Instruction::PtrToInt) {
      return PtrToIntExpr->getOperand(0);
    }
  }

  // Looks like it's not been dependence-tainted. Returns itself.
  return CurrentAddress;
}

MemoryLocation GetUntaintedMemoryLocation(StoreInst* SI) {
  AAMDNodes AATags;
  SI->getAAMetadata(AATags);
  const auto& DL = SI->getModule()->getDataLayout();
  const auto* OriginalAddr = GetUntaintedAddress(SI->getPointerOperand());
  DEBUG(if (OriginalAddr != SI->getPointerOperand()) {
    dbgs() << "[GetUntaintedMemoryLocation]\n"
           << "Storing address: " << *SI->getPointerOperand()
           << "\nUntainted address: " << *OriginalAddr << "\n";
  });
  return MemoryLocation(OriginalAddr,
                        DL.getTypeStoreSize(SI->getValueOperand()->getType()),
                        AATags);
}

bool TaintDependenceToStore(StoreInst* SI, Value* DepVal) {
  if (dependenceSetInclusion(SI, DepVal)) {
    return false;
  }

  bool tainted = taintStoreAddress(SI, DepVal);
  assert(tainted);
  return tainted;
}

bool TaintDependenceToStoreAddress(StoreInst* SI, Value* DepVal) {
  if (dependenceSetInclusion(SI->getPointerOperand(), DepVal)) {
    return false;
  }

  bool tainted = taintStoreAddress(SI, DepVal);
  assert(tainted);
  return tainted;
}

bool CompressTaintedStore(BasicBlock* BB) {
  // This function looks for windows of adajcent stores in 'BB' that satisfy the
  // following condition (and then do optimization):
  // *Addr(d1) = v1, d1 is a condition and is the only dependence the store's
  //                 address depends on && Dep(v1) includes Dep(d1);
  // *Addr(d2) = v2, d2 is a condition and is the only dependnece the store's
  //                 address depends on && Dep(v2) includes Dep(d2) &&
  //                 Dep(d2) includes Dep(d1);
  // ...
  // *Addr(dN) = vN, dN is a condition and is the only dependence the store's
  //                 address depends on && Dep(dN) includes Dep(d"N-1").
  //
  // As a result, Dep(dN) includes [Dep(d1) V ... V Dep(d"N-1")], so we can
  // safely transform the above to the following. In between these stores, we
  // can omit untainted stores to the same address 'Addr' since they internally
  // have dependence on the previous stores on the same address.
  // =>
  // *Addr = v1
  // *Addr = v2
  // *Addr(d3) = v3
  for (auto BI = BB->begin(), BE = BB->end(); BI != BE; BI++) {
    // Look for the first store in such a window of adajacent stores.
    auto* FirstSI = dyn_cast<StoreInst>(&*BI);
    if (!FirstSI) {
      continue;
    }

    // The first store in the window must be tainted.
    auto* UntaintedAddress = GetUntaintedAddress(FirstSI->getPointerOperand());
    if (UntaintedAddress == FirstSI->getPointerOperand()) {
      continue;
    }

    // The first store's address must directly depend on and only depend on a
    // condition.
    auto* FirstSIDepCond = getConditionDependence(FirstSI->getPointerOperand());
    if (nullptr == FirstSIDepCond) {
      continue;
    }

    // Dep(first store's storing value) includes Dep(tainted dependence).
    if (!dependenceSetInclusion(FirstSI->getValueOperand(), FirstSIDepCond)) {
      continue;
    }

    // Look for subsequent stores to the same address that satisfy the condition
    // of "compressing the dependence".
    SmallVector<StoreInst*, 8> AdajacentStores;
    AdajacentStores.push_back(FirstSI);
    auto BII = BasicBlock::iterator(FirstSI);
    for (BII++; BII != BE; BII++) {
      auto* CurrSI = dyn_cast<StoreInst>(&*BII);
      if (!CurrSI) {
        if (BII->mayHaveSideEffects()) {
          // Be conservative. Instructions with side effects are similar to
          // stores.
          break;
        }
        continue;
      }

      auto* OrigAddress = GetUntaintedAddress(CurrSI->getPointerOperand());
      auto* CurrSIDepCond = getConditionDependence(CurrSI->getPointerOperand());
      // All other stores must satisfy either:
      // A. 'CurrSI' is an untainted store to the same address, or
      // B. the combination of the following 5 subconditions:
      // 1. Tainted;
      // 2. Untainted address is the same as the group's address;
      // 3. The address is tainted with a sole value which is a condition;
      // 4. The storing value depends on the condition in 3.
      // 5. The condition in 3 depends on the previous stores dependence
      // condition.

      // Condition A. Should ignore this store directly.
      if (OrigAddress == CurrSI->getPointerOperand() &&
          OrigAddress == UntaintedAddress) {
        continue;
      }
      // Check condition B.
      Value* Cond = nullptr;
      if (OrigAddress == CurrSI->getPointerOperand() ||
          OrigAddress != UntaintedAddress || CurrSIDepCond == nullptr ||
          !dependenceSetInclusion(CurrSI->getValueOperand(), CurrSIDepCond)) {
        // Check condition 1, 2, 3 & 4.
        break;
      }

      // Check condition 5.
      StoreInst* PrevSI = AdajacentStores[AdajacentStores.size() - 1];
      auto* PrevSIDepCond = getConditionDependence(PrevSI->getPointerOperand());
      assert(PrevSIDepCond &&
             "Store in the group must already depend on a condtion");
      if (!dependenceSetInclusion(CurrSIDepCond, PrevSIDepCond)) {
        break;
      }

      AdajacentStores.push_back(CurrSI);
    }

    if (AdajacentStores.size() == 1) {
      // The outer loop should keep looking from the next store.
      continue;
    }

    // Now we have such a group of tainted stores to the same address.
    DEBUG(dbgs() << "[CompressTaintedStore]\n");
    DEBUG(dbgs() << "Original BB\n");
    DEBUG(dbgs() << *BB << '\n');
    auto* LastSI = AdajacentStores[AdajacentStores.size() - 1];
    for (unsigned i = 0; i < AdajacentStores.size() - 1; ++i) {
      auto* SI = AdajacentStores[i];

      // Use the original address for stores before the last one.
      SI->setOperand(1, UntaintedAddress);

      DEBUG(dbgs() << "Store address has been reversed: " << *SI << '\n';);
    }
    // XXX-comment: Try to make the last store use fewer registers.
    // If LastSI's storing value is a select based on the condition with which
    // its address is tainted, transform the tainted address to a select
    // instruction, as follows:
    // r1 = Select Cond ? A : B
    // r2 = Cond & 0
    // r3 = Addr | r2
    // *r3 = r1
    // ==>
    // r1 = Select Cond ? A : B
    // r2 = Select Cond ? Addr : Addr
    // *r2 = r1
    // The idea is that both Select instructions depend on the same condition,
    // so hopefully the backend can generate two cmov instructions for them (and
    // this saves the number of registers needed).
    auto* LastSIDep = getConditionDependence(LastSI->getPointerOperand());
    auto* LastSIValue = dyn_cast<Instruction>(LastSI->getValueOperand());
    if (LastSIValue && LastSIValue->getOpcode() == Instruction::Select &&
        LastSIValue->getOperand(0) == LastSIDep) {
      // XXX-comment: Maybe it's better for us to just leave it as an and/or
      // dependence pattern.
      //      /*
      IRBuilder<true, NoFolder> Builder(LastSI);
      auto* Address =
          Builder.CreateSelect(LastSIDep, UntaintedAddress, UntaintedAddress);
      LastSI->setOperand(1, Address);
      DEBUG(dbgs() << "The last store becomes :" << *LastSI << "\n\n";);
      //      */
    }
  }

  return true;
}

bool PassDependenceToStore(Value* OldAddress, StoreInst* NewStore) {
  Value* OldDep = getDependence(OldAddress);
  // Return false when there's no dependence to pass from the OldAddress.
  if (!OldDep) {
    return false;
  }

  // No need to pass the dependence to NewStore's address if it already depends
  // on whatever 'OldAddress' depends on.
  if (StoreAddressDependOnValue(NewStore, OldDep)) {
    return false;
  }
  return taintStoreAddress(NewStore, OldAddress);
}

SmallSet<Value*, 8> FindDependence(Value* Val) {
  SmallSet<Value*, 8> DepSet;
  recursivelyFindDependence(&DepSet, Val, true /*Only insert leaf nodes*/);
  return DepSet;
}

bool StoreAddressDependOnValue(StoreInst* SI, Value* DepVal) {
  return dependenceSetInclusion(SI->getPointerOperand(), DepVal);
}

bool StoreDependOnValue(StoreInst* SI, Value* Dep) {
  return dependenceSetInclusion(SI, Dep);
}

} // namespace



bool CodeGenPrepare::runOnFunction(Function &F) {
  bool EverMadeChange = false;

  if (skipOptnoneFunction(F))
    return false;

  DL = &F.getParent()->getDataLayout();

  // Clear per function information.
  InsertedInsts.clear();
  PromotedInsts.clear();

  ModifiedDT = false;
  if (TM)
    TLI = TM->getSubtargetImpl(F)->getTargetLowering();
  TLInfo = &getAnalysis<TargetLibraryInfoWrapperPass>().getTLI();
  TTI = &getAnalysis<TargetTransformInfoWrapperPass>().getTTI(F);
  OptSize = F.optForSize();

  /// This optimization identifies DIV instructions that can be
  /// profitably bypassed and carried out with a shorter, faster divide.
  if (!OptSize && TLI && TLI->isSlowDivBypassed()) {
    const DenseMap<unsigned int, unsigned int> &BypassWidths =
       TLI->getBypassSlowDivWidths();
    BasicBlock* BB = &*F.begin();
    while (BB != nullptr) {
      // bypassSlowDivision may create new BBs, but we don't want to reapply the
      // optimization to those blocks.
      BasicBlock* Next = BB->getNextNode();
      EverMadeChange |= bypassSlowDivision(BB, BypassWidths);
      BB = Next;
    }
  }

  // Eliminate blocks that contain only PHI nodes and an
  // unconditional branch.
  EverMadeChange |= eliminateMostlyEmptyBlocks(F);

  // llvm.dbg.value is far away from the value then iSel may not be able
  // handle it properly. iSel will drop llvm.dbg.value if it can not
  // find a node corresponding to the value.
  EverMadeChange |= placeDbgValues(F);

  // If there is a mask, compare against zero, and branch that can be combined
  // into a single target instruction, push the mask and compare into branch
  // users. Do this before OptimizeBlock -> OptimizeInst ->
  // OptimizeCmpExpression, which perturbs the pattern being searched for.
  if (!DisableBranchOpts) {
    EverMadeChange |= sinkAndCmp(F);
    EverMadeChange |= splitBranchCondition(F);
  }

  bool MadeChange = true;
  while (MadeChange) {
    MadeChange = false;
    for (Function::iterator I = F.begin(); I != F.end(); ) {
      BasicBlock *BB = &*I++;
      bool ModifiedDTOnIteration = false;
      MadeChange |= optimizeBlock(*BB, ModifiedDTOnIteration);

      // Restart BB iteration if the dominator tree of the Function was changed
      if (ModifiedDTOnIteration)
        break;
    }
    EverMadeChange |= MadeChange;
  }

  SunkAddrs.clear();

  if (!DisableBranchOpts) {
    MadeChange = false;
    SmallPtrSet<BasicBlock*, 8> WorkList;
    for (BasicBlock &BB : F) {
      SmallVector<BasicBlock *, 2> Successors(succ_begin(&BB), succ_end(&BB));
      MadeChange |= ConstantFoldTerminator(&BB, true);
      if (!MadeChange) continue;

      for (SmallVectorImpl<BasicBlock*>::iterator
             II = Successors.begin(), IE = Successors.end(); II != IE; ++II)
        if (pred_begin(*II) == pred_end(*II))
          WorkList.insert(*II);
    }

    // Delete the dead blocks and any of their dead successors.
    MadeChange |= !WorkList.empty();
    while (!WorkList.empty()) {
      BasicBlock *BB = *WorkList.begin();
      WorkList.erase(BB);
      SmallVector<BasicBlock*, 2> Successors(succ_begin(BB), succ_end(BB));

      DeleteDeadBlock(BB);

      for (SmallVectorImpl<BasicBlock*>::iterator
             II = Successors.begin(), IE = Successors.end(); II != IE; ++II)
        if (pred_begin(*II) == pred_end(*II))
          WorkList.insert(*II);
    }

    // Merge pairs of basic blocks with unconditional branches, connected by
    // a single edge.
    if (EverMadeChange || MadeChange)
      MadeChange |= eliminateFallThrough(F);

    EverMadeChange |= MadeChange;
  }

  if (!DisableGCOpts) {
    SmallVector<Instruction *, 2> Statepoints;
    for (BasicBlock &BB : F)
      for (Instruction &I : BB)
        if (isStatepoint(I))
          Statepoints.push_back(&I);
    for (auto &I : Statepoints)
      EverMadeChange |= simplifyOffsetableRelocate(*I);
  }

  // XXX-comment: Delay dealing with relaxed loads in this function to avoid
  // further changes done by other passes (e.g., SimplifyCFG).
  // Collect all the relaxed loads.
  SmallVector<LoadInst*, 1> MonotonicLoadInsts;
  for (inst_iterator I = inst_begin(F), E = inst_end(F); I != E; ++I) {
    if (I->isAtomic()) {
      switch (I->getOpcode()) {
        case Instruction::Load: {
          auto* LI = dyn_cast<LoadInst>(&*I);
          if (LI->getOrdering() == Monotonic) {
            MonotonicLoadInsts.push_back(LI);
          }
          break;
        }
        default: {
          break;
        }
      }
    }
  }
  EverMadeChange |=
      AddFakeConditionalBranchAfterMonotonicLoads(MonotonicLoadInsts);

  return EverMadeChange;
}

/// Merge basic blocks which are connected by a single edge, where one of the
/// basic blocks has a single successor pointing to the other basic block,
/// which has a single predecessor.
bool CodeGenPrepare::eliminateFallThrough(Function &F) {
  bool Changed = false;
  // Scan all of the blocks in the function, except for the entry block.
  for (Function::iterator I = std::next(F.begin()), E = F.end(); I != E;) {
    BasicBlock *BB = &*I++;
    // If the destination block has a single pred, then this is a trivial
    // edge, just collapse it.
    BasicBlock *SinglePred = BB->getSinglePredecessor();

    // Don't merge if BB's address is taken.
    if (!SinglePred || SinglePred == BB || BB->hasAddressTaken()) continue;

    BranchInst *Term = dyn_cast<BranchInst>(SinglePred->getTerminator());
    if (Term && !Term->isConditional()) {
      Changed = true;
      DEBUG(dbgs() << "To merge:\n"<< *SinglePred << "\n\n\n");
      // Remember if SinglePred was the entry block of the function.
      // If so, we will need to move BB back to the entry position.
      bool isEntry = SinglePred == &SinglePred->getParent()->getEntryBlock();
      MergeBasicBlockIntoOnlyPred(BB, nullptr);

      if (isEntry && BB != &BB->getParent()->getEntryBlock())
        BB->moveBefore(&BB->getParent()->getEntryBlock());

      // We have erased a block. Update the iterator.
      I = BB->getIterator();
    }
  }
  return Changed;
}

/// Eliminate blocks that contain only PHI nodes, debug info directives, and an
/// unconditional branch. Passes before isel (e.g. LSR/loopsimplify) often split
/// edges in ways that are non-optimal for isel. Start by eliminating these
/// blocks so we can split them the way we want them.
bool CodeGenPrepare::eliminateMostlyEmptyBlocks(Function &F) {
  bool MadeChange = false;
  // Note that this intentionally skips the entry block.
  for (Function::iterator I = std::next(F.begin()), E = F.end(); I != E;) {
    BasicBlock *BB = &*I++;
    // If this block doesn't end with an uncond branch, ignore it.
    BranchInst *BI = dyn_cast<BranchInst>(BB->getTerminator());
    if (!BI || !BI->isUnconditional())
      continue;

    // If the instruction before the branch (skipping debug info) isn't a phi
    // node, then other stuff is happening here.
    BasicBlock::iterator BBI = BI->getIterator();
    if (BBI != BB->begin()) {
      --BBI;
      while (isa<DbgInfoIntrinsic>(BBI)) {
        if (BBI == BB->begin())
          break;
        --BBI;
      }
      if (!isa<DbgInfoIntrinsic>(BBI) && !isa<PHINode>(BBI))
        continue;
    }

    // Do not break infinite loops.
    BasicBlock *DestBB = BI->getSuccessor(0);
    if (DestBB == BB)
      continue;

    if (!canMergeBlocks(BB, DestBB))
      continue;

    eliminateMostlyEmptyBlock(BB);
    MadeChange = true;
  }
  return MadeChange;
}

/// Return true if we can merge BB into DestBB if there is a single
/// unconditional branch between them, and BB contains no other non-phi
/// instructions.
bool CodeGenPrepare::canMergeBlocks(const BasicBlock *BB,
                                    const BasicBlock *DestBB) const {
  // We only want to eliminate blocks whose phi nodes are used by phi nodes in
  // the successor.  If there are more complex condition (e.g. preheaders),
  // don't mess around with them.
  BasicBlock::const_iterator BBI = BB->begin();
  while (const PHINode *PN = dyn_cast<PHINode>(BBI++)) {
    for (const User *U : PN->users()) {
      const Instruction *UI = cast<Instruction>(U);
      if (UI->getParent() != DestBB || !isa<PHINode>(UI))
        return false;
      // If User is inside DestBB block and it is a PHINode then check
      // incoming value. If incoming value is not from BB then this is
      // a complex condition (e.g. preheaders) we want to avoid here.
      if (UI->getParent() == DestBB) {
        if (const PHINode *UPN = dyn_cast<PHINode>(UI))
          for (unsigned I = 0, E = UPN->getNumIncomingValues(); I != E; ++I) {
            Instruction *Insn = dyn_cast<Instruction>(UPN->getIncomingValue(I));
            if (Insn && Insn->getParent() == BB &&
                Insn->getParent() != UPN->getIncomingBlock(I))
              return false;
          }
      }
    }
  }

  // If BB and DestBB contain any common predecessors, then the phi nodes in BB
  // and DestBB may have conflicting incoming values for the block.  If so, we
  // can't merge the block.
  const PHINode *DestBBPN = dyn_cast<PHINode>(DestBB->begin());
  if (!DestBBPN) return true;  // no conflict.

  // Collect the preds of BB.
  SmallPtrSet<const BasicBlock*, 16> BBPreds;
  if (const PHINode *BBPN = dyn_cast<PHINode>(BB->begin())) {
    // It is faster to get preds from a PHI than with pred_iterator.
    for (unsigned i = 0, e = BBPN->getNumIncomingValues(); i != e; ++i)
      BBPreds.insert(BBPN->getIncomingBlock(i));
  } else {
    BBPreds.insert(pred_begin(BB), pred_end(BB));
  }

  // Walk the preds of DestBB.
  for (unsigned i = 0, e = DestBBPN->getNumIncomingValues(); i != e; ++i) {
    BasicBlock *Pred = DestBBPN->getIncomingBlock(i);
    if (BBPreds.count(Pred)) {   // Common predecessor?
      BBI = DestBB->begin();
      while (const PHINode *PN = dyn_cast<PHINode>(BBI++)) {
        const Value *V1 = PN->getIncomingValueForBlock(Pred);
        const Value *V2 = PN->getIncomingValueForBlock(BB);

        // If V2 is a phi node in BB, look up what the mapped value will be.
        if (const PHINode *V2PN = dyn_cast<PHINode>(V2))
          if (V2PN->getParent() == BB)
            V2 = V2PN->getIncomingValueForBlock(Pred);

        // If there is a conflict, bail out.
        if (V1 != V2) return false;
      }
    }
  }

  return true;
}


/// Eliminate a basic block that has only phi's and an unconditional branch in
/// it.
void CodeGenPrepare::eliminateMostlyEmptyBlock(BasicBlock *BB) {
  BranchInst *BI = cast<BranchInst>(BB->getTerminator());
  BasicBlock *DestBB = BI->getSuccessor(0);

  DEBUG(dbgs() << "MERGING MOSTLY EMPTY BLOCKS - BEFORE:\n" << *BB << *DestBB);

  // If the destination block has a single pred, then this is a trivial edge,
  // just collapse it.
  if (BasicBlock *SinglePred = DestBB->getSinglePredecessor()) {
    if (SinglePred != DestBB) {
      // Remember if SinglePred was the entry block of the function.  If so, we
      // will need to move BB back to the entry position.
      bool isEntry = SinglePred == &SinglePred->getParent()->getEntryBlock();
      MergeBasicBlockIntoOnlyPred(DestBB, nullptr);

      if (isEntry && BB != &BB->getParent()->getEntryBlock())
        BB->moveBefore(&BB->getParent()->getEntryBlock());

      DEBUG(dbgs() << "AFTER:\n" << *DestBB << "\n\n\n");
      return;
    }
  }

  // Otherwise, we have multiple predecessors of BB.  Update the PHIs in DestBB
  // to handle the new incoming edges it is about to have.
  PHINode *PN;
  for (BasicBlock::iterator BBI = DestBB->begin();
       (PN = dyn_cast<PHINode>(BBI)); ++BBI) {
    // Remove the incoming value for BB, and remember it.
    Value *InVal = PN->removeIncomingValue(BB, false);

    // Two options: either the InVal is a phi node defined in BB or it is some
    // value that dominates BB.
    PHINode *InValPhi = dyn_cast<PHINode>(InVal);
    if (InValPhi && InValPhi->getParent() == BB) {
      // Add all of the input values of the input PHI as inputs of this phi.
      for (unsigned i = 0, e = InValPhi->getNumIncomingValues(); i != e; ++i)
        PN->addIncoming(InValPhi->getIncomingValue(i),
                        InValPhi->getIncomingBlock(i));
    } else {
      // Otherwise, add one instance of the dominating value for each edge that
      // we will be adding.
      if (PHINode *BBPN = dyn_cast<PHINode>(BB->begin())) {
        for (unsigned i = 0, e = BBPN->getNumIncomingValues(); i != e; ++i)
          PN->addIncoming(InVal, BBPN->getIncomingBlock(i));
      } else {
        for (pred_iterator PI = pred_begin(BB), E = pred_end(BB); PI != E; ++PI)
          PN->addIncoming(InVal, *PI);
      }
    }
  }

  // The PHIs are now updated, change everything that refers to BB to use
  // DestBB and remove BB.
  BB->replaceAllUsesWith(DestBB);
  BB->eraseFromParent();
  ++NumBlocksElim;

  DEBUG(dbgs() << "AFTER:\n" << *DestBB << "\n\n\n");
}

// Computes a map of base pointer relocation instructions to corresponding
// derived pointer relocation instructions given a vector of all relocate calls
static void computeBaseDerivedRelocateMap(
    const SmallVectorImpl<GCRelocateInst *> &AllRelocateCalls,
    DenseMap<GCRelocateInst *, SmallVector<GCRelocateInst *, 2>>
        &RelocateInstMap) {
  // Collect information in two maps: one primarily for locating the base object
  // while filling the second map; the second map is the final structure holding
  // a mapping between Base and corresponding Derived relocate calls
  DenseMap<std::pair<unsigned, unsigned>, GCRelocateInst *> RelocateIdxMap;
  for (auto *ThisRelocate : AllRelocateCalls) {
    auto K = std::make_pair(ThisRelocate->getBasePtrIndex(),
                            ThisRelocate->getDerivedPtrIndex());
    RelocateIdxMap.insert(std::make_pair(K, ThisRelocate));
  }
  for (auto &Item : RelocateIdxMap) {
    std::pair<unsigned, unsigned> Key = Item.first;
    if (Key.first == Key.second)
      // Base relocation: nothing to insert
      continue;

    GCRelocateInst *I = Item.second;
    auto BaseKey = std::make_pair(Key.first, Key.first);

    // We're iterating over RelocateIdxMap so we cannot modify it.
    auto MaybeBase = RelocateIdxMap.find(BaseKey);
    if (MaybeBase == RelocateIdxMap.end())
      // TODO: We might want to insert a new base object relocate and gep off
      // that, if there are enough derived object relocates.
      continue;

    RelocateInstMap[MaybeBase->second].push_back(I);
  }
}

// Accepts a GEP and extracts the operands into a vector provided they're all
// small integer constants
static bool getGEPSmallConstantIntOffsetV(GetElementPtrInst *GEP,
                                          SmallVectorImpl<Value *> &OffsetV) {
  for (unsigned i = 1; i < GEP->getNumOperands(); i++) {
    // Only accept small constant integer operands
    auto Op = dyn_cast<ConstantInt>(GEP->getOperand(i));
    if (!Op || Op->getZExtValue() > 20)
      return false;
  }

  for (unsigned i = 1; i < GEP->getNumOperands(); i++)
    OffsetV.push_back(GEP->getOperand(i));
  return true;
}

// Takes a RelocatedBase (base pointer relocation instruction) and Targets to
// replace, computes a replacement, and affects it.
static bool
simplifyRelocatesOffABase(GCRelocateInst *RelocatedBase,
                          const SmallVectorImpl<GCRelocateInst *> &Targets) {
  bool MadeChange = false;
  for (GCRelocateInst *ToReplace : Targets) {
    assert(ToReplace->getBasePtrIndex() == RelocatedBase->getBasePtrIndex() &&
           "Not relocating a derived object of the original base object");
    if (ToReplace->getBasePtrIndex() == ToReplace->getDerivedPtrIndex()) {
      // A duplicate relocate call. TODO: coalesce duplicates.
      continue;
    }

    if (RelocatedBase->getParent() != ToReplace->getParent()) {
      // Base and derived relocates are in different basic blocks.
      // In this case transform is only valid when base dominates derived
      // relocate. However it would be too expensive to check dominance
      // for each such relocate, so we skip the whole transformation.
      continue;
    }

    Value *Base = ToReplace->getBasePtr();
    auto Derived = dyn_cast<GetElementPtrInst>(ToReplace->getDerivedPtr());
    if (!Derived || Derived->getPointerOperand() != Base)
      continue;

    SmallVector<Value *, 2> OffsetV;
    if (!getGEPSmallConstantIntOffsetV(Derived, OffsetV))
      continue;

    // Create a Builder and replace the target callsite with a gep
    assert(RelocatedBase->getNextNode() && "Should always have one since it's not a terminator");

    // Insert after RelocatedBase
    IRBuilder<> Builder(RelocatedBase->getNextNode());
    Builder.SetCurrentDebugLocation(ToReplace->getDebugLoc());

    // If gc_relocate does not match the actual type, cast it to the right type.
    // In theory, there must be a bitcast after gc_relocate if the type does not
    // match, and we should reuse it to get the derived pointer. But it could be
    // cases like this:
    // bb1:
    //  ...
    //  %g1 = call coldcc i8 addrspace(1)* @llvm.experimental.gc.relocate.p1i8(...)
    //  br label %merge
    //
    // bb2:
    //  ...
    //  %g2 = call coldcc i8 addrspace(1)* @llvm.experimental.gc.relocate.p1i8(...)
    //  br label %merge
    //
    // merge:
    //  %p1 = phi i8 addrspace(1)* [ %g1, %bb1 ], [ %g2, %bb2 ]
    //  %cast = bitcast i8 addrspace(1)* %p1 in to i32 addrspace(1)*
    //
    // In this case, we can not find the bitcast any more. So we insert a new bitcast
    // no matter there is already one or not. In this way, we can handle all cases, and
    // the extra bitcast should be optimized away in later passes.
    Value *ActualRelocatedBase = RelocatedBase;
    if (RelocatedBase->getType() != Base->getType()) {
      ActualRelocatedBase =
          Builder.CreateBitCast(RelocatedBase, Base->getType());
    }
    Value *Replacement = Builder.CreateGEP(
        Derived->getSourceElementType(), ActualRelocatedBase, makeArrayRef(OffsetV));
    Replacement->takeName(ToReplace);
    // If the newly generated derived pointer's type does not match the original derived
    // pointer's type, cast the new derived pointer to match it. Same reasoning as above.
    Value *ActualReplacement = Replacement;
    if (Replacement->getType() != ToReplace->getType()) {
      ActualReplacement =
          Builder.CreateBitCast(Replacement, ToReplace->getType());
    }
    ToReplace->replaceAllUsesWith(ActualReplacement);
    ToReplace->eraseFromParent();

    MadeChange = true;
  }
  return MadeChange;
}

// Turns this:
//
// %base = ...
// %ptr = gep %base + 15
// %tok = statepoint (%fun, i32 0, i32 0, i32 0, %base, %ptr)
// %base' = relocate(%tok, i32 4, i32 4)
// %ptr' = relocate(%tok, i32 4, i32 5)
// %val = load %ptr'
//
// into this:
//
// %base = ...
// %ptr = gep %base + 15
// %tok = statepoint (%fun, i32 0, i32 0, i32 0, %base, %ptr)
// %base' = gc.relocate(%tok, i32 4, i32 4)
// %ptr' = gep %base' + 15
// %val = load %ptr'
bool CodeGenPrepare::simplifyOffsetableRelocate(Instruction &I) {
  bool MadeChange = false;
  SmallVector<GCRelocateInst *, 2> AllRelocateCalls;

  for (auto *U : I.users())
    if (GCRelocateInst *Relocate = dyn_cast<GCRelocateInst>(U))
      // Collect all the relocate calls associated with a statepoint
      AllRelocateCalls.push_back(Relocate);

  // We need atleast one base pointer relocation + one derived pointer
  // relocation to mangle
  if (AllRelocateCalls.size() < 2)
    return false;

  // RelocateInstMap is a mapping from the base relocate instruction to the
  // corresponding derived relocate instructions
  DenseMap<GCRelocateInst *, SmallVector<GCRelocateInst *, 2>> RelocateInstMap;
  computeBaseDerivedRelocateMap(AllRelocateCalls, RelocateInstMap);
  if (RelocateInstMap.empty())
    return false;

  for (auto &Item : RelocateInstMap)
    // Item.first is the RelocatedBase to offset against
    // Item.second is the vector of Targets to replace
    MadeChange = simplifyRelocatesOffABase(Item.first, Item.second);
  return MadeChange;
}

/// SinkCast - Sink the specified cast instruction into its user blocks
static bool SinkCast(CastInst *CI) {
  BasicBlock *DefBB = CI->getParent();

  /// InsertedCasts - Only insert a cast in each block once.
  DenseMap<BasicBlock*, CastInst*> InsertedCasts;

  bool MadeChange = false;
  for (Value::user_iterator UI = CI->user_begin(), E = CI->user_end();
       UI != E; ) {
    Use &TheUse = UI.getUse();
    Instruction *User = cast<Instruction>(*UI);

    // Figure out which BB this cast is used in.  For PHI's this is the
    // appropriate predecessor block.
    BasicBlock *UserBB = User->getParent();
    if (PHINode *PN = dyn_cast<PHINode>(User)) {
      UserBB = PN->getIncomingBlock(TheUse);
    }

    // Preincrement use iterator so we don't invalidate it.
    ++UI;

    // If the block selected to receive the cast is an EH pad that does not
    // allow non-PHI instructions before the terminator, we can't sink the
    // cast.
    if (UserBB->getTerminator()->isEHPad())
      continue;

    // If this user is in the same block as the cast, don't change the cast.
    if (UserBB == DefBB) continue;

    // If we have already inserted a cast into this block, use it.
    CastInst *&InsertedCast = InsertedCasts[UserBB];

    if (!InsertedCast) {
      BasicBlock::iterator InsertPt = UserBB->getFirstInsertionPt();
      assert(InsertPt != UserBB->end());
      InsertedCast = CastInst::Create(CI->getOpcode(), CI->getOperand(0),
                                      CI->getType(), "", &*InsertPt);
    }

    // Replace a use of the cast with a use of the new cast.
    TheUse = InsertedCast;
    MadeChange = true;
    ++NumCastUses;
  }

  // If we removed all uses, nuke the cast.
  if (CI->use_empty()) {
    CI->eraseFromParent();
    MadeChange = true;
  }

  return MadeChange;
}

/// If the specified cast instruction is a noop copy (e.g. it's casting from
/// one pointer type to another, i32->i8 on PPC), sink it into user blocks to
/// reduce the number of virtual registers that must be created and coalesced.
///
/// Return true if any changes are made.
///
static bool OptimizeNoopCopyExpression(CastInst *CI, const TargetLowering &TLI,
                                       const DataLayout &DL) {
  // If this is a noop copy,
  EVT SrcVT = TLI.getValueType(DL, CI->getOperand(0)->getType());
  EVT DstVT = TLI.getValueType(DL, CI->getType());

  // This is an fp<->int conversion?
  if (SrcVT.isInteger() != DstVT.isInteger())
    return false;

  // If this is an extension, it will be a zero or sign extension, which
  // isn't a noop.
  if (SrcVT.bitsLT(DstVT)) return false;

  // If these values will be promoted, find out what they will be promoted
  // to.  This helps us consider truncates on PPC as noop copies when they
  // are.
  if (TLI.getTypeAction(CI->getContext(), SrcVT) ==
      TargetLowering::TypePromoteInteger)
    SrcVT = TLI.getTypeToTransformTo(CI->getContext(), SrcVT);
  if (TLI.getTypeAction(CI->getContext(), DstVT) ==
      TargetLowering::TypePromoteInteger)
    DstVT = TLI.getTypeToTransformTo(CI->getContext(), DstVT);

  // If, after promotion, these are the same types, this is a noop copy.
  if (SrcVT != DstVT)
    return false;

  return SinkCast(CI);
}

/// Try to combine CI into a call to the llvm.uadd.with.overflow intrinsic if
/// possible.
///
/// Return true if any changes were made.
static bool CombineUAddWithOverflow(CmpInst *CI) {
  Value *A, *B;
  Instruction *AddI;
  if (!match(CI,
             m_UAddWithOverflow(m_Value(A), m_Value(B), m_Instruction(AddI))))
    return false;

  Type *Ty = AddI->getType();
  if (!isa<IntegerType>(Ty))
    return false;

  // We don't want to move around uses of condition values this late, so we we
  // check if it is legal to create the call to the intrinsic in the basic
  // block containing the icmp:

  if (AddI->getParent() != CI->getParent() && !AddI->hasOneUse())
    return false;

#ifndef NDEBUG
  // Someday m_UAddWithOverflow may get smarter, but this is a safe assumption
  // for now:
  if (AddI->hasOneUse())
    assert(*AddI->user_begin() == CI && "expected!");
#endif

  Module *M = CI->getModule();
  Value *F = Intrinsic::getDeclaration(M, Intrinsic::uadd_with_overflow, Ty);

  auto *InsertPt = AddI->hasOneUse() ? CI : AddI;

  auto *UAddWithOverflow =
      CallInst::Create(F, {A, B}, "uadd.overflow", InsertPt);
  auto *UAdd = ExtractValueInst::Create(UAddWithOverflow, 0, "uadd", InsertPt);
  auto *Overflow =
      ExtractValueInst::Create(UAddWithOverflow, 1, "overflow", InsertPt);

  CI->replaceAllUsesWith(Overflow);
  AddI->replaceAllUsesWith(UAdd);
  CI->eraseFromParent();
  AddI->eraseFromParent();
  return true;
}

/// Sink the given CmpInst into user blocks to reduce the number of virtual
/// registers that must be created and coalesced. This is a clear win except on
/// targets with multiple condition code registers (PowerPC), where it might
/// lose; some adjustment may be wanted there.
///
/// Return true if any changes are made.
static bool SinkCmpExpression(CmpInst *CI) {
  BasicBlock *DefBB = CI->getParent();

  /// Only insert a cmp in each block once.
  DenseMap<BasicBlock*, CmpInst*> InsertedCmps;

  bool MadeChange = false;
  for (Value::user_iterator UI = CI->user_begin(), E = CI->user_end();
       UI != E; ) {
    Use &TheUse = UI.getUse();
    Instruction *User = cast<Instruction>(*UI);

    // Preincrement use iterator so we don't invalidate it.
    ++UI;

    // Don't bother for PHI nodes.
    if (isa<PHINode>(User))
      continue;

    // Figure out which BB this cmp is used in.
    BasicBlock *UserBB = User->getParent();

    // If this user is in the same block as the cmp, don't change the cmp.
    if (UserBB == DefBB) continue;

    // If we have already inserted a cmp into this block, use it.
    CmpInst *&InsertedCmp = InsertedCmps[UserBB];

    if (!InsertedCmp) {
      BasicBlock::iterator InsertPt = UserBB->getFirstInsertionPt();
      assert(InsertPt != UserBB->end());
      InsertedCmp =
          CmpInst::Create(CI->getOpcode(), CI->getPredicate(),
                          CI->getOperand(0), CI->getOperand(1), "", &*InsertPt);
    }

    // Replace a use of the cmp with a use of the new cmp.
    TheUse = InsertedCmp;
    MadeChange = true;
    ++NumCmpUses;
  }

  // If we removed all uses, nuke the cmp.
  if (CI->use_empty()) {
    CI->eraseFromParent();
    MadeChange = true;
  }

  return MadeChange;
}

static bool OptimizeCmpExpression(CmpInst *CI) {
  if (SinkCmpExpression(CI))
    return true;

  if (CombineUAddWithOverflow(CI))
    return true;

  return false;
}

/// Check if the candidates could be combined with a shift instruction, which
/// includes:
/// 1. Truncate instruction
/// 2. And instruction and the imm is a mask of the low bits:
/// imm & (imm+1) == 0
static bool isExtractBitsCandidateUse(Instruction *User) {
  if (!isa<TruncInst>(User)) {
    if (User->getOpcode() != Instruction::And ||
        !isa<ConstantInt>(User->getOperand(1)))
      return false;

    const APInt &Cimm = cast<ConstantInt>(User->getOperand(1))->getValue();

    if ((Cimm & (Cimm + 1)).getBoolValue())
      return false;
  }
  return true;
}

/// Sink both shift and truncate instruction to the use of truncate's BB.
static bool
SinkShiftAndTruncate(BinaryOperator *ShiftI, Instruction *User, ConstantInt *CI,
                     DenseMap<BasicBlock *, BinaryOperator *> &InsertedShifts,
                     const TargetLowering &TLI, const DataLayout &DL) {
  BasicBlock *UserBB = User->getParent();
  DenseMap<BasicBlock *, CastInst *> InsertedTruncs;
  TruncInst *TruncI = dyn_cast<TruncInst>(User);
  bool MadeChange = false;

  for (Value::user_iterator TruncUI = TruncI->user_begin(),
                            TruncE = TruncI->user_end();
       TruncUI != TruncE;) {

    Use &TruncTheUse = TruncUI.getUse();
    Instruction *TruncUser = cast<Instruction>(*TruncUI);
    // Preincrement use iterator so we don't invalidate it.

    ++TruncUI;

    int ISDOpcode = TLI.InstructionOpcodeToISD(TruncUser->getOpcode());
    if (!ISDOpcode)
      continue;

    // If the use is actually a legal node, there will not be an
    // implicit truncate.
    // FIXME: always querying the result type is just an
    // approximation; some nodes' legality is determined by the
    // operand or other means. There's no good way to find out though.
    if (TLI.isOperationLegalOrCustom(
            ISDOpcode, TLI.getValueType(DL, TruncUser->getType(), true)))
      continue;

    // Don't bother for PHI nodes.
    if (isa<PHINode>(TruncUser))
      continue;

    BasicBlock *TruncUserBB = TruncUser->getParent();

    if (UserBB == TruncUserBB)
      continue;

    BinaryOperator *&InsertedShift = InsertedShifts[TruncUserBB];
    CastInst *&InsertedTrunc = InsertedTruncs[TruncUserBB];

    if (!InsertedShift && !InsertedTrunc) {
      BasicBlock::iterator InsertPt = TruncUserBB->getFirstInsertionPt();
      assert(InsertPt != TruncUserBB->end());
      // Sink the shift
      if (ShiftI->getOpcode() == Instruction::AShr)
        InsertedShift = BinaryOperator::CreateAShr(ShiftI->getOperand(0), CI,
                                                   "", &*InsertPt);
      else
        InsertedShift = BinaryOperator::CreateLShr(ShiftI->getOperand(0), CI,
                                                   "", &*InsertPt);

      // Sink the trunc
      BasicBlock::iterator TruncInsertPt = TruncUserBB->getFirstInsertionPt();
      TruncInsertPt++;
      assert(TruncInsertPt != TruncUserBB->end());

      InsertedTrunc = CastInst::Create(TruncI->getOpcode(), InsertedShift,
                                       TruncI->getType(), "", &*TruncInsertPt);

      MadeChange = true;

      TruncTheUse = InsertedTrunc;
    }
  }
  return MadeChange;
}

/// Sink the shift *right* instruction into user blocks if the uses could
/// potentially be combined with this shift instruction and generate BitExtract
/// instruction. It will only be applied if the architecture supports BitExtract
/// instruction. Here is an example:
/// BB1:
///   %x.extract.shift = lshr i64 %arg1, 32
/// BB2:
///   %x.extract.trunc = trunc i64 %x.extract.shift to i16
/// ==>
///
/// BB2:
///   %x.extract.shift.1 = lshr i64 %arg1, 32
///   %x.extract.trunc = trunc i64 %x.extract.shift.1 to i16
///
/// CodeGen will recoginze the pattern in BB2 and generate BitExtract
/// instruction.
/// Return true if any changes are made.
static bool OptimizeExtractBits(BinaryOperator *ShiftI, ConstantInt *CI,
                                const TargetLowering &TLI,
                                const DataLayout &DL) {
  BasicBlock *DefBB = ShiftI->getParent();

  /// Only insert instructions in each block once.
  DenseMap<BasicBlock *, BinaryOperator *> InsertedShifts;

  bool shiftIsLegal = TLI.isTypeLegal(TLI.getValueType(DL, ShiftI->getType()));

  bool MadeChange = false;
  for (Value::user_iterator UI = ShiftI->user_begin(), E = ShiftI->user_end();
       UI != E;) {
    Use &TheUse = UI.getUse();
    Instruction *User = cast<Instruction>(*UI);
    // Preincrement use iterator so we don't invalidate it.
    ++UI;

    // Don't bother for PHI nodes.
    if (isa<PHINode>(User))
      continue;

    if (!isExtractBitsCandidateUse(User))
      continue;

    BasicBlock *UserBB = User->getParent();

    if (UserBB == DefBB) {
      // If the shift and truncate instruction are in the same BB. The use of
      // the truncate(TruncUse) may still introduce another truncate if not
      // legal. In this case, we would like to sink both shift and truncate
      // instruction to the BB of TruncUse.
      // for example:
      // BB1:
      // i64 shift.result = lshr i64 opnd, imm
      // trunc.result = trunc shift.result to i16
      //
      // BB2:
      //   ----> We will have an implicit truncate here if the architecture does
      //   not have i16 compare.
      // cmp i16 trunc.result, opnd2
      //
      if (isa<TruncInst>(User) && shiftIsLegal
          // If the type of the truncate is legal, no trucate will be
          // introduced in other basic blocks.
          &&
          (!TLI.isTypeLegal(TLI.getValueType(DL, User->getType()))))
        MadeChange =
            SinkShiftAndTruncate(ShiftI, User, CI, InsertedShifts, TLI, DL);

      continue;
    }
    // If we have already inserted a shift into this block, use it.
    BinaryOperator *&InsertedShift = InsertedShifts[UserBB];

    if (!InsertedShift) {
      BasicBlock::iterator InsertPt = UserBB->getFirstInsertionPt();
      assert(InsertPt != UserBB->end());

      if (ShiftI->getOpcode() == Instruction::AShr)
        InsertedShift = BinaryOperator::CreateAShr(ShiftI->getOperand(0), CI,
                                                   "", &*InsertPt);
      else
        InsertedShift = BinaryOperator::CreateLShr(ShiftI->getOperand(0), CI,
                                                   "", &*InsertPt);

      MadeChange = true;
    }

    // Replace a use of the shift with a use of the new shift.
    TheUse = InsertedShift;
  }

  // If we removed all uses, nuke the shift.
  if (ShiftI->use_empty())
    ShiftI->eraseFromParent();

  return MadeChange;
}

// Translate a masked load intrinsic like
// <16 x i32 > @llvm.masked.load( <16 x i32>* %addr, i32 align,
//                               <16 x i1> %mask, <16 x i32> %passthru)
// to a chain of basic blocks, with loading element one-by-one if
// the appropriate mask bit is set
//
//  %1 = bitcast i8* %addr to i32*
//  %2 = extractelement <16 x i1> %mask, i32 0
//  %3 = icmp eq i1 %2, true
//  br i1 %3, label %cond.load, label %else
//
//cond.load:                                        ; preds = %0
//  %4 = getelementptr i32* %1, i32 0
//  %5 = load i32* %4
//  %6 = insertelement <16 x i32> undef, i32 %5, i32 0
//  br label %else
//
//else:                                             ; preds = %0, %cond.load
//  %res.phi.else = phi <16 x i32> [ %6, %cond.load ], [ undef, %0 ]
//  %7 = extractelement <16 x i1> %mask, i32 1
//  %8 = icmp eq i1 %7, true
//  br i1 %8, label %cond.load1, label %else2
//
//cond.load1:                                       ; preds = %else
//  %9 = getelementptr i32* %1, i32 1
//  %10 = load i32* %9
//  %11 = insertelement <16 x i32> %res.phi.else, i32 %10, i32 1
//  br label %else2
//
//else2:                                            ; preds = %else, %cond.load1
//  %res.phi.else3 = phi <16 x i32> [ %11, %cond.load1 ], [ %res.phi.else, %else ]
//  %12 = extractelement <16 x i1> %mask, i32 2
//  %13 = icmp eq i1 %12, true
//  br i1 %13, label %cond.load4, label %else5
//
static void ScalarizeMaskedLoad(CallInst *CI) {
  Value *Ptr  = CI->getArgOperand(0);
  Value *Alignment = CI->getArgOperand(1);
  Value *Mask = CI->getArgOperand(2);
  Value *Src0 = CI->getArgOperand(3);

  unsigned AlignVal = cast<ConstantInt>(Alignment)->getZExtValue();
  VectorType *VecType = dyn_cast<VectorType>(CI->getType());
  assert(VecType && "Unexpected return type of masked load intrinsic");

  Type *EltTy = CI->getType()->getVectorElementType();

  IRBuilder<> Builder(CI->getContext());
  Instruction *InsertPt = CI;
  BasicBlock *IfBlock = CI->getParent();
  BasicBlock *CondBlock = nullptr;
  BasicBlock *PrevIfBlock = CI->getParent();

  Builder.SetInsertPoint(InsertPt);
  Builder.SetCurrentDebugLocation(CI->getDebugLoc());

  // Short-cut if the mask is all-true.
  bool IsAllOnesMask = isa<Constant>(Mask) &&
    cast<Constant>(Mask)->isAllOnesValue();

  if (IsAllOnesMask) {
    Value *NewI = Builder.CreateAlignedLoad(Ptr, AlignVal);
    CI->replaceAllUsesWith(NewI);
    CI->eraseFromParent();
    return;
  }

  // Adjust alignment for the scalar instruction.
  AlignVal = std::min(AlignVal, VecType->getScalarSizeInBits()/8);
  // Bitcast %addr fron i8* to EltTy*
  Type *NewPtrType =
    EltTy->getPointerTo(cast<PointerType>(Ptr->getType())->getAddressSpace());
  Value *FirstEltPtr = Builder.CreateBitCast(Ptr, NewPtrType);
  unsigned VectorWidth = VecType->getNumElements();

  Value *UndefVal = UndefValue::get(VecType);

  // The result vector
  Value *VResult = UndefVal;

  if (isa<ConstantVector>(Mask)) {
    for (unsigned Idx = 0; Idx < VectorWidth; ++Idx) {
      if (cast<ConstantVector>(Mask)->getOperand(Idx)->isNullValue())
          continue;
      Value *Gep =
          Builder.CreateInBoundsGEP(EltTy, FirstEltPtr, Builder.getInt32(Idx));
      LoadInst* Load = Builder.CreateAlignedLoad(Gep, AlignVal);
      VResult = Builder.CreateInsertElement(VResult, Load,
                                            Builder.getInt32(Idx));
    }
    Value *NewI = Builder.CreateSelect(Mask, VResult, Src0);
    CI->replaceAllUsesWith(NewI);
    CI->eraseFromParent();
    return;
  }

  PHINode *Phi = nullptr;
  Value *PrevPhi = UndefVal;

  for (unsigned Idx = 0; Idx < VectorWidth; ++Idx) {

    // Fill the "else" block, created in the previous iteration
    //
    //  %res.phi.else3 = phi <16 x i32> [ %11, %cond.load1 ], [ %res.phi.else, %else ]
    //  %mask_1 = extractelement <16 x i1> %mask, i32 Idx
    //  %to_load = icmp eq i1 %mask_1, true
    //  br i1 %to_load, label %cond.load, label %else
    //
    if (Idx > 0) {
      Phi = Builder.CreatePHI(VecType, 2, "res.phi.else");
      Phi->addIncoming(VResult, CondBlock);
      Phi->addIncoming(PrevPhi, PrevIfBlock);
      PrevPhi = Phi;
      VResult = Phi;
    }

    Value *Predicate = Builder.CreateExtractElement(Mask, Builder.getInt32(Idx));
    Value *Cmp = Builder.CreateICmp(ICmpInst::ICMP_EQ, Predicate,
                                    ConstantInt::get(Predicate->getType(), 1));

    // Create "cond" block
    //
    //  %EltAddr = getelementptr i32* %1, i32 0
    //  %Elt = load i32* %EltAddr
    //  VResult = insertelement <16 x i32> VResult, i32 %Elt, i32 Idx
    //
    CondBlock = IfBlock->splitBasicBlock(InsertPt->getIterator(), "cond.load");
    Builder.SetInsertPoint(InsertPt);

    Value *Gep =
        Builder.CreateInBoundsGEP(EltTy, FirstEltPtr, Builder.getInt32(Idx));
    LoadInst *Load = Builder.CreateAlignedLoad(Gep, AlignVal);
    VResult = Builder.CreateInsertElement(VResult, Load, Builder.getInt32(Idx));

    // Create "else" block, fill it in the next iteration
    BasicBlock *NewIfBlock =
        CondBlock->splitBasicBlock(InsertPt->getIterator(), "else");
    Builder.SetInsertPoint(InsertPt);
    Instruction *OldBr = IfBlock->getTerminator();
    BranchInst::Create(CondBlock, NewIfBlock, Cmp, OldBr);
    OldBr->eraseFromParent();
    PrevIfBlock = IfBlock;
    IfBlock = NewIfBlock;
  }

  Phi = Builder.CreatePHI(VecType, 2, "res.phi.select");
  Phi->addIncoming(VResult, CondBlock);
  Phi->addIncoming(PrevPhi, PrevIfBlock);
  Value *NewI = Builder.CreateSelect(Mask, Phi, Src0);
  CI->replaceAllUsesWith(NewI);
  CI->eraseFromParent();
}

// Translate a masked store intrinsic, like
// void @llvm.masked.store(<16 x i32> %src, <16 x i32>* %addr, i32 align,
//                               <16 x i1> %mask)
// to a chain of basic blocks, that stores element one-by-one if
// the appropriate mask bit is set
//
//   %1 = bitcast i8* %addr to i32*
//   %2 = extractelement <16 x i1> %mask, i32 0
//   %3 = icmp eq i1 %2, true
//   br i1 %3, label %cond.store, label %else
//
// cond.store:                                       ; preds = %0
//   %4 = extractelement <16 x i32> %val, i32 0
//   %5 = getelementptr i32* %1, i32 0
//   store i32 %4, i32* %5
//   br label %else
//
// else:                                             ; preds = %0, %cond.store
//   %6 = extractelement <16 x i1> %mask, i32 1
//   %7 = icmp eq i1 %6, true
//   br i1 %7, label %cond.store1, label %else2
//
// cond.store1:                                      ; preds = %else
//   %8 = extractelement <16 x i32> %val, i32 1
//   %9 = getelementptr i32* %1, i32 1
//   store i32 %8, i32* %9
//   br label %else2
//   . . .
static void ScalarizeMaskedStore(CallInst *CI) {
  Value *Src = CI->getArgOperand(0);
  Value *Ptr  = CI->getArgOperand(1);
  Value *Alignment = CI->getArgOperand(2);
  Value *Mask = CI->getArgOperand(3);

  unsigned AlignVal = cast<ConstantInt>(Alignment)->getZExtValue();
  VectorType *VecType = dyn_cast<VectorType>(Src->getType());
  assert(VecType && "Unexpected data type in masked store intrinsic");

  Type *EltTy = VecType->getElementType();

  IRBuilder<> Builder(CI->getContext());
  Instruction *InsertPt = CI;
  BasicBlock *IfBlock = CI->getParent();
  Builder.SetInsertPoint(InsertPt);
  Builder.SetCurrentDebugLocation(CI->getDebugLoc());

  // Short-cut if the mask is all-true.
  bool IsAllOnesMask = isa<Constant>(Mask) &&
    cast<Constant>(Mask)->isAllOnesValue();

  if (IsAllOnesMask) {
    Builder.CreateAlignedStore(Src, Ptr, AlignVal);
    CI->eraseFromParent();
    return;
  }

  // Adjust alignment for the scalar instruction.
  AlignVal = std::max(AlignVal, VecType->getScalarSizeInBits()/8);
  // Bitcast %addr fron i8* to EltTy*
  Type *NewPtrType =
    EltTy->getPointerTo(cast<PointerType>(Ptr->getType())->getAddressSpace());
  Value *FirstEltPtr = Builder.CreateBitCast(Ptr, NewPtrType);
  unsigned VectorWidth = VecType->getNumElements();

  if (isa<ConstantVector>(Mask)) {
    for (unsigned Idx = 0; Idx < VectorWidth; ++Idx) {
      if (cast<ConstantVector>(Mask)->getOperand(Idx)->isNullValue())
          continue;
      Value *OneElt = Builder.CreateExtractElement(Src, Builder.getInt32(Idx));
      Value *Gep =
          Builder.CreateInBoundsGEP(EltTy, FirstEltPtr, Builder.getInt32(Idx));
      Builder.CreateAlignedStore(OneElt, Gep, AlignVal);
    }
    CI->eraseFromParent();
    return;
  }

  for (unsigned Idx = 0; Idx < VectorWidth; ++Idx) {

    // Fill the "else" block, created in the previous iteration
    //
    //  %mask_1 = extractelement <16 x i1> %mask, i32 Idx
    //  %to_store = icmp eq i1 %mask_1, true
    //  br i1 %to_store, label %cond.store, label %else
    //
    Value *Predicate = Builder.CreateExtractElement(Mask, Builder.getInt32(Idx));
    Value *Cmp = Builder.CreateICmp(ICmpInst::ICMP_EQ, Predicate,
                                    ConstantInt::get(Predicate->getType(), 1));

    // Create "cond" block
    //
    //  %OneElt = extractelement <16 x i32> %Src, i32 Idx
    //  %EltAddr = getelementptr i32* %1, i32 0
    //  %store i32 %OneElt, i32* %EltAddr
    //
    BasicBlock *CondBlock =
        IfBlock->splitBasicBlock(InsertPt->getIterator(), "cond.store");
    Builder.SetInsertPoint(InsertPt);

    Value *OneElt = Builder.CreateExtractElement(Src, Builder.getInt32(Idx));
    Value *Gep =
        Builder.CreateInBoundsGEP(EltTy, FirstEltPtr, Builder.getInt32(Idx));
    Builder.CreateAlignedStore(OneElt, Gep, AlignVal);

    // Create "else" block, fill it in the next iteration
    BasicBlock *NewIfBlock =
        CondBlock->splitBasicBlock(InsertPt->getIterator(), "else");
    Builder.SetInsertPoint(InsertPt);
    Instruction *OldBr = IfBlock->getTerminator();
    BranchInst::Create(CondBlock, NewIfBlock, Cmp, OldBr);
    OldBr->eraseFromParent();
    IfBlock = NewIfBlock;
  }
  CI->eraseFromParent();
}

// Translate a masked gather intrinsic like
// <16 x i32 > @llvm.masked.gather.v16i32( <16 x i32*> %Ptrs, i32 4,
//                               <16 x i1> %Mask, <16 x i32> %Src)
// to a chain of basic blocks, with loading element one-by-one if
// the appropriate mask bit is set
//
// % Ptrs = getelementptr i32, i32* %base, <16 x i64> %ind
// % Mask0 = extractelement <16 x i1> %Mask, i32 0
// % ToLoad0 = icmp eq i1 % Mask0, true
// br i1 % ToLoad0, label %cond.load, label %else
//
// cond.load:
// % Ptr0 = extractelement <16 x i32*> %Ptrs, i32 0
// % Load0 = load i32, i32* % Ptr0, align 4
// % Res0 = insertelement <16 x i32> undef, i32 % Load0, i32 0
// br label %else
//
// else:
// %res.phi.else = phi <16 x i32>[% Res0, %cond.load], [undef, % 0]
// % Mask1 = extractelement <16 x i1> %Mask, i32 1
// % ToLoad1 = icmp eq i1 % Mask1, true
// br i1 % ToLoad1, label %cond.load1, label %else2
//
// cond.load1:
// % Ptr1 = extractelement <16 x i32*> %Ptrs, i32 1
// % Load1 = load i32, i32* % Ptr1, align 4
// % Res1 = insertelement <16 x i32> %res.phi.else, i32 % Load1, i32 1
// br label %else2
// . . .
// % Result = select <16 x i1> %Mask, <16 x i32> %res.phi.select, <16 x i32> %Src
// ret <16 x i32> %Result
static void ScalarizeMaskedGather(CallInst *CI) {
  Value *Ptrs = CI->getArgOperand(0);
  Value *Alignment = CI->getArgOperand(1);
  Value *Mask = CI->getArgOperand(2);
  Value *Src0 = CI->getArgOperand(3);

  VectorType *VecType = dyn_cast<VectorType>(CI->getType());

  assert(VecType && "Unexpected return type of masked load intrinsic");

  IRBuilder<> Builder(CI->getContext());
  Instruction *InsertPt = CI;
  BasicBlock *IfBlock = CI->getParent();
  BasicBlock *CondBlock = nullptr;
  BasicBlock *PrevIfBlock = CI->getParent();
  Builder.SetInsertPoint(InsertPt);
  unsigned AlignVal = cast<ConstantInt>(Alignment)->getZExtValue();

  Builder.SetCurrentDebugLocation(CI->getDebugLoc());

  Value *UndefVal = UndefValue::get(VecType);

  // The result vector
  Value *VResult = UndefVal;
  unsigned VectorWidth = VecType->getNumElements();

  // Shorten the way if the mask is a vector of constants.
  bool IsConstMask = isa<ConstantVector>(Mask);

  if (IsConstMask) {
    for (unsigned Idx = 0; Idx < VectorWidth; ++Idx) {
      if (cast<ConstantVector>(Mask)->getOperand(Idx)->isNullValue())
        continue;
      Value *Ptr = Builder.CreateExtractElement(Ptrs, Builder.getInt32(Idx),
                                                "Ptr" + Twine(Idx));
      LoadInst *Load = Builder.CreateAlignedLoad(Ptr, AlignVal,
                                                 "Load" + Twine(Idx));
      VResult = Builder.CreateInsertElement(VResult, Load,
                                            Builder.getInt32(Idx),
                                            "Res" + Twine(Idx));
    }
    Value *NewI = Builder.CreateSelect(Mask, VResult, Src0);
    CI->replaceAllUsesWith(NewI);
    CI->eraseFromParent();
    return;
  }

  PHINode *Phi = nullptr;
  Value *PrevPhi = UndefVal;

  for (unsigned Idx = 0; Idx < VectorWidth; ++Idx) {

    // Fill the "else" block, created in the previous iteration
    //
    //  %Mask1 = extractelement <16 x i1> %Mask, i32 1
    //  %ToLoad1 = icmp eq i1 %Mask1, true
    //  br i1 %ToLoad1, label %cond.load, label %else
    //
    if (Idx > 0) {
      Phi = Builder.CreatePHI(VecType, 2, "res.phi.else");
      Phi->addIncoming(VResult, CondBlock);
      Phi->addIncoming(PrevPhi, PrevIfBlock);
      PrevPhi = Phi;
      VResult = Phi;
    }

    Value *Predicate = Builder.CreateExtractElement(Mask,
                                                    Builder.getInt32(Idx),
                                                    "Mask" + Twine(Idx));
    Value *Cmp = Builder.CreateICmp(ICmpInst::ICMP_EQ, Predicate,
                                    ConstantInt::get(Predicate->getType(), 1),
                                    "ToLoad" + Twine(Idx));

    // Create "cond" block
    //
    //  %EltAddr = getelementptr i32* %1, i32 0
    //  %Elt = load i32* %EltAddr
    //  VResult = insertelement <16 x i32> VResult, i32 %Elt, i32 Idx
    //
    CondBlock = IfBlock->splitBasicBlock(InsertPt, "cond.load");
    Builder.SetInsertPoint(InsertPt);

    Value *Ptr = Builder.CreateExtractElement(Ptrs, Builder.getInt32(Idx),
                                              "Ptr" + Twine(Idx));
    LoadInst *Load = Builder.CreateAlignedLoad(Ptr, AlignVal,
                                               "Load" + Twine(Idx));
    VResult = Builder.CreateInsertElement(VResult, Load, Builder.getInt32(Idx),
                                          "Res" + Twine(Idx));

    // Create "else" block, fill it in the next iteration
    BasicBlock *NewIfBlock = CondBlock->splitBasicBlock(InsertPt, "else");
    Builder.SetInsertPoint(InsertPt);
    Instruction *OldBr = IfBlock->getTerminator();
    BranchInst::Create(CondBlock, NewIfBlock, Cmp, OldBr);
    OldBr->eraseFromParent();
    PrevIfBlock = IfBlock;
    IfBlock = NewIfBlock;
  }

  Phi = Builder.CreatePHI(VecType, 2, "res.phi.select");
  Phi->addIncoming(VResult, CondBlock);
  Phi->addIncoming(PrevPhi, PrevIfBlock);
  Value *NewI = Builder.CreateSelect(Mask, Phi, Src0);
  CI->replaceAllUsesWith(NewI);
  CI->eraseFromParent();
}

// Translate a masked scatter intrinsic, like
// void @llvm.masked.scatter.v16i32(<16 x i32> %Src, <16 x i32*>* %Ptrs, i32 4,
//                                  <16 x i1> %Mask)
// to a chain of basic blocks, that stores element one-by-one if
// the appropriate mask bit is set.
//
// % Ptrs = getelementptr i32, i32* %ptr, <16 x i64> %ind
// % Mask0 = extractelement <16 x i1> % Mask, i32 0
// % ToStore0 = icmp eq i1 % Mask0, true
// br i1 %ToStore0, label %cond.store, label %else
//
// cond.store:
// % Elt0 = extractelement <16 x i32> %Src, i32 0
// % Ptr0 = extractelement <16 x i32*> %Ptrs, i32 0
// store i32 %Elt0, i32* % Ptr0, align 4
// br label %else
//
// else:
// % Mask1 = extractelement <16 x i1> % Mask, i32 1
// % ToStore1 = icmp eq i1 % Mask1, true
// br i1 % ToStore1, label %cond.store1, label %else2
//
// cond.store1:
// % Elt1 = extractelement <16 x i32> %Src, i32 1
// % Ptr1 = extractelement <16 x i32*> %Ptrs, i32 1
// store i32 % Elt1, i32* % Ptr1, align 4
// br label %else2
//   . . .
static void ScalarizeMaskedScatter(CallInst *CI) {
  Value *Src = CI->getArgOperand(0);
  Value *Ptrs = CI->getArgOperand(1);
  Value *Alignment = CI->getArgOperand(2);
  Value *Mask = CI->getArgOperand(3);

  assert(isa<VectorType>(Src->getType()) &&
         "Unexpected data type in masked scatter intrinsic");
  assert(isa<VectorType>(Ptrs->getType()) &&
         isa<PointerType>(Ptrs->getType()->getVectorElementType()) &&
         "Vector of pointers is expected in masked scatter intrinsic");

  IRBuilder<> Builder(CI->getContext());
  Instruction *InsertPt = CI;
  BasicBlock *IfBlock = CI->getParent();
  Builder.SetInsertPoint(InsertPt);
  Builder.SetCurrentDebugLocation(CI->getDebugLoc());

  unsigned AlignVal = cast<ConstantInt>(Alignment)->getZExtValue();
  unsigned VectorWidth = Src->getType()->getVectorNumElements();

  // Shorten the way if the mask is a vector of constants.
  bool IsConstMask = isa<ConstantVector>(Mask);

  if (IsConstMask) {
    for (unsigned Idx = 0; Idx < VectorWidth; ++Idx) {
      if (cast<ConstantVector>(Mask)->getOperand(Idx)->isNullValue())
        continue;
      Value *OneElt = Builder.CreateExtractElement(Src, Builder.getInt32(Idx),
                                                   "Elt" + Twine(Idx));
      Value *Ptr = Builder.CreateExtractElement(Ptrs, Builder.getInt32(Idx),
                                                "Ptr" + Twine(Idx));
      Builder.CreateAlignedStore(OneElt, Ptr, AlignVal);
    }
    CI->eraseFromParent();
    return;
  }
  for (unsigned Idx = 0; Idx < VectorWidth; ++Idx) {
    // Fill the "else" block, created in the previous iteration
    //
    //  % Mask1 = extractelement <16 x i1> % Mask, i32 Idx
    //  % ToStore = icmp eq i1 % Mask1, true
    //  br i1 % ToStore, label %cond.store, label %else
    //
    Value *Predicate = Builder.CreateExtractElement(Mask,
                                                    Builder.getInt32(Idx),
                                                    "Mask" + Twine(Idx));
    Value *Cmp =
       Builder.CreateICmp(ICmpInst::ICMP_EQ, Predicate,
                          ConstantInt::get(Predicate->getType(), 1),
                          "ToStore" + Twine(Idx));

    // Create "cond" block
    //
    //  % Elt1 = extractelement <16 x i32> %Src, i32 1
    //  % Ptr1 = extractelement <16 x i32*> %Ptrs, i32 1
    //  %store i32 % Elt1, i32* % Ptr1
    //
    BasicBlock *CondBlock = IfBlock->splitBasicBlock(InsertPt, "cond.store");
    Builder.SetInsertPoint(InsertPt);

    Value *OneElt = Builder.CreateExtractElement(Src, Builder.getInt32(Idx),
                                                 "Elt" + Twine(Idx));
    Value *Ptr = Builder.CreateExtractElement(Ptrs, Builder.getInt32(Idx),
                                              "Ptr" + Twine(Idx));
    Builder.CreateAlignedStore(OneElt, Ptr, AlignVal);

    // Create "else" block, fill it in the next iteration
    BasicBlock *NewIfBlock = CondBlock->splitBasicBlock(InsertPt, "else");
    Builder.SetInsertPoint(InsertPt);
    Instruction *OldBr = IfBlock->getTerminator();
    BranchInst::Create(CondBlock, NewIfBlock, Cmp, OldBr);
    OldBr->eraseFromParent();
    IfBlock = NewIfBlock;
  }
  CI->eraseFromParent();
}

/// If counting leading or trailing zeros is an expensive operation and a zero
/// input is defined, add a check for zero to avoid calling the intrinsic.
///
/// We want to transform:
///     %z = call i64 @llvm.cttz.i64(i64 %A, i1 false)
///
/// into:
///   entry:
///     %cmpz = icmp eq i64 %A, 0
///     br i1 %cmpz, label %cond.end, label %cond.false
///   cond.false:
///     %z = call i64 @llvm.cttz.i64(i64 %A, i1 true)
///     br label %cond.end
///   cond.end:
///     %ctz = phi i64 [ 64, %entry ], [ %z, %cond.false ]
///
/// If the transform is performed, return true and set ModifiedDT to true.
static bool despeculateCountZeros(IntrinsicInst *CountZeros,
                                  const TargetLowering *TLI,
                                  const DataLayout *DL,
                                  bool &ModifiedDT) {
  if (!TLI || !DL)
    return false;

  // If a zero input is undefined, it doesn't make sense to despeculate that.
  if (match(CountZeros->getOperand(1), m_One()))
    return false;

  // If it's cheap to speculate, there's nothing to do.
  auto IntrinsicID = CountZeros->getIntrinsicID();
  if ((IntrinsicID == Intrinsic::cttz && TLI->isCheapToSpeculateCttz()) ||
      (IntrinsicID == Intrinsic::ctlz && TLI->isCheapToSpeculateCtlz()))
    return false;

  // Only handle legal scalar cases. Anything else requires too much work.
  Type *Ty = CountZeros->getType();
  unsigned SizeInBits = Ty->getPrimitiveSizeInBits();
  if (Ty->isVectorTy() || SizeInBits > DL->getLargestLegalIntTypeSize())
    return false;

  // The intrinsic will be sunk behind a compare against zero and branch.
  BasicBlock *StartBlock = CountZeros->getParent();
  BasicBlock *CallBlock = StartBlock->splitBasicBlock(CountZeros, "cond.false");

  // Create another block after the count zero intrinsic. A PHI will be added
  // in this block to select the result of the intrinsic or the bit-width
  // constant if the input to the intrinsic is zero.
  BasicBlock::iterator SplitPt = ++(BasicBlock::iterator(CountZeros));
  BasicBlock *EndBlock = CallBlock->splitBasicBlock(SplitPt, "cond.end");

  // Set up a builder to create a compare, conditional branch, and PHI.
  IRBuilder<> Builder(CountZeros->getContext());
  Builder.SetInsertPoint(StartBlock->getTerminator());
  Builder.SetCurrentDebugLocation(CountZeros->getDebugLoc());

  // Replace the unconditional branch that was created by the first split with
  // a compare against zero and a conditional branch.
  Value *Zero = Constant::getNullValue(Ty);
  Value *Cmp = Builder.CreateICmpEQ(CountZeros->getOperand(0), Zero, "cmpz");
  Builder.CreateCondBr(Cmp, EndBlock, CallBlock);
  StartBlock->getTerminator()->eraseFromParent();

  // Create a PHI in the end block to select either the output of the intrinsic
  // or the bit width of the operand.
  Builder.SetInsertPoint(&EndBlock->front());
  PHINode *PN = Builder.CreatePHI(Ty, 2, "ctz");
  CountZeros->replaceAllUsesWith(PN);
  Value *BitWidth = Builder.getInt(APInt(SizeInBits, SizeInBits));
  PN->addIncoming(BitWidth, StartBlock);
  PN->addIncoming(CountZeros, CallBlock);

  // We are explicitly handling the zero case, so we can set the intrinsic's
  // undefined zero argument to 'true'. This will also prevent reprocessing the
  // intrinsic; we only despeculate when a zero input is defined.
  CountZeros->setArgOperand(1, Builder.getTrue());
  ModifiedDT = true;
  return true;
}

bool CodeGenPrepare::optimizeCallInst(CallInst *CI, bool& ModifiedDT) {
  BasicBlock *BB = CI->getParent();

  // Lower inline assembly if we can.
  // If we found an inline asm expession, and if the target knows how to
  // lower it to normal LLVM code, do so now.
  if (TLI && isa<InlineAsm>(CI->getCalledValue())) {
    if (TLI->ExpandInlineAsm(CI)) {
      // Avoid invalidating the iterator.
      CurInstIterator = BB->begin();
      // Avoid processing instructions out of order, which could cause
      // reuse before a value is defined.
      SunkAddrs.clear();
      return true;
    }
    // Sink address computing for memory operands into the block.
    if (optimizeInlineAsmInst(CI))
      return true;
  }

  // Align the pointer arguments to this call if the target thinks it's a good
  // idea
  unsigned MinSize, PrefAlign;
  if (TLI && TLI->shouldAlignPointerArgs(CI, MinSize, PrefAlign)) {
    for (auto &Arg : CI->arg_operands()) {
      // We want to align both objects whose address is used directly and
      // objects whose address is used in casts and GEPs, though it only makes
      // sense for GEPs if the offset is a multiple of the desired alignment and
      // if size - offset meets the size threshold.
      if (!Arg->getType()->isPointerTy())
        continue;
      APInt Offset(DL->getPointerSizeInBits(
                       cast<PointerType>(Arg->getType())->getAddressSpace()),
                   0);
      Value *Val = Arg->stripAndAccumulateInBoundsConstantOffsets(*DL, Offset);
      uint64_t Offset2 = Offset.getLimitedValue();
      if ((Offset2 & (PrefAlign-1)) != 0)
        continue;
      AllocaInst *AI;
      if ((AI = dyn_cast<AllocaInst>(Val)) && AI->getAlignment() < PrefAlign &&
          DL->getTypeAllocSize(AI->getAllocatedType()) >= MinSize + Offset2)
        AI->setAlignment(PrefAlign);
      // Global variables can only be aligned if they are defined in this
      // object (i.e. they are uniquely initialized in this object), and
      // over-aligning global variables that have an explicit section is
      // forbidden.
      GlobalVariable *GV;
      if ((GV = dyn_cast<GlobalVariable>(Val)) && GV->canIncreaseAlignment() &&
          GV->getAlignment() < PrefAlign &&
          DL->getTypeAllocSize(GV->getType()->getElementType()) >=
              MinSize + Offset2)
        GV->setAlignment(PrefAlign);
    }
    // If this is a memcpy (or similar) then we may be able to improve the
    // alignment
    if (MemIntrinsic *MI = dyn_cast<MemIntrinsic>(CI)) {
      unsigned Align = getKnownAlignment(MI->getDest(), *DL);
      if (MemTransferInst *MTI = dyn_cast<MemTransferInst>(MI))
        Align = std::min(Align, getKnownAlignment(MTI->getSource(), *DL));
      if (Align > MI->getAlignment())
        MI->setAlignment(ConstantInt::get(MI->getAlignmentType(), Align));
    }
  }

  IntrinsicInst *II = dyn_cast<IntrinsicInst>(CI);
  if (II) {
    switch (II->getIntrinsicID()) {
    default: break;
    case Intrinsic::objectsize: {
      // Lower all uses of llvm.objectsize.*
      bool Min = (cast<ConstantInt>(II->getArgOperand(1))->getZExtValue() == 1);
      Type *ReturnTy = CI->getType();
      Constant *RetVal = ConstantInt::get(ReturnTy, Min ? 0 : -1ULL);

      // Substituting this can cause recursive simplifications, which can
      // invalidate our iterator.  Use a WeakVH to hold onto it in case this
      // happens.
      WeakVH IterHandle(&*CurInstIterator);

      replaceAndRecursivelySimplify(CI, RetVal,
                                    TLInfo, nullptr);

      // If the iterator instruction was recursively deleted, start over at the
      // start of the block.
      if (IterHandle != CurInstIterator.getNodePtrUnchecked()) {
        CurInstIterator = BB->begin();
        SunkAddrs.clear();
      }
      return true;
    }
    case Intrinsic::masked_load: {
      // Scalarize unsupported vector masked load
      if (!TTI->isLegalMaskedLoad(CI->getType())) {
        ScalarizeMaskedLoad(CI);
        ModifiedDT = true;
        return true;
      }
      return false;
    }
    case Intrinsic::masked_store: {
      if (!TTI->isLegalMaskedStore(CI->getArgOperand(0)->getType())) {
        ScalarizeMaskedStore(CI);
        ModifiedDT = true;
        return true;
      }
      return false;
    }
    case Intrinsic::masked_gather: {
      if (!TTI->isLegalMaskedGather(CI->getType())) {
        ScalarizeMaskedGather(CI);
        ModifiedDT = true;
        return true;
      }
      return false;
    }
    case Intrinsic::masked_scatter: {
      if (!TTI->isLegalMaskedScatter(CI->getArgOperand(0)->getType())) {
        ScalarizeMaskedScatter(CI);
        ModifiedDT = true;
        return true;
      }
      return false;
    }
    case Intrinsic::aarch64_stlxr:
    case Intrinsic::aarch64_stxr: {
      ZExtInst *ExtVal = dyn_cast<ZExtInst>(CI->getArgOperand(0));
      if (!ExtVal || !ExtVal->hasOneUse() ||
          ExtVal->getParent() == CI->getParent())
        return false;
      // Sink a zext feeding stlxr/stxr before it, so it can be folded into it.
      ExtVal->moveBefore(CI);
      // Mark this instruction as "inserted by CGP", so that other
      // optimizations don't touch it.
      InsertedInsts.insert(ExtVal);
      return true;
    }
    case Intrinsic::invariant_group_barrier:
      II->replaceAllUsesWith(II->getArgOperand(0));
      II->eraseFromParent();
      return true;

    case Intrinsic::cttz:
    case Intrinsic::ctlz:
      // If counting zeros is expensive, try to avoid it.
      return despeculateCountZeros(II, TLI, DL, ModifiedDT);
    }

    if (TLI) {
      // Unknown address space.
      // TODO: Target hook to pick which address space the intrinsic cares
      // about?
      unsigned AddrSpace = ~0u;
      SmallVector<Value*, 2> PtrOps;
      Type *AccessTy;
      if (TLI->GetAddrModeArguments(II, PtrOps, AccessTy, AddrSpace))
        while (!PtrOps.empty())
          if (optimizeMemoryInst(II, PtrOps.pop_back_val(), AccessTy, AddrSpace))
            return true;
    }
  }

  // From here on out we're working with named functions.
  if (!CI->getCalledFunction()) return false;

  // Lower all default uses of _chk calls.  This is very similar
  // to what InstCombineCalls does, but here we are only lowering calls
  // to fortified library functions (e.g. __memcpy_chk) that have the default
  // "don't know" as the objectsize.  Anything else should be left alone.
  FortifiedLibCallSimplifier Simplifier(TLInfo, true);
  if (Value *V = Simplifier.optimizeCall(CI)) {
    CI->replaceAllUsesWith(V);
    CI->eraseFromParent();
    return true;
  }
  return false;
}

/// Look for opportunities to duplicate return instructions to the predecessor
/// to enable tail call optimizations. The case it is currently looking for is:
/// @code
/// bb0:
///   %tmp0 = tail call i32 @f0()
///   br label %return
/// bb1:
///   %tmp1 = tail call i32 @f1()
///   br label %return
/// bb2:
///   %tmp2 = tail call i32 @f2()
///   br label %return
/// return:
///   %retval = phi i32 [ %tmp0, %bb0 ], [ %tmp1, %bb1 ], [ %tmp2, %bb2 ]
///   ret i32 %retval
/// @endcode
///
/// =>
///
/// @code
/// bb0:
///   %tmp0 = tail call i32 @f0()
///   ret i32 %tmp0
/// bb1:
///   %tmp1 = tail call i32 @f1()
///   ret i32 %tmp1
/// bb2:
///   %tmp2 = tail call i32 @f2()
///   ret i32 %tmp2
/// @endcode
bool CodeGenPrepare::dupRetToEnableTailCallOpts(BasicBlock *BB) {
  if (!TLI)
    return false;

  ReturnInst *RI = dyn_cast<ReturnInst>(BB->getTerminator());
  if (!RI)
    return false;

  PHINode *PN = nullptr;
  BitCastInst *BCI = nullptr;
  Value *V = RI->getReturnValue();
  if (V) {
    BCI = dyn_cast<BitCastInst>(V);
    if (BCI)
      V = BCI->getOperand(0);

    PN = dyn_cast<PHINode>(V);
    if (!PN)
      return false;
  }

  if (PN && PN->getParent() != BB)
    return false;

  // It's not safe to eliminate the sign / zero extension of the return value.
  // See llvm::isInTailCallPosition().
  const Function *F = BB->getParent();
  AttributeSet CallerAttrs = F->getAttributes();
  if (CallerAttrs.hasAttribute(AttributeSet::ReturnIndex, Attribute::ZExt) ||
      CallerAttrs.hasAttribute(AttributeSet::ReturnIndex, Attribute::SExt))
    return false;

  // Make sure there are no instructions between the PHI and return, or that the
  // return is the first instruction in the block.
  if (PN) {
    BasicBlock::iterator BI = BB->begin();
    do { ++BI; } while (isa<DbgInfoIntrinsic>(BI));
    if (&*BI == BCI)
      // Also skip over the bitcast.
      ++BI;
    if (&*BI != RI)
      return false;
  } else {
    BasicBlock::iterator BI = BB->begin();
    while (isa<DbgInfoIntrinsic>(BI)) ++BI;
    if (&*BI != RI)
      return false;
  }

  /// Only dup the ReturnInst if the CallInst is likely to be emitted as a tail
  /// call.
  SmallVector<CallInst*, 4> TailCalls;
  if (PN) {
    for (unsigned I = 0, E = PN->getNumIncomingValues(); I != E; ++I) {
      CallInst *CI = dyn_cast<CallInst>(PN->getIncomingValue(I));
      // Make sure the phi value is indeed produced by the tail call.
      if (CI && CI->hasOneUse() && CI->getParent() == PN->getIncomingBlock(I) &&
          TLI->mayBeEmittedAsTailCall(CI))
        TailCalls.push_back(CI);
    }
  } else {
    SmallPtrSet<BasicBlock*, 4> VisitedBBs;
    for (pred_iterator PI = pred_begin(BB), PE = pred_end(BB); PI != PE; ++PI) {
      if (!VisitedBBs.insert(*PI).second)
        continue;

      BasicBlock::InstListType &InstList = (*PI)->getInstList();
      BasicBlock::InstListType::reverse_iterator RI = InstList.rbegin();
      BasicBlock::InstListType::reverse_iterator RE = InstList.rend();
      do { ++RI; } while (RI != RE && isa<DbgInfoIntrinsic>(&*RI));
      if (RI == RE)
        continue;

      CallInst *CI = dyn_cast<CallInst>(&*RI);
      if (CI && CI->use_empty() && TLI->mayBeEmittedAsTailCall(CI))
        TailCalls.push_back(CI);
    }
  }

  bool Changed = false;
  for (unsigned i = 0, e = TailCalls.size(); i != e; ++i) {
    CallInst *CI = TailCalls[i];
    CallSite CS(CI);

    // Conservatively require the attributes of the call to match those of the
    // return. Ignore noalias because it doesn't affect the call sequence.
    AttributeSet CalleeAttrs = CS.getAttributes();
    if (AttrBuilder(CalleeAttrs, AttributeSet::ReturnIndex).
          removeAttribute(Attribute::NoAlias) !=
        AttrBuilder(CalleeAttrs, AttributeSet::ReturnIndex).
          removeAttribute(Attribute::NoAlias))
      continue;

    // Make sure the call instruction is followed by an unconditional branch to
    // the return block.
    BasicBlock *CallBB = CI->getParent();
    BranchInst *BI = dyn_cast<BranchInst>(CallBB->getTerminator());
    if (!BI || !BI->isUnconditional() || BI->getSuccessor(0) != BB)
      continue;

    // Duplicate the return into CallBB.
    (void)FoldReturnIntoUncondBranch(RI, BB, CallBB);
    ModifiedDT = Changed = true;
    ++NumRetsDup;
  }

  // If we eliminated all predecessors of the block, delete the block now.
  if (Changed && !BB->hasAddressTaken() && pred_begin(BB) == pred_end(BB))
    BB->eraseFromParent();

  return Changed;
}

//===----------------------------------------------------------------------===//
// Memory Optimization
//===----------------------------------------------------------------------===//

namespace {

/// This is an extended version of TargetLowering::AddrMode
/// which holds actual Value*'s for register values.
struct ExtAddrMode : public TargetLowering::AddrMode {
  Value *BaseReg;
  Value *ScaledReg;
  ExtAddrMode() : BaseReg(nullptr), ScaledReg(nullptr) {}
  void print(raw_ostream &OS) const;
  void dump() const;

  bool operator==(const ExtAddrMode& O) const {
    return (BaseReg == O.BaseReg) && (ScaledReg == O.ScaledReg) &&
           (BaseGV == O.BaseGV) && (BaseOffs == O.BaseOffs) &&
           (HasBaseReg == O.HasBaseReg) && (Scale == O.Scale);
  }
};

#ifndef NDEBUG
static inline raw_ostream &operator<<(raw_ostream &OS, const ExtAddrMode &AM) {
  AM.print(OS);
  return OS;
}
#endif

void ExtAddrMode::print(raw_ostream &OS) const {
  bool NeedPlus = false;
  OS << "[";
  if (BaseGV) {
    OS << (NeedPlus ? " + " : "")
       << "GV:";
    BaseGV->printAsOperand(OS, /*PrintType=*/false);
    NeedPlus = true;
  }

  if (BaseOffs) {
    OS << (NeedPlus ? " + " : "")
       << BaseOffs;
    NeedPlus = true;
  }

  if (BaseReg) {
    OS << (NeedPlus ? " + " : "")
       << "Base:";
    BaseReg->printAsOperand(OS, /*PrintType=*/false);
    NeedPlus = true;
  }
  if (Scale) {
    OS << (NeedPlus ? " + " : "")
       << Scale << "*";
    ScaledReg->printAsOperand(OS, /*PrintType=*/false);
  }

  OS << ']';
}

#if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
void ExtAddrMode::dump() const {
  print(dbgs());
  dbgs() << '\n';
}
#endif

/// \brief This class provides transaction based operation on the IR.
/// Every change made through this class is recorded in the internal state and
/// can be undone (rollback) until commit is called.
class TypePromotionTransaction {

  /// \brief This represents the common interface of the individual transaction.
  /// Each class implements the logic for doing one specific modification on
  /// the IR via the TypePromotionTransaction.
  class TypePromotionAction {
  protected:
    /// The Instruction modified.
    Instruction *Inst;

  public:
    /// \brief Constructor of the action.
    /// The constructor performs the related action on the IR.
    TypePromotionAction(Instruction *Inst) : Inst(Inst) {}

    virtual ~TypePromotionAction() {}

    /// \brief Undo the modification done by this action.
    /// When this method is called, the IR must be in the same state as it was
    /// before this action was applied.
    /// \pre Undoing the action works if and only if the IR is in the exact same
    /// state as it was directly after this action was applied.
    virtual void undo() = 0;

    /// \brief Advocate every change made by this action.
    /// When the results on the IR of the action are to be kept, it is important
    /// to call this function, otherwise hidden information may be kept forever.
    virtual void commit() {
      // Nothing to be done, this action is not doing anything.
    }
  };

  /// \brief Utility to remember the position of an instruction.
  class InsertionHandler {
    /// Position of an instruction.
    /// Either an instruction:
    /// - Is the first in a basic block: BB is used.
    /// - Has a previous instructon: PrevInst is used.
    union {
      Instruction *PrevInst;
      BasicBlock *BB;
    } Point;
    /// Remember whether or not the instruction had a previous instruction.
    bool HasPrevInstruction;

  public:
    /// \brief Record the position of \p Inst.
    InsertionHandler(Instruction *Inst) {
      BasicBlock::iterator It = Inst->getIterator();
      HasPrevInstruction = (It != (Inst->getParent()->begin()));
      if (HasPrevInstruction)
        Point.PrevInst = &*--It;
      else
        Point.BB = Inst->getParent();
    }

    /// \brief Insert \p Inst at the recorded position.
    void insert(Instruction *Inst) {
      if (HasPrevInstruction) {
        if (Inst->getParent())
          Inst->removeFromParent();
        Inst->insertAfter(Point.PrevInst);
      } else {
        Instruction *Position = &*Point.BB->getFirstInsertionPt();
        if (Inst->getParent())
          Inst->moveBefore(Position);
        else
          Inst->insertBefore(Position);
      }
    }
  };

  /// \brief Move an instruction before another.
  class InstructionMoveBefore : public TypePromotionAction {
    /// Original position of the instruction.
    InsertionHandler Position;

  public:
    /// \brief Move \p Inst before \p Before.
    InstructionMoveBefore(Instruction *Inst, Instruction *Before)
        : TypePromotionAction(Inst), Position(Inst) {
      DEBUG(dbgs() << "Do: move: " << *Inst << "\nbefore: " << *Before << "\n");
      Inst->moveBefore(Before);
    }

    /// \brief Move the instruction back to its original position.
    void undo() override {
      DEBUG(dbgs() << "Undo: moveBefore: " << *Inst << "\n");
      Position.insert(Inst);
    }
  };

  /// \brief Set the operand of an instruction with a new value.
  class OperandSetter : public TypePromotionAction {
    /// Original operand of the instruction.
    Value *Origin;
    /// Index of the modified instruction.
    unsigned Idx;

  public:
    /// \brief Set \p Idx operand of \p Inst with \p NewVal.
    OperandSetter(Instruction *Inst, unsigned Idx, Value *NewVal)
        : TypePromotionAction(Inst), Idx(Idx) {
      DEBUG(dbgs() << "Do: setOperand: " << Idx << "\n"
                   << "for:" << *Inst << "\n"
                   << "with:" << *NewVal << "\n");
      Origin = Inst->getOperand(Idx);
      Inst->setOperand(Idx, NewVal);
    }

    /// \brief Restore the original value of the instruction.
    void undo() override {
      DEBUG(dbgs() << "Undo: setOperand:" << Idx << "\n"
                   << "for: " << *Inst << "\n"
                   << "with: " << *Origin << "\n");
      Inst->setOperand(Idx, Origin);
    }
  };

  /// \brief Hide the operands of an instruction.
  /// Do as if this instruction was not using any of its operands.
  class OperandsHider : public TypePromotionAction {
    /// The list of original operands.
    SmallVector<Value *, 4> OriginalValues;

  public:
    /// \brief Remove \p Inst from the uses of the operands of \p Inst.
    OperandsHider(Instruction *Inst) : TypePromotionAction(Inst) {
      DEBUG(dbgs() << "Do: OperandsHider: " << *Inst << "\n");
      unsigned NumOpnds = Inst->getNumOperands();
      OriginalValues.reserve(NumOpnds);
      for (unsigned It = 0; It < NumOpnds; ++It) {
        // Save the current operand.
        Value *Val = Inst->getOperand(It);
        OriginalValues.push_back(Val);
        // Set a dummy one.
        // We could use OperandSetter here, but that would imply an overhead
        // that we are not willing to pay.
        Inst->setOperand(It, UndefValue::get(Val->getType()));
      }
    }

    /// \brief Restore the original list of uses.
    void undo() override {
      DEBUG(dbgs() << "Undo: OperandsHider: " << *Inst << "\n");
      for (unsigned It = 0, EndIt = OriginalValues.size(); It != EndIt; ++It)
        Inst->setOperand(It, OriginalValues[It]);
    }
  };

  /// \brief Build a truncate instruction.
  class TruncBuilder : public TypePromotionAction {
    Value *Val;
  public:
    /// \brief Build a truncate instruction of \p Opnd producing a \p Ty
    /// result.
    /// trunc Opnd to Ty.
    TruncBuilder(Instruction *Opnd, Type *Ty) : TypePromotionAction(Opnd) {
      IRBuilder<> Builder(Opnd);
      Val = Builder.CreateTrunc(Opnd, Ty, "promoted");
      DEBUG(dbgs() << "Do: TruncBuilder: " << *Val << "\n");
    }

    /// \brief Get the built value.
    Value *getBuiltValue() { return Val; }

    /// \brief Remove the built instruction.
    void undo() override {
      DEBUG(dbgs() << "Undo: TruncBuilder: " << *Val << "\n");
      if (Instruction *IVal = dyn_cast<Instruction>(Val))
        IVal->eraseFromParent();
    }
  };

  /// \brief Build a sign extension instruction.
  class SExtBuilder : public TypePromotionAction {
    Value *Val;
  public:
    /// \brief Build a sign extension instruction of \p Opnd producing a \p Ty
    /// result.
    /// sext Opnd to Ty.
    SExtBuilder(Instruction *InsertPt, Value *Opnd, Type *Ty)
        : TypePromotionAction(InsertPt) {
      IRBuilder<> Builder(InsertPt);
      Val = Builder.CreateSExt(Opnd, Ty, "promoted");
      DEBUG(dbgs() << "Do: SExtBuilder: " << *Val << "\n");
    }

    /// \brief Get the built value.
    Value *getBuiltValue() { return Val; }

    /// \brief Remove the built instruction.
    void undo() override {
      DEBUG(dbgs() << "Undo: SExtBuilder: " << *Val << "\n");
      if (Instruction *IVal = dyn_cast<Instruction>(Val))
        IVal->eraseFromParent();
    }
  };

  /// \brief Build a zero extension instruction.
  class ZExtBuilder : public TypePromotionAction {
    Value *Val;
  public:
    /// \brief Build a zero extension instruction of \p Opnd producing a \p Ty
    /// result.
    /// zext Opnd to Ty.
    ZExtBuilder(Instruction *InsertPt, Value *Opnd, Type *Ty)
        : TypePromotionAction(InsertPt) {
      IRBuilder<> Builder(InsertPt);
      Val = Builder.CreateZExt(Opnd, Ty, "promoted");
      DEBUG(dbgs() << "Do: ZExtBuilder: " << *Val << "\n");
    }

    /// \brief Get the built value.
    Value *getBuiltValue() { return Val; }

    /// \brief Remove the built instruction.
    void undo() override {
      DEBUG(dbgs() << "Undo: ZExtBuilder: " << *Val << "\n");
      if (Instruction *IVal = dyn_cast<Instruction>(Val))
        IVal->eraseFromParent();
    }
  };

  /// \brief Mutate an instruction to another type.
  class TypeMutator : public TypePromotionAction {
    /// Record the original type.
    Type *OrigTy;

  public:
    /// \brief Mutate the type of \p Inst into \p NewTy.
    TypeMutator(Instruction *Inst, Type *NewTy)
        : TypePromotionAction(Inst), OrigTy(Inst->getType()) {
      DEBUG(dbgs() << "Do: MutateType: " << *Inst << " with " << *NewTy
                   << "\n");
      Inst->mutateType(NewTy);
    }

    /// \brief Mutate the instruction back to its original type.
    void undo() override {
      DEBUG(dbgs() << "Undo: MutateType: " << *Inst << " with " << *OrigTy
                   << "\n");
      Inst->mutateType(OrigTy);
    }
  };

  /// \brief Replace the uses of an instruction by another instruction.
  class UsesReplacer : public TypePromotionAction {
    /// Helper structure to keep track of the replaced uses.
    struct InstructionAndIdx {
      /// The instruction using the instruction.
      Instruction *Inst;
      /// The index where this instruction is used for Inst.
      unsigned Idx;
      InstructionAndIdx(Instruction *Inst, unsigned Idx)
          : Inst(Inst), Idx(Idx) {}
    };

    /// Keep track of the original uses (pair Instruction, Index).
    SmallVector<InstructionAndIdx, 4> OriginalUses;
    typedef SmallVectorImpl<InstructionAndIdx>::iterator use_iterator;

  public:
    /// \brief Replace all the use of \p Inst by \p New.
    UsesReplacer(Instruction *Inst, Value *New) : TypePromotionAction(Inst) {
      DEBUG(dbgs() << "Do: UsersReplacer: " << *Inst << " with " << *New
                   << "\n");
      // Record the original uses.
      for (Use &U : Inst->uses()) {
        Instruction *UserI = cast<Instruction>(U.getUser());
        OriginalUses.push_back(InstructionAndIdx(UserI, U.getOperandNo()));
      }
      // Now, we can replace the uses.
      Inst->replaceAllUsesWith(New);
    }

    /// \brief Reassign the original uses of Inst to Inst.
    void undo() override {
      DEBUG(dbgs() << "Undo: UsersReplacer: " << *Inst << "\n");
      for (use_iterator UseIt = OriginalUses.begin(),
                        EndIt = OriginalUses.end();
           UseIt != EndIt; ++UseIt) {
        UseIt->Inst->setOperand(UseIt->Idx, Inst);
      }
    }
  };

  /// \brief Remove an instruction from the IR.
  class InstructionRemover : public TypePromotionAction {
    /// Original position of the instruction.
    InsertionHandler Inserter;
    /// Helper structure to hide all the link to the instruction. In other
    /// words, this helps to do as if the instruction was removed.
    OperandsHider Hider;
    /// Keep track of the uses replaced, if any.
    UsesReplacer *Replacer;

  public:
    /// \brief Remove all reference of \p Inst and optinally replace all its
    /// uses with New.
    /// \pre If !Inst->use_empty(), then New != nullptr
    InstructionRemover(Instruction *Inst, Value *New = nullptr)
        : TypePromotionAction(Inst), Inserter(Inst), Hider(Inst),
          Replacer(nullptr) {
      if (New)
        Replacer = new UsesReplacer(Inst, New);
      DEBUG(dbgs() << "Do: InstructionRemover: " << *Inst << "\n");
      Inst->removeFromParent();
    }

    ~InstructionRemover() override { delete Replacer; }

    /// \brief Really remove the instruction.
    void commit() override { delete Inst; }

    /// \brief Resurrect the instruction and reassign it to the proper uses if
    /// new value was provided when build this action.
    void undo() override {
      DEBUG(dbgs() << "Undo: InstructionRemover: " << *Inst << "\n");
      Inserter.insert(Inst);
      if (Replacer)
        Replacer->undo();
      Hider.undo();
    }
  };

public:
  /// Restoration point.
  /// The restoration point is a pointer to an action instead of an iterator
  /// because the iterator may be invalidated but not the pointer.
  typedef const TypePromotionAction *ConstRestorationPt;
  /// Advocate every changes made in that transaction.
  void commit();
  /// Undo all the changes made after the given point.
  void rollback(ConstRestorationPt Point);
  /// Get the current restoration point.
  ConstRestorationPt getRestorationPoint() const;

  /// \name API for IR modification with state keeping to support rollback.
  /// @{
  /// Same as Instruction::setOperand.
  void setOperand(Instruction *Inst, unsigned Idx, Value *NewVal);
  /// Same as Instruction::eraseFromParent.
  void eraseInstruction(Instruction *Inst, Value *NewVal = nullptr);
  /// Same as Value::replaceAllUsesWith.
  void replaceAllUsesWith(Instruction *Inst, Value *New);
  /// Same as Value::mutateType.
  void mutateType(Instruction *Inst, Type *NewTy);
  /// Same as IRBuilder::createTrunc.
  Value *createTrunc(Instruction *Opnd, Type *Ty);
  /// Same as IRBuilder::createSExt.
  Value *createSExt(Instruction *Inst, Value *Opnd, Type *Ty);
  /// Same as IRBuilder::createZExt.
  Value *createZExt(Instruction *Inst, Value *Opnd, Type *Ty);
  /// Same as Instruction::moveBefore.
  void moveBefore(Instruction *Inst, Instruction *Before);
  /// @}

private:
  /// The ordered list of actions made so far.
  SmallVector<std::unique_ptr<TypePromotionAction>, 16> Actions;
  typedef SmallVectorImpl<std::unique_ptr<TypePromotionAction>>::iterator CommitPt;
};

void TypePromotionTransaction::setOperand(Instruction *Inst, unsigned Idx,
                                          Value *NewVal) {
  Actions.push_back(
      make_unique<TypePromotionTransaction::OperandSetter>(Inst, Idx, NewVal));
}

void TypePromotionTransaction::eraseInstruction(Instruction *Inst,
                                                Value *NewVal) {
  Actions.push_back(
      make_unique<TypePromotionTransaction::InstructionRemover>(Inst, NewVal));
}

void TypePromotionTransaction::replaceAllUsesWith(Instruction *Inst,
                                                  Value *New) {
  Actions.push_back(make_unique<TypePromotionTransaction::UsesReplacer>(Inst, New));
}

void TypePromotionTransaction::mutateType(Instruction *Inst, Type *NewTy) {
  Actions.push_back(make_unique<TypePromotionTransaction::TypeMutator>(Inst, NewTy));
}

Value *TypePromotionTransaction::createTrunc(Instruction *Opnd,
                                             Type *Ty) {
  std::unique_ptr<TruncBuilder> Ptr(new TruncBuilder(Opnd, Ty));
  Value *Val = Ptr->getBuiltValue();
  Actions.push_back(std::move(Ptr));
  return Val;
}

Value *TypePromotionTransaction::createSExt(Instruction *Inst,
                                            Value *Opnd, Type *Ty) {
  std::unique_ptr<SExtBuilder> Ptr(new SExtBuilder(Inst, Opnd, Ty));
  Value *Val = Ptr->getBuiltValue();
  Actions.push_back(std::move(Ptr));
  return Val;
}

Value *TypePromotionTransaction::createZExt(Instruction *Inst,
                                            Value *Opnd, Type *Ty) {
  std::unique_ptr<ZExtBuilder> Ptr(new ZExtBuilder(Inst, Opnd, Ty));
  Value *Val = Ptr->getBuiltValue();
  Actions.push_back(std::move(Ptr));
  return Val;
}

void TypePromotionTransaction::moveBefore(Instruction *Inst,
                                          Instruction *Before) {
  Actions.push_back(
      make_unique<TypePromotionTransaction::InstructionMoveBefore>(Inst, Before));
}

TypePromotionTransaction::ConstRestorationPt
TypePromotionTransaction::getRestorationPoint() const {
  return !Actions.empty() ? Actions.back().get() : nullptr;
}

void TypePromotionTransaction::commit() {
  for (CommitPt It = Actions.begin(), EndIt = Actions.end(); It != EndIt;
       ++It)
    (*It)->commit();
  Actions.clear();
}

void TypePromotionTransaction::rollback(
    TypePromotionTransaction::ConstRestorationPt Point) {
  while (!Actions.empty() && Point != Actions.back().get()) {
    std::unique_ptr<TypePromotionAction> Curr = Actions.pop_back_val();
    Curr->undo();
  }
}

/// \brief A helper class for matching addressing modes.
///
/// This encapsulates the logic for matching the target-legal addressing modes.
class AddressingModeMatcher {
  SmallVectorImpl<Instruction*> &AddrModeInsts;
  const TargetMachine &TM;
  const TargetLowering &TLI;
  const DataLayout &DL;

  /// AccessTy/MemoryInst - This is the type for the access (e.g. double) and
  /// the memory instruction that we're computing this address for.
  Type *AccessTy;
  unsigned AddrSpace;
  Instruction *MemoryInst;

  /// This is the addressing mode that we're building up. This is
  /// part of the return value of this addressing mode matching stuff.
  ExtAddrMode &AddrMode;

  /// The instructions inserted by other CodeGenPrepare optimizations.
  const SetOfInstrs &InsertedInsts;
  /// A map from the instructions to their type before promotion.
  InstrToOrigTy &PromotedInsts;
  /// The ongoing transaction where every action should be registered.
  TypePromotionTransaction &TPT;

  /// This is set to true when we should not do profitability checks.
  /// When true, IsProfitableToFoldIntoAddressingMode always returns true.
  bool IgnoreProfitability;

  AddressingModeMatcher(SmallVectorImpl<Instruction *> &AMI,
                        const TargetMachine &TM, Type *AT, unsigned AS,
                        Instruction *MI, ExtAddrMode &AM,
                        const SetOfInstrs &InsertedInsts,
                        InstrToOrigTy &PromotedInsts,
                        TypePromotionTransaction &TPT)
      : AddrModeInsts(AMI), TM(TM),
        TLI(*TM.getSubtargetImpl(*MI->getParent()->getParent())
                 ->getTargetLowering()),
        DL(MI->getModule()->getDataLayout()), AccessTy(AT), AddrSpace(AS),
        MemoryInst(MI), AddrMode(AM), InsertedInsts(InsertedInsts),
        PromotedInsts(PromotedInsts), TPT(TPT) {
    IgnoreProfitability = false;
  }
public:

  /// Find the maximal addressing mode that a load/store of V can fold,
  /// give an access type of AccessTy.  This returns a list of involved
  /// instructions in AddrModeInsts.
  /// \p InsertedInsts The instructions inserted by other CodeGenPrepare
  /// optimizations.
  /// \p PromotedInsts maps the instructions to their type before promotion.
  /// \p The ongoing transaction where every action should be registered.
  static ExtAddrMode Match(Value *V, Type *AccessTy, unsigned AS,
                           Instruction *MemoryInst,
                           SmallVectorImpl<Instruction*> &AddrModeInsts,
                           const TargetMachine &TM,
                           const SetOfInstrs &InsertedInsts,
                           InstrToOrigTy &PromotedInsts,
                           TypePromotionTransaction &TPT) {
    ExtAddrMode Result;

    bool Success = AddressingModeMatcher(AddrModeInsts, TM, AccessTy, AS,
                                         MemoryInst, Result, InsertedInsts,
                                         PromotedInsts, TPT).matchAddr(V, 0);
    (void)Success; assert(Success && "Couldn't select *anything*?");
    return Result;
  }
private:
  bool matchScaledValue(Value *ScaleReg, int64_t Scale, unsigned Depth);
  bool matchAddr(Value *V, unsigned Depth);
  bool matchOperationAddr(User *Operation, unsigned Opcode, unsigned Depth,
                          bool *MovedAway = nullptr);
  bool isProfitableToFoldIntoAddressingMode(Instruction *I,
                                            ExtAddrMode &AMBefore,
                                            ExtAddrMode &AMAfter);
  bool valueAlreadyLiveAtInst(Value *Val, Value *KnownLive1, Value *KnownLive2);
  bool isPromotionProfitable(unsigned NewCost, unsigned OldCost,
                             Value *PromotedOperand) const;
};

/// Try adding ScaleReg*Scale to the current addressing mode.
/// Return true and update AddrMode if this addr mode is legal for the target,
/// false if not.
bool AddressingModeMatcher::matchScaledValue(Value *ScaleReg, int64_t Scale,
                                             unsigned Depth) {
  // If Scale is 1, then this is the same as adding ScaleReg to the addressing
  // mode.  Just process that directly.
  if (Scale == 1)
    return matchAddr(ScaleReg, Depth);

  // If the scale is 0, it takes nothing to add this.
  if (Scale == 0)
    return true;

  // If we already have a scale of this value, we can add to it, otherwise, we
  // need an available scale field.
  if (AddrMode.Scale != 0 && AddrMode.ScaledReg != ScaleReg)
    return false;

  ExtAddrMode TestAddrMode = AddrMode;

  // Add scale to turn X*4+X*3 -> X*7.  This could also do things like
  // [A+B + A*7] -> [B+A*8].
  TestAddrMode.Scale += Scale;
  TestAddrMode.ScaledReg = ScaleReg;

  // If the new address isn't legal, bail out.
  if (!TLI.isLegalAddressingMode(DL, TestAddrMode, AccessTy, AddrSpace))