CGBuiltin.cpp

Go to the documentation of this file.

//===---- CGBuiltin.cpp - Emit LLVM Code for builtins ---------------------===//
//
//                     The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This contains code to emit Builtin calls as LLVM code.
//
//===----------------------------------------------------------------------===//

#include "CGCXXABI.h"
#include "CGObjCRuntime.h"
#include "CGOpenCLRuntime.h"
#include "CGRecordLayout.h"
#include "CodeGenFunction.h"
#include "CodeGenModule.h"
#include "ConstantEmitter.h"
#include "TargetInfo.h"
#include "clang/AST/ASTContext.h"
#include "clang/AST/Decl.h"
#include "clang/Analysis/Analyses/OSLog.h"
#include "clang/Basic/TargetBuiltins.h"
#include "clang/Basic/TargetInfo.h"
#include "clang/CodeGen/CGFunctionInfo.h"
#include "llvm/ADT/StringExtras.h"
#include "llvm/IR/CallSite.h"
#include "llvm/IR/DataLayout.h"
#include "llvm/IR/InlineAsm.h"
#include "llvm/IR/Intrinsics.h"
#include "llvm/IR/MDBuilder.h"
#include "llvm/Support/ConvertUTF.h"
#include "llvm/Support/ScopedPrinter.h"
#include "llvm/Support/TargetParser.h"
#include <sstream>

using namespace clang;
using namespace CodeGen;
using namespace llvm;

static
int64_t clamp(int64_t Value, int64_t Low, int64_t High) {
  return std::min(High, std::max(Low, Value));
}

/// getBuiltinLibFunction - Given a builtin id for a function like
/// "__builtin_fabsf", return a Function* for "fabsf".
llvm::Constant *CodeGenModule::getBuiltinLibFunction(const FunctionDecl *FD,
                                                     unsigned BuiltinID) {
  assert(Context.BuiltinInfo.isLibFunction(BuiltinID));

  // Get the name, skip over the __builtin_ prefix (if necessary).
  StringRef Name;
  GlobalDecl D(FD);

  // If the builtin has been declared explicitly with an assembler label,
  // use the mangled name. This differs from the plain label on platforms
  // that prefix labels.
  if (FD->hasAttr<AsmLabelAttr>())
    Name = getMangledName(D);
  else
    Name = Context.BuiltinInfo.getName(BuiltinID) + 10;

  llvm::FunctionType *Ty =
    cast<llvm::FunctionType>(getTypes().ConvertType(FD->getType()));

  return GetOrCreateLLVMFunction(Name, Ty, D, /*ForVTable=*/false);
}

/// Emit the conversions required to turn the given value into an
/// integer of the given size.
static Value *EmitToInt(CodeGenFunction &CGF, llvm::Value *V,
                        QualType T, llvm::IntegerType *IntType) {
  V = CGF.EmitToMemory(V, T);

  if (V->getType()->isPointerTy())
    return CGF.Builder.CreatePtrToInt(V, IntType);

  assert(V->getType() == IntType);
  return V;
}

static Value *EmitFromInt(CodeGenFunction &CGF, llvm::Value *V,
                          QualType T, llvm::Type *ResultType) {
  V = CGF.EmitFromMemory(V, T);

  if (ResultType->isPointerTy())
    return CGF.Builder.CreateIntToPtr(V, ResultType);

  assert(V->getType() == ResultType);
  return V;
}

/// Utility to insert an atomic instruction based on Instrinsic::ID
/// and the expression node.
static Value *MakeBinaryAtomicValue(CodeGenFunction &CGF,
                                    llvm::AtomicRMWInst::BinOp Kind,
                                    const CallExpr *E) {
  QualType T = E->getType();
  assert(E->getArg(0)->getType()->isPointerType());
  assert(CGF.getContext().hasSameUnqualifiedType(T,
                                  E->getArg(0)->getType()->getPointeeType()));
  assert(CGF.getContext().hasSameUnqualifiedType(T, E->getArg(1)->getType()));

  llvm::Value *DestPtr = CGF.EmitScalarExpr(E->getArg(0));
  unsigned AddrSpace = DestPtr->getType()->getPointerAddressSpace();

  llvm::IntegerType *IntType =
    llvm::IntegerType::get(CGF.getLLVMContext(),
                           CGF.getContext().getTypeSize(T));
  llvm::Type *IntPtrType = IntType->getPointerTo(AddrSpace);

  llvm::Value *Args[2];
  Args[0] = CGF.Builder.CreateBitCast(DestPtr, IntPtrType);
  Args[1] = CGF.EmitScalarExpr(E->getArg(1));
  llvm::Type *ValueType = Args[1]->getType();
  Args[1] = EmitToInt(CGF, Args[1], T, IntType);

  llvm::Value *Result = CGF.Builder.CreateAtomicRMW(
      Kind, Args[0], Args[1], llvm::AtomicOrdering::SequentiallyConsistent);
  return EmitFromInt(CGF, Result, T, ValueType);
}

static Value *EmitNontemporalStore(CodeGenFunction &CGF, const CallExpr *E) {
  Value *Val = CGF.EmitScalarExpr(E->getArg(0));
  Value *Address = CGF.EmitScalarExpr(E->getArg(1));

  // Convert the type of the pointer to a pointer to the stored type.
  Val = CGF.EmitToMemory(Val, E->getArg(0)->getType());
  Value *BC = CGF.Builder.CreateBitCast(
      Address, llvm::PointerType::getUnqual(Val->getType()), "cast");
  LValue LV = CGF.MakeNaturalAlignAddrLValue(BC, E->getArg(0)->getType());
  LV.setNontemporal(true);
  CGF.EmitStoreOfScalar(Val, LV, false);
  return nullptr;
}

static Value *EmitNontemporalLoad(CodeGenFunction &CGF, const CallExpr *E) {
  Value *Address = CGF.EmitScalarExpr(E->getArg(0));

  LValue LV = CGF.MakeNaturalAlignAddrLValue(Address, E->getType());
  LV.setNontemporal(true);
  return CGF.EmitLoadOfScalar(LV, E->getExprLoc());
}

static RValue EmitBinaryAtomic(CodeGenFunction &CGF,
                               llvm::AtomicRMWInst::BinOp Kind,
                               const CallExpr *E) {
  return RValue::get(MakeBinaryAtomicValue(CGF, Kind, E));
}

/// Utility to insert an atomic instruction based Instrinsic::ID and
/// the expression node, where the return value is the result of the
/// operation.
static RValue EmitBinaryAtomicPost(CodeGenFunction &CGF,
                                   llvm::AtomicRMWInst::BinOp Kind,
                                   const CallExpr *E,
                                   Instruction::BinaryOps Op,
                                   bool Invert = false) {
  QualType T = E->getType();
  assert(E->getArg(0)->getType()->isPointerType());
  assert(CGF.getContext().hasSameUnqualifiedType(T,
                                  E->getArg(0)->getType()->getPointeeType()));
  assert(CGF.getContext().hasSameUnqualifiedType(T, E->getArg(1)->getType()));

  llvm::Value *DestPtr = CGF.EmitScalarExpr(E->getArg(0));
  unsigned AddrSpace = DestPtr->getType()->getPointerAddressSpace();

  llvm::IntegerType *IntType =
    llvm::IntegerType::get(CGF.getLLVMContext(),
                           CGF.getContext().getTypeSize(T));
  llvm::Type *IntPtrType = IntType->getPointerTo(AddrSpace);

  llvm::Value *Args[2];
  Args[1] = CGF.EmitScalarExpr(E->getArg(1));
  llvm::Type *ValueType = Args[1]->getType();
  Args[1] = EmitToInt(CGF, Args[1], T, IntType);
  Args[0] = CGF.Builder.CreateBitCast(DestPtr, IntPtrType);

  llvm::Value *Result = CGF.Builder.CreateAtomicRMW(
      Kind, Args[0], Args[1], llvm::AtomicOrdering::SequentiallyConsistent);
  Result = CGF.Builder.CreateBinOp(Op, Result, Args[1]);
  if (Invert)
    Result = CGF.Builder.CreateBinOp(llvm::Instruction::Xor, Result,
                                     llvm::ConstantInt::get(IntType, -1));
  Result = EmitFromInt(CGF, Result, T, ValueType);
  return RValue::get(Result);
}

/// Utility to insert an atomic cmpxchg instruction.
///
/// @param CGF The current codegen function.
/// @param E   Builtin call expression to convert to cmpxchg.
///            arg0 - address to operate on
///            arg1 - value to compare with
///            arg2 - new value
/// @param ReturnBool Specifies whether to return success flag of
///                   cmpxchg result or the old value.
///
/// @returns result of cmpxchg, according to ReturnBool
static Value *MakeAtomicCmpXchgValue(CodeGenFunction &CGF, const CallExpr *E,
                                     bool ReturnBool) {
  QualType T = ReturnBool ? E->getArg(1)->getType() : E->getType();
  llvm::Value *DestPtr = CGF.EmitScalarExpr(E->getArg(0));
  unsigned AddrSpace = DestPtr->getType()->getPointerAddressSpace();

  llvm::IntegerType *IntType = llvm::IntegerType::get(
      CGF.getLLVMContext(), CGF.getContext().getTypeSize(T));
  llvm::Type *IntPtrType = IntType->getPointerTo(AddrSpace);

  Value *Args[3];
  Args[0] = CGF.Builder.CreateBitCast(DestPtr, IntPtrType);
  Args[1] = CGF.EmitScalarExpr(E->getArg(1));
  llvm::Type *ValueType = Args[1]->getType();
  Args[1] = EmitToInt(CGF, Args[1], T, IntType);
  Args[2] = EmitToInt(CGF, CGF.EmitScalarExpr(E->getArg(2)), T, IntType);

  Value *Pair = CGF.Builder.CreateAtomicCmpXchg(
      Args[0], Args[1], Args[2], llvm::AtomicOrdering::SequentiallyConsistent,
      llvm::AtomicOrdering::SequentiallyConsistent);
  if (ReturnBool)
    // Extract boolean success flag and zext it to int.
    return CGF.Builder.CreateZExt(CGF.Builder.CreateExtractValue(Pair, 1),
                                  CGF.ConvertType(E->getType()));
  else
    // Extract old value and emit it using the same type as compare value.
    return EmitFromInt(CGF, CGF.Builder.CreateExtractValue(Pair, 0), T,
                       ValueType);
}

// Emit a simple mangled intrinsic that has 1 argument and a return type
// matching the argument type.
static Value *emitUnaryBuiltin(CodeGenFunction &CGF,
                               const CallExpr *E,
                               unsigned IntrinsicID) {
  llvm::Value *Src0 = CGF.EmitScalarExpr(E->getArg(0));

  Value *F = CGF.CGM.getIntrinsic(IntrinsicID, Src0->getType());
  return CGF.Builder.CreateCall(F, Src0);
}

// Emit an intrinsic that has 2 operands of the same type as its result.
static Value *emitBinaryBuiltin(CodeGenFunction &CGF,
                                const CallExpr *E,
                                unsigned IntrinsicID) {
  llvm::Value *Src0 = CGF.EmitScalarExpr(E->getArg(0));
  llvm::Value *Src1 = CGF.EmitScalarExpr(E->getArg(1));

  Value *F = CGF.CGM.getIntrinsic(IntrinsicID, Src0->getType());
  return CGF.Builder.CreateCall(F, { Src0, Src1 });
}

// Emit an intrinsic that has 3 operands of the same type as its result.
static Value *emitTernaryBuiltin(CodeGenFunction &CGF,
                                 const CallExpr *E,
                                 unsigned IntrinsicID) {
  llvm::Value *Src0 = CGF.EmitScalarExpr(E->getArg(0));
  llvm::Value *Src1 = CGF.EmitScalarExpr(E->getArg(1));
  llvm::Value *Src2 = CGF.EmitScalarExpr(E->getArg(2));

  Value *F = CGF.CGM.getIntrinsic(IntrinsicID, Src0->getType());
  return CGF.Builder.CreateCall(F, { Src0, Src1, Src2 });
}

// Emit an intrinsic that has 1 float or double operand, and 1 integer.
static Value *emitFPIntBuiltin(CodeGenFunction &CGF,
                               const CallExpr *E,
                               unsigned IntrinsicID) {
  llvm::Value *Src0 = CGF.EmitScalarExpr(E->getArg(0));
  llvm::Value *Src1 = CGF.EmitScalarExpr(E->getArg(1));

  Value *F = CGF.CGM.getIntrinsic(IntrinsicID, Src0->getType());
  return CGF.Builder.CreateCall(F, {Src0, Src1});
}

/// EmitFAbs - Emit a call to @llvm.fabs().
static Value *EmitFAbs(CodeGenFunction &CGF, Value *V) {
  Value *F = CGF.CGM.getIntrinsic(Intrinsic::fabs, V->getType());
  llvm::CallInst *Call = CGF.Builder.CreateCall(F, V);
  Call->setDoesNotAccessMemory();
  return Call;
}

/// Emit the computation of the sign bit for a floating point value. Returns
/// the i1 sign bit value.
static Value *EmitSignBit(CodeGenFunction &CGF, Value *V) {
  LLVMContext &C = CGF.CGM.getLLVMContext();

  llvm::Type *Ty = V->getType();
  int Width = Ty->getPrimitiveSizeInBits();
  llvm::Type *IntTy = llvm::IntegerType::get(C, Width);
  V = CGF.Builder.CreateBitCast(V, IntTy);
  if (Ty->isPPC_FP128Ty()) {
    // We want the sign bit of the higher-order double. The bitcast we just
    // did works as if the double-double was stored to memory and then
    // read as an i128. The "store" will put the higher-order double in the
    // lower address in both little- and big-Endian modes, but the "load"
    // will treat those bits as a different part of the i128: the low bits in
    // little-Endian, the high bits in big-Endian. Therefore, on big-Endian
    // we need to shift the high bits down to the low before truncating.
    Width >>= 1;
    if (CGF.getTarget().isBigEndian()) {
      Value *ShiftCst = llvm::ConstantInt::get(IntTy, Width);
      V = CGF.Builder.CreateLShr(V, ShiftCst);
    }
    // We are truncating value in order to extract the higher-order
    // double, which we will be using to extract the sign from.
    IntTy = llvm::IntegerType::get(C, Width);
    V = CGF.Builder.CreateTrunc(V, IntTy);
  }
  Value *Zero = llvm::Constant::getNullValue(IntTy);
  return CGF.Builder.CreateICmpSLT(V, Zero);
}

static RValue emitLibraryCall(CodeGenFunction &CGF, const FunctionDecl *FD,
                              const CallExpr *E, llvm::Constant *calleeValue) {
  CGCallee callee = CGCallee::forDirect(calleeValue, FD);
  return CGF.EmitCall(E->getCallee()->getType(), callee, E, ReturnValueSlot());
}

/// Emit a call to llvm.{sadd,uadd,ssub,usub,smul,umul}.with.overflow.*
/// depending on IntrinsicID.
///
/// \arg CGF The current codegen function.
/// \arg IntrinsicID The ID for the Intrinsic we wish to generate.
/// \arg X The first argument to the llvm.*.with.overflow.*.
/// \arg Y The second argument to the llvm.*.with.overflow.*.
/// \arg Carry The carry returned by the llvm.*.with.overflow.*.
/// \returns The result (i.e. sum/product) returned by the intrinsic.
static llvm::Value *EmitOverflowIntrinsic(CodeGenFunction &CGF,
                                          const llvm::Intrinsic::ID IntrinsicID,
                                          llvm::Value *X, llvm::Value *Y,
                                          llvm::Value *&Carry) {
  // Make sure we have integers of the same width.
  assert(X->getType() == Y->getType() &&
         "Arguments must be the same type. (Did you forget to make sure both "
         "arguments have the same integer width?)");

  llvm::Value *Callee = CGF.CGM.getIntrinsic(IntrinsicID, X->getType());
  llvm::Value *Tmp = CGF.Builder.CreateCall(Callee, {X, Y});
  Carry = CGF.Builder.CreateExtractValue(Tmp, 1);
  return CGF.Builder.CreateExtractValue(Tmp, 0);
}

static Value *emitRangedBuiltin(CodeGenFunction &CGF,
                                unsigned IntrinsicID,
                                int low, int high) {
    llvm::MDBuilder MDHelper(CGF.getLLVMContext());
    llvm::MDNode *RNode = MDHelper.createRange(APInt(32, low), APInt(32, high));
    Value *F = CGF.CGM.getIntrinsic(IntrinsicID, {});
    llvm::Instruction *Call = CGF.Builder.CreateCall(F);
    Call->setMetadata(llvm::LLVMContext::MD_range, RNode);
    return Call;
}

namespace {
  struct WidthAndSignedness {
    unsigned Width;
    bool Signed;
  };
}

static WidthAndSignedness
getIntegerWidthAndSignedness(const clang::ASTContext &context,
                             const clang::QualType Type) {
  assert(Type->isIntegerType() && "Given type is not an integer.");
  unsigned Width = Type->isBooleanType() ? 1 : context.getTypeInfo(Type).Width;
  bool Signed = Type->isSignedIntegerType();
  return {Width, Signed};
}

// Given one or more integer types, this function produces an integer type that
// encompasses them: any value in one of the given types could be expressed in
// the encompassing type.
static struct WidthAndSignedness
EncompassingIntegerType(ArrayRef<struct WidthAndSignedness> Types) {
  assert(Types.size() > 0 && "Empty list of types.");

  // If any of the given types is signed, we must return a signed type.
  bool Signed = false;
  for (const auto &Type : Types) {
    Signed |= Type.Signed;
  }

  // The encompassing type must have a width greater than or equal to the width
  // of the specified types.  Additionally, if the encompassing type is signed,
  // its width must be strictly greater than the width of any unsigned types
  // given.
  unsigned Width = 0;
  for (const auto &Type : Types) {
    unsigned MinWidth = Type.Width + (Signed && !Type.Signed);
    if (Width < MinWidth) {
      Width = MinWidth;
    }
  }

  return {Width, Signed};
}

Value *CodeGenFunction::EmitVAStartEnd(Value *ArgValue, bool IsStart) {
  llvm::Type *DestType = Int8PtrTy;
  if (ArgValue->getType() != DestType)
    ArgValue =
        Builder.CreateBitCast(ArgValue, DestType, ArgValue->getName().data());

  Intrinsic::ID inst = IsStart ? Intrinsic::vastart : Intrinsic::vaend;
  return Builder.CreateCall(CGM.getIntrinsic(inst), ArgValue);
}

/// Checks if using the result of __builtin_object_size(p, @p From) in place of
/// __builtin_object_size(p, @p To) is correct
static bool areBOSTypesCompatible(int From, int To) {
  // Note: Our __builtin_object_size implementation currently treats Type=0 and
  // Type=2 identically. Encoding this implementation detail here may make
  // improving __builtin_object_size difficult in the future, so it's omitted.
  return From == To || (From == 0 && To == 1) || (From == 3 && To == 2);
}

static llvm::Value *
getDefaultBuiltinObjectSizeResult(unsigned Type, llvm::IntegerType *ResType) {
  return ConstantInt::get(ResType, (Type & 2) ? 0 : -1, /*isSigned=*/true);
}

llvm::Value *
CodeGenFunction::evaluateOrEmitBuiltinObjectSize(const Expr *E, unsigned Type,
                                                 llvm::IntegerType *ResType,
                                                 llvm::Value *EmittedE) {
  uint64_t ObjectSize;
  if (!E->tryEvaluateObjectSize(ObjectSize, getContext(), Type))
    return emitBuiltinObjectSize(E, Type, ResType, EmittedE);
  return ConstantInt::get(ResType, ObjectSize, /*isSigned=*/true);
}

/// Returns a Value corresponding to the size of the given expression.
/// This Value may be either of the following:
///   - A llvm::Argument (if E is a param with the pass_object_size attribute on
///     it)
///   - A call to the @llvm.objectsize intrinsic
///
/// EmittedE is the result of emitting `E` as a scalar expr. If it's non-null
/// and we wouldn't otherwise try to reference a pass_object_size parameter,
/// we'll call @llvm.objectsize on EmittedE, rather than emitting E.
llvm::Value *
CodeGenFunction::emitBuiltinObjectSize(const Expr *E, unsigned Type,
                                       llvm::IntegerType *ResType,
                                       llvm::Value *EmittedE) {
  // We need to reference an argument if the pointer is a parameter with the
  // pass_object_size attribute.
  if (auto *D = dyn_cast<DeclRefExpr>(E->IgnoreParenImpCasts())) {
    auto *Param = dyn_cast<ParmVarDecl>(D->getDecl());
    auto *PS = D->getDecl()->getAttr<PassObjectSizeAttr>();
    if (Param != nullptr && PS != nullptr &&
        areBOSTypesCompatible(PS->getType(), Type)) {
      auto Iter = SizeArguments.find(Param);
      assert(Iter != SizeArguments.end());

      const ImplicitParamDecl *D = Iter->second;
      auto DIter = LocalDeclMap.find(D);
      assert(DIter != LocalDeclMap.end());

      return EmitLoadOfScalar(DIter->second, /*volatile=*/false,
                              getContext().getSizeType(), E->getBeginLoc());
    }
  }

  // LLVM can't handle Type=3 appropriately, and __builtin_object_size shouldn't
  // evaluate E for side-effects. In either case, we shouldn't lower to
  // @llvm.objectsize.
  if (Type == 3 || (!EmittedE && E->HasSideEffects(getContext())))
    return getDefaultBuiltinObjectSizeResult(Type, ResType);

  Value *Ptr = EmittedE ? EmittedE : EmitScalarExpr(E);
  assert(Ptr->getType()->isPointerTy() &&
         "Non-pointer passed to __builtin_object_size?");

  Value *F = CGM.getIntrinsic(Intrinsic::objectsize, {ResType, Ptr->getType()});

  // LLVM only supports 0 and 2, make sure that we pass along that as a boolean.
  Value *Min = Builder.getInt1((Type & 2) != 0);
  // For GCC compatibility, __builtin_object_size treat NULL as unknown size.
  Value *NullIsUnknown = Builder.getTrue();
  return Builder.CreateCall(F, {Ptr, Min, NullIsUnknown});
}

namespace {
/// A struct to generically desribe a bit test intrinsic.
struct BitTest {
  enum ActionKind : uint8_t { TestOnly, Complement, Reset, Set };
  enum InterlockingKind : uint8_t {
    Unlocked,
    Sequential,
    Acquire,
    Release,
    NoFence
  };

  ActionKind Action;
  InterlockingKind Interlocking;
  bool Is64Bit;

  static BitTest decodeBitTestBuiltin(unsigned BuiltinID);
};
} // namespace

BitTest BitTest::decodeBitTestBuiltin(unsigned BuiltinID) {
  switch (BuiltinID) {
    // Main portable variants.
  case Builtin::BI_bittest:
    return {TestOnly, Unlocked, false};
  case Builtin::BI_bittestandcomplement:
    return {Complement, Unlocked, false};
  case Builtin::BI_bittestandreset:
    return {Reset, Unlocked, false};
  case Builtin::BI_bittestandset:
    return {Set, Unlocked, false};
  case Builtin::BI_interlockedbittestandreset:
    return {Reset, Sequential, false};
  case Builtin::BI_interlockedbittestandset:
    return {Set, Sequential, false};

    // X86-specific 64-bit variants.
  case Builtin::BI_bittest64:
    return {TestOnly, Unlocked, true};
  case Builtin::BI_bittestandcomplement64:
    return {Complement, Unlocked, true};
  case Builtin::BI_bittestandreset64:
    return {Reset, Unlocked, true};
  case Builtin::BI_bittestandset64:
    return {Set, Unlocked, true};
  case Builtin::BI_interlockedbittestandreset64:
    return {Reset, Sequential, true};
  case Builtin::BI_interlockedbittestandset64:
    return {Set, Sequential, true};

    // ARM/AArch64-specific ordering variants.
  case Builtin::BI_interlockedbittestandset_acq:
    return {Set, Acquire, false};
  case Builtin::BI_interlockedbittestandset_rel:
    return {Set, Release, false};
  case Builtin::BI_interlockedbittestandset_nf:
    return {Set, NoFence, false};
  case Builtin::BI_interlockedbittestandreset_acq:
    return {Reset, Acquire, false};
  case Builtin::BI_interlockedbittestandreset_rel:
    return {Reset, Release, false};
  case Builtin::BI_interlockedbittestandreset_nf:
    return {Reset, NoFence, false};
  }
  llvm_unreachable("expected only bittest intrinsics");
}

static char bitActionToX86BTCode(BitTest::ActionKind A) {
  switch (A) {
  case BitTest::TestOnly:   return '\0';
  case BitTest::Complement: return 'c';
  case BitTest::Reset:      return 'r';
  case BitTest::Set:        return 's';
  }
  llvm_unreachable("invalid action");
}

static llvm::Value *EmitX86BitTestIntrinsic(CodeGenFunction &CGF,
                                            BitTest BT,
                                            const CallExpr *E, Value *BitBase,
                                            Value *BitPos) {
  char Action = bitActionToX86BTCode(BT.Action);
  char SizeSuffix = BT.Is64Bit ? 'q' : 'l';

  // Build the assembly.
  SmallString<64> Asm;
  raw_svector_ostream AsmOS(Asm);
  if (BT.Interlocking != BitTest::Unlocked)
    AsmOS << "lock ";
  AsmOS << "bt";
  if (Action)
    AsmOS << Action;
  AsmOS << SizeSuffix << " $2, ($1)\n\tsetc ${0:b}";

  // Build the constraints. FIXME: We should support immediates when possible.
  std::string Constraints = "=r,r,r,~{cc},~{flags},~{fpsr}";
  llvm::IntegerType *IntType = llvm::IntegerType::get(
      CGF.getLLVMContext(),
      CGF.getContext().getTypeSize(E->getArg(1)->getType()));
  llvm::Type *IntPtrType = IntType->getPointerTo();
  llvm::FunctionType *FTy =
      llvm::FunctionType::get(CGF.Int8Ty, {IntPtrType, IntType}, false);

  llvm::InlineAsm *IA =
      llvm::InlineAsm::get(FTy, Asm, Constraints, /*SideEffects=*/true);
  return CGF.Builder.CreateCall(IA, {BitBase, BitPos});
}

static llvm::AtomicOrdering
getBitTestAtomicOrdering(BitTest::InterlockingKind I) {
  switch (I) {
  case BitTest::Unlocked:   return llvm::AtomicOrdering::NotAtomic;
  case BitTest::Sequential: return llvm::AtomicOrdering::SequentiallyConsistent;
  case BitTest::Acquire:    return llvm::AtomicOrdering::Acquire;
  case BitTest::Release:    return llvm::AtomicOrdering::Release;
  case BitTest::NoFence:    return llvm::AtomicOrdering::Monotonic;
  }
  llvm_unreachable("invalid interlocking");
}

/// Emit a _bittest* intrinsic. These intrinsics take a pointer to an array of
/// bits and a bit position and read and optionally modify the bit at that
/// position. The position index can be arbitrarily large, i.e. it can be larger
/// than 31 or 63, so we need an indexed load in the general case.
static llvm::Value *EmitBitTestIntrinsic(CodeGenFunction &CGF,
                                         unsigned BuiltinID,
                                         const CallExpr *E) {
  Value *BitBase = CGF.EmitScalarExpr(E->getArg(0));
  Value *BitPos = CGF.EmitScalarExpr(E->getArg(1));

  BitTest BT = BitTest::decodeBitTestBuiltin(BuiltinID);

  // X86 has special BT, BTC, BTR, and BTS instructions that handle the array
  // indexing operation internally. Use them if possible.
  llvm::Triple::ArchType Arch = CGF.getTarget().getTriple().getArch();
  if (Arch == llvm::Triple::x86 || Arch == llvm::Triple::x86_64)
    return EmitX86BitTestIntrinsic(CGF, BT, E, BitBase, BitPos);

  // Otherwise, use generic code to load one byte and test the bit. Use all but
  // the bottom three bits as the array index, and the bottom three bits to form
  // a mask.
  // Bit = BitBaseI8[BitPos >> 3] & (1 << (BitPos & 0x7)) != 0;
  Value *ByteIndex = CGF.Builder.CreateAShr(
      BitPos, llvm::ConstantInt::get(BitPos->getType(), 3), "bittest.byteidx");
  Value *BitBaseI8 = CGF.Builder.CreatePointerCast(BitBase, CGF.Int8PtrTy);
  Address ByteAddr(CGF.Builder.CreateInBoundsGEP(CGF.Int8Ty, BitBaseI8,
                                                 ByteIndex, "bittest.byteaddr"),
                   CharUnits::One());
  Value *PosLow =
      CGF.Builder.CreateAnd(CGF.Builder.CreateTrunc(BitPos, CGF.Int8Ty),
                            llvm::ConstantInt::get(CGF.Int8Ty, 0x7));

  // The updating instructions will need a mask.
  Value *Mask = nullptr;
  if (BT.Action != BitTest::TestOnly) {
    Mask = CGF.Builder.CreateShl(llvm::ConstantInt::get(CGF.Int8Ty, 1), PosLow,
                                 "bittest.mask");
  }

  // Check the action and ordering of the interlocked intrinsics.
  llvm::AtomicOrdering Ordering = getBitTestAtomicOrdering(BT.Interlocking);

  Value *OldByte = nullptr;
  if (Ordering != llvm::AtomicOrdering::NotAtomic) {
    // Emit a combined atomicrmw load/store operation for the interlocked
    // intrinsics.
    llvm::AtomicRMWInst::BinOp RMWOp = llvm::AtomicRMWInst::Or;
    if (BT.Action == BitTest::Reset) {
      Mask = CGF.Builder.CreateNot(Mask);
      RMWOp = llvm::AtomicRMWInst::And;
    }
    OldByte = CGF.Builder.CreateAtomicRMW(RMWOp, ByteAddr.getPointer(), Mask,
                                          Ordering);
  } else {
    // Emit a plain load for the non-interlocked intrinsics.
    OldByte = CGF.Builder.CreateLoad(ByteAddr, "bittest.byte");
    Value *NewByte = nullptr;
    switch (BT.Action) {
    case BitTest::TestOnly:
      // Don't store anything.
      break;
    case BitTest::Complement:
      NewByte = CGF.Builder.CreateXor(OldByte, Mask);
      break;
    case BitTest::Reset:
      NewByte = CGF.Builder.CreateAnd(OldByte, CGF.Builder.CreateNot(Mask));
      break;
    case BitTest::Set:
      NewByte = CGF.Builder.CreateOr(OldByte, Mask);
      break;
    }
    if (NewByte)
      CGF.Builder.CreateStore(NewByte, ByteAddr);
  }

  // However we loaded the old byte, either by plain load or atomicrmw, shift
  // the bit into the low position and mask it to 0 or 1.
  Value *ShiftedByte = CGF.Builder.CreateLShr(OldByte, PosLow, "bittest.shr");
  return CGF.Builder.CreateAnd(
      ShiftedByte, llvm::ConstantInt::get(CGF.Int8Ty, 1), "bittest.res");
}

namespace {
enum class MSVCSetJmpKind {
  _setjmpex,
  _setjmp3,
  _setjmp
};
}

/// MSVC handles setjmp a bit differently on different platforms. On every
/// architecture except 32-bit x86, the frame address is passed. On x86, extra
/// parameters can be passed as variadic arguments, but we always pass none.
static RValue EmitMSVCRTSetJmp(CodeGenFunction &CGF, MSVCSetJmpKind SJKind,
                               const CallExpr *E) {
  llvm::Value *Arg1 = nullptr;
  llvm::Type *Arg1Ty = nullptr;
  StringRef Name;
  bool IsVarArg = false;
  if (SJKind == MSVCSetJmpKind::_setjmp3) {
    Name = "_setjmp3";
    Arg1Ty = CGF.Int32Ty;
    Arg1 = llvm::ConstantInt::get(CGF.IntTy, 0);
    IsVarArg = true;
  } else {
    Name = SJKind == MSVCSetJmpKind::_setjmp ? "_setjmp" : "_setjmpex";
    Arg1Ty = CGF.Int8PtrTy;
    Arg1 = CGF.Builder.CreateCall(CGF.CGM.getIntrinsic(Intrinsic::frameaddress),
                                  llvm::ConstantInt::get(CGF.Int32Ty, 0));
  }

  // Mark the call site and declaration with ReturnsTwice.
  llvm::Type *ArgTypes[2] = {CGF.Int8PtrTy, Arg1Ty};
  llvm::AttributeList ReturnsTwiceAttr = llvm::AttributeList::get(
      CGF.getLLVMContext(), llvm::AttributeList::FunctionIndex,
      llvm::Attribute::ReturnsTwice);
  llvm::Constant *SetJmpFn = CGF.CGM.CreateRuntimeFunction(
      llvm::FunctionType::get(CGF.IntTy, ArgTypes, IsVarArg), Name,
      ReturnsTwiceAttr, /*Local=*/true);

  llvm::Value *Buf = CGF.Builder.CreateBitOrPointerCast(
      CGF.EmitScalarExpr(E->getArg(0)), CGF.Int8PtrTy);
  llvm::Value *Args[] = {Buf, Arg1};
  llvm::CallSite CS = CGF.EmitRuntimeCallOrInvoke(SetJmpFn, Args);
  CS.setAttributes(ReturnsTwiceAttr);
  return RValue::get(CS.getInstruction());
}

// Many of MSVC builtins are on x64, ARM and AArch64; to avoid repeating code,
// we handle them here.
enum class CodeGenFunction::MSVCIntrin {
  _BitScanForward,
  _BitScanReverse,
  _InterlockedAnd,
  _InterlockedDecrement,
  _InterlockedExchange,
  _InterlockedExchangeAdd,
  _InterlockedExchangeSub,
  _InterlockedIncrement,
  _InterlockedOr,
  _InterlockedXor,
  __fastfail,
};

Value *CodeGenFunction::EmitMSVCBuiltinExpr(MSVCIntrin BuiltinID,
                                            const CallExpr *E) {
  switch (BuiltinID) {
  case MSVCIntrin::_BitScanForward:
  case MSVCIntrin::_BitScanReverse: {
    Value *ArgValue = EmitScalarExpr(E->getArg(1));

    llvm::Type *ArgType = ArgValue->getType();
    llvm::Type *IndexType =
      EmitScalarExpr(E->getArg(0))->getType()->getPointerElementType();
    llvm::Type *ResultType = ConvertType(E->getType());

    Value *ArgZero = llvm::Constant::getNullValue(ArgType);
    Value *ResZero = llvm::Constant::getNullValue(ResultType);
    Value *ResOne = llvm::ConstantInt::get(ResultType, 1);

    BasicBlock *Begin = Builder.GetInsertBlock();
    BasicBlock *End = createBasicBlock("bitscan_end", this->CurFn);
    Builder.SetInsertPoint(End);
    PHINode *Result = Builder.CreatePHI(ResultType, 2, "bitscan_result");

    Builder.SetInsertPoint(Begin);
    Value *IsZero = Builder.CreateICmpEQ(ArgValue, ArgZero);
    BasicBlock *NotZero = createBasicBlock("bitscan_not_zero", this->CurFn);
    Builder.CreateCondBr(IsZero, End, NotZero);
    Result->addIncoming(ResZero, Begin);

    Builder.SetInsertPoint(NotZero);
    Address IndexAddress = EmitPointerWithAlignment(E->getArg(0));

    if (BuiltinID == MSVCIntrin::_BitScanForward) {
      Value *F = CGM.getIntrinsic(Intrinsic::cttz, ArgType);
      Value *ZeroCount = Builder.CreateCall(F, {ArgValue, Builder.getTrue()});
      ZeroCount = Builder.CreateIntCast(ZeroCount, IndexType, false);
      Builder.CreateStore(ZeroCount, IndexAddress, false);
    } else {
      unsigned ArgWidth = cast<llvm::IntegerType>(ArgType)->getBitWidth();
      Value *ArgTypeLastIndex = llvm::ConstantInt::get(IndexType, ArgWidth - 1);

      Value *F = CGM.getIntrinsic(Intrinsic::ctlz, ArgType);
      Value *ZeroCount = Builder.CreateCall(F, {ArgValue, Builder.getTrue()});
      ZeroCount = Builder.CreateIntCast(ZeroCount, IndexType, false);
      Value *Index = Builder.CreateNSWSub(ArgTypeLastIndex, ZeroCount);
      Builder.CreateStore(Index, IndexAddress, false);
    }
    Builder.CreateBr(End);
    Result->addIncoming(ResOne, NotZero);

    Builder.SetInsertPoint(End);
    return Result;
  }
  case MSVCIntrin::_InterlockedAnd:
    return MakeBinaryAtomicValue(*this, AtomicRMWInst::And, E);
  case MSVCIntrin::_InterlockedExchange:
    return MakeBinaryAtomicValue(*this, AtomicRMWInst::Xchg, E);
  case MSVCIntrin::_InterlockedExchangeAdd:
    return MakeBinaryAtomicValue(*this, AtomicRMWInst::Add, E);
  case MSVCIntrin::_InterlockedExchangeSub:
    return MakeBinaryAtomicValue(*this, AtomicRMWInst::Sub, E);
  case MSVCIntrin::_InterlockedOr:
    return MakeBinaryAtomicValue(*this, AtomicRMWInst::Or, E);
  case MSVCIntrin::_InterlockedXor:
    return MakeBinaryAtomicValue(*this, AtomicRMWInst::Xor, E);

  case MSVCIntrin::_InterlockedDecrement: {
    llvm::Type *IntTy = ConvertType(E->getType());
    AtomicRMWInst *RMWI = Builder.CreateAtomicRMW(
      AtomicRMWInst::Sub,
      EmitScalarExpr(E->getArg(0)),
      ConstantInt::get(IntTy, 1),
      llvm::AtomicOrdering::SequentiallyConsistent);
    return Builder.CreateSub(RMWI, ConstantInt::get(IntTy, 1));
  }
  case MSVCIntrin::_InterlockedIncrement: {
    llvm::Type *IntTy = ConvertType(E->getType());
    AtomicRMWInst *RMWI = Builder.CreateAtomicRMW(
      AtomicRMWInst::Add,
      EmitScalarExpr(E->getArg(0)),
      ConstantInt::get(IntTy, 1),
      llvm::AtomicOrdering::SequentiallyConsistent);
    return Builder.CreateAdd(RMWI, ConstantInt::get(IntTy, 1));
  }

  case MSVCIntrin::__fastfail: {
    // Request immediate process termination from the kernel. The instruction
    // sequences to do this are documented on MSDN:
    // https://msdn.microsoft.com/en-us/library/dn774154.aspx
    llvm::Triple::ArchType ISA = getTarget().getTriple().getArch();
    StringRef Asm, Constraints;
    switch (ISA) {
    default:
      ErrorUnsupported(E, "__fastfail call for this architecture");
      break;
    case llvm::Triple::x86:
    case llvm::Triple::x86_64:
      Asm = "int $$0x29";
      Constraints = "{cx}";
      break;
    case llvm::Triple::thumb:
      Asm = "udf #251";
      Constraints = "{r0}";
      break;
    }
    llvm::FunctionType *FTy = llvm::FunctionType::get(VoidTy, {Int32Ty}, false);
    llvm::InlineAsm *IA =
        llvm::InlineAsm::get(FTy, Asm, Constraints, /*SideEffects=*/true);
    llvm::AttributeList NoReturnAttr = llvm::AttributeList::get(
        getLLVMContext(), llvm::AttributeList::FunctionIndex,
        llvm::Attribute::NoReturn);
    CallSite CS = Builder.CreateCall(IA, EmitScalarExpr(E->getArg(0)));
    CS.setAttributes(NoReturnAttr);
    return CS.getInstruction();
  }
  }
  llvm_unreachable("Incorrect MSVC intrinsic!");
}

namespace {
// ARC cleanup for __builtin_os_log_format
struct CallObjCArcUse final : EHScopeStack::Cleanup {
  CallObjCArcUse(llvm::Value *object) : object(object) {}
  llvm::Value *object;

  void Emit(CodeGenFunction &CGF, Flags flags) override {
    CGF.EmitARCIntrinsicUse(object);
  }
};
}

Value *CodeGenFunction::EmitCheckedArgForBuiltin(const Expr *E,
                                                 BuiltinCheckKind Kind) {
  assert((Kind == BCK_CLZPassedZero || Kind == BCK_CTZPassedZero)
          && "Unsupported builtin check kind");

  Value *ArgValue = EmitScalarExpr(E);
  if (!SanOpts.has(SanitizerKind::Builtin) || !getTarget().isCLZForZeroUndef())
    return ArgValue;

  SanitizerScope SanScope(this);
  Value *Cond = Builder.CreateICmpNE(
      ArgValue, llvm::Constant::getNullValue(ArgValue->getType()));
  EmitCheck(std::make_pair(Cond, SanitizerKind::Builtin),
            SanitizerHandler::InvalidBuiltin,
            {EmitCheckSourceLocation(E->getExprLoc()),
             llvm::ConstantInt::get(Builder.getInt8Ty(), Kind)},
            None);
  return ArgValue;
}

/// Get the argument type for arguments to os_log_helper.
static CanQualType getOSLogArgType(ASTContext &C, int Size) {
  QualType UnsignedTy = C.getIntTypeForBitwidth(Size * 8, /*Signed=*/false);
  return C.getCanonicalType(UnsignedTy);
}

llvm::Function *CodeGenFunction::generateBuiltinOSLogHelperFunction(
    const analyze_os_log::OSLogBufferLayout &Layout,
    CharUnits BufferAlignment) {
  ASTContext &Ctx = getContext();

  llvm::SmallString<64> Name;
  {
    raw_svector_ostream OS(Name);
    OS << "__os_log_helper";
    OS << "_" << BufferAlignment.getQuantity();
    OS << "_" << int(Layout.getSummaryByte());
    OS << "_" << int(Layout.getNumArgsByte());
    for (const auto &Item : Layout.Items)
      OS << "_" << int(Item.getSizeByte()) << "_"
         << int(Item.getDescriptorByte());
  }

  if (llvm::Function *F = CGM.getModule().getFunction(Name))
    return F;

  llvm::SmallVector<ImplicitParamDecl, 4> Params;
  Params.emplace_back(Ctx, nullptr, SourceLocation(), &Ctx.Idents.get("buffer"),
                      Ctx.VoidPtrTy, ImplicitParamDecl::Other);

  for (unsigned int I = 0, E = Layout.Items.size(); I < E; ++I) {
    char Size = Layout.Items[I].getSizeByte();
    if (!Size)
      continue;

    Params.emplace_back(
        Ctx, nullptr, SourceLocation(),
        &Ctx.Idents.get(std::string("arg") + llvm::to_string(I)),
        getOSLogArgType(Ctx, Size), ImplicitParamDecl::Other);
  }

  FunctionArgList Args;
  for (auto &P : Params)
    Args.push_back(&P);

  // The helper function has linkonce_odr linkage to enable the linker to merge
  // identical functions. To ensure the merging always happens, 'noinline' is
  // attached to the function when compiling with -Oz.
  const CGFunctionInfo &FI =
      CGM.getTypes().arrangeBuiltinFunctionDeclaration(Ctx.VoidTy, Args);
  llvm::FunctionType *FuncTy = CGM.getTypes().GetFunctionType(FI);
  llvm::Function *Fn = llvm::Function::Create(
      FuncTy, llvm::GlobalValue::LinkOnceODRLinkage, Name, &CGM.getModule());
  Fn->setVisibility(llvm::GlobalValue::HiddenVisibility);
  CGM.SetLLVMFunctionAttributes(nullptr, FI, Fn);
  CGM.SetLLVMFunctionAttributesForDefinition(nullptr, Fn);

  // Attach 'noinline' at -Oz.
  if (CGM.getCodeGenOpts().OptimizeSize == 2)
    Fn->addFnAttr(llvm::Attribute::NoInline);

  auto NL = ApplyDebugLocation::CreateEmpty(*this);
  IdentifierInfo *II = &Ctx.Idents.get(Name);
  FunctionDecl *FD = FunctionDecl::Create(
      Ctx, Ctx.getTranslationUnitDecl(), SourceLocation(), SourceLocation(), II,
      Ctx.VoidTy, nullptr, SC_PrivateExtern, false, false);

  StartFunction(FD, Ctx.VoidTy, Fn, FI, Args);

  // Create a scope with an artificial location for the body of this function.
  auto AL = ApplyDebugLocation::CreateArtificial(*this);

  CharUnits Offset;
  Address BufAddr(Builder.CreateLoad(GetAddrOfLocalVar(&Params[0]), "buf"),
                  BufferAlignment);
  Builder.CreateStore(Builder.getInt8(Layout.getSummaryByte()),
                      Builder.CreateConstByteGEP(BufAddr, Offset++, "summary"));
  Builder.CreateStore(Builder.getInt8(Layout.getNumArgsByte()),
                      Builder.CreateConstByteGEP(BufAddr, Offset++, "numArgs"));

  unsigned I = 1;
  for (const auto &Item : Layout.Items) {
    Builder.CreateStore(
        Builder.getInt8(Item.getDescriptorByte()),
        Builder.CreateConstByteGEP(BufAddr, Offset++, "argDescriptor"));
    Builder.CreateStore(
        Builder.getInt8(Item.getSizeByte()),
        Builder.CreateConstByteGEP(BufAddr, Offset++, "argSize"));

    CharUnits Size = Item.size();
    if (!Size.getQuantity())
      continue;

    Address Arg = GetAddrOfLocalVar(&Params[I]);
    Address Addr = Builder.CreateConstByteGEP(BufAddr, Offset, "argData");
    Addr = Builder.CreateBitCast(Addr, Arg.getPointer()->getType(),
                                 "argDataCast");
    Builder.CreateStore(Builder.CreateLoad(Arg), Addr);
    Offset += Size;
    ++I;
  }

  FinishFunction();

  return Fn;
}

RValue CodeGenFunction::emitBuiltinOSLogFormat(const CallExpr &E) {
  assert(E.getNumArgs() >= 2 &&
         "__builtin_os_log_format takes at least 2 arguments");
  ASTContext &Ctx = getContext();
  analyze_os_log::OSLogBufferLayout Layout;
  analyze_os_log::computeOSLogBufferLayout(Ctx, &E, Layout);
  Address BufAddr = EmitPointerWithAlignment(E.getArg(0));
  llvm::SmallVector<llvm::Value *, 4> RetainableOperands;

  // Ignore argument 1, the format string. It is not currently used.
  CallArgList Args;
  Args.add(RValue::get(BufAddr.getPointer()), Ctx.VoidPtrTy);

  for (const auto &Item : Layout.Items) {
    int Size = Item.getSizeByte();
    if (!Size)
      continue;

    llvm::Value *ArgVal;

    if (const Expr *TheExpr = Item.getExpr()) {
      ArgVal = EmitScalarExpr(TheExpr, /*Ignore*/ false);

      // Check if this is a retainable type.
      if (TheExpr->getType()->isObjCRetainableType()) {
        assert(getEvaluationKind(TheExpr->getType()) == TEK_Scalar &&
               "Only scalar can be a ObjC retainable type");
        // Check if the object is constant, if not, save it in
        // RetainableOperands.
        if (!isa<Constant>(ArgVal))
          RetainableOperands.push_back(ArgVal);
      }
    } else {
      ArgVal = Builder.getInt32(Item.getConstValue().getQuantity());
    }

    unsigned ArgValSize =
        CGM.getDataLayout().getTypeSizeInBits(ArgVal->getType());
    llvm::IntegerType *IntTy = llvm::Type::getIntNTy(getLLVMContext(),
                                                     ArgValSize);
    ArgVal = Builder.CreateBitOrPointerCast(ArgVal, IntTy);
    CanQualType ArgTy = getOSLogArgType(Ctx, Size);
    // If ArgVal has type x86_fp80, zero-extend ArgVal.
    ArgVal = Builder.CreateZExtOrBitCast(ArgVal, ConvertType(ArgTy));
    Args.add(RValue::get(ArgVal), ArgTy);
  }

  const CGFunctionInfo &FI =
      CGM.getTypes().arrangeBuiltinFunctionCall(Ctx.VoidTy, Args);
  llvm::Function *F = CodeGenFunction(CGM).generateBuiltinOSLogHelperFunction(
      Layout, BufAddr.getAlignment());
  EmitCall(FI, CGCallee::forDirect(F), ReturnValueSlot(), Args);

  // Push a clang.arc.use cleanup for each object in RetainableOperands. The
  // cleanup will cause the use to appear after the final log call, keeping
  // the object valid while it’s held in the log buffer.  Note that if there’s
  // a release cleanup on the object, it will already be active; since
  // cleanups are emitted in reverse order, the use will occur before the
  // object is released.
  if (!RetainableOperands.empty() && getLangOpts().ObjCAutoRefCount &&
      CGM.getCodeGenOpts().OptimizationLevel != 0)
    for (llvm::Value *Object : RetainableOperands)
      pushFullExprCleanup<CallObjCArcUse>(getARCCleanupKind(), Object);

  return RValue::get(BufAddr.getPointer());
}

/// Determine if a binop is a checked mixed-sign multiply we can specialize.
static bool isSpecialMixedSignMultiply(unsigned BuiltinID,
                                       WidthAndSignedness Op1Info,
                                       WidthAndSignedness Op2Info,
                                       WidthAndSignedness ResultInfo) {
  return BuiltinID == Builtin::BI__builtin_mul_overflow &&
         Op1Info.Width == Op2Info.Width && Op1Info.Width >= ResultInfo.Width &&
         Op1Info.Signed != Op2Info.Signed;
}

/// Emit a checked mixed-sign multiply. This is a cheaper specialization of
/// the generic checked-binop irgen.
static RValue
EmitCheckedMixedSignMultiply(CodeGenFunction &CGF, const clang::Expr *Op1,
                             WidthAndSignedness Op1Info, const clang::Expr *Op2,
                             WidthAndSignedness Op2Info,
                             const clang::Expr *ResultArg, QualType ResultQTy,
                             WidthAndSignedness ResultInfo) {
  assert(isSpecialMixedSignMultiply(Builtin::BI__builtin_mul_overflow, Op1Info,
                                    Op2Info, ResultInfo) &&
         "Not a mixed-sign multipliction we can specialize");

  // Emit the signed and unsigned operands.
  const clang::Expr *SignedOp = Op1Info.Signed ? Op1 : Op2;
  const clang::Expr *UnsignedOp = Op1Info.Signed ? Op2 : Op1;
  llvm::Value *Signed = CGF.EmitScalarExpr(SignedOp);
  llvm::Value *Unsigned = CGF.EmitScalarExpr(UnsignedOp);

  llvm::Type *OpTy = Signed->getType();
  llvm::Value *Zero = llvm::Constant::getNullValue(OpTy);
  Address ResultPtr = CGF.EmitPointerWithAlignment(ResultArg);
  llvm::Type *ResTy = ResultPtr.getElementType();

  // Take the absolute value of the signed operand.
  llvm::Value *IsNegative = CGF.Builder.CreateICmpSLT(Signed, Zero);
  llvm::Value *AbsOfNegative = CGF.Builder.CreateSub(Zero, Signed);
  llvm::Value *AbsSigned =
      CGF.Builder.CreateSelect(IsNegative, AbsOfNegative, Signed);

  // Perform a checked unsigned multiplication.
  llvm::Value *UnsignedOverflow;
  llvm::Value *UnsignedResult =
      EmitOverflowIntrinsic(CGF, llvm::Intrinsic::umul_with_overflow, AbsSigned,
                            Unsigned, UnsignedOverflow);

  llvm::Value *Overflow, *Result;
  if (ResultInfo.Signed) {
    // Signed overflow occurs if the result is greater than INT_MAX or lesser
    // than INT_MIN, i.e when |Result| > (INT_MAX + IsNegative).
    auto IntMax = llvm::APInt::getSignedMaxValue(ResultInfo.Width)
                      .zextOrSelf(Op1Info.Width);
    llvm::Value *MaxResult =
        CGF.Builder.CreateAdd(llvm::ConstantInt::get(OpTy, IntMax),
                              CGF.Builder.CreateZExt(IsNegative, OpTy));
    llvm::Value *SignedOverflow =
        CGF.Builder.CreateICmpUGT(UnsignedResult, MaxResult);
    Overflow = CGF.Builder.CreateOr(UnsignedOverflow, SignedOverflow);

    // Prepare the signed result (possibly by negating it).
    llvm::Value *NegativeResult = CGF.Builder.CreateNeg(UnsignedResult);
    llvm::Value *SignedResult =
        CGF.Builder.CreateSelect(IsNegative, NegativeResult, UnsignedResult);
    Result = CGF.Builder.CreateTrunc(SignedResult, ResTy);
  } else {
    // Unsigned overflow occurs if the result is < 0 or greater than UINT_MAX.
    llvm::Value *Underflow = CGF.Builder.CreateAnd(
        IsNegative, CGF.Builder.CreateIsNotNull(UnsignedResult));
    Overflow = CGF.Builder.CreateOr(UnsignedOverflow, Underflow);
    if (ResultInfo.Width < Op1Info.Width) {
      auto IntMax =
          llvm::APInt::getMaxValue(ResultInfo.Width).zext(Op1Info.Width);
      llvm::Value *TruncOverflow = CGF.Builder.CreateICmpUGT(
          UnsignedResult, llvm::ConstantInt::get(OpTy, IntMax));
      Overflow = CGF.Builder.CreateOr(Overflow, TruncOverflow);
    }

    // Negate the product if it would be negative in infinite precision.
    Result = CGF.Builder.CreateSelect(
        IsNegative, CGF.Builder.CreateNeg(UnsignedResult), UnsignedResult);

    Result = CGF.Builder.CreateTrunc(Result, ResTy);
  }
  assert(Overflow && Result && "Missing overflow or result");

  bool isVolatile =
      ResultArg->getType()->getPointeeType().isVolatileQualified();
  CGF.Builder.CreateStore(CGF.EmitToMemory(Result, ResultQTy), ResultPtr,
                          isVolatile);
  return RValue::get(Overflow);
}

static llvm::Value *dumpRecord(CodeGenFunction &CGF, QualType RType,
                               Value *&RecordPtr, CharUnits Align, Value *Func,
                               int Lvl) {
  const auto *RT = RType->getAs<RecordType>();
  ASTContext &Context = CGF.getContext();
  RecordDecl *RD = RT->getDecl()->getDefinition();
  ASTContext &Ctx = RD->getASTContext();
  const ASTRecordLayout &RL = Ctx.getASTRecordLayout(RD);
  std::string Pad = std::string(Lvl * 4, ' ');

  Value *GString =
      CGF.Builder.CreateGlobalStringPtr(RType.getAsString() + " {\n");
  Value *Res = CGF.Builder.CreateCall(Func, {GString});

  static llvm::DenseMap<QualType, const char *> Types;
  if (Types.empty()) {
    Types[Context.CharTy] = "%c";
    Types[Context.BoolTy] = "%d";
    Types[Context.SignedCharTy] = "%hhd";
    Types[Context.UnsignedCharTy] = "%hhu";
    Types[Context.IntTy] = "%d";
    Types[Context.UnsignedIntTy] = "%u";
    Types[Context.LongTy] = "%ld";
    Types[Context.UnsignedLongTy] = "%lu";
    Types[Context.LongLongTy] = "%lld";
    Types[Context.UnsignedLongLongTy] = "%llu";
    Types[Context.ShortTy] = "%hd";
    Types[Context.UnsignedShortTy] = "%hu";
    Types[Context.VoidPtrTy] = "%p";
    Types[Context.FloatTy] = "%f";
    Types[Context.DoubleTy] = "%f";
    Types[Context.LongDoubleTy] = "%Lf";
    Types[Context.getPointerType(Context.CharTy)] = "%s";
    Types[Context.getPointerType(Context.getConstType(Context.CharTy))] = "%s";
  }

  for (const auto *FD : RD->fields()) {
    uint64_t Off = RL.getFieldOffset(FD->getFieldIndex());
    Off = Ctx.toCharUnitsFromBits(Off).getQuantity();

    Value *FieldPtr = RecordPtr;
    if (RD->isUnion())
      FieldPtr = CGF.Builder.CreatePointerCast(
          FieldPtr, CGF.ConvertType(Context.getPointerType(FD->getType())));
    else
      FieldPtr = CGF.Builder.CreateStructGEP(CGF.ConvertType(RType), FieldPtr,
                                             FD->getFieldIndex());

    GString = CGF.Builder.CreateGlobalStringPtr(
        llvm::Twine(Pad)
            .concat(FD->getType().getAsString())
            .concat(llvm::Twine(' '))
            .concat(FD->getNameAsString())
            .concat(" : ")
            .str());
    Value *TmpRes = CGF.Builder.CreateCall(Func, {GString});
    Res = CGF.Builder.CreateAdd(Res, TmpRes);

    QualType CanonicalType =
        FD->getType().getUnqualifiedType().getCanonicalType();

    // We check whether we are in a recursive type
    if (CanonicalType->isRecordType()) {
      Value *TmpRes =
          dumpRecord(CGF, CanonicalType, FieldPtr, Align, Func, Lvl + 1);
      Res = CGF.Builder.CreateAdd(TmpRes, Res);
      continue;
    }

    // We try to determine the best format to print the current field
    llvm::Twine Format = Types.find(CanonicalType) == Types.end()
                             ? Types[Context.VoidPtrTy]
                             : Types[CanonicalType];

    Address FieldAddress = Address(FieldPtr, Align);
    FieldPtr = CGF.Builder.CreateLoad(FieldAddress);

    // FIXME Need to handle bitfield here
    GString = CGF.Builder.CreateGlobalStringPtr(
        Format.concat(llvm::Twine('\n')).str());
    TmpRes = CGF.Builder.CreateCall(Func, {GString, FieldPtr});
    Res = CGF.Builder.CreateAdd(Res, TmpRes);
  }

  GString = CGF.Builder.CreateGlobalStringPtr(Pad + "}\n");
  Value *TmpRes = CGF.Builder.CreateCall(Func, {GString});
  Res = CGF.Builder.CreateAdd(Res, TmpRes);
  return Res;
}

RValue CodeGenFunction::EmitBuiltinExpr(const FunctionDecl *FD,
                                        unsigned BuiltinID, const CallExpr *E,
                                        ReturnValueSlot ReturnValue) {
  // See if we can constant fold this builtin.  If so, don't emit it at all.
  Expr::EvalResult Result;
  if (E->EvaluateAsRValue(Result, CGM.getContext()) &&
      !Result.hasSideEffects()) {
    if (Result.Val.isInt())
      return RValue::get(llvm::ConstantInt::get(getLLVMContext(),
                                                Result.Val.getInt()));
    if (Result.Val.isFloat())
      return RValue::get(llvm::ConstantFP::get(getLLVMContext(),
                                               Result.Val.getFloat()));
  }

  // There are LLVM math intrinsics/instructions corresponding to math library
  // functions except the LLVM op will never set errno while the math library
  // might. Also, math builtins have the same semantics as their math library
  // twins. Thus, we can transform math library and builtin calls to their
  // LLVM counterparts if the call is marked 'const' (known to never set errno).
  if (FD->hasAttr<ConstAttr>()) {
    switch (BuiltinID) {
    case Builtin::BIceil:
    case Builtin::BIceilf:
    case Builtin::BIceill:
    case Builtin::BI__builtin_ceil:
    case Builtin::BI__builtin_ceilf:
    case Builtin::BI__builtin_ceill:
      return RValue::get(emitUnaryBuiltin(*this, E, Intrinsic::ceil));

    case Builtin::BIcopysign:
    case Builtin::BIcopysignf:
    case Builtin::BIcopysignl:
    case Builtin::BI__builtin_copysign:
    case Builtin::BI__builtin_copysignf:
    case Builtin::BI__builtin_copysignl:
    case Builtin::BI__builtin_copysignf128:
      return RValue::get(emitBinaryBuiltin(*this, E, Intrinsic::copysign));

    case Builtin::BIcos:
    case Builtin::BIcosf:
    case Builtin::BIcosl:
    case Builtin::BI__builtin_cos:
    case Builtin::BI__builtin_cosf:
    case Builtin::BI__builtin_cosl:
      return RValue::get(emitUnaryBuiltin(*this, E, Intrinsic::cos));

    case Builtin::BIexp:
    case Builtin::BIexpf:
    case Builtin::BIexpl:
    case Builtin::BI__builtin_exp:
    case Builtin::BI__builtin_expf:
    case Builtin::BI__builtin_expl:
      return RValue::get(emitUnaryBuiltin(*this, E, Intrinsic::exp));

    case Builtin::BIexp2:
    case Builtin::BIexp2f:
    case Builtin::BIexp2l:
    case Builtin::BI__builtin_exp2:
    case Builtin::BI__builtin_exp2f:
    case Builtin::BI__builtin_exp2l:
      return RValue::get(emitUnaryBuiltin(*this, E, Intrinsic::exp2));

    case Builtin::BIfabs:
    case Builtin::BIfabsf:
    case Builtin::BIfabsl:
    case Builtin::BI__builtin_fabs:
    case Builtin::BI__builtin_fabsf:
    case Builtin::BI__builtin_fabsl:
    case Builtin::BI__builtin_fabsf128:
      return RValue::get(emitUnaryBuiltin(*this, E, Intrinsic::fabs));

    case Builtin::BIfloor:
    case Builtin::BIfloorf:
    case Builtin::BIfloorl:
    case Builtin::BI__builtin_floor:
    case Builtin::BI__builtin_floorf:
    case Builtin::BI__builtin_floorl:
      return RValue::get(emitUnaryBuiltin(*this, E, Intrinsic::floor));

    case Builtin::BIfma:
    case Builtin::BIfmaf:
    case Builtin::BIfmal:
    case Builtin::BI__builtin_fma:
    case Builtin::BI__builtin_fmaf:
    case Builtin::BI__builtin_fmal:
      return RValue::get(emitTernaryBuiltin(*this, E, Intrinsic::fma));

    case Builtin::BIfmax:
    case Builtin::BIfmaxf:
    case Builtin::BIfmaxl:
    case Builtin::BI__builtin_fmax:
    case Builtin::BI__builtin_fmaxf:
    case Builtin::BI__builtin_fmaxl:
      return RValue::get(emitBinaryBuiltin(*this, E, Intrinsic::maxnum));

    case Builtin::BIfmin:
    case Builtin::BIfminf:
    case Builtin::BIfminl:
    case Builtin::BI__builtin_fmin:
    case Builtin::BI__builtin_fminf:
    case Builtin::BI__builtin_fminl:
      return RValue::get(emitBinaryBuiltin(*this, E, Intrinsic::minnum));

    // fmod() is a special-case. It maps to the frem instruction rather than an
    // LLVM intrinsic.
    case Builtin::BIfmod:
    case Builtin::BIfmodf:
    case Builtin::BIfmodl:
    case Builtin::BI__builtin_fmod:
    case Builtin::BI__builtin_fmodf:
    case Builtin::BI__builtin_fmodl: {
      Value *Arg1 = EmitScalarExpr(E->getArg(0));
      Value *Arg2 = EmitScalarExpr(E->getArg(1));
      return RValue::get(Builder.CreateFRem(Arg1, Arg2, "fmod"));
    }

    case Builtin::BIlog:
    case Builtin::BIlogf:
    case Builtin::BIlogl:
    case Builtin::BI__builtin_log:
    case Builtin::BI__builtin_logf:
    case Builtin::BI__builtin_logl:
      return RValue::get(emitUnaryBuiltin(*this, E, Intrinsic::log));

    case Builtin::BIlog10:
    case Builtin::BIlog10f:
    case Builtin::BIlog10l:
    case Builtin::BI__builtin_log10:
    case Builtin::BI__builtin_log10f:
    case Builtin::BI__builtin_log10l:
      return RValue::get(emitUnaryBuiltin(*this, E, Intrinsic::log10));

    case Builtin::BIlog2:
    case Builtin::BIlog2f:
    case Builtin::BIlog2l:
    case Builtin::BI__builtin_log2:
    case Builtin::BI__builtin_log2f:
    case Builtin::BI__builtin_log2l:
      return RValue::get(emitUnaryBuiltin(*this, E, Intrinsic::log2));

    case Builtin::BInearbyint:
    case Builtin::BInearbyintf:
    case Builtin::BInearbyintl:
    case Builtin::BI__builtin_nearbyint:
    case Builtin::BI__builtin_nearbyintf:
    case Builtin::BI__builtin_nearbyintl:
      return RValue::get(emitUnaryBuiltin(*this, E, Intrinsic::nearbyint));

    case Builtin::BIpow:
    case Builtin::BIpowf:
    case Builtin::BIpowl:
    case Builtin::BI__builtin_pow:
    case Builtin::BI__builtin_powf:
    case Builtin::BI__builtin_powl:
      return RValue::get(emitBinaryBuiltin(*this, E, Intrinsic::pow));

    case Builtin::BIrint:
    case Builtin::BIrintf:
    case Builtin::BIrintl:
    case Builtin::BI__builtin_rint:
    case Builtin::BI__builtin_rintf:
    case Builtin::BI__builtin_rintl:
      return RValue::get(emitUnaryBuiltin(*this, E, Intrinsic::rint));

    case Builtin::BIround:
    case Builtin::BIroundf:
    case Builtin::BIroundl:
    case Builtin::BI__builtin_round:
    case Builtin::BI__builtin_roundf:
    case Builtin::BI__builtin_roundl:
      return RValue::get(emitUnaryBuiltin(*this, E, Intrinsic::round));

    case Builtin::BIsin:
    case Builtin::BIsinf:
    case Builtin::BIsinl:
    case Builtin::BI__builtin_sin:
    case Builtin::BI__builtin_sinf:
    case Builtin::BI__builtin_sinl:
      return RValue::get(emitUnaryBuiltin(*this, E, Intrinsic::sin));

    case Builtin::BIsqrt:
    case Builtin::BIsqrtf:
    case Builtin::BIsqrtl:
    case Builtin::BI__builtin_sqrt:
    case Builtin::BI__builtin_sqrtf:
    case Builtin::BI__builtin_sqrtl:
      return RValue::get(emitUnaryBuiltin(*this, E, Intrinsic::sqrt));

    case Builtin::BItrunc:
    case Builtin::BItruncf:
    case Builtin::BItruncl:
    case Builtin::BI__builtin_trunc:
    case Builtin::BI__builtin_truncf:
    case Builtin::BI__builtin_truncl:
      return RValue::get(emitUnaryBuiltin(*this, E, Intrinsic::trunc));

    default:
      break;
    }
  }

  switch (BuiltinID) {
  default: break;
  case Builtin::BI__builtin___CFStringMakeConstantString:
  case Builtin::BI__builtin___NSStringMakeConstantString:
    return RValue::get(ConstantEmitter(*this).emitAbstract(E, E->getType()));
  case Builtin::BI__builtin_stdarg_start:
  case Builtin::BI__builtin_va_start:
  case Builtin::BI__va_start:
  case Builtin::BI__builtin_va_end:
    return RValue::get(
        EmitVAStartEnd(BuiltinID == Builtin::BI__va_start
                           ? EmitScalarExpr(E->getArg(0))
                           : EmitVAListRef(E->getArg(0)).getPointer(),
                       BuiltinID != Builtin::BI__builtin_va_end));
  case Builtin::BI__builtin_va_copy: {
    Value *DstPtr = EmitVAListRef(E->getArg(0)).getPointer();
    Value *SrcPtr = EmitVAListRef(E->getArg(1)).getPointer();

    llvm::Type *Type = Int8PtrTy;

    DstPtr = Builder.CreateBitCast(DstPtr, Type);
    SrcPtr = Builder.CreateBitCast(SrcPtr, Type);
    return RValue::get(Builder.CreateCall(CGM.getIntrinsic(Intrinsic::vacopy),
                                          {DstPtr, SrcPtr}));
  }
  case Builtin::BI__builtin_abs:
  case Builtin::BI__builtin_labs:
  case Builtin::BI__builtin_llabs: {
    // X < 0 ? -X : X
    // The negation has 'nsw' because abs of INT_MIN is undefined.
    Value *ArgValue = EmitScalarExpr(E->getArg(0));
    Value *NegOp = Builder.CreateNSWNeg(ArgValue, "neg");
    Constant *Zero = llvm::Constant::getNullValue(ArgValue->getType());
    Value *CmpResult = Builder.CreateICmpSLT(ArgValue, Zero, "abscond");
    Value *Result = Builder.CreateSelect(CmpResult, NegOp, ArgValue, "abs");
    return RValue::get(Result);
  }
  case Builtin::BI__builtin_conj:
  case Builtin::BI__builtin_conjf:
  case Builtin::BI__builtin_conjl: {
    ComplexPairTy ComplexVal = EmitComplexExpr(E->getArg(0));
    Value *Real = ComplexVal.first;
    Value *Imag = ComplexVal.second;
    Value *Zero =
      Imag->getType()->isFPOrFPVectorTy()
        ? llvm::ConstantFP::getZeroValueForNegation(Imag->getType())
        : llvm::Constant::getNullValue(Imag->getType());

    Imag = Builder.CreateFSub(Zero, Imag, "sub");
    return RValue::getComplex(std::make_pair(Real, Imag));
  }
  case Builtin::BI__builtin_creal:
  case Builtin::BI__builtin_crealf:
  case Builtin::BI__builtin_creall:
  case Builtin::BIcreal:
  case Builtin::BIcrealf:
  case Builtin::BIcreall: {
    ComplexPairTy ComplexVal = EmitComplexExpr(E->getArg(0));
    return RValue::get(ComplexVal.first);
  }

  case Builtin::BI__builtin_dump_struct: {
    Value *Func = EmitScalarExpr(E->getArg(1)->IgnoreImpCasts());
    CharUnits Arg0Align = EmitPointerWithAlignment(E->getArg(0)).getAlignment();

    const Expr *Arg0 = E->getArg(0)->IgnoreImpCasts();
    QualType Arg0Type = Arg0->getType()->getPointeeType();

    Value *RecordPtr = EmitScalarExpr(Arg0);
    Value *Res = dumpRecord(*this, Arg0Type, RecordPtr, Arg0Align, Func, 0);
    return RValue::get(Res);
  }

  case Builtin::BI__builtin_cimag:
  case Builtin::BI__builtin_cimagf:
  case Builtin::BI__builtin_cimagl:
  case Builtin::BIcimag:
  case Builtin::BIcimagf:
  case Builtin::BIcimagl: {
    ComplexPairTy ComplexVal = EmitComplexExpr(E->getArg(0));
    return RValue::get(ComplexVal.second);
  }

  case Builtin::BI__builtin_clrsb:
  case Builtin::BI__builtin_clrsbl:
  case Builtin::BI__builtin_clrsbll: {
    // clrsb(x) -> clz(x < 0 ? ~x : x) - 1 or
    Value *ArgValue = EmitScalarExpr(E->getArg(0));

    llvm::Type *ArgType = ArgValue->getType();
    Value *F = CGM.getIntrinsic(Intrinsic::ctlz, ArgType);

    llvm::Type *ResultType = ConvertType(E->getType());
    Value *Zero = llvm::Constant::getNullValue(ArgType);
    Value *IsNeg = Builder.CreateICmpSLT(ArgValue, Zero, "isneg");
    Value *Inverse = Builder.CreateNot(ArgValue, "not");
    Value *Tmp = Builder.CreateSelect(IsNeg, Inverse, ArgValue);
    Value *Ctlz = Builder.CreateCall(F, {Tmp, Builder.getFalse()});
    Value *Result = Builder.CreateSub(Ctlz, llvm::ConstantInt::get(ArgType, 1));
    Result = Builder.CreateIntCast(Result, ResultType, /*isSigned*/true,
                                   "cast");
    return RValue::get(Result);
  }
  case Builtin::BI__builtin_ctzs:
  case Builtin::BI__builtin_ctz:
  case Builtin::BI__builtin_ctzl:
  case Builtin::BI__builtin_ctzll: {
    Value *ArgValue = EmitCheckedArgForBuiltin(E->getArg(0), BCK_CTZPassedZero);

    llvm::Type *ArgType = ArgValue->getType();
    Value *F = CGM.getIntrinsic(Intrinsic::cttz, ArgType);

    llvm::Type *ResultType = ConvertType(E->getType());
    Value *ZeroUndef = Builder.getInt1(getTarget().isCLZForZeroUndef());
    Value *Result = Builder.CreateCall(F, {ArgValue, ZeroUndef});
    if (Result->getType() != ResultType)
      Result = Builder.CreateIntCast(Result, ResultType, /*isSigned*/true,
                                     "cast");
    return RValue::get(Result);
  }
  case Builtin::BI__builtin_clzs:
  case Builtin::BI__builtin_clz:
  case Builtin::BI__builtin_clzl:
  case Builtin::BI__builtin_clzll: {
    Value *ArgValue = EmitCheckedArgForBuiltin(E->getArg(0), BCK_CLZPassedZero);

    llvm::Type *ArgType = ArgValue->getType();
    Value *F = CGM.getIntrinsic(Intrinsic::ctlz, ArgType);

    llvm::Type *ResultType = ConvertType(E->getType());
    Value *ZeroUndef = Builder.getInt1(getTarget().isCLZForZeroUndef());
    Value *Result = Builder.CreateCall(F, {ArgValue, ZeroUndef});
    if (Result->getType() != ResultType)
      Result = Builder.CreateIntCast(Result, ResultType, /*isSigned*/true,
                                     "cast");
    return RValue::get(Result);
  }
  case Builtin::BI__builtin_ffs:
  case Builtin::BI__builtin_ffsl:
  case Builtin::BI__builtin_ffsll: {
    // ffs(x) -> x ? cttz(x) + 1 : 0
    Value *ArgValue = EmitScalarExpr(E->getArg(0));

    llvm::Type *ArgType = ArgValue->getType();
    Value *F = CGM.getIntrinsic(Intrinsic::cttz, ArgType);

    llvm::Type *ResultType = ConvertType(E->getType());
    Value *Tmp =
        Builder.CreateAdd(Builder.CreateCall(F, {ArgValue, Builder.getTrue()}),
                          llvm::ConstantInt::get(ArgType, 1));
    Value *Zero = llvm::Constant::getNullValue(ArgType);
    Value *IsZero = Builder.CreateICmpEQ(ArgValue, Zero, "iszero");
    Value *Result = Builder.CreateSelect(IsZero, Zero, Tmp, "ffs");
    if (Result->getType() != ResultType)
      Result = Builder.CreateIntCast(Result, ResultType, /*isSigned*/true,
                                     "cast");
    return RValue::get(Result);
  }
  case Builtin::BI__builtin_parity:
  case Builtin::BI__builtin_parityl:
  case Builtin::BI__builtin_parityll: {
    // parity(x) -> ctpop(x) & 1
    Value *ArgValue = EmitScalarExpr(E->getArg(0));

    llvm::Type *ArgType = ArgValue->getType();
    Value *F = CGM.getIntrinsic(Intrinsic::ctpop, ArgType);

    llvm::Type *ResultType = ConvertType(E->getType());
    Value *Tmp = Builder.CreateCall(F, ArgValue);
    Value *Result = Builder.CreateAnd(Tmp, llvm::ConstantInt::get(ArgType, 1));
    if (Result->getType() != ResultType)
      Result = Builder.CreateIntCast(Result, ResultType, /*isSigned*/true,
                                     "cast");
    return RValue::get(Result);
  }
  case Builtin::BI__popcnt16:
  case Builtin::BI__popcnt:
  case Builtin::BI__popcnt64:
  case Builtin::BI__builtin_popcount:
  case Builtin::BI__builtin_popcountl:
  case Builtin::BI__builtin_popcountll: {
    Value *ArgValue = EmitScalarExpr(E->getArg(0));

    llvm::Type *ArgType = ArgValue->getType();
    Value *F = CGM.getIntrinsic(Intrinsic::ctpop, ArgType);

    llvm::Type *ResultType = ConvertType(E->getType());
    Value *Result = Builder.CreateCall(F, ArgValue);
    if (Result->getType() != ResultType)
      Result = Builder.CreateIntCast(Result, ResultType, /*isSigned*/true,
                                     "cast");
    return RValue::get(Result);
  }
  case Builtin::BI_rotr8:
  case Builtin::BI_rotr16:
  case Builtin::BI_rotr:
  case Builtin::BI_lrotr:
  case Builtin::BI_rotr64: {
    Value *Val = EmitScalarExpr(E->getArg(0));
    Value *Shift = EmitScalarExpr(E->getArg(1));

    llvm::Type *ArgType = Val->getType();
    Shift = Builder.CreateIntCast(Shift, ArgType, false);
    unsigned ArgWidth = ArgType->getIntegerBitWidth();
    Value *Mask = llvm::ConstantInt::get(ArgType, ArgWidth - 1);

    Value *RightShiftAmt = Builder.CreateAnd(Shift, Mask);
    Value *RightShifted = Builder.CreateLShr(Val, RightShiftAmt);
    Value *LeftShiftAmt = Builder.CreateAnd(Builder.CreateNeg(Shift), Mask);
    Value *LeftShifted = Builder.CreateShl(Val, LeftShiftAmt);
    Value *Result = Builder.CreateOr(LeftShifted, RightShifted);
    return RValue::get(Result);
  }
  case Builtin::BI_rotl8:
  case Builtin::BI_rotl16:
  case Builtin::BI_rotl:
  case Builtin::BI_lrotl:
  case Builtin::BI_rotl64: {
    Value *Val = EmitScalarExpr(E->getArg(0));
    Value *Shift = EmitScalarExpr(E->getArg(1));

    llvm::Type *ArgType = Val->getType();
    Shift = Builder.CreateIntCast(Shift, ArgType, false);
    unsigned ArgWidth = ArgType->getIntegerBitWidth();
    Value *Mask = llvm::ConstantInt::get(ArgType, ArgWidth - 1);

    Value *LeftShiftAmt = Builder.CreateAnd(Shift, Mask);
    Value *LeftShifted = Builder.CreateShl(Val, LeftShiftAmt);
    Value *RightShiftAmt = Builder.CreateAnd(Builder.CreateNeg(Shift), Mask);
    Value *RightShifted = Builder.CreateLShr(Val, RightShiftAmt);
    Value *Result = Builder.CreateOr(LeftShifted, RightShifted);
    return RValue::get(Result);
  }
  case Builtin::BI__builtin_unpredictable: {
    // Always return the argument of __builtin_unpredictable. LLVM does not
    // handle this builtin. Metadata for this builtin should be added directly
    // to instructions such as branches or switches that use it.
    return RValue::get(EmitScalarExpr(E->getArg(0)));
  }
  case Builtin::BI__builtin_expect: {
    Value *ArgValue = EmitScalarExpr(E->getArg(0));
    llvm::Type *ArgType = ArgValue->getType();

    Value *ExpectedValue = EmitScalarExpr(E->getArg(1));
    // Don't generate llvm.expect on -O0 as the backend won't use it for
    // anything.
    // Note, we still IRGen ExpectedValue because it could have side-effects.
    if (CGM.getCodeGenOpts().OptimizationLevel == 0)
      return RValue::get(ArgValue);

    Value *FnExpect = CGM.getIntrinsic(Intrinsic::expect, ArgType);
    Value *Result =
        Builder.CreateCall(FnExpect, {ArgValue, ExpectedValue}, "expval");
    return RValue::get(Result);
  }
  case Builtin::BI__builtin_assume_aligned: {
    Value *PtrValue = EmitScalarExpr(E->getArg(0));
    Value *OffsetValue =
      (E->getNumArgs() > 2) ? EmitScalarExpr(E->getArg(2)) : nullptr;

    Value *AlignmentValue = EmitScalarExpr(E->getArg(1));
    ConstantInt *AlignmentCI = cast<ConstantInt>(AlignmentValue);
    unsigned Alignment = (unsigned) AlignmentCI->getZExtValue();

    EmitAlignmentAssumption(PtrValue, Alignment, OffsetValue);
    return RValue::get(PtrValue);
  }
  case Builtin::BI__assume:
  case Builtin::BI__builtin_assume: {
    if (E->getArg(0)->HasSideEffects(getContext()))
      return RValue::get(nullptr);

    Value *ArgValue = EmitScalarExpr(E->getArg(0));
    Value *FnAssume = CGM.getIntrinsic(Intrinsic::assume);
    return RValue::get(Builder.CreateCall(FnAssume, ArgValue));
  }
  case Builtin::BI__builtin_bswap16:
  case Builtin::BI__builtin_bswap32:
  case Builtin::BI__builtin_bswap64: {
    return RValue::get(emitUnaryBuiltin(*this, E, Intrinsic::bswap));
  }
  case Builtin::BI__builtin_bitreverse8:
  case Builtin::BI__builtin_bitreverse16:
  case Builtin::BI__builtin_bitreverse32:
  case Builtin::BI__builtin_bitreverse64: {
    return RValue::get(emitUnaryBuiltin(*this, E, Intrinsic::bitreverse));
  }
  case Builtin::BI__builtin_object_size: {
    unsigned Type =
        E->getArg(1)->EvaluateKnownConstInt(getContext()).getZExtValue();
    auto *ResType = cast<llvm::IntegerType>(ConvertType(E->getType()));

    // We pass this builtin onto the optimizer so that it can figure out the
    // object size in more complex cases.
    return RValue::get(emitBuiltinObjectSize(E->getArg(0), Type, ResType,
                                             /*EmittedE=*/nullptr));
  }
  case Builtin::BI__builtin_prefetch: {
    Value *Locality, *RW, *Address = EmitScalarExpr(E->getArg(0));
    // FIXME: Technically these constants should of type 'int', yes?
    RW = (E->getNumArgs() > 1) ? EmitScalarExpr(E->getArg(1)) :
      llvm::ConstantInt::get(Int32Ty, 0);
    Locality = (E->getNumArgs() > 2) ? EmitScalarExpr(E->getArg(2)) :
      llvm::ConstantInt::get(Int32Ty, 3);
    Value *Data = llvm::ConstantInt::get(Int32Ty, 1);
    Value *F = CGM.getIntrinsic(Intrinsic::prefetch);
    return RValue::get(Builder.CreateCall(F, {Address, RW, Locality, Data}));
  }
  case Builtin::BI__builtin_readcyclecounter: {
    Value *F = CGM.getIntrinsic(Intrinsic::readcyclecounter);
    return RValue::get(Builder.CreateCall(F));
  }
  case Builtin::BI__builtin___clear_cache: {
    Value *Begin = EmitScalarExpr(E->getArg(0));
    Value *End = EmitScalarExpr(E->getArg(1));
    Value *F = CGM.getIntrinsic(Intrinsic::clear_cache);
    return RValue::get(Builder.CreateCall(F, {Begin, End}));
  }
  case Builtin::BI__builtin_trap:
    return RValue::get(EmitTrapCall(Intrinsic::trap));
  case Builtin::BI__debugbreak:
    return RValue::get(EmitTrapCall(Intrinsic::debugtrap));
  case Builtin::BI__builtin_unreachable: {
    EmitUnreachable(E->getExprLoc());

    // We do need to preserve an insertion point.
    EmitBlock(createBasicBlock("unreachable.cont"));

    return RValue::get(nullptr);
  }

  case Builtin::BI__builtin_powi:
  case Builtin::BI__builtin_powif:
  case Builtin::BI__builtin_powil: {
    Value *Base = EmitScalarExpr(E->getArg(0));
    Value *Exponent = EmitScalarExpr(E->getArg(1));
    llvm::Type *ArgType = Base->getType();
    Value *F = CGM.getIntrinsic(Intrinsic::powi, ArgType);
    return RValue::get(Builder.CreateCall(F, {Base, Exponent}));
  }

  case Builtin::BI__builtin_isgreater:
  case Builtin::BI__builtin_isgreaterequal:
  case Builtin::BI__builtin_isless:
  case Builtin::BI__builtin_islessequal:
  case Builtin::BI__builtin_islessgreater:
  case Builtin::BI__builtin_isunordered: {
    // Ordered comparisons: we know the arguments to these are matching scalar
    // floating point values.
    Value *LHS = EmitScalarExpr(E->getArg(0));
    Value *RHS = EmitScalarExpr(E->getArg(1));

    switch (BuiltinID) {
    default: llvm_unreachable("Unknown ordered comparison");
    case Builtin::BI__builtin_isgreater:
      LHS = Builder.CreateFCmpOGT(LHS, RHS, "cmp");
      break;
    case Builtin::BI__builtin_isgreaterequal:
      LHS = Builder.CreateFCmpOGE(LHS, RHS, "cmp");
      break;
    case Builtin::BI__builtin_isless:
      LHS = Builder.CreateFCmpOLT(LHS, RHS, "cmp");
      break;
    case Builtin::BI__builtin_islessequal:
      LHS = Builder.CreateFCmpOLE(LHS, RHS, "cmp");
      break;
    case Builtin::BI__builtin_islessgreater:
      LHS = Builder.CreateFCmpONE(LHS, RHS, "cmp");
      break;
    case Builtin::BI__builtin_isunordered:
      LHS = Builder.CreateFCmpUNO(LHS, RHS, "cmp");
      break;
    }
    // ZExt bool to int type.
    return RValue::get(Builder.CreateZExt(LHS, ConvertType(E->getType())));
  }
  case Builtin::BI__builtin_isnan: {
    Value *V = EmitScalarExpr(E->getArg(0));
    V = Builder.CreateFCmpUNO(V, V, "cmp");
    return RValue::get(Builder.CreateZExt(V, ConvertType(E->getType())));
  }

  case Builtin::BIfinite:
  case Builtin::BI__finite:
  case Builtin::BIfinitef:
  case Builtin::BI__finitef:
  case Builtin::BIfinitel:
  case Builtin::BI__finitel:
  case Builtin::BI__builtin_isinf:
  case Builtin::BI__builtin_isfinite: {
    // isinf(x)    --> fabs(x) == infinity
    // isfinite(x) --> fabs(x) != infinity
    // x != NaN via the ordered compare in either case.
    Value *V = EmitScalarExpr(E->getArg(0));
    Value *Fabs = EmitFAbs(*this, V);
    Constant *Infinity = ConstantFP::getInfinity(V->getType());
    CmpInst::Predicate Pred = (BuiltinID == Builtin::BI__builtin_isinf)
                                  ? CmpInst::FCMP_OEQ
                                  : CmpInst::FCMP_ONE;
    Value *FCmp = Builder.CreateFCmp(Pred, Fabs, Infinity, "cmpinf");
    return RValue::get(Builder.CreateZExt(FCmp, ConvertType(E->getType())));
  }

  case Builtin::BI__builtin_isinf_sign: {
    // isinf_sign(x) -> fabs(x) == infinity ? (signbit(x) ? -1 : 1) : 0
    Value *Arg = EmitScalarExpr(E->getArg(0));
    Value *AbsArg = EmitFAbs(*this, Arg);
    Value *IsInf = Builder.CreateFCmpOEQ(
        AbsArg, ConstantFP::getInfinity(Arg->getType()), "isinf");
    Value *IsNeg = EmitSignBit(*this, Arg);

    llvm::Type *IntTy = ConvertType(E->getType());
    Value *Zero = Constant::getNullValue(IntTy);
    Value *One = ConstantInt::get(IntTy, 1);
    Value *NegativeOne = ConstantInt::get(IntTy, -1);
    Value *SignResult = Builder.CreateSelect(IsNeg, NegativeOne, One);
    Value *Result = Builder.CreateSelect(IsInf, SignResult, Zero);
    return RValue::get(Result);
  }

  case Builtin::BI__builtin_isnormal: {
    // isnormal(x) --> x == x && fabsf(x) < infinity && fabsf(x) >= float_min
    Value *V = EmitScalarExpr(E->getArg(0));
    Value *Eq = Builder.CreateFCmpOEQ(V, V, "iseq");

    Value *Abs = EmitFAbs(*this, V);
    Value *IsLessThanInf =
      Builder.CreateFCmpULT(Abs, ConstantFP::getInfinity(V->getType()),"isinf");
    APFloat Smallest = APFloat::getSmallestNormalized(
                   getContext().getFloatTypeSemantics(E->getArg(0)->getType()));
    Value *IsNormal =
      Builder.CreateFCmpUGE(Abs, ConstantFP::get(V->getContext(), Smallest),
                            "isnormal");
    V = Builder.CreateAnd(Eq, IsLessThanInf, "and");
    V = Builder.CreateAnd(V, IsNormal, "and");
    return RValue::get(Builder.CreateZExt(V, ConvertType(E->getType())));
  }

  case Builtin::BI__builtin_fpclassify: {
    Value *V = EmitScalarExpr(E->getArg(5));
    llvm::Type *Ty = ConvertType(E->getArg(5)->getType());

    // Create Result
    BasicBlock *Begin = Builder.GetInsertBlock();
    BasicBlock *End = createBasicBlock("fpclassify_end", this->CurFn);
    Builder.SetInsertPoint(End);
    PHINode *Result =
      Builder.CreatePHI(ConvertType(E->getArg(0)->getType()), 4,
                        "fpclassify_result");

    // if (V==0) return FP_ZERO
    Builder.SetInsertPoint(Begin);
    Value *IsZero = Builder.CreateFCmpOEQ(V, Constant::getNullValue(Ty),
                                          "iszero");
    Value *ZeroLiteral = EmitScalarExpr(E->getArg(4));
    BasicBlock *NotZero = createBasicBlock("fpclassify_not_zero", this->CurFn);
    Builder.CreateCondBr(IsZero, End, NotZero);
    Result->addIncoming(ZeroLiteral, Begin);

    // if (V != V) return FP_NAN
    Builder.SetInsertPoint(NotZero);
    Value *IsNan = Builder.CreateFCmpUNO(V, V, "cmp");
    Value *NanLiteral = EmitScalarExpr(E->getArg(0));
    BasicBlock *NotNan = createBasicBlock("fpclassify_not_nan", this->CurFn);
    Builder.CreateCondBr(IsNan, End, NotNan);
    Result->addIncoming(NanLiteral, NotZero);

    // if (fabs(V) == infinity) return FP_INFINITY
    Builder.SetInsertPoint(NotNan);
    Value *VAbs = EmitFAbs(*this, V);
    Value *IsInf =
      Builder.CreateFCmpOEQ(VAbs, ConstantFP::getInfinity(V->getType()),
                            "isinf");
    Value *InfLiteral = EmitScalarExpr(E->getArg(1));
    BasicBlock *NotInf = createBasicBlock("fpclassify_not_inf", this->CurFn);
    Builder.CreateCondBr(IsInf, End, NotInf);
    Result->addIncoming(InfLiteral, NotNan);

    // if (fabs(V) >= MIN_NORMAL) return FP_NORMAL else FP_SUBNORMAL
    Builder.SetInsertPoint(NotInf);
    APFloat Smallest = APFloat::getSmallestNormalized(
        getContext().getFloatTypeSemantics(E->getArg(5)->getType()));
    Value *IsNormal =
      Builder.CreateFCmpUGE(VAbs, ConstantFP::get(V->getContext(), Smallest),
                            "isnormal");
    Value *NormalResult =
      Builder.CreateSelect(IsNormal, EmitScalarExpr(E->getArg(2)),
                           EmitScalarExpr(E->getArg(3)));
    Builder.CreateBr(End);
    Result->addIncoming(NormalResult, NotInf);

    // return Result
    Builder.SetInsertPoint(End);
    return RValue::get(Result);
  }

  case Builtin::BIalloca:
  case Builtin::BI_alloca:
  case Builtin::BI__builtin_alloca: {
    Value *Size = EmitScalarExpr(E->getArg(0));
    const TargetInfo &TI = getContext().getTargetInfo();
    // The alignment of the alloca should correspond to __BIGGEST_ALIGNMENT__.
    unsigned SuitableAlignmentInBytes =
        CGM.getContext()
            .toCharUnitsFromBits(TI.getSuitableAlign())
            .getQuantity();
    AllocaInst *AI = Builder.CreateAlloca(Builder.getInt8Ty(), Size);
    AI->setAlignment(SuitableAlignmentInBytes);
    return RValue::get(AI);
  }

  case Builtin::BI__builtin_alloca_with_align: {
    Value *Size = EmitScalarExpr(E->getArg(0));
    Value *AlignmentInBitsValue = EmitScalarExpr(E->getArg(1));
    auto *AlignmentInBitsCI = cast<ConstantInt>(AlignmentInBitsValue);
    unsigned AlignmentInBits = AlignmentInBitsCI->getZExtValue();
    unsigned AlignmentInBytes =
        CGM.getContext().toCharUnitsFromBits(AlignmentInBits).getQuantity();
    AllocaInst *AI = Builder.CreateAlloca(Builder.getInt8Ty(), Size);
    AI->setAlignment(AlignmentInBytes);
    return RValue::get(AI);
  }

  case Builtin::BIbzero:
  case Builtin::BI__builtin_bzero: {
    Address Dest = EmitPointerWithAlignment(E->getArg(0));
    Value *SizeVal = EmitScalarExpr(E->getArg(1));
    EmitNonNullArgCheck(RValue::get(Dest.getPointer()), E->getArg(0)->getType(),
                        E->getArg(0)->getExprLoc(), FD, 0);
    Builder.CreateMemSet(Dest, Builder.getInt8(0), SizeVal, false);
    return RValue::get(nullptr);
  }
  case Builtin::BImemcpy:
  case Builtin::BI__builtin_memcpy: {
    Address Dest = EmitPointerWithAlignment(E->getArg(0));
    Address Src = EmitPointerWithAlignment(E->getArg(1));
    Value *SizeVal = EmitScalarExpr(E->getArg(2));
    EmitNonNullArgCheck(RValue::get(Dest.getPointer()), E->getArg(0)->getType(),
                        E->getArg(0)->getExprLoc(), FD, 0);
    EmitNonNullArgCheck(RValue::get(Src.getPointer()), E->getArg(1)->getType(),
                        E->getArg(1)->getExprLoc(), FD, 1);
    Builder.CreateMemCpy(Dest, Src, SizeVal, false);
    return RValue::get(Dest.getPointer());
  }

  case Builtin::BI__builtin_char_memchr:
    BuiltinID = Builtin::BI__builtin_memchr;
    break;

  case Builtin::BI__builtin___memcpy_chk: {
    // fold __builtin_memcpy_chk(x, y, cst1, cst2) to memcpy iff cst1<=cst2.
    llvm::APSInt Size, DstSize;
    if (!E->getArg(2)->EvaluateAsInt(Size, CGM.getContext()) ||
        !E->getArg(3)->EvaluateAsInt(DstSize, CGM.getContext()))
      break;
    if (Size.ugt(DstSize))
      break;
    Address Dest = EmitPointerWithAlignment(E->getArg(0));
    Address Src = EmitPointerWithAlignment(E->getArg(1));
    Value *SizeVal = llvm::ConstantInt::get(Builder.getContext(), Size);
    Builder.CreateMemCpy(Dest, Src, SizeVal, false);
    return RValue::get(Dest.getPointer());
  }

  case Builtin::BI__builtin_objc_memmove_collectable: {
    Address DestAddr = EmitPointerWithAlignment(E->getArg(0));
    Address SrcAddr = EmitPointerWithAlignment(E->getArg(1));
    Value *SizeVal = EmitScalarExpr(E->getArg(2));
    CGM.getObjCRuntime().EmitGCMemmoveCollectable(*this,
                                                  DestAddr, SrcAddr, SizeVal);
    return RValue::get(DestAddr.getPointer());
  }

  case Builtin::BI__builtin___memmove_chk: {
    // fold __builtin_memmove_chk(x, y, cst1, cst2) to memmove iff cst1<=cst2.
    llvm::APSInt Size, DstSize;
    if (!E->getArg(2)->EvaluateAsInt(Size, CGM.getContext()) ||
        !E->getArg(3)->EvaluateAsInt(DstSize, CGM.getContext()))
      break;
    if (Size.ugt(DstSize))
      break;
    Address Dest = EmitPointerWithAlignment(E->getArg(0));
    Address Src = EmitPointerWithAlignment(E->getArg(1));
    Value *SizeVal = llvm::ConstantInt::get(Builder.getContext(), Size);
    Builder.CreateMemMove(Dest, Src, SizeVal, false);
    return RValue::get(Dest.getPointer());
  }

  case Builtin::BImemmove:
  case Builtin::BI__builtin_memmove: {
    Address Dest = EmitPointerWithAlignment(E->getArg(0));
    Address Src = EmitPointerWithAlignment(E->getArg(1));
    Value *SizeVal = EmitScalarExpr(E->getArg(2));
    EmitNonNullArgCheck(RValue::get(Dest.getPointer()), E->getArg(0)->getType(),
                        E->getArg(0)->getExprLoc(), FD, 0);
    EmitNonNullArgCheck(RValue::get(Src.getPointer()), E->getArg(1)->getType(),
                        E->getArg(1)->getExprLoc(), FD, 1);
    Builder.CreateMemMove(Dest, Src, SizeVal, false);
    return RValue::get(Dest.getPointer());
  }
  case Builtin::BImemset:
  case Builtin::BI__builtin_memset: {
    Address Dest = EmitPointerWithAlignment(E->getArg(0));
    Value *ByteVal = Builder.CreateTrunc(EmitScalarExpr(E->getArg(1)),
                                         Builder.getInt8Ty());
    Value *SizeVal = EmitScalarExpr(E->getArg(2));
    EmitNonNullArgCheck(RValue::get(Dest.getPointer()), E->getArg(0)->getType(),
                        E->getArg(0)->getExprLoc(), FD, 0);
    Builder.CreateMemSet(Dest, ByteVal, SizeVal, false);
    return RValue::get(Dest.getPointer());
  }
  case Builtin::BI__builtin___memset_chk: {
    // fold __builtin_memset_chk(x, y, cst1, cst2) to memset iff cst1<=cst2.
    llvm::APSInt Size, DstSize;
    if (!E->getArg(2)->EvaluateAsInt(Size, CGM.getContext()) ||
        !E->getArg(3)->EvaluateAsInt(DstSize, CGM.getContext()))
      break;
    if (Size.ugt(DstSize))
      break;
    Address Dest = EmitPointerWithAlignment(E->getArg(0));
    Value *ByteVal = Builder.CreateTrunc(EmitScalarExpr(E->getArg(1)),
                                         Builder.getInt8Ty());
    Value *SizeVal = llvm::ConstantInt::get(Builder.getContext(), Size);
    Builder.CreateMemSet(Dest, ByteVal, SizeVal, false);
    return RValue::get(Dest.getPointer());
  }
  case Builtin::BI__builtin_wmemcmp: {
    // The MSVC runtime library does not provide a definition of wmemcmp, so we
    // need an inline implementation.
    if (!getTarget().getTriple().isOSMSVCRT())
      break;

    llvm::Type *WCharTy = ConvertType(getContext().WCharTy);

    Value *Dst = EmitScalarExpr(E->getArg(0));
    Value *Src = EmitScalarExpr(E->getArg(1));
    Value *Size = EmitScalarExpr(E->getArg(2));

    BasicBlock *Entry = Builder.GetInsertBlock();
    BasicBlock *CmpGT = createBasicBlock("wmemcmp.gt");
    BasicBlock *CmpLT = createBasicBlock("wmemcmp.lt");
    BasicBlock *Next = createBasicBlock("wmemcmp.next");
    BasicBlock *Exit = createBasicBlock("wmemcmp.exit");
    Value *SizeEq0 = Builder.CreateICmpEQ(Size, ConstantInt::get(SizeTy, 0));
    Builder.CreateCondBr(SizeEq0, Exit, CmpGT);

    EmitBlock(CmpGT);
    PHINode *DstPhi = Builder.CreatePHI(Dst->getType(), 2);
    DstPhi->addIncoming(Dst, Entry);
    PHINode *SrcPhi = Builder.CreatePHI(Src->getType(), 2);
    SrcPhi->addIncoming(Src, Entry);
    PHINode *SizePhi = Builder.CreatePHI(SizeTy, 2);
    SizePhi->addIncoming(Size, Entry);
    CharUnits WCharAlign =
        getContext().getTypeAlignInChars(getContext().WCharTy);
    Value *DstCh = Builder.CreateAlignedLoad(WCharTy, DstPhi, WCharAlign);
    Value *SrcCh = Builder.CreateAlignedLoad(WCharTy, SrcPhi, WCharAlign);
    Value *DstGtSrc = Builder.CreateICmpUGT(DstCh, SrcCh);
    Builder.CreateCondBr(DstGtSrc, Exit, CmpLT);

    EmitBlock(CmpLT);
    Value *DstLtSrc = Builder.CreateICmpULT(DstCh, SrcCh);
    Builder.CreateCondBr(DstLtSrc, Exit, Next);

    EmitBlock(Next);
    Value *NextDst = Builder.CreateConstInBoundsGEP1_32(WCharTy, DstPhi, 1);
    Value *NextSrc = Builder.CreateConstInBoundsGEP1_32(WCharTy, SrcPhi, 1);
    Value *NextSize = Builder.CreateSub(SizePhi, ConstantInt::get(SizeTy, 1));
    Value *NextSizeEq0 =
        Builder.CreateICmpEQ(NextSize, ConstantInt::get(SizeTy, 0));
    Builder.CreateCondBr(NextSizeEq0, Exit, CmpGT);
    DstPhi->addIncoming(NextDst, Next);
    SrcPhi->addIncoming(NextSrc, Next);
    SizePhi->addIncoming(NextSize, Next);

    EmitBlock(Exit);
    PHINode *Ret = Builder.CreatePHI(IntTy, 4);
    Ret->addIncoming(ConstantInt::get(IntTy, 0), Entry);
    Ret->addIncoming(ConstantInt::get(IntTy, 1), CmpGT);
    Ret->addIncoming(ConstantInt::get(IntTy, -1), CmpLT);
    Ret->addIncoming(ConstantInt::get(IntTy, 0), Next);
    return RValue::get(Ret);
  }
  case Builtin::BI__builtin_dwarf_cfa: {
    // The offset in bytes from the first argument to the CFA.
    //
    // Why on earth is this in the frontend?  Is there any reason at
    // all that the backend can't reasonably determine this while
    // lowering llvm.eh.dwarf.cfa()?
    //
    // TODO: If there's a satisfactory reason, add a target hook for
    // this instead of hard-coding 0, which is correct for most targets.
    int32_t Offset = 0;

    Value *F = CGM.getIntrinsic(Intrinsic::eh_dwarf_cfa);
    return RValue::get(Builder.CreateCall(F,
                                      llvm::ConstantInt::get(Int32Ty, Offset)));
  }
  case Builtin::BI__builtin_return_address: {
    Value *Depth = ConstantEmitter(*this).emitAbstract(E->getArg(0),
                                                   getContext().UnsignedIntTy);
    Value *F = CGM.getIntrinsic(Intrinsic::returnaddress);
    return RValue::get(Builder.CreateCall(F, Depth));
  }
  case Builtin::BI_ReturnAddress: {
    Value *F = CGM.getIntrinsic(Intrinsic::returnaddress);
    return RValue::get(Builder.CreateCall(F, Builder.getInt32(0)));
  }
  case Builtin::BI__builtin_frame_address: {
    Value *Depth = ConstantEmitter(*this).emitAbstract(E->getArg(0),
                                                   getContext().UnsignedIntTy);
    Value *F = CGM.getIntrinsic(Intrinsic::frameaddress);
    return RValue::get(Builder.CreateCall(F, Depth));
  }
  case Builtin::BI__builtin_extract_return_addr: {
    Value *Address = EmitScalarExpr(E->getArg(0));
    Value *Result = getTargetHooks().decodeReturnAddress(*this, Address);
    return RValue::get(Result);
  }
  case Builtin::BI__builtin_frob_return_addr: {
    Value *Address = EmitScalarExpr(E->getArg(0));
    Value *Result = getTargetHooks().encodeReturnAddress(*this, Address);
    return RValue::get(Result);
  }
  case Builtin::BI__builtin_dwarf_sp_column: {
    llvm::IntegerType *Ty
      = cast<llvm::IntegerType>(ConvertType(E->getType()));
    int Column = getTargetHooks().getDwarfEHStackPointer(CGM);
    if (Column == -1) {
      CGM.ErrorUnsupported(E, "__builtin_dwarf_sp_column");
      return RValue::get(llvm::UndefValue::get(Ty));
    }
    return RValue::get(llvm::ConstantInt::get(Ty, Column, true));
  }
  case Builtin::BI__builtin_init_dwarf_reg_size_table: {
    Value *Address = EmitScalarExpr(E->getArg(0));
    if (getTargetHooks().initDwarfEHRegSizeTable(*this, Address))
      CGM.ErrorUnsupported(E, "__builtin_init_dwarf_reg_size_table");
    return RValue::get(llvm::UndefValue::get(ConvertType(E->getType())));
  }
  case Builtin::BI__builtin_eh_return: {
    Value *Int = EmitScalarExpr(E->getArg(0));
    Value *Ptr = EmitScalarExpr(E->getArg(1));

    llvm::IntegerType *IntTy = cast<llvm::IntegerType>(Int->getType());
    assert((IntTy->getBitWidth() == 32 || IntTy->getBitWidth() == 64) &&
           "LLVM's __builtin_eh_return only supports 32- and 64-bit variants");
    Value *F = CGM.getIntrinsic(IntTy->getBitWidth() == 32
                                  ? Intrinsic::eh_return_i32
                                  : Intrinsic::eh_return_i64);
    Builder.CreateCall(F, {Int, Ptr});
    Builder.CreateUnreachable();

    // We do need to preserve an insertion point.
    EmitBlock(createBasicBlock("builtin_eh_return.cont"));

    return RValue::get(nullptr);
  }
  case Builtin::BI__builtin_unwind_init: {
    Value *F = CGM.getIntrinsic(Intrinsic::eh_unwind_init);
    return RValue::get(Builder.CreateCall(F));
  }
  case Builtin::BI__builtin_extend_pointer: {
    // Extends a pointer to the size of an _Unwind_Word, which is
    // uint64_t on all platforms.  Generally this gets poked into a
    // register and eventually used as an address, so if the
    // addressing registers are wider than pointers and the platform
    // doesn't implicitly ignore high-order bits when doing
    // addressing, we need to make sure we zext / sext based on
    // the platform's expectations.
    //
    // See: http://gcc.gnu.org/ml/gcc-bugs/2002-02/msg00237.html

    // Cast the pointer to intptr_t.
    Value *Ptr = EmitScalarExpr(E->getArg(0));
    Value *Result = Builder.CreatePtrToInt(Ptr, IntPtrTy, "extend.cast");

    // If that's 64 bits, we're done.
    if (IntPtrTy->getBitWidth() == 64)
      return RValue::get(Result);

    // Otherwise, ask the codegen data what to do.
    if (getTargetHooks().extendPointerWithSExt())
      return RValue::get(Builder.CreateSExt(Result, Int64Ty, "extend.sext"));
    else
      return RValue::get(Builder.CreateZExt(Result, Int64Ty, "extend.zext"));
  }
  case Builtin::BI__builtin_setjmp: {
    // Buffer is a void**.
    Address Buf = EmitPointerWithAlignment(E->getArg(0));

    // Store the frame pointer to the setjmp buffer.
    Value *FrameAddr =
      Builder.CreateCall(CGM.getIntrinsic(Intrinsic::frameaddress),
                         ConstantInt::get(Int32Ty, 0));
    Builder.CreateStore(FrameAddr, Buf);

    // Store the stack pointer to the setjmp buffer.
    Value *StackAddr =
        Builder.CreateCall(CGM.getIntrinsic(Intrinsic::stacksave));
    Address StackSaveSlot =
      Builder.CreateConstInBoundsGEP(Buf, 2, getPointerSize());
    Builder.CreateStore(StackAddr, StackSaveSlot);

    // Call LLVM's EH setjmp, which is lightweight.
    Value *F = CGM.getIntrinsic(Intrinsic::eh_sjlj_setjmp);
    Buf = Builder.CreateBitCast(Buf, Int8PtrTy);
    return RValue::get(Builder.CreateCall(F, Buf.getPointer()));
  }
  case Builtin::BI__builtin_longjmp: {
    Value *Buf = EmitScalarExpr(E->getArg(0));
    Buf = Builder.CreateBitCast(Buf, Int8PtrTy);

    // Call LLVM's EH longjmp, which is lightweight.
    Builder.CreateCall(CGM.getIntrinsic(Intrinsic::eh_sjlj_longjmp), Buf);

    // longjmp doesn't return; mark this as unreachable.
    Builder.CreateUnreachable();

    // We do need to preserve an insertion point.
    EmitBlock(createBasicBlock("longjmp.cont"));

    return RValue::get(nullptr);
  }
  case Builtin::BI__sync_fetch_and_add:
  case Builtin::BI__sync_fetch_and_sub:
  case Builtin::BI__sync_fetch_and_or:
  case Builtin::BI__sync_fetch_and_and:
  case Builtin::BI__sync_fetch_and_xor:
  case Builtin::BI__sync_fetch_and_nand:
  case Builtin::BI__sync_add_and_fetch:
  case Builtin::BI__sync_sub_and_fetch:
  case Builtin::BI__sync_and_and_fetch:
  case Builtin::BI__sync_or_and_fetch:
  case Builtin::BI__sync_xor_and_fetch:
  case Builtin::BI__sync_nand_and_fetch:
  case Builtin::BI__sync_val_compare_and_swap:
  case Builtin::BI__sync_bool_compare_and_swap:
  case Builtin::BI__sync_lock_test_and_set:
  case Builtin::BI__sync_lock_release:
  case Builtin::BI__sync_swap:
    llvm_unreachable("Shouldn't make it through sema");
  case Builtin::BI__sync_fetch_and_add_1:
  case Builtin::BI__sync_fetch_and_add_2:
  case Builtin::BI__sync_fetch_and_add_4:
  case Builtin::BI__sync_fetch_and_add_8:
  case Builtin::BI__sync_fetch_and_add_16:
    return EmitBinaryAtomic(*this, llvm::AtomicRMWInst::Add, E);
  case Builtin::BI__sync_fetch_and_sub_1:
  case Builtin::BI__sync_fetch_and_sub_2:
  case Builtin::BI__sync_fetch_and_sub_4:
  case Builtin::BI__sync_fetch_and_sub_8:
  case Builtin::BI__sync_fetch_and_sub_16:
    return EmitBinaryAtomic(*this, llvm::AtomicRMWInst::Sub, E);
  case Builtin::BI__sync_fetch_and_or_1:
  case Builtin::BI__sync_fetch_and_or_2:
  case Builtin::BI__sync_fetch_and_or_4:
  case Builtin::BI__sync_fetch_and_or_8:
  case Builtin::BI__sync_fetch_and_or_16:
    return EmitBinaryAtomic(*this, llvm::AtomicRMWInst::Or, E);
  case Builtin::BI__sync_fetch_and_and_1:
  case Builtin::BI__sync_fetch_and_and_2:
  case Builtin::BI__sync_fetch_and_and_4:
  case Builtin::BI__sync_fetch_and_and_8:
  case Builtin::BI__sync_fetch_and_and_16:
    return EmitBinaryAtomic(*this, llvm::AtomicRMWInst::And, E);
  case Builtin::BI__sync_fetch_and_xor_1:
  case Builtin::BI__sync_fetch_and_xor_2:
  case Builtin::BI__sync_fetch_and_xor_4:
  case Builtin::BI__sync_fetch_and_xor_8:
  case Builtin::BI__sync_fetch_and_xor_16:
    return EmitBinaryAtomic(*this, llvm::AtomicRMWInst::Xor, E);
  case Builtin::BI__sync_fetch_and_nand_1:
  case Builtin::BI__sync_fetch_and_nand_2:
  case Builtin::BI__sync_fetch_and_nand_4:
  case Builtin::BI__sync_fetch_and_nand_8:
  case Builtin::BI__sync_fetch_and_nand_16:
    return EmitBinaryAtomic(*this, llvm::AtomicRMWInst::Nand, E);

  // Clang extensions: not overloaded yet.
  case Builtin::BI__sync_fetch_and_min:
    return EmitBinaryAtomic(*this, llvm::AtomicRMWInst::Min, E);
  case Builtin::BI__sync_fetch_and_max:
    return EmitBinaryAtomic(*this, llvm::AtomicRMWInst::Max, E);
  case Builtin::BI__sync_fetch_and_umin:
    return EmitBinaryAtomic(*this, llvm::AtomicRMWInst::UMin, E);
  case Builtin::BI__sync_fetch_and_umax:
    return EmitBinaryAtomic(*this, llvm::AtomicRMWInst::UMax, E);

  case Builtin::BI__sync_add_and_fetch_1:
  case Builtin::BI__sync_add_and_fetch_2:
  case Builtin::BI__sync_add_and_fetch_4:
  case Builtin::BI__sync_add_and_fetch_8:
  case Builtin::BI__sync_add_and_fetch_16:
    return EmitBinaryAtomicPost(*this, llvm::AtomicRMWInst::Add, E,
                                llvm::Instruction::Add);
  case Builtin::BI__sync_sub_and_fetch_1:
  case Builtin::BI__sync_sub_and_fetch_2:
  case Builtin::BI__sync_sub_and_fetch_4:
  case Builtin::BI__sync_sub_and_fetch_8:
  case Builtin::BI__sync_sub_and_fetch_16:
    return EmitBinaryAtomicPost(*this, llvm::AtomicRMWInst::Sub, E,
                                llvm::Instruction::Sub);
  case Builtin::BI__sync_and_and_fetch_1:
  case Builtin::BI__sync_and_and_fetch_2:
  case Builtin::BI__sync_and_and_fetch_4:
  case Builtin::BI__sync_and_and_fetch_8:
  case Builtin::BI__sync_and_and_fetch_16:
    return EmitBinaryAtomicPost(*this, llvm::AtomicRMWInst::And, E,
                                llvm::Instruction::And);
  case Builtin::BI__sync_or_and_fetch_1:
  case Builtin::BI__sync_or_and_fetch_2:
  case Builtin::BI__sync_or_and_fetch_4:
  case Builtin::BI__sync_or_and_fetch_8:
  case Builtin::BI__sync_or_and_fetch_16:
    return EmitBinaryAtomicPost(*this, llvm::AtomicRMWInst::Or, E,
                                llvm::Instruction::Or);
  case Builtin::BI__sync_xor_and_fetch_1:
  case Builtin::BI__sync_xor_and_fetch_2:
  case Builtin::BI__sync_xor_and_fetch_4:
  case Builtin::BI__sync_xor_and_fetch_8:
  case Builtin::BI__sync_xor_and_fetch_16:
    return EmitBinaryAtomicPost(*this, llvm::AtomicRMWInst::Xor, E,
                                llvm::Instruction::Xor);
  case Builtin::BI__sync_nand_and_fetch_1:
  case Builtin::BI__sync_nand_and_fetch_2:
  case Builtin::BI__sync_nand_and_fetch_4:
  case Builtin::BI__sync_nand_and_fetch_8:
  case Builtin::BI__sync_nand_and_fetch_16:
    return EmitBinaryAtomicPost(*this, llvm::AtomicRMWInst::Nand, E,
                                llvm::Instruction::And, true);

  case Builtin::BI__sync_val_compare_and_swap_1:
  case Builtin::BI__sync_val_compare_and_swap_2:
  case Builtin::BI__sync_val_compare_and_swap_4:
  case Builtin::BI__sync_val_compare_and_swap_8:
  case Builtin::BI__sync_val_compare_and_swap_16:
    return RValue::get(MakeAtomicCmpXchgValue(*this, E, false));

  case Builtin::BI__sync_bool_compare_and_swap_1:
  case Builtin::BI__sync_bool_compare_and_swap_2:
  case Builtin::BI__sync_bool_compare_and_swap_4:
  case Builtin::BI__sync_bool_compare_and_swap_8:
  case Builtin::BI__sync_bool_compare_and_swap_16:
    return RValue::get(MakeAtomicCmpXchgValue(*this, E, true));

  case Builtin::BI__sync_swap_1:
  case Builtin::BI__sync_swap_2:
  case Builtin::BI__sync_swap_4:
  case Builtin::BI__sync_swap_8:
  case Builtin::BI__sync_swap_16:
    return EmitBinaryAtomic(*this, llvm::AtomicRMWInst::Xchg, E);

  case Builtin::BI__sync_lock_test_and_set_1:
  case Builtin::BI__sync_lock_test_and_set_2:
  case Builtin::BI__sync_lock_test_and_set_4:
  case Builtin::BI__sync_lock_test_and_set_8:
  case Builtin::BI__sync_lock_test_and_set_16:
    return EmitBinaryAtomic(*this, llvm::AtomicRMWInst::Xchg, E);

  case Builtin::BI__sync_lock_release_1:
  case Builtin::BI__sync_lock_release_2:
  case Builtin::BI__sync_lock_release_4:
  case Builtin::BI__sync_lock_release_8:
  case Builtin::BI__sync_lock_release_16: {
    Value *Ptr = EmitScalarExpr(E->getArg(0));
    QualType ElTy = E->getArg(0)->getType()->getPointeeType();
    CharUnits StoreSize = getContext().getTypeSizeInChars(ElTy);
    llvm::Type *ITy = llvm::IntegerType::get(getLLVMContext(),
                                             StoreSize.getQuantity() * 8);
    Ptr = Builder.CreateBitCast(Ptr, ITy->getPointerTo());
    llvm::StoreInst *Store =
      Builder.CreateAlignedStore(llvm::Constant::getNullValue(ITy), Ptr,
                                 StoreSize);
    Store->setAtomic(llvm::AtomicOrdering::Release);
    return RValue::get(nullptr);
  }

  case Builtin::BI__sync_synchronize: {
    // We assume this is supposed to correspond to a C++0x-style
    // sequentially-consistent fence (i.e. this is only usable for
    // synchronization, not device I/O or anything like that). This intrinsic
    // is really badly designed in the sense that in theory, there isn't
    // any way to safely use it... but in practice, it mostly works
    // to use it with non-atomic loads and stores to get acquire/release
    // semantics.
    Builder.CreateFence(llvm::AtomicOrdering::SequentiallyConsistent);
    return RValue::get(nullptr);
  }

  case Builtin::BI__builtin_nontemporal_load:
    return RValue::get(EmitNontemporalLoad(*this, E));
  case Builtin::BI__builtin_nontemporal_store:
    return RValue::get(EmitNontemporalStore(*this, E));
  case Builtin::BI__c11_atomic_is_lock_free:
  case Builtin::BI__atomic_is_lock_free: {
    // Call "bool __atomic_is_lock_free(size_t size, void *ptr)". For the
    // __c11 builtin, ptr is 0 (indicating a properly-aligned object), since
    // _Atomic(T) is always properly-aligned.
    const char *LibCallName = "__atomic_is_lock_free";
    CallArgList Args;
    Args.add(RValue::get(EmitScalarExpr(E->getArg(0))),
             getContext().getSizeType());
    if (BuiltinID == Builtin::BI__atomic_is_lock_free)
      Args.add(RValue::get(EmitScalarExpr(E->getArg(1))),
               getContext().VoidPtrTy);
    else
      Args.add(RValue::get(llvm::Constant::getNullValue(VoidPtrTy)),
               getContext().VoidPtrTy);
    const CGFunctionInfo &FuncInfo =
        CGM.getTypes().arrangeBuiltinFunctionCall(E->getType(), Args);
    llvm::FunctionType *FTy = CGM.getTypes().GetFunctionType(FuncInfo);
    llvm::Constant *Func = CGM.CreateRuntimeFunction(FTy, LibCallName);
    return EmitCall(FuncInfo, CGCallee::forDirect(Func),
                    ReturnValueSlot(), Args);
  }

  case Builtin::BI__atomic_test_and_set: {
    // Look at the argument type to determine whether this is a volatile
    // operation. The parameter type is always volatile.
    QualType PtrTy = E->getArg(0)->IgnoreImpCasts()->getType();
    bool Volatile =
        PtrTy->castAs<PointerType>()->getPointeeType().isVolatileQualified();

    Value *Ptr = EmitScalarExpr(E->getArg(0));
    unsigned AddrSpace = Ptr->getType()->getPointerAddressSpace();
    Ptr = Builder.CreateBitCast(Ptr, Int8Ty->getPointerTo(AddrSpace));
    Value *NewVal = Builder.getInt8(1);
    Value *Order = EmitScalarExpr(E->getArg(1));
    if (isa<llvm::ConstantInt>(Order)) {
      int ord = cast<llvm::ConstantInt>(Order)->getZExtValue();
      AtomicRMWInst *Result = nullptr;
      switch (ord) {
      case 0:  // memory_order_relaxed
      default: // invalid order
        Result = Builder.CreateAtomicRMW(llvm::AtomicRMWInst::Xchg, Ptr, NewVal,
                                         llvm::AtomicOrdering::Monotonic);
        break;
      case 1: // memory_order_consume
      case 2: // memory_order_acquire
        Result = Builder.CreateAtomicRMW(llvm::AtomicRMWInst::Xchg, Ptr, NewVal,
                                         llvm::AtomicOrdering::Acquire);
        break;
      case 3: // memory_order_release
        Result = Builder.CreateAtomicRMW(llvm::AtomicRMWInst::Xchg, Ptr, NewVal,
                                         llvm::AtomicOrdering::Release);
        break;
      case 4: // memory_order_acq_rel

        Result = Builder.CreateAtomicRMW(llvm::AtomicRMWInst::Xchg, Ptr, NewVal,
                                         llvm::AtomicOrdering::AcquireRelease);
        break;
      case 5: // memory_order_seq_cst
        Result = Builder.CreateAtomicRMW(
            llvm::AtomicRMWInst::Xchg, Ptr, NewVal,
            llvm::AtomicOrdering::SequentiallyConsistent);
        break;
      }
      Result->setVolatile(Volatile);
      return RValue::get(Builder.CreateIsNotNull(Result, "tobool"));
    }

    llvm::BasicBlock *ContBB = createBasicBlock("atomic.continue", CurFn);

    llvm::BasicBlock *BBs[5] = {
      createBasicBlock("monotonic", CurFn),
      createBasicBlock("acquire", CurFn),
      createBasicBlock("release", CurFn),
      createBasicBlock("acqrel", CurFn),
      createBasicBlock("seqcst", CurFn)
    };
    llvm::AtomicOrdering Orders[5] = {
        llvm::AtomicOrdering::Monotonic, llvm::AtomicOrdering::Acquire,
        llvm::AtomicOrdering::Release, llvm::AtomicOrdering::AcquireRelease,
        llvm::AtomicOrdering::SequentiallyConsistent};

    Order = Builder.CreateIntCast(Order, Builder.getInt32Ty(), false);
    llvm::SwitchInst *SI = Builder.CreateSwitch(Order, BBs[0]);

    Builder.SetInsertPoint(ContBB);
    PHINode *Result = Builder.CreatePHI(Int8Ty, 5, "was_set");

    for (unsigned i = 0; i < 5; ++i) {
      Builder.SetInsertPoint(BBs[i]);
      AtomicRMWInst *RMW = Builder.CreateAtomicRMW(llvm::AtomicRMWInst::Xchg,
                                                   Ptr, NewVal, Orders[i]);
      RMW->setVolatile(Volatile);
      Result->addIncoming(RMW, BBs[i]);
      Builder.CreateBr(ContBB);
    }

    SI->addCase(Builder.getInt32(0), BBs[0]);
    SI->addCase(Builder.getInt32(1), BBs[1]);
    SI->addCase(Builder.getInt32(2), BBs[1]);
    SI->addCase(Builder.getInt32(3), BBs[2]);
    SI->addCase(Builder.getInt32(4), BBs[3]);
    SI->addCase(Builder.getInt32(5), BBs[4]);

    Builder.SetInsertPoint(ContBB);
    return RValue::get(Builder.CreateIsNotNull(Result, "tobool"));
  }

  case Builtin::BI__atomic_clear: {
    QualType PtrTy = E->getArg(0)->IgnoreImpCasts()->getType();
    bool Volatile =
        PtrTy->castAs<PointerType>()->getPointeeType().isVolatileQualified();

    Address Ptr = EmitPointerWithAlignment(E->getArg(0));
    unsigned AddrSpace = Ptr.getPointer()->getType()->getPointerAddressSpace();
    Ptr = Builder.CreateBitCast(Ptr, Int8Ty->getPointerTo(AddrSpace));
    Value *NewVal = Builder.getInt8(0);
    Value *Order = EmitScalarExpr(E->getArg(1));
    if (isa<llvm::ConstantInt>(Order)) {
      int ord = cast<llvm::ConstantInt>(Order)->getZExtValue();
      StoreInst *Store = Builder.CreateStore(NewVal, Ptr, Volatile);
      switch (ord) {
      case 0:  // memory_order_relaxed
      default: // invalid order
        Store->setOrdering(llvm::AtomicOrdering::Monotonic);
        break;
      case 3:  // memory_order_release
        Store->setOrdering(llvm::AtomicOrdering::Release);
        break;
      case 5:  // memory_order_seq_cst
        Store->setOrdering(llvm::AtomicOrdering::SequentiallyConsistent);
        break;
      }
      return RValue::get(nullptr);
    }

    llvm::BasicBlock *ContBB = createBasicBlock("atomic.continue", CurFn);

    llvm::BasicBlock *BBs[3] = {
      createBasicBlock("monotonic", CurFn),
      createBasicBlock("release", CurFn),
      createBasicBlock("seqcst", CurFn)
    };
    llvm::AtomicOrdering Orders[3] = {
        llvm::AtomicOrdering::Monotonic, llvm::AtomicOrdering::Release,
        llvm::AtomicOrdering::SequentiallyConsistent};

    Order = Builder.CreateIntCast(Order, Builder.getInt32Ty(), false);
    llvm::SwitchInst *SI = Builder.CreateSwitch(Order, BBs[0]);

    for (unsigned i = 0; i < 3; ++i) {
      Builder.SetInsertPoint(BBs[i]);
      StoreInst *Store = Builder.CreateStore(NewVal, Ptr, Volatile);
      Store->setOrdering(Orders[i]);
      Builder.CreateBr(ContBB);
    }

    SI->addCase(Builder.getInt32(0), BBs[0]);
    SI->addCase(Builder.getInt32(3), BBs[1]);
    SI->addCase(Builder.getInt32(5), BBs[2]);

    Builder.SetInsertPoint(ContBB);
    return RValue::get(nullptr);
  }

  case Builtin::BI__atomic_thread_fence:
  case Builtin::BI__atomic_signal_fence:
  case Builtin::BI__c11_atomic_thread_fence:
  case Builtin::BI__c11_atomic_signal_fence: {
    llvm::SyncScope::ID SSID;
    if (BuiltinID == Builtin::BI__atomic_signal_fence ||
        BuiltinID == Builtin::BI__c11_atomic_signal_fence)
      SSID = llvm::SyncScope::SingleThread;
    else
      SSID = llvm::SyncScope::System;
    Value *Order = EmitScalarExpr(E->getArg(0));
    if (isa<llvm::ConstantInt>(Order)) {
      int ord = cast<llvm::ConstantInt>(Order)->getZExtValue();
      switch (ord) {
      case 0:  // memory_order_relaxed
      default: // invalid order
        break;
      case 1:  // memory_order_consume
      case 2:  // memory_order_acquire
        Builder.CreateFence(llvm::AtomicOrdering::Acquire, SSID);
        break;
      case 3:  // memory_order_release
        Builder.CreateFence(llvm::AtomicOrdering::Release, SSID);
        break;
      case 4:  // memory_order_acq_rel
        Builder.CreateFence(llvm::AtomicOrdering::AcquireRelease, SSID);
        break;
      case 5:  // memory_order_seq_cst
        Builder.CreateFence(llvm::AtomicOrdering::SequentiallyConsistent, SSID);
        break;
      }
      return RValue::get(nullptr);
    }

    llvm::BasicBlock *AcquireBB, *ReleaseBB, *AcqRelBB, *SeqCstBB;
    AcquireBB = createBasicBlock("acquire", CurFn);
    ReleaseBB = createBasicBlock("release", CurFn);
    AcqRelBB = createBasicBlock("acqrel", CurFn);
    SeqCstBB = createBasicBlock("seqcst", CurFn);
    llvm::BasicBlock *ContBB = createBasicBlock("atomic.continue", CurFn);

    Order = Builder.CreateIntCast(Order, Builder.getInt32Ty(), false);
    llvm::SwitchInst *SI = Builder.CreateSwitch(Order, ContBB);

    Builder.SetInsertPoint(AcquireBB);
    Builder.CreateFence(llvm::AtomicOrdering::Acquire, SSID);
    Builder.CreateBr(ContBB);
    SI->addCase(Builder.getInt32(1), AcquireBB);
    SI->addCase(Builder.getInt32(2), AcquireBB);

    Builder.SetInsertPoint(ReleaseBB);
    Builder.CreateFence(llvm::AtomicOrdering::Release, SSID);
    Builder.CreateBr(ContBB);
    SI->addCase(Builder.getInt32(3), ReleaseBB);

    Builder.SetInsertPoint(AcqRelBB);
    Builder.CreateFence(llvm::AtomicOrdering::AcquireRelease, SSID);
    Builder.CreateBr(ContBB);
    SI->addCase(Builder.getInt32(4), AcqRelBB);

    Builder.SetInsertPoint(SeqCstBB);
    Builder.CreateFence(llvm::AtomicOrdering::SequentiallyConsistent, SSID);
    Builder.CreateBr(ContBB);
    SI->addCase(Builder.getInt32(5), SeqCstBB);

    Builder.SetInsertPoint(ContBB);
    return RValue::get(nullptr);
  }

  case Builtin::BI__builtin_signbit:
  case Builtin::BI__builtin_signbitf:
  case Builtin::BI__builtin_signbitl: {
    return RValue::get(
        Builder.CreateZExt(EmitSignBit(*this, EmitScalarExpr(E->getArg(0))),
                           ConvertType(E->getType())));
  }
  case Builtin::BI__annotation: {
    // Re-encode each wide string to UTF8 and make an MDString.
    SmallVector<Metadata *, 1> Strings;
    for (const Expr *Arg : E->arguments()) {
      const auto *Str = cast<StringLiteral>(Arg->IgnoreParenCasts());
      assert(Str->getCharByteWidth() == 2);
      StringRef WideBytes = Str->getBytes();
      std::string StrUtf8;
      if (!convertUTF16ToUTF8String(
              makeArrayRef(WideBytes.data(), WideBytes.size()), StrUtf8)) {
        CGM.ErrorUnsupported(E, "non-UTF16 __annotation argument");
        continue;
      }
      Strings.push_back(llvm::MDString::get(getLLVMContext(), StrUtf8));
    }

    // Build and MDTuple of MDStrings and emit the intrinsic call.
    llvm::Value *F = CGM.getIntrinsic(llvm::Intrinsic::codeview_annotation, {});
    MDTuple *StrTuple = MDTuple::get(getLLVMContext(), Strings);
    Builder.CreateCall(F, MetadataAsValue::get(getLLVMContext(), StrTuple));
    return RValue::getIgnored();
  }
  case Builtin::BI__builtin_annotation: {
    llvm::Value *AnnVal = EmitScalarExpr(E->getArg(0));
    llvm::Value *F = CGM.getIntrinsic(llvm::Intrinsic::annotation,
                                      AnnVal->getType());

    // Get the annotation string, go through casts. Sema requires this to be a
    // non-wide string literal, potentially casted, so the cast<> is safe.
    const Expr *AnnotationStrExpr = E->getArg(1)->IgnoreParenCasts();
    StringRef Str = cast<StringLiteral>(AnnotationStrExpr)->getString();
    return RValue::get(EmitAnnotationCall(F, AnnVal, Str, E->getExprLoc()));
  }
  case Builtin::BI__builtin_addcb:
  case Builtin::BI__builtin_addcs:
  case Builtin::BI__builtin_addc:
  case Builtin::BI__builtin_addcl:
  case Builtin::BI__builtin_addcll:
  case Builtin::BI__builtin_subcb:
  case Builtin::BI__builtin_subcs:
  case Builtin::BI__builtin_subc:
  case Builtin::BI__builtin_subcl:
  case Builtin::BI__builtin_subcll: {

    // We translate all of these builtins from expressions of the form:
    //   int x = ..., y = ..., carryin = ..., carryout, result;
    //   result = __builtin_addc(x, y, carryin, &carryout);
    //
    // to LLVM IR of the form:
    //
    //   %tmp1 = call {i32, i1} @llvm.uadd.with.overflow.i32(i32 %x, i32 %y)
    //   %tmpsum1 = extractvalue {i32, i1} %tmp1, 0
    //   %carry1 = extractvalue {i32, i1} %tmp1, 1
    //   %tmp2 = call {i32, i1} @llvm.uadd.with.overflow.i32(i32 %tmpsum1,
    //                                                       i32 %carryin)
    //   %result = extractvalue {i32, i1} %tmp2, 0
    //   %carry2 = extractvalue {i32, i1} %tmp2, 1
    //   %tmp3 = or i1 %carry1, %carry2
    //   %tmp4 = zext i1 %tmp3 to i32
    //   store i32 %tmp4, i32* %carryout

    // Scalarize our inputs.
    llvm::Value *X = EmitScalarExpr(E->getArg(0));
    llvm::Value *Y = EmitScalarExpr(E->getArg(1));
    llvm::Value *Carryin = EmitScalarExpr(E->getArg(2));
    Address CarryOutPtr = EmitPointerWithAlignment(E->getArg(3));

    // Decide if we are lowering to a uadd.with.overflow or usub.with.overflow.
    llvm::Intrinsic::ID IntrinsicId;
    switch (BuiltinID) {
    default: llvm_unreachable("Unknown multiprecision builtin id.");
    case Builtin::BI__builtin_addcb:
    case Builtin::BI__builtin_addcs:
    case Builtin::BI__builtin_addc:
    case Builtin::BI__builtin_addcl:
    case Builtin::BI__builtin_addcll:
      IntrinsicId = llvm::Intrinsic::uadd_with_overflow;
      break;
    case Builtin::BI__builtin_subcb:
    case Builtin::BI__builtin_subcs:
    case Builtin::BI__builtin_subc:
    case Builtin::BI__builtin_subcl:
    case Builtin::BI__builtin_subcll:
      IntrinsicId = llvm::Intrinsic::usub_with_overflow;
      break;
    }

    // Construct our resulting LLVM IR expression.
    llvm::Value *Carry1;
    llvm::Value *Sum1 = EmitOverflowIntrinsic(*this, IntrinsicId,
                                              X, Y, Carry1);
    llvm::Value *Carry2;
    llvm::Value *Sum2 = EmitOverflowIntrinsic(*this, IntrinsicId,
                                              Sum1, Carryin, Carry2);
    llvm::Value *CarryOut = Builder.CreateZExt(Builder.CreateOr(Carry1, Carry2),
                                               X->getType());
    Builder.CreateStore(CarryOut, CarryOutPtr);
    return RValue::get(Sum2);
  }

  case Builtin::BI__builtin_add_overflow:
  case Builtin::BI__builtin_sub_overflow:
  case Builtin::BI__builtin_mul_overflow: {
    const clang::Expr *LeftArg = E->getArg(0);
    const clang::Expr *RightArg = E->getArg(1);
    const clang::Expr *ResultArg = E->getArg(2);

    clang::QualType ResultQTy =
        ResultArg->getType()->castAs<PointerType>()->getPointeeType();

    WidthAndSignedness LeftInfo =
        getIntegerWidthAndSignedness(CGM.getContext(), LeftArg->getType());
    WidthAndSignedness RightInfo =
        getIntegerWidthAndSignedness(CGM.getContext(), RightArg->getType());
    WidthAndSignedness ResultInfo =
        getIntegerWidthAndSignedness(CGM.getContext(), ResultQTy);

    // Handle mixed-sign multiplication as a special case, because adding
    // runtime or backend support for our generic irgen would be too expensive.
    if (isSpecialMixedSignMultiply(BuiltinID, LeftInfo, RightInfo, ResultInfo))
      return EmitCheckedMixedSignMultiply(*this, LeftArg, LeftInfo, RightArg,
                                          RightInfo, ResultArg, ResultQTy,
                                          ResultInfo);

    WidthAndSignedness EncompassingInfo =
        EncompassingIntegerType({LeftInfo, RightInfo, ResultInfo});

    llvm::Type *EncompassingLLVMTy =
        llvm::IntegerType::get(CGM.getLLVMContext(), EncompassingInfo.Width);

    llvm::Type *ResultLLVMTy = CGM.getTypes().ConvertType(ResultQTy);

    llvm::Intrinsic::ID IntrinsicId;
    switch (BuiltinID) {
    default:
      llvm_unreachable("Unknown overflow builtin id.");
    case Builtin::BI__builtin_add_overflow:
      IntrinsicId = EncompassingInfo.Signed
                        ? llvm::Intrinsic::sadd_with_overflow
                        : llvm::Intrinsic::uadd_with_overflow;
      break;
    case Builtin::BI__builtin_sub_overflow:
      IntrinsicId = EncompassingInfo.Signed
                        ? llvm::Intrinsic::ssub_with_overflow
                        : llvm::Intrinsic::usub_with_overflow;
      break;
    case Builtin::BI__builtin_mul_overflow:
      IntrinsicId = EncompassingInfo.Signed
                        ? llvm::Intrinsic::smul_with_overflow
                        : llvm::Intrinsic::umul_with_overflow;
      break;
    }

    llvm::Value *Left = EmitScalarExpr(LeftArg);
    llvm::Value *Right = EmitScalarExpr(RightArg);
    Address ResultPtr = EmitPointerWithAlignment(ResultArg);

    // Extend each operand to the encompassing type.
    Left = Builder.CreateIntCast(Left, EncompassingLLVMTy, LeftInfo.Signed);
    Right = Builder.CreateIntCast(Right, EncompassingLLVMTy, RightInfo.Signed);

    // Perform the operation on the extended values.
    llvm::Value *Overflow, *Result;
    Result = EmitOverflowIntrinsic(*this, IntrinsicId, Left, Right, Overflow);

    if (EncompassingInfo.Width > ResultInfo.Width) {
      // The encompassing type is wider than the result type, so we need to
      // truncate it.
      llvm::Value *ResultTrunc = Builder.CreateTrunc(Result, ResultLLVMTy);

      // To see if the truncation caused an overflow, we will extend
      // the result and then compare it to the original result.
      llvm::Value *ResultTruncExt = Builder.CreateIntCast(
          ResultTrunc, EncompassingLLVMTy, ResultInfo.Signed);
      llvm::Value *TruncationOverflow =
          Builder.CreateICmpNE(Result, ResultTruncExt);

      Overflow = Builder.CreateOr(Overflow, TruncationOverflow);
      Result = ResultTrunc;
    }

    // Finally, store the result using the pointer.
    bool isVolatile =
      ResultArg->getType()->getPointeeType().isVolatileQualified();
    Builder.CreateStore(EmitToMemory(Result, ResultQTy), ResultPtr, isVolatile);

    return RValue::get(Overflow);
  }

  case Builtin::BI__builtin_uadd_overflow:
  case Builtin::BI__builtin_uaddl_overflow:
  case Builtin::BI__builtin_uaddll_overflow:
  case Builtin::BI__builtin_usub_overflow:
  case Builtin::BI__builtin_usubl_overflow:
  case Builtin::BI__builtin_usubll_overflow:
  case Builtin::BI__builtin_umul_overflow:
  case Builtin::BI__builtin_umull_overflow:
  case Builtin::BI__builtin_umulll_overflow:
  case Builtin::BI__builtin_sadd_overflow:
  case Builtin::BI__builtin_saddl_overflow:
  case Builtin::BI__builtin_saddll_overflow:
  case Builtin::BI__builtin_ssub_overflow:
  case Builtin::BI__builtin_ssubl_overflow:
  case Builtin::BI__builtin_ssubll_overflow:
  case Builtin::BI__builtin_smul_overflow:
  case Builtin::BI__builtin_smull_overflow:
  case Builtin::BI__builtin_smulll_overflow: {

    // We translate all of these builtins directly to the relevant llvm IR node.

    // Scalarize our inputs.
    llvm::Value *X = EmitScalarExpr(E->getArg(0));
    llvm::Value *Y = EmitScalarExpr(E->getArg(1));
    Address SumOutPtr = EmitPointerWithAlignment(E->getArg(2));

    // Decide which of the overflow intrinsics we are lowering to:
    llvm::Intrinsic::ID IntrinsicId;
    switch (BuiltinID) {
    default: llvm_unreachable("Unknown overflow builtin id.");
    case Builtin::BI__builtin_uadd_overflow:
    case Builtin::BI__builtin_uaddl_overflow:
    case Builtin::BI__builtin_uaddll_overflow:
      IntrinsicId = llvm::Intrinsic::uadd_with_overflow;
      break;
    case Builtin::BI__builtin_usub_overflow:
    case Builtin::BI__builtin_usubl_overflow:
    case Builtin::BI__builtin_usubll_overflow:
      IntrinsicId = llvm::Intrinsic::usub_with_overflow;
      break;
    case Builtin::BI__builtin_umul_overflow:
    case Builtin::BI__builtin_umull_overflow:
    case Builtin::BI__builtin_umulll_overflow:
      IntrinsicId = llvm::Intrinsic::umul_with_overflow;
      break;
    case Builtin::BI__builtin_sadd_overflow:
    case Builtin::BI__builtin_saddl_overflow:
    case Builtin::BI__builtin_saddll_overflow:
      IntrinsicId = llvm::Intrinsic::sadd_with_overflow;
      break;
    case Builtin::BI__builtin_ssub_overflow:
    case Builtin::BI__builtin_ssubl_overflow:
    case Builtin::BI__builtin_ssubll_overflow:
      IntrinsicId = llvm::Intrinsic::ssub_with_overflow;
      break;
    case Builtin::BI__builtin_smul_overflow:
    case Builtin::BI__builtin_smull_overflow:
    case Builtin::BI__builtin_smulll_overflow:
      IntrinsicId = llvm::Intrinsic::smul_with_overflow;
      break;
    }


    llvm::Value *Carry;
    llvm::Value *Sum = EmitOverflowIntrinsic(*this, IntrinsicId, X, Y, Carry);
    Builder.CreateStore(Sum, SumOutPtr);

    return RValue::get(Carry);
  }
  case Builtin::BI__builtin_addressof:
    return RValue::get(EmitLValue(E->getArg(0)).getPointer());
  case Builtin::BI__builtin_operator_new:
    return EmitBuiltinNewDeleteCall(
        E->getCallee()->getType()->castAs<FunctionProtoType>(), E, false);
  case Builtin::BI__builtin_operator_delete:
    return EmitBuiltinNewDeleteCall(
        E->getCallee()->getType()->castAs<FunctionProtoType>(), E, true);

  case Builtin::BI__noop:
    // __noop always evaluates to an integer literal zero.
    return RValue::get(ConstantInt::get(IntTy, 0));
  case Builtin::BI__builtin_call_with_static_chain: {
    const CallExpr *Call = cast<CallExpr>(E->getArg(0));
    const Expr *Chain = E->getArg(1);
    return EmitCall(Call->getCallee()->getType(),
                    EmitCallee(Call->getCallee()), Call, ReturnValue,
                    EmitScalarExpr(Chain));
  }
  case Builtin::BI_InterlockedExchange8:
  case Builtin::BI_InterlockedExchange16:
  case Builtin::BI_InterlockedExchange:
  case Builtin::BI_InterlockedExchangePointer:
    return RValue::get(
        EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedExchange, E));
  case Builtin::BI_InterlockedCompareExchangePointer: {
    llvm::Type *RTy;
    llvm::IntegerType *IntType =
      IntegerType::get(getLLVMContext(),
                       getContext().getTypeSize(E->getType()));
    llvm::Type *IntPtrType = IntType->getPointerTo();

    llvm::Value *Destination =
      Builder.CreateBitCast(EmitScalarExpr(E->getArg(0)), IntPtrType);

    llvm::Value *Exchange = EmitScalarExpr(E->getArg(1));
    RTy = Exchange->getType();
    Exchange = Builder.CreatePtrToInt(Exchange, IntType);

    llvm::Value *Comparand =
      Builder.CreatePtrToInt(EmitScalarExpr(E->getArg(2)), IntType);

    auto Result =
        Builder.CreateAtomicCmpXchg(Destination, Comparand, Exchange,
                                    AtomicOrdering::SequentiallyConsistent,
                                    AtomicOrdering::SequentiallyConsistent);
    Result->setVolatile(true);

    return RValue::get(Builder.CreateIntToPtr(Builder.CreateExtractValue(Result,
                                                                         0),
                                              RTy));
  }
  case Builtin::BI_InterlockedCompareExchange8:
  case Builtin::BI_InterlockedCompareExchange16:
  case Builtin::BI_InterlockedCompareExchange:
  case Builtin::BI_InterlockedCompareExchange64: {
    AtomicCmpXchgInst *CXI = Builder.CreateAtomicCmpXchg(
        EmitScalarExpr(E->getArg(0)),
        EmitScalarExpr(E->getArg(2)),
        EmitScalarExpr(E->getArg(1)),
        AtomicOrdering::SequentiallyConsistent,
        AtomicOrdering::SequentiallyConsistent);
      CXI->setVolatile(true);
      return RValue::get(Builder.CreateExtractValue(CXI, 0));
  }
  case Builtin::BI_InterlockedIncrement16:
  case Builtin::BI_InterlockedIncrement:
    return RValue::get(
        EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedIncrement, E));
  case Builtin::BI_InterlockedDecrement16:
  case Builtin::BI_InterlockedDecrement:
    return RValue::get(
        EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedDecrement, E));
  case Builtin::BI_InterlockedAnd8:
  case Builtin::BI_InterlockedAnd16:
  case Builtin::BI_InterlockedAnd:
    return RValue::get(EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedAnd, E));
  case Builtin::BI_InterlockedExchangeAdd8:
  case Builtin::BI_InterlockedExchangeAdd16:
  case Builtin::BI_InterlockedExchangeAdd:
    return RValue::get(
        EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedExchangeAdd, E));
  case Builtin::BI_InterlockedExchangeSub8:
  case Builtin::BI_InterlockedExchangeSub16:
  case Builtin::BI_InterlockedExchangeSub:
    return RValue::get(
        EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedExchangeSub, E));
  case Builtin::BI_InterlockedOr8:
  case Builtin::BI_InterlockedOr16:
  case Builtin::BI_InterlockedOr:
    return RValue::get(EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedOr, E));
  case Builtin::BI_InterlockedXor8:
  case Builtin::BI_InterlockedXor16:
  case Builtin::BI_InterlockedXor:
    return RValue::get(EmitMSVCBuiltinExpr(MSVCIntrin::_InterlockedXor, E));

  case Builtin::BI_bittest64:
  case Builtin::BI_bittest:
  case Builtin::BI_bittestandcomplement64:
  case Builtin::BI_bittestandcomplement:
  case Builtin::BI_bittestandreset64:
  case Builtin::BI_bittestandreset:
  case Builtin::BI_bittestandset64:
  case Builtin::BI_bittestandset:
  case Builtin::BI_interlockedbittestandreset:
  case Builtin::BI_interlockedbittestandreset64:
  case Builtin::BI_interlockedbittestandset64:
  case Builtin::BI_interlockedbittestandset:
  case Builtin::BI_interlockedbittestandset_acq:
  case Builtin::BI_interlockedbittestandset_rel:
  case Builtin::BI_interlockedbittestandset_nf:
  case Builtin::BI_interlockedbittestandreset_acq:
  case Builtin::BI_interlockedbittestandreset_rel:
  case Builtin::BI_interlockedbittestandreset_nf:
    return RValue::get(EmitBitTestIntrinsic(*this, BuiltinID, E));

  case Builtin::BI__exception_code:
  case Builtin::BI_exception_code:
    return RValue::get(EmitSEHExceptionCode());
  case Builtin::BI__exception_info:
  case Builtin::BI_exception_info:
    return RValue::get(EmitSEHExceptionInfo());
  case Builtin::BI__abnormal_termination:
  case Builtin::BI_abnormal_termination:
    return RValue::get(EmitSEHAbnormalTermination());
  case Builtin::BI_setjmpex:
    if (getTarget().getTriple().isOSMSVCRT())
      return EmitMSVCRTSetJmp(*this, MSVCSetJmpKind::_setjmpex, E);
    break;
  case Builtin::BI_setjmp:
    if (getTarget().getTriple().isOSMSVCRT()) {
      if (getTarget().getTriple().getArch() == llvm::Triple::x86)
        return EmitMSVCRTSetJmp(*this, MSVCSetJmpKind::_setjmp3, E);
      else if (getTarget().getTriple().getArch() == llvm::Triple::aarch64)
        return EmitMSVCRTSetJmp(*this, MSVCSetJmpKind::_setjmpex, E);
      return EmitMSVCRTSetJmp(*this, MSVCSetJmpKind::_setjmp, E);
    }
    break;

  case Builtin::BI__GetExceptionInfo: {
    if (llvm::GlobalVariable *GV =
            CGM.getCXXABI().getThrowInfo(FD->getParamDecl(0)->getType()))
      return RValue::get(llvm::ConstantExpr::getBitCast(GV, CGM.Int8PtrTy));
    break;
  }

  case Builtin::BI__fastfail:
    return RValue::get(EmitMSVCBuiltinExpr(MSVCIntrin::__fastfail, E));

  case Builtin::BI__builtin_coro_size: {
    auto & Context = getContext();
    auto SizeTy = Context.getSizeType();
    auto T = Builder.getIntNTy(Context.getTypeSize(SizeTy));
    Value *F = CGM.getIntrinsic(Intrinsic::coro_size, T);
    return RValue::get(Builder.CreateCall(F));
  }

  case Builtin::BI__builtin_coro_id:
    return EmitCoroutineIntrinsic(E, Intrinsic::coro_id);
  case Builtin::BI__builtin_coro_promise:
    return EmitCoroutineIntrinsic(E, Intrinsic::coro_promise);
  case Builtin::BI__builtin_coro_resume:
    return EmitCoroutineIntrinsic(E, Intrinsic::coro_resume);
  case Builtin::BI__builtin_coro_frame:
    return EmitCoroutineIntrinsic(E, Intrinsic::coro_frame);
  case Builtin::BI__builtin_coro_noop:
    return EmitCoroutineIntrinsic(E, Intrinsic::coro_noop);
  case Builtin::BI__builtin_coro_free:
    return EmitCoroutineIntrinsic(E, Intrinsic::coro_free);
  case Builtin::BI__builtin_coro_destroy:
    return EmitCoroutineIntrinsic(E, Intrinsic::coro_destroy);
  case Builtin::BI__builtin_coro_done:
    return EmitCoroutineIntrinsic(E, Intrinsic::coro_done);
  case Builtin::BI__builtin_coro_alloc:
    return EmitCoroutineIntrinsic(E, Intrinsic::coro_alloc);
  case Builtin::BI__builtin_coro_begin:
    return EmitCoroutineIntrinsic(E, Intrinsic::coro_begin);
  case Builtin::BI__builtin_coro_end:
    return EmitCoroutineIntrinsic(E, Intrinsic::coro_end);
  case Builtin::BI__builtin_coro_suspend:
    return EmitCoroutineIntrinsic(E, Intrinsic::coro_suspend);
  case Builtin::BI__builtin_coro_param:
    return EmitCoroutineIntrinsic(E, Intrinsic::coro_param);

  // OpenCL v2.0 s6.13.16.2, Built-in pipe read and write functions
  case Builtin::BIread_pipe:
  case Builtin::BIwrite_pipe: {
    Value *Arg0 = EmitScalarExpr(E->getArg(0)),
          *Arg1 = EmitScalarExpr(E->getArg(1));
    CGOpenCLRuntime OpenCLRT(CGM);
    Value *PacketSize = OpenCLRT.getPipeElemSize(E->getArg(0));
    Value *PacketAlign = OpenCLRT.getPipeElemAlign(E->getArg(0));

    // Type of the generic packet parameter.
    unsigned GenericAS =
        getContext().getTargetAddressSpace(LangAS::opencl_generic);
    llvm::Type *I8PTy = llvm::PointerType::get(
        llvm::Type::getInt8Ty(getLLVMContext()), GenericAS);

    // Testing which overloaded version we should generate the call for.
    if (2U == E->getNumArgs()) {
      const char *Name = (BuiltinID == Builtin::BIread_pipe) ? "__read_pipe_2"
                                                             : "__write_pipe_2";
      // Creating a generic function type to be able to call with any builtin or
      // user defined type.
      llvm::Type *ArgTys[] = {Arg0->getType(), I8PTy, Int32Ty, Int32Ty};
      llvm::FunctionType *FTy = llvm::FunctionType::get(
          Int32Ty, llvm::ArrayRef<llvm::Type *>(ArgTys), false);
      Value *BCast = Builder.CreatePointerCast(Arg1, I8PTy);
      return RValue::get(
          Builder.CreateCall(CGM.CreateRuntimeFunction(FTy, Name),
                             {Arg0, BCast, PacketSize, PacketAlign}));
    } else {
      assert(4 == E->getNumArgs() &&
             "Illegal number of parameters to pipe function");
      const char *Name = (BuiltinID == Builtin::BIread_pipe) ? "__read_pipe_4"
                                                             : "__write_pipe_4";

      llvm::Type *ArgTys[] = {Arg0->getType(), Arg1->getType(), Int32Ty, I8PTy,
                              Int32Ty, Int32Ty};
      Value *Arg2 = EmitScalarExpr(E->getArg(2)),
            *Arg3 = EmitScalarExpr(E->getArg(3));
      llvm::FunctionType *FTy = llvm::FunctionType::get(
          Int32Ty, llvm::ArrayRef<llvm::Type *>(ArgTys), false);
      Value *BCast = Builder.CreatePointerCast(Arg3, I8PTy);
      // We know the third argument is an integer type, but we may need to cast
      // it to i32.
      if (Arg2->getType() != Int32Ty)
        Arg2 = Builder.CreateZExtOrTrunc(Arg2, Int32Ty);
      return RValue::get(Builder.CreateCall(
          CGM.CreateRuntimeFunction(FTy, Name),
          {Arg0, Arg1, Arg2, BCast, PacketSize, PacketAlign}));
    }
  }
  // OpenCL v2.0 s6.13.16 ,s9.17.3.5 - Built-in pipe reserve read and write
  // functions
  case Builtin::BIreserve_read_pipe:
  case Builtin::BIreserve_write_pipe:
  case Builtin::BIwork_group_reserve_read_pipe:
  case Builtin::BIwork_group_reserve_write_pipe:
  case Builtin::BIsub_group_reserve_read_pipe:
  case Builtin::BIsub_group_reserve_write_pipe: {
    // Composing the mangled name for the function.
    const char *Name;
    if (BuiltinID == Builtin::BIreserve_read_pipe)
      Name = "__reserve_read_pipe";
    else if (BuiltinID == Builtin::BIreserve_write_pipe)
      Name = "__reserve_write_pipe";
    else if (BuiltinID == Builtin::BIwork_group_reserve_read_pipe)
      Name = "__work_group_reserve_read_pipe";
    else if (BuiltinID == Builtin::BIwork_group_reserve_write_pipe)
      Name = "__work_group_reserve_write_pipe";
    else if (BuiltinID == Builtin::BIsub_group_reserve_read_pipe)
      Name = "__sub_group_reserve_read_pipe";
    else
      Name = "__sub_group_reserve_write_pipe";

    Value *Arg0 = EmitScalarExpr(E->getArg(0)),
          *Arg1 = EmitScalarExpr(E->getArg(1));
    llvm::Type *ReservedIDTy = ConvertType(getContext().OCLReserveIDTy);
    CGOpenCLRuntime OpenCLRT(CGM);
    Value *PacketSize = OpenCLRT.getPipeElemSize(E->getArg(0));
    Value *PacketAlign = OpenCLRT.getPipeElemAlign(E->getArg(0));

    // Building the generic function prototype.
    llvm::Type *ArgTys[] = {Arg0->getType(), Int32Ty, Int32Ty, Int32Ty};
    llvm::FunctionType *FTy = llvm::FunctionType::get(
        ReservedIDTy, llvm::ArrayRef<llvm::Type *>(ArgTys), false);
    // We know the second argument is an integer type, but we may need to cast
    // it to i32.
    if (Arg1->getType() != Int32Ty)
      Arg1 = Builder.CreateZExtOrTrunc(Arg1, Int32Ty);
    return RValue::get(
        Builder.CreateCall(CGM.CreateRuntimeFunction(FTy, Name),
                           {Arg0, Arg1, PacketSize, PacketAlign}));
  }
  // OpenCL v2.0 s6.13.16, s9.17.3.5 - Built-in pipe commit read and write
  // functions
  case Builtin::BIcommit_read_pipe:
  case Builtin::BIcommit_write_pipe:
  case Builtin::BIwork_group_commit_read_pipe:
  case Builtin::BIwork_group_commit_write_pipe:
  case Builtin::BIsub_group_commit_read_pipe:
  case Builtin::BIsub_group_commit_write_pipe: {
    const char *Name;
    if (BuiltinID == Builtin::BIcommit_read_pipe)
      Name = "__commit_read_pipe";
    else if (BuiltinID == Builtin::BIcommit_write_pipe)
      Name = "__commit_write_pipe";
    else if (BuiltinID == Builtin::BIwork_group_commit_read_pipe)
      Name = "__work_group_commit_read_pipe";
    else if (BuiltinID == Builtin::BIwork_group_commit_write_pipe)
      Name = "__work_group_commit_write_pipe";
    else if (BuiltinID == Builtin::BIsub_group_commit_read_pipe)
      Name = "__sub_group_commit_read_pipe";
    else
      Name = "__sub_group_commit_write_pipe";

    Value *Arg0 = EmitScalarExpr(E->getArg(0)),
          *Arg1 = EmitScalarExpr(E->getArg(1));
    CGOpenCLRuntime OpenCLRT(CGM);
    Value *PacketSize = OpenCLRT.getPipeElemSize(E->getArg(0));
    Value *PacketAlign = OpenCLRT.getPipeElemAlign(E->getArg(0));

    // Building the generic function prototype.
    llvm::Type *ArgTys[] = {Arg0->getType(), Arg1->getType(), Int32Ty, Int32Ty};
    llvm::FunctionType *FTy =
        llvm::FunctionType::get(llvm::Type::getVoidTy(getLLVMContext()),
                                llvm::ArrayRef<llvm::Type *>(ArgTys), false);

    return RValue::get(
        Builder.CreateCall(CGM.CreateRuntimeFunction(FTy, Name),
                           {Arg0, Arg1, PacketSize, PacketAlign}));
  }
  // OpenCL v2.0 s6.13.16.4 Built-in pipe query functions
  case Builtin::BIget_pipe_num_packets:
  case Builtin::BIget_pipe_max_packets: {
    const char *BaseName;
    const PipeType *PipeTy = E->getArg(0)->getType()->getAs<PipeType>();
    if (BuiltinID == Builtin::BIget_pipe_num_packets)
      BaseName = "__get_pipe_num_packets";
    else
      BaseName = "__get_pipe_max_packets";
    auto Name = std::string(BaseName) +
                std::string(PipeTy->isReadOnly() ? "_ro" : "_wo");

    // Building the generic function prototype.
    Value *Arg0 = EmitScalarExpr(E->getArg(0));
    CGOpenCLRuntime OpenCLRT(CGM);
    Value *PacketSize = OpenCLRT.getPipeElemSize(E->getArg(0));
    Value *PacketAlign = OpenCLRT.getPipeElemAlign(E->getArg(0));
    llvm::Type *ArgTys[] = {Arg0->getType(), Int32Ty, Int32Ty};
    llvm::FunctionType *FTy = llvm::FunctionType::get(
        Int32Ty, llvm::ArrayRef<llvm::Type *>(ArgTys), false);

    return RValue::get(Builder.CreateCall(CGM.CreateRuntimeFunction(FTy, Name),
                                          {Arg0, PacketSize, PacketAlign}));
  }

  // OpenCL v2.0 s6.13.9 - Address space qualifier functions.
  case Builtin::BIto_global:
  case Builtin::BIto_local:
  case Builtin::BIto_private: {
    auto Arg0 = EmitScalarExpr(E->getArg(0));
    auto NewArgT = llvm::PointerType::get(Int8Ty,
      CGM.getContext().getTargetAddressSpace(LangAS::opencl_generic));
    auto NewRetT = llvm::PointerType::get(Int8Ty,
      CGM.getContext().getTargetAddressSpace(
        E->getType()->getPointeeType().getAddressSpace()));
    auto FTy = llvm::FunctionType::get(NewRetT, {NewArgT}, false);
    llvm::Value *NewArg;
    if (Arg0->getType()->getPointerAddressSpace() !=
        NewArgT->getPointerAddressSpace())
      NewArg = Builder.CreateAddrSpaceCast(Arg0, NewArgT);
    else
      NewArg = Builder.CreateBitOrPointerCast(Arg0, NewArgT);
    auto NewName = std::string("__") + E->getDirectCallee()->getName().str();
    auto NewCall =
        Builder.CreateCall(CGM.CreateRuntimeFunction(FTy, NewName), {NewArg});
    return RValue::get(Builder.CreateBitOrPointerCast(NewCall,
      ConvertType(E->getType())));
  }

  // OpenCL v2.0, s6.13.17 - Enqueue kernel function.
  // It contains four different overload formats specified in Table 6.13.17.1.
  case Builtin::BIenqueue_kernel: {
    StringRef Name; // Generated function call name
    unsigned NumArgs = E->getNumArgs();

    llvm::Type *QueueTy = ConvertType(getContext().OCLQueueTy);
    llvm::Type *GenericVoidPtrTy = Builder.getInt8PtrTy(
        getContext().getTargetAddressSpace(LangAS::opencl_generic));

    llvm::Value *Queue = EmitScalarExpr(E->getArg(0));
    llvm::Value *Flags = EmitScalarExpr(E->getArg(1));
    LValue NDRangeL = EmitAggExprToLValue(E->getArg(2));
    llvm::Value *Range = NDRangeL.getAddress().getPointer();
    llvm::Type *RangeTy = NDRangeL.getAddress().getType();

    if (NumArgs == 4) {
      // The most basic form of the call with parameters:
      // queue_t, kernel_enqueue_flags_t, ndrange_t, block(void)
      Name = "__enqueue_kernel_basic";
      llvm::Type *ArgTys[] = {QueueTy, Int32Ty, RangeTy, GenericVoidPtrTy,
                              GenericVoidPtrTy};
      llvm::FunctionType *FTy = llvm::FunctionType::get(
          Int32Ty, llvm::ArrayRef<llvm::Type *>(ArgTys), false);

      auto Info =
          CGM.getOpenCLRuntime().emitOpenCLEnqueuedBlock(*this, E->getArg(3));
      llvm::Value *Kernel =
          Builder.CreatePointerCast(Info.Kernel, GenericVoidPtrTy);
      llvm::Value *Block =
          Builder.CreatePointerCast(Info.BlockArg, GenericVoidPtrTy);

      AttrBuilder B;
      B.addAttribute(Attribute::ByVal);
      llvm::AttributeList ByValAttrSet =
          llvm::AttributeList::get(CGM.getModule().getContext(), 3U, B);

      auto RTCall =
          Builder.CreateCall(CGM.CreateRuntimeFunction(FTy, Name, ByValAttrSet),
                             {Queue, Flags, Range, Kernel, Block});
      RTCall->setAttributes(ByValAttrSet);
      return RValue::get(RTCall);
    }
    assert(NumArgs >= 5 && "Invalid enqueue_kernel signature");

    // Create a temporary array to hold the sizes of local pointer arguments
    // for the block. \p First is the position of the first size argument.
    auto CreateArrayForSizeVar = [=](unsigned First)
        -> std::tuple<llvm::Value *, llvm::Value *, llvm::Value *> {
      llvm::APInt ArraySize(32, NumArgs - First);
      QualType SizeArrayTy = getContext().getConstantArrayType(
          getContext().getSizeType(), ArraySize, ArrayType::Normal,
          /*IndexTypeQuals=*/0);
      auto Tmp = CreateMemTemp(SizeArrayTy, "block_sizes");
      llvm::Value *TmpPtr = Tmp.getPointer();
      llvm::Value *TmpSize = EmitLifetimeStart(
          CGM.getDataLayout().getTypeAllocSize(Tmp.getElementType()), TmpPtr);
      llvm::Value *ElemPtr;
      // Each of the following arguments specifies the size of the corresponding
      // argument passed to the enqueued block.
      auto *Zero = llvm::ConstantInt::get(IntTy, 0);
      for (unsigned I = First; I < NumArgs; ++I) {
        auto *Index = llvm::ConstantInt::get(IntTy, I - First);
        auto *GEP = Builder.CreateGEP(TmpPtr, {Zero, Index});
        if (I == First)
          ElemPtr = GEP;
        auto *V =
            Builder.CreateZExtOrTrunc(EmitScalarExpr(E->getArg(I)), SizeTy);
        Builder.CreateAlignedStore(
            V, GEP, CGM.getDataLayout().getPrefTypeAlignment(SizeTy));
      }
      return std::tie(ElemPtr, TmpSize, TmpPtr);
    };

    // Could have events and/or varargs.
    if (E->getArg(3)->getType()->isBlockPointerType()) {
      // No events passed, but has variadic arguments.
      Name = "__enqueue_kernel_varargs";
      auto Info =
          CGM.getOpenCLRuntime().emitOpenCLEnqueuedBlock(*this, E->getArg(3));
      llvm::Value *Kernel =
          Builder.CreatePointerCast(Info.Kernel, GenericVoidPtrTy);
      auto *Block = Builder.CreatePointerCast(Info.BlockArg, GenericVoidPtrTy);
      llvm::Value *ElemPtr, *TmpSize, *TmpPtr;
      std::tie(ElemPtr, TmpSize, TmpPtr) = CreateArrayForSizeVar(4);

      // Create a vector of the arguments, as well as a constant value to
      // express to the runtime the number of variadic arguments.
      std::vector<llvm::Value *> Args = {
          Queue,  Flags, Range,
          Kernel, Block, ConstantInt::get(IntTy, NumArgs - 4),
          ElemPtr};
      std::vector<llvm::Type *> ArgTys = {
          QueueTy,          IntTy, RangeTy,           GenericVoidPtrTy,
          GenericVoidPtrTy, IntTy, ElemPtr->getType()};

      llvm::FunctionType *FTy = llvm::FunctionType::get(
          Int32Ty, llvm::ArrayRef<llvm::Type *>(ArgTys), false);
      auto Call =
          RValue::get(Builder.CreateCall(CGM.CreateRuntimeFunction(FTy, Name),
                                         llvm::ArrayRef<llvm::Value *>(Args)));
      if (TmpSize)
        EmitLifetimeEnd(TmpSize, TmpPtr);
      return Call;
    }
    // Any calls now have event arguments passed.
    if (NumArgs >= 7) {
      llvm::Type *EventTy = ConvertType(getContext().OCLClkEventTy);
      llvm::Type *EventPtrTy = EventTy->getPointerTo(
          CGM.getContext().getTargetAddressSpace(LangAS::opencl_generic));

      llvm::Value *NumEvents =
          Builder.CreateZExtOrTrunc(EmitScalarExpr(E->getArg(3)), Int32Ty);
      llvm::Value *EventList =
          E->getArg(4)->getType()->isArrayType()
              ? EmitArrayToPointerDecay(E->getArg(4)).getPointer()
              : EmitScalarExpr(E->getArg(4));
      llvm::Value *ClkEvent = EmitScalarExpr(E->getArg(5));
      // Convert to generic address space.
      EventList = Builder.CreatePointerCast(EventList, EventPtrTy);
      ClkEvent = Builder.CreatePointerCast(ClkEvent, EventPtrTy);
      auto Info =
          CGM.getOpenCLRuntime().emitOpenCLEnqueuedBlock(*this, E->getArg(6));
      llvm::Value *Kernel =
          Builder.CreatePointerCast(Info.Kernel, GenericVoidPtrTy);
      llvm::Value *Block =
          Builder.CreatePointerCast(Info.BlockArg, GenericVoidPtrTy);

      std::vector<llvm::Type *> ArgTys = {
          QueueTy,    Int32Ty,    RangeTy,          Int32Ty,
          EventPtrTy, EventPtrTy, GenericVoidPtrTy, GenericVoidPtrTy};

      std::vector<llvm::Value *> Args = {Queue,     Flags,    Range,  NumEvents,
                                         EventList, ClkEvent, Kernel, Block};

      if (NumArgs == 7) {
        // Has events but no variadics.
        Name = "__enqueue_kernel_basic_events";
        llvm::FunctionType *FTy = llvm::FunctionType::get(
            Int32Ty, llvm::ArrayRef<llvm::Type *>(ArgTys), false);
        return RValue::get(
            Builder.CreateCall(CGM.CreateRuntimeFunction(FTy, Name),
                               llvm::ArrayRef<llvm::Value *>(Args)));
      }
      // Has event info and variadics
      // Pass the number of variadics to the runtime function too.
      Args.push_back(ConstantInt::get(Int32Ty, NumArgs - 7));
      ArgTys.push_back(Int32Ty);
      Name = "__enqueue_kernel_events_varargs";

      llvm::Value *ElemPtr, *TmpSize, *TmpPtr;
      std::tie(ElemPtr, TmpSize, TmpPtr) = CreateArrayForSizeVar(7);
      Args.push_back(ElemPtr);
      ArgTys.push_back(ElemPtr->getType());

      llvm::FunctionType *FTy = llvm::FunctionType::get(
          Int32Ty, llvm::ArrayRef<llvm::Type *>(ArgTys), false);
      auto Call =
          RValue::get(Builder.CreateCall(CGM.CreateRuntimeFunction(FTy, Name),
                                         llvm::ArrayRef<llvm::Value *>(Args)));
      if (TmpSize)
        EmitLifetimeEnd(TmpSize, TmpPtr);
      return Call;
    }
    LLVM_FALLTHROUGH;
  }
  // OpenCL v2.0 s6.13.17.6 - Kernel query functions need bitcast of block
  // parameter.
  case Builtin::BIget_kernel_work_group_size: {
    llvm::Type *GenericVoidPtrTy = Builder.getInt8PtrTy(
        getContext().getTargetAddressSpace(LangAS::opencl_generic));
    auto Info =
        CGM.getOpenCLRuntime().emitOpenCLEnqueuedBlock(*this, E->getArg(0));
    Value *Kernel = Builder.CreatePointerCast(Info.Kernel, GenericVoidPtrTy);
    Value *Arg = Builder.CreatePointerCast(Info.BlockArg, GenericVoidPtrTy);
    return RValue::get(Builder.CreateCall(
        CGM.CreateRuntimeFunction(
            llvm::FunctionType::get(IntTy, {GenericVoidPtrTy, GenericVoidPtrTy},
                                    false),
            "__get_kernel_work_group_size_impl"),
        {Kernel, Arg}));
  }
  case Builtin::BIget_kernel_preferred_work_group_size_multiple: {
    llvm::Type *GenericVoidPtrTy = Builder.getInt8PtrTy(
        getContext().getTargetAddressSpace(LangAS::opencl_generic));
    auto Info =
        CGM.getOpenCLRuntime().emitOpenCLEnqueuedBlock(*this, E->getArg(0));
    Value *Kernel = Builder.CreatePointerCast(Info.Kernel, GenericVoidPtrTy);
    Value *Arg = Builder.CreatePointerCast(Info.BlockArg, GenericVoidPtrTy);
    return RValue::get(Builder.CreateCall(
        CGM.CreateRuntimeFunction(
            llvm::FunctionType::get(IntTy, {GenericVoidPtrTy, GenericVoidPtrTy},
                                    false),
            "__get_kernel_preferred_work_group_size_multiple_impl"),
        {Kernel, Arg}));
  }
  case Builtin::BIget_kernel_max_sub_group_size_for_ndrange:
  case Builtin::BIget_kernel_sub_group_count_for_ndrange: {
    llvm::Type *GenericVoidPtrTy = Builder.getInt8PtrTy(
        getContext().getTargetAddressSpace(LangAS::opencl_generic));
    LValue NDRangeL = EmitAggExprToLValue(E->getArg(0));
    llvm::Value *NDRange = NDRangeL.getAddress().getPointer();
    auto Info =
        CGM.getOpenCLRuntime().emitOpenCLEnqueuedBlock(*this, E->getArg(1));
    Value *Kernel = Builder.CreatePointerCast(Info.Kernel, GenericVoidPtrTy);
    Value *Block = Builder.CreatePointerCast(Info.BlockArg, GenericVoidPtrTy);
    const char *Name =
        BuiltinID == Builtin::BIget_kernel_max_sub_group_size_for_ndrange
            ? "__get_kernel_max_sub_group_size_for_ndrange_impl"
            : "__get_kernel_sub_group_count_for_ndrange_impl";
    return RValue::get(Builder.CreateCall(
        CGM.CreateRuntimeFunction(
            llvm::FunctionType::get(
                IntTy, {NDRange->getType(), GenericVoidPtrTy, GenericVoidPtrTy},
                false),
            Name),
        {NDRange, Kernel, Block}));
  }

  case Builtin::BI__builtin_store_half:
  case Builtin::BI__builtin_store_halff: {
    Value *Val = EmitScalarExpr(E->getArg(0));
    Address Address = EmitPointerWithAlignment(E->getArg(1));
    Value *HalfVal = Builder.CreateFPTrunc(Val, Builder.getHalfTy());
    return RValue::get(Builder.CreateStore(HalfVal, Address));
  }
  case Builtin::BI__builtin_load_half: {
    Address Address = EmitPointerWithAlignment(E->getArg(0));
    Value *HalfVal = Builder.CreateLoad(Address);
    return RValue::get(Builder.CreateFPExt(HalfVal, Builder.getDoubleTy()));
  }
  case Builtin::BI__builtin_load_halff: {
    Address Address = EmitPointerWithAlignment(E->getArg(0));
    Value *HalfVal = Builder.CreateLoad(Address);
    return RValue::get(Builder.CreateFPExt(HalfVal, Builder.getFloatTy()));
  }
  case Builtin::BIprintf:
    if (getTarget().getTriple().isNVPTX())
      return EmitNVPTXDevicePrintfCallExpr(E, ReturnValue);
    break;
  case Builtin::BI__builtin_canonicalize:
  case Builtin::BI__builtin_canonicalizef:
  case Builtin::BI__builtin_canonicalizel:
    return RValue::get(emitUnaryBuiltin(*this, E, Intrinsic::canonicalize));

  case Builtin::BI__builtin_thread_pointer: {
    if (!getContext().getTargetInfo().isTLSSupported())
      CGM.ErrorUnsupported(E, "__builtin_thread_pointer");
    // Fall through - it's already mapped to the intrinsic by GCCBuiltin.
    break;
  }
  case Builtin::BI__builtin_os_log_format:
    return emitBuiltinOSLogFormat(*E);

  case Builtin::BI__builtin_os_log_format_buffer_size: {
    analyze_os_log::OSLogBufferLayout Layout;
    analyze_os_log::computeOSLogBufferLayout(CGM.getContext(), E, Layout);
    return RValue::get(ConstantInt::get(ConvertType(E->getType()),
                                        Layout.size().getQuantity()));
  }

  case Builtin::BI__xray_customevent: {
    if (!ShouldXRayInstrumentFunction())
      return RValue::getIgnored();

    if (!CGM.getCodeGenOpts().XRayInstrumentationBundle.has(
            XRayInstrKind::Custom))
      return RValue::getIgnored();

    if (const auto *XRayAttr = CurFuncDecl->getAttr<XRayInstrumentAttr>())
      if (XRayAttr->neverXRayInstrument() && !AlwaysEmitXRayCustomEvents())
        return RValue::getIgnored();

    Function *F = CGM.getIntrinsic(Intrinsic::xray_customevent);
    auto FTy = F->getFunctionType();
    auto Arg0 = E->getArg(0);
    auto Arg0Val = EmitScalarExpr(Arg0);
    auto Arg0Ty = Arg0->getType();
    auto PTy0 = FTy->getParamType(0);
    if (PTy0 != Arg0Val->getType()) {
      if (Arg0Ty->isArrayType())
        Arg0Val = EmitArrayToPointerDecay(Arg0).getPointer();
      else
        Arg0Val = Builder.CreatePointerCast(Arg0Val, PTy0);
    }
    auto Arg1 = EmitScalarExpr(E->getArg(1));
    auto PTy1 = FTy->getParamType(1);
    if (PTy1 != Arg1->getType())
      Arg1 = Builder.CreateTruncOrBitCast(Arg1, PTy1);
    return RValue::get(Builder.CreateCall(F, {Arg0Val, Arg1}));
  }

  case Builtin::BI__xray_typedevent: {
    // TODO: There should be a way to always emit events even if the current
    // function is not instrumented. Losing events in a stream can cripple
    // a trace.
    if (!ShouldXRayInstrumentFunction())
      return RValue::getIgnored();

    if (!CGM.getCodeGenOpts().XRayInstrumentationBundle.has(
            XRayInstrKind::Typed))
      return RValue::getIgnored();

    if (const auto *XRayAttr = CurFuncDecl->getAttr<XRayInstrumentAttr>())
      if (XRayAttr->neverXRayInstrument() && !AlwaysEmitXRayTypedEvents())
        return RValue::getIgnored();

    Function *F = CGM.getIntrinsic(Intrinsic::xray_typedevent);
    auto FTy = F->getFunctionType();
    auto Arg0 = EmitScalarExpr(E->getArg(0));
    auto PTy0 = FTy->getParamType(0);
    if (PTy0 != Arg0->getType())
      Arg0 = Builder.CreateTruncOrBitCast(Arg0, PTy0);
    auto Arg1 = E->getArg(1);
    auto Arg1Val = EmitScalarExpr(Arg1);
    auto Arg1Ty = Arg1->getType();
    auto PTy1 = FTy->getParamType(1);
    if (PTy1 != Arg1Val->getType()) {
      if (Arg1Ty->isArrayType())
        Arg1Val = EmitArrayToPointerDecay(Arg1).getPointer();
      else
        Arg1Val = Builder.CreatePointerCast(Arg1Val, PTy1);
    }
    auto Arg2 = EmitScalarExpr(E->getArg(2));
    auto PTy2 = FTy->getParamType(2);
    if (PTy2 != Arg2->getType())
      Arg2 = Builder.CreateTruncOrBitCast(Arg2, PTy2);
    return RValue::get(Builder.CreateCall(F, {Arg0, Arg1Val, Arg2}));
  }

  case Builtin::BI__builtin_ms_va_start:
  case Builtin::BI__builtin_ms_va_end:
    return RValue::get(
        EmitVAStartEnd(EmitMSVAListRef(E->getArg(0)).getPointer(),
                       BuiltinID == Builtin::BI__builtin_ms_va_start));

  case Builtin::BI__builtin_ms_va_copy: {
    // Lower this manually. We can't reliably determine whether or not any
    // given va_copy() is for a Win64 va_list from the calling convention
    // alone, because it's legal to do this from a System V ABI function.
    // With opaque pointer types, we won't have enough information in LLVM
    // IR to determine this from the argument types, either. Best to do it
    // now, while we have enough information.
    Address DestAddr = EmitMSVAListRef(E->getArg(0));
    Address SrcAddr = EmitMSVAListRef(E->getArg(1));

    llvm::Type *BPP = Int8PtrPtrTy;

    DestAddr = Address(Builder.CreateBitCast(DestAddr.getPointer(), BPP, "cp"),
                       DestAddr.getAlignment());
    SrcAddr = Address(Builder.CreateBitCast(SrcAddr.getPointer(), BPP, "ap"),
                      SrcAddr.getAlignment());

    Value *ArgPtr = Builder.CreateLoad(SrcAddr, "ap.val");
    return RValue::get(Builder.CreateStore(ArgPtr, DestAddr));
  }
  }

  // If this is an alias for a lib function (e.g. __builtin_sin), emit
  // the call using the normal call path, but using the unmangled
  // version of the function name.
  if (getContext().BuiltinInfo.isLibFunction(BuiltinID))
    return emitLibraryCall(*this, FD, E,
                           CGM.getBuiltinLibFunction(FD, BuiltinID));

  // If this is a predefined lib function (e.g. malloc), emit the call
  // using exactly the normal call path.
  if (getContext().BuiltinInfo.isPredefinedLibFunction(BuiltinID))
    return emitLibraryCall(*this, FD, E,
                      cast<llvm::Constant>(EmitScalarExpr(E->getCallee())));

  // Check that a call to a target specific builtin has the correct target
  // features.
  // This is down here to avoid non-target specific builtins, however, if
  // generic builtins start to require generic target features then we
  // can move this up to the beginning of the function.
  checkTargetFeatures(E, FD);

  if (unsigned VectorWidth = getContext().BuiltinInfo.getRequiredVectorWidth(BuiltinID))
    LargestVectorWidth = std::max(LargestVectorWidth, VectorWidth);

  // See if we have a target specific intrinsic.
  const char *Name = getContext().BuiltinInfo.getName(BuiltinID);
  Intrinsic::ID IntrinsicID = Intrinsic::not_intrinsic;
  StringRef Prefix =
      llvm::Triple::getArchTypePrefix(getTarget().getTriple().getArch());
  if (!Prefix.empty()) {
    IntrinsicID = Intrinsic::getIntrinsicForGCCBuiltin(Prefix.data(), Name);
    // NOTE we don't need to perform a compatibility flag check here since the
    // intrinsics are declared in Builtins*.def via LANGBUILTIN which filter the
    // MS builtins via ALL_MS_LANGUAGES and are filtered earlier.
    if (IntrinsicID == Intrinsic::not_intrinsic)
      IntrinsicID = Intrinsic::getIntrinsicForMSBuiltin(Prefix.data(), Name);
  }

  if (IntrinsicID != Intrinsic::not_intrinsic) {
    SmallVector<Value*, 16> Args;

    // Find out if any arguments are required to be integer constant
    // expressions.
    unsigned ICEArguments = 0;
    ASTContext::GetBuiltinTypeError Error;
    getContext().GetBuiltinType(BuiltinID, Error, &ICEArguments);
    assert(Error == ASTContext::GE_None && "Should not codegen an error");

    Function *F = CGM.getIntrinsic(IntrinsicID);
    llvm::FunctionType *FTy = F->getFunctionType();

    for (unsigned i = 0, e = E->getNumArgs(); i != e; ++i) {
      Value *ArgValue;
      // If this is a normal argument, just emit it as a scalar.
      if ((ICEArguments & (1 << i)) == 0) {
        ArgValue = EmitScalarExpr(E->getArg(i));
      } else {
        // If this is required to be a constant, constant fold it so that we
        // know that the generated intrinsic gets a ConstantInt.
        llvm::APSInt Result;
        bool IsConst = E->getArg(i)->isIntegerConstantExpr(Result,getContext());
        assert(IsConst && "Constant arg isn't actually constant?");
        (void)IsConst;
        ArgValue = llvm::ConstantInt::get(getLLVMContext(), Result);
      }

      // If the intrinsic arg type is different from the builtin arg type
      // we need to do a bit cast.
      llvm::Type *PTy = FTy->getParamType(i);
      if (PTy != ArgValue->getType()) {
        // XXX - vector of pointers?
        if (auto *PtrTy = dyn_cast<llvm::PointerType>(PTy)) {
          if (PtrTy->getAddressSpace() !=
              ArgValue->getType()->getPointerAddressSpace()) {
            ArgValue = Builder.CreateAddrSpaceCast(
              ArgValue,
              ArgValue->getType()->getPointerTo(PtrTy->getAddressSpace()));
          }
        }

        assert(PTy->canLosslesslyBitCastTo(FTy->getParamType(i)) &&
               "Must be able to losslessly bit cast to param");
        ArgValue = Builder.CreateBitCast(ArgValue, PTy);
      }

      Args.push_back(ArgValue);
    }

    Value *V = Builder.CreateCall(F, Args);
    QualType BuiltinRetType = E->getType();

    llvm::Type *RetTy = VoidTy;
    if (!BuiltinRetType->isVoidType())
      RetTy = ConvertType(BuiltinRetType);

    if (RetTy != V->getType()) {
      // XXX - vector of pointers?
      if (auto *PtrTy = dyn_cast<llvm::PointerType>(RetTy)) {
        if (PtrTy->getAddressSpace() != V->getType()->getPointerAddressSpace()) {
          V = Builder.CreateAddrSpaceCast(
            V, V->getType()->getPointerTo(PtrTy->getAddressSpace()));
        }
      }

      assert(V->getType()->canLosslesslyBitCastTo(RetTy) &&
             "Must be able to losslessly bit cast result type");
      V = Builder.CreateBitCast(V, RetTy);
    }

    return RValue::get(V);
  }

  // See if we have a target specific builtin that needs to be lowered.
  if (Value *V = EmitTargetBuiltinExpr(BuiltinID, E))
    return RValue::get(V);

  ErrorUnsupported(E, "builtin function");

  // Unknown builtin, for now just dump it out and return undef.
  return GetUndefRValue(E->getType());
}

static Value *EmitTargetArchBuiltinExpr(CodeGenFunction *CGF,
                                        unsigned BuiltinID, const CallExpr *E,
                                        llvm::Triple::ArchType Arch) {
  switch (Arch) {
  case llvm::Triple::arm:
  case llvm::Triple::armeb:
  case llvm::Triple::thumb:
  case llvm::Triple::thumbeb:
    return CGF->EmitARMBuiltinExpr(BuiltinID, E, Arch);
  case llvm::Triple::aarch64:
  case llvm::Triple::aarch64_be:
    return CGF->EmitAArch64BuiltinExpr(BuiltinID, E, Arch);
  case llvm::Triple::x86:
  case llvm::Triple::x86_64:
    return CGF->EmitX86BuiltinExpr(BuiltinID, E);
  case llvm::Triple::ppc:
  case llvm::Triple::ppc64:
  case llvm::Triple::ppc64le:
    return CGF->EmitPPCBuiltinExpr(BuiltinID, E);
  case llvm::Triple::r600:
  case llvm::Triple::amdgcn:
    return CGF->EmitAMDGPUBuiltinExpr(BuiltinID, E);
  case llvm::Triple::systemz:
    return CGF->EmitSystemZBuiltinExpr(BuiltinID, E);
  case llvm::Triple::nvptx:
  case llvm::Triple::nvptx64:
    return CGF->EmitNVPTXBuiltinExpr(BuiltinID, E);
  case llvm::Triple::wasm32:
  case llvm::Triple::wasm64:
    return CGF->EmitWebAssemblyBuiltinExpr(BuiltinID, E);
  case llvm::Triple::hexagon:
    return CGF->EmitHexagonBuiltinExpr(BuiltinID, E);
  default:
    return nullptr;
  }
}

Value *CodeGenFunction::EmitTargetBuiltinExpr(unsigned BuiltinID,
                                              const CallExpr *E) {
  if (getContext().BuiltinInfo.isAuxBuiltinID(BuiltinID)) {
    assert(getContext().getAuxTargetInfo() && "Missing aux target info");
    return EmitTargetArchBuiltinExpr(
        this, getContext().BuiltinInfo.getAuxBuiltinID(BuiltinID), E,
        getContext().getAuxTargetInfo()->getTriple().getArch());
  }

  return EmitTargetArchBuiltinExpr(this, BuiltinID, E,
                                   getTarget().getTriple().getArch());
}

static llvm::VectorType *GetNeonType(CodeGenFunction *CGF,
                                     NeonTypeFlags TypeFlags,
                                     bool HasLegalHalfType=true,
                                     bool V1Ty=false) {
  int IsQuad = TypeFlags.isQuad();
  switch (TypeFlags.getEltType()) {
  case NeonTypeFlags::Int8:
  case NeonTypeFlags::Poly8:
    return llvm::VectorType::get(CGF->Int8Ty, V1Ty ? 1 : (8 << IsQuad));
  case NeonTypeFlags::Int16:
  case NeonTypeFlags::Poly16:
    return llvm::VectorType::get(CGF->Int16Ty, V1Ty ? 1 : (4 << IsQuad));
  case NeonTypeFlags::Float16:
    if (HasLegalHalfType)
      return llvm::VectorType::get(CGF->HalfTy, V1Ty ? 1 : (4 << IsQuad));
    else
      return llvm::VectorType::get(CGF->Int16Ty, V1Ty ? 1 : (4 << IsQuad));
  case NeonTypeFlags::Int32:
    return llvm::VectorType::get(CGF->Int32Ty, V1Ty ? 1 : (2 << IsQuad));
  case NeonTypeFlags::Int64:
  case NeonTypeFlags::Poly64:
    return llvm::VectorType::get(CGF->Int64Ty, V1Ty ? 1 : (1 << IsQuad));
  case NeonTypeFlags::Poly128:
    // FIXME: i128 and f128 doesn't get fully support in Clang and llvm.
    // There is a lot of i128 and f128 API missing.
    // so we use v16i8 to represent poly128 and get pattern matched.
    return llvm::VectorType::get(CGF->Int8Ty, 16);
  case NeonTypeFlags::Float32:
    return llvm::VectorType::get(CGF->FloatTy, V1Ty ? 1 : (2 << IsQuad));
  case NeonTypeFlags::Float64:
    return llvm::VectorType::get(CGF->DoubleTy, V1Ty ? 1 : (1 << IsQuad));
  }
  llvm_unreachable("Unknown vector element type!");
}

static llvm::VectorType *GetFloatNeonType(CodeGenFunction *CGF,
                                          NeonTypeFlags IntTypeFlags) {
  int IsQuad = IntTypeFlags.isQuad();
  switch (IntTypeFlags.getEltType()) {
  case NeonTypeFlags::Int16:
    return llvm::VectorType::get(CGF->HalfTy, (4 << IsQuad));
  case NeonTypeFlags::Int32:
    return llvm::VectorType::get(CGF->FloatTy, (2 << IsQuad));
  case NeonTypeFlags::Int64:
    return llvm::VectorType::get(CGF->DoubleTy, (1 << IsQuad));
  default:
    llvm_unreachable("Type can't be converted to floating-point!");
  }
}

Value *CodeGenFunction::EmitNeonSplat(Value *V, Constant *C) {
  unsigned nElts = V->getType()->getVectorNumElements();
  Value* SV = llvm::ConstantVector::getSplat(nElts, C);
  return Builder.CreateShuffleVector(V, V, SV, "lane");
}

Value *CodeGenFunction::EmitNeonCall(Function *F, SmallVectorImpl<Value*> &Ops,
                                     const char *name,
                                     unsigned shift, bool rightshift) {
  unsigned j = 0;
  for (Function::const_arg_iterator ai = F->arg_begin(), ae = F->arg_end();
       ai != ae; ++ai, ++j)
    if (shift > 0 && shift == j)
      Ops[j] = EmitNeonShiftVector(Ops[j], ai->getType(), rightshift);
    else
      Ops[j] = Builder.CreateBitCast(Ops[j], ai->getType(), name);

  return Builder.CreateCall(F, Ops, name);
}

Value *CodeGenFunction::EmitNeonShiftVector(Value *V, llvm::Type *Ty,
                                            bool neg) {
  int SV = cast<ConstantInt>(V)->getSExtValue();
  return ConstantInt::get(Ty, neg ? -SV : SV);
}

// Right-shift a vector by a constant.
Value *CodeGenFunction::EmitNeonRShiftImm(Value *Vec, Value *Shift,
                                          llvm::Type *Ty, bool usgn,
                                          const char *name) {
  llvm::VectorType *VTy = cast<llvm::VectorType>(Ty);

  int ShiftAmt = cast<ConstantInt>(Shift)->getSExtValue();
  int EltSize = VTy->getScalarSizeInBits();

  Vec = Builder.CreateBitCast(Vec, Ty);

  // lshr/ashr are undefined when the shift amount is equal to the vector
  // element size.
  if (ShiftAmt == EltSize) {
    if (usgn) {
      // Right-shifting an unsigned value by its size yields 0.
      return llvm::ConstantAggregateZero::get(VTy);
    } else {
      // Right-shifting a signed value by its size is equivalent
      // to a shift of size-1.
      --ShiftAmt;
      Shift = ConstantInt::get(VTy->getElementType(), ShiftAmt);
    }
  }

  Shift = EmitNeonShiftVector(Shift, Ty, false);
  if (usgn)
    return Builder.CreateLShr(Vec, Shift, name);
  else
    return Builder.CreateAShr(Vec, Shift, name);
}

enum {
  AddRetType = (1 << 0),
  Add1ArgType = (1 << 1),
  Add2ArgTypes = (1 << 2),

  VectorizeRetType = (1 << 3),
  VectorizeArgTypes = (1 << 4),

  InventFloatType = (1 << 5),
  UnsignedAlts = (1 << 6),

  Use64BitVectors = (1 << 7),
  Use128BitVectors = (1 << 8),

  Vectorize1ArgType = Add1ArgType | VectorizeArgTypes,
  VectorRet = AddRetType | VectorizeRetType,
  VectorRetGetArgs01 =
      AddRetType | Add2ArgTypes | VectorizeRetType | VectorizeArgTypes,
  FpCmpzModifiers =
      AddRetType | VectorizeRetType | Add1ArgType | InventFloatType
};

namespace {
struct NeonIntrinsicInfo {
  const char *NameHint;
  unsigned BuiltinID;
  unsigned LLVMIntrinsic;
  unsigned AltLLVMIntrinsic;
  unsigned TypeModifier;

  bool operator<(unsigned RHSBuiltinID) const {
    return BuiltinID < RHSBuiltinID;
  }
  bool operator<(const NeonIntrinsicInfo &TE) const {
    return BuiltinID < TE.BuiltinID;
  }
};
} // end anonymous namespace

#define NEONMAP0(NameBase) \
  { #NameBase, NEON::BI__builtin_neon_ ## NameBase, 0, 0, 0 }

#define NEONMAP1(NameBase, LLVMIntrinsic, TypeModifier) \
  { #NameBase, NEON:: BI__builtin_neon_ ## NameBase, \
      Intrinsic::LLVMIntrinsic, 0, TypeModifier }

#define NEONMAP2(NameBase, LLVMIntrinsic, AltLLVMIntrinsic, TypeModifier) \
  { #NameBase, NEON:: BI__builtin_neon_ ## NameBase, \
      Intrinsic::LLVMIntrinsic, Intrinsic::AltLLVMIntrinsic, \
      TypeModifier }

static const NeonIntrinsicInfo ARMSIMDIntrinsicMap [] = {
  NEONMAP2(vabd_v, arm_neon_vabdu, arm_neon_vabds, Add1ArgType | UnsignedAlts),
  NEONMAP2(vabdq_v, arm_neon_vabdu, arm_neon_vabds, Add1ArgType | UnsignedAlts),
  NEONMAP1(vabs_v, arm_neon_vabs, 0),
  NEONMAP1(vabsq_v, arm_neon_vabs, 0),
  NEONMAP0(vaddhn_v),
  NEONMAP1(vaesdq_v, arm_neon_aesd, 0),
  NEONMAP1(vaeseq_v, arm_neon_aese, 0),
  NEONMAP1(vaesimcq_v, arm_neon_aesimc, 0),
  NEONMAP1(vaesmcq_v, arm_neon_aesmc, 0),
  NEONMAP1(vbsl_v, arm_neon_vbsl, AddRetType),
  NEONMAP1(vbslq_v, arm_neon_vbsl, AddRetType),
  NEONMAP1(vcage_v, arm_neon_vacge, 0),
  NEONMAP1(vcageq_v, arm_neon_vacge, 0),
  NEONMAP1(vcagt_v, arm_neon_vacgt, 0),
  NEONMAP1(vcagtq_v, arm_neon_vacgt, 0),
  NEONMAP1(vcale_v, arm_neon_vacge, 0),
  NEONMAP1(vcaleq_v, arm_neon_vacge, 0),
  NEONMAP1(vcalt_v, arm_neon_vacgt, 0),
  NEONMAP1(vcaltq_v, arm_neon_vacgt, 0),
  NEONMAP0(vceqz_v),
  NEONMAP0(vceqzq_v),
  NEONMAP0(vcgez_v),
  NEONMAP0(vcgezq_v),
  NEONMAP0(vcgtz_v),
  NEONMAP0(vcgtzq_v),
  NEONMAP0(vclez_v),
  NEONMAP0(vclezq_v),
  NEONMAP1(vcls_v, arm_neon_vcls, Add1ArgType),
  NEONMAP1(vclsq_v, arm_neon_vcls, Add1ArgType),
  NEONMAP0(vcltz_v),
  NEONMAP0(vcltzq_v),
  NEONMAP1(vclz_v, ctlz, Add1ArgType),
  NEONMAP1(vclzq_v, ctlz, Add1ArgType),
  NEONMAP1(vcnt_v, ctpop, Add1ArgType),
  NEONMAP1(vcntq_v, ctpop, Add1ArgType),
  NEONMAP1(vcvt_f16_f32, arm_neon_vcvtfp2hf, 0),
  NEONMAP0(vcvt_f16_v),
  NEONMAP1(vcvt_f32_f16, arm_neon_vcvthf2fp, 0),
  NEONMAP0(vcvt_f32_v),
  NEONMAP2(vcvt_n_f16_v, arm_neon_vcvtfxu2fp, arm_neon_vcvtfxs2fp, 0),
  NEONMAP2(vcvt_n_f32_v, arm_neon_vcvtfxu2fp, arm_neon_vcvtfxs2fp, 0),
  NEONMAP1(vcvt_n_s16_v, arm_neon_vcvtfp2fxs, 0),
  NEONMAP1(vcvt_n_s32_v, arm_neon_vcvtfp2fxs, 0),
  NEONMAP1(vcvt_n_s64_v, arm_neon_vcvtfp2fxs, 0),
  NEONMAP1(vcvt_n_u16_v, arm_neon_vcvtfp2fxu, 0),
  NEONMAP1(vcvt_n_u32_v, arm_neon_vcvtfp2fxu, 0),
  NEONMAP1(vcvt_n_u64_v, arm_neon_vcvtfp2fxu, 0),
  NEONMAP0(vcvt_s16_v),
  NEONMAP0(vcvt_s32_v),
  NEONMAP0(vcvt_s64_v),
  NEONMAP0(vcvt_u16_v),
  NEONMAP0(vcvt_u32_v),
  NEONMAP0(vcvt_u64_v),
  NEONMAP1(vcvta_s16_v, arm_neon_vcvtas, 0),
  NEONMAP1(vcvta_s32_v, arm_neon_vcvtas, 0),
  NEONMAP1(vcvta_s64_v, arm_neon_vcvtas, 0),
  NEONMAP1(vcvta_u16_v, arm_neon_vcvtau, 0),
  NEONMAP1(vcvta_u32_v, arm_neon_vcvtau, 0),
  NEONMAP1(vcvta_u64_v, arm_neon_vcvtau, 0),
  NEONMAP1(vcvtaq_s16_v, arm_neon_vcvtas, 0),
  NEONMAP1(vcvtaq_s32_v, arm_neon_vcvtas, 0),
  NEONMAP1(vcvtaq_s64_v, arm_neon_vcvtas, 0),
  NEONMAP1(vcvtaq_u16_v, arm_neon_vcvtau, 0),
  NEONMAP1(vcvtaq_u32_v, arm_neon_vcvtau, 0),
  NEONMAP1(vcvtaq_u64_v, arm_neon_vcvtau, 0),
  NEONMAP1(vcvtm_s16_v, arm_neon_vcvtms, 0),
  NEONMAP1(vcvtm_s32_v, arm_neon_vcvtms, 0),
  NEONMAP1(vcvtm_s64_v, arm_neon_vcvtms, 0),
  NEONMAP1(vcvtm_u16_v, arm_neon_vcvtmu, 0),
  NEONMAP1(vcvtm_u32_v, arm_neon_vcvtmu, 0),
  NEONMAP1(vcvtm_u64_v, arm_neon_vcvtmu, 0),
  NEONMAP1(vcvtmq_s16_v, arm_neon_vcvtms, 0),
  NEONMAP1(vcvtmq_s32_v, arm_neon_vcvtms, 0),
  NEONMAP1(vcvtmq_s64_v, arm_neon_vcvtms, 0),
  NEONMAP1(vcvtmq_u16_v, arm_neon_vcvtmu, 0),
  NEONMAP1(vcvtmq_u32_v, arm_neon_vcvtmu, 0),
  NEONMAP1(vcvtmq_u64_v, arm_neon_vcvtmu, 0),
  NEONMAP1(vcvtn_s16_v, arm_neon_vcvtns, 0),
  NEONMAP1(vcvtn_s32_v, arm_neon_vcvtns, 0),
  NEONMAP1(vcvtn_s64_v, arm_neon_vcvtns, 0),
  NEONMAP1(vcvtn_u16_v, arm_neon_vcvtnu, 0),
  NEONMAP1(vcvtn_u32_v, arm_neon_vcvtnu, 0),
  NEONMAP1(vcvtn_u64_v, arm_neon_vcvtnu, 0),
  NEONMAP1(vcvtnq_s16_v, arm_neon_vcvtns, 0),
  NEONMAP1(vcvtnq_s32_v, arm_neon_vcvtns, 0),
  NEONMAP1(vcvtnq_s64_v, arm_neon_vcvtns, 0),
  NEONMAP1(vcvtnq_u16_v, arm_neon_vcvtnu, 0),
  NEONMAP1(vcvtnq_u32_v, arm_neon_vcvtnu, 0),
  NEONMAP1(vcvtnq_u64_v, arm_neon_vcvtnu, 0),
  NEONMAP1(vcvtp_s16_v, arm_neon_vcvtps, 0),
  NEONMAP1(vcvtp_s32_v, arm_neon_vcvtps, 0),
  NEONMAP1(vcvtp_s64_v, arm_neon_vcvtps, 0),
  NEONMAP1(vcvtp_u16_v, arm_neon_vcvtpu, 0),
  NEONMAP1(vcvtp_u32_v, arm_neon_vcvtpu, 0),
  NEONMAP1(vcvtp_u64_v, arm_neon_vcvtpu, 0),
  NEONMAP1(vcvtpq_s16_v, arm_neon_vcvtps, 0),
  NEONMAP1(vcvtpq_s32_v, arm_neon_vcvtps, 0),
  NEONMAP1(vcvtpq_s64_v, arm_neon_vcvtps, 0),
  NEONMAP1(vcvtpq_u16_v, arm_neon_vcvtpu, 0),
  NEONMAP1(vcvtpq_u32_v, arm_neon_vcvtpu, 0),
  NEONMAP1(vcvtpq_u64_v, arm_neon_vcvtpu, 0),
  NEONMAP0(vcvtq_f16_v),
  NEONMAP0(vcvtq_f32_v),
  NEONMAP2(vcvtq_n_f16_v, arm_neon_vcvtfxu2fp, arm_neon_vcvtfxs2fp, 0),
  NEONMAP2(vcvtq_n_f32_v, arm_neon_vcvtfxu2fp, arm_neon_vcvtfxs2fp, 0),
  NEONMAP1(vcvtq_n_s16_v, arm_neon_vcvtfp2fxs, 0),
  NEONMAP1(vcvtq_n_s32_v, arm_neon_vcvtfp2fxs, 0),
  NEONMAP1(vcvtq_n_s64_v, arm_neon_vcvtfp2fxs, 0),
  NEONMAP1(vcvtq_n_u16_v, arm_neon_vcvtfp2fxu, 0),
  NEONMAP1(vcvtq_n_u32_v, arm_neon_vcvtfp2fxu, 0),
  NEONMAP1(vcvtq_n_u64_v, arm_neon_vcvtfp2fxu, 0),
  NEONMAP0(vcvtq_s16_v),
  NEONMAP0(vcvtq_s32_v),
  NEONMAP0(vcvtq_s64_v),
  NEONMAP0(vcvtq_u16_v),
  NEONMAP0(vcvtq_u32_v),
  NEONMAP0(vcvtq_u64_v),
  NEONMAP2(vdot_v, arm_neon_udot, arm_neon_sdot, 0),
  NEONMAP2(vdotq_v, arm_neon_udot, arm_neon_sdot, 0),
  NEONMAP0(vext_v),
  NEONMAP0(vextq_v),
  NEONMAP0(vfma_v),
  NEONMAP0(vfmaq_v),
  NEONMAP2(vhadd_v, arm_neon_vhaddu, arm_neon_vhadds, Add1ArgType | UnsignedAlts),
  NEONMAP2(vhaddq_v, arm_neon_vhaddu, arm_neon_vhadds, Add1ArgType | UnsignedAlts),
  NEONMAP2(vhsub_v, arm_neon_vhsubu, arm_neon_vhsubs, Add1ArgType | UnsignedAlts),
  NEONMAP2(vhsubq_v, arm_neon_vhsubu, arm_neon_vhsubs, Add1ArgType | UnsignedAlts),
  NEONMAP0(vld1_dup_v),
  NEONMAP1(vld1_v, arm_neon_vld1, 0),
  NEONMAP1(vld1_x2_v, arm_neon_vld1x2, 0),
  NEONMAP1(vld1_x3_v, arm_neon_vld1x3, 0),
  NEONMAP1(vld1_x4_v, arm_neon_vld1x4, 0),
  NEONMAP0(vld1q_dup_v),
  NEONMAP1(vld1q_v, arm_neon_vld1, 0),
  NEONMAP1(vld1q_x2_v, arm_neon_vld1x2, 0),
  NEONMAP1(vld1q_x3_v, arm_neon_vld1x3, 0),
  NEONMAP1(vld1q_x4_v, arm_neon_vld1x4, 0),
  NEONMAP1(vld2_dup_v, arm_neon_vld2dup, 0),
  NEONMAP1(vld2_lane_v, arm_neon_vld2lane, 0),
  NEONMAP1(vld2_v, arm_neon_vld2, 0),
  NEONMAP1(vld2q_dup_v, arm_neon_vld2dup, 0),
  NEONMAP1(vld2q_lane_v, arm_neon_vld2lane, 0),
  NEONMAP1(vld2q_v, arm_neon_vld2, 0),
  NEONMAP1(vld3_dup_v, arm_neon_vld3dup, 0),
  NEONMAP1(vld3_lane_v, arm_neon_vld3lane, 0),
  NEONMAP1(vld3_v, arm_neon_vld3, 0),
  NEONMAP1(vld3q_dup_v, arm_neon_vld3dup, 0),
  NEONMAP1(vld3q_lane_v, arm_neon_vld3lane, 0),
  NEONMAP1(vld3q_v, arm_neon_vld3, 0),
  NEONMAP1(vld4_dup_v, arm_neon_vld4dup, 0),
  NEONMAP1(vld4_lane_v, arm_neon_vld4lane, 0),
  NEONMAP1(vld4_v, arm_neon_vld4, 0),
  NEONMAP1(vld4q_dup_v, arm_neon_vld4dup, 0),
  NEONMAP1(vld4q_lane_v, arm_neon_vld4lane, 0),
  NEONMAP1(vld4q_v, arm_neon_vld4, 0),
  NEONMAP2(vmax_v, arm_neon_vmaxu, arm_neon_vmaxs, Add1ArgType | UnsignedAlts),
  NEONMAP1(vmaxnm_v, arm_neon_vmaxnm, Add1ArgType),
  NEONMAP1(vmaxnmq_v, arm_neon_vmaxnm, Add1ArgType),
  NEONMAP2(vmaxq_v, arm_neon_vmaxu, arm_neon_vmaxs, Add1ArgType | UnsignedAlts),
  NEONMAP2(vmin_v, arm_neon_vminu, arm_neon_vmins, Add1ArgType | UnsignedAlts),
  NEONMAP1(vminnm_v, arm_neon_vminnm, Add1ArgType),
  NEONMAP1(vminnmq_v, arm_neon_vminnm, Add1ArgType),
  NEONMAP2(vminq_v, arm_neon_vminu, arm_neon_vmins, Add1ArgType | UnsignedAlts),
  NEONMAP0(vmovl_v),
  NEONMAP0(vmovn_v),
  NEONMAP1(vmul_v, arm_neon_vmulp, Add1ArgType),
  NEONMAP0(vmull_v),
  NEONMAP1(vmulq_v, arm_neon_vmulp, Add1ArgType),
  NEONMAP2(vpadal_v, arm_neon_vpadalu, arm_neon_vpadals, UnsignedAlts),
  NEONMAP2(vpadalq_v, arm_neon_vpadalu, arm_neon_vpadals, UnsignedAlts),
  NEONMAP1(vpadd_v, arm_neon_vpadd, Add1ArgType),
  NEONMAP2(vpaddl_v, arm_neon_vpaddlu, arm_neon_vpaddls, UnsignedAlts),
  NEONMAP2(vpaddlq_v, arm_neon_vpaddlu, arm_neon_vpaddls, UnsignedAlts),
  NEONMAP1(vpaddq_v, arm_neon_vpadd, Add1ArgType),
  NEONMAP2(vpmax_v, arm_neon_vpmaxu, arm_neon_vpmaxs, Add1ArgType | UnsignedAlts),
  NEONMAP2(vpmin_v, arm_neon_vpminu, arm_neon_vpmins, Add1ArgType | UnsignedAlts),
  NEONMAP1(vqabs_v, arm_neon_vqabs, Add1ArgType),
  NEONMAP1(vqabsq_v, arm_neon_vqabs, Add1ArgType),
  NEONMAP2(vqadd_v, arm_neon_vqaddu, arm_neon_vqadds, Add1ArgType | UnsignedAlts),
  NEONMAP2(vqaddq_v, arm_neon_vqaddu, arm_neon_vqadds, Add1ArgType | UnsignedAlts),
  NEONMAP2(vqdmlal_v, arm_neon_vqdmull, arm_neon_vqadds, 0),
  NEONMAP2(vqdmlsl_v, arm_neon_vqdmull, arm_neon_vqsubs, 0),
  NEONMAP1(vqdmulh_v, arm_neon_vqdmulh, Add1ArgType),
  NEONMAP1(vqdmulhq_v, arm_neon_vqdmulh, Add1ArgType),
  NEONMAP1(vqdmull_v, arm_neon_vqdmull, Add1ArgType),
  NEONMAP2(vqmovn_v, arm_neon_vqmovnu, arm_neon_vqmovns, Add1ArgType | UnsignedAlts),
  NEONMAP1(vqmovun_v, arm_neon_vqmovnsu, Add1ArgType),
  NEONMAP1(vqneg_v, arm_neon_vqneg, Add1ArgType),
  NEONMAP1(vqnegq_v, arm_neon_vqneg, Add1ArgType),
  NEONMAP1(vqrdmulh_v, arm_neon_vqrdmulh, Add1ArgType),
  NEONMAP1(vqrdmulhq_v, arm_neon_vqrdmulh, Add1ArgType),
  NEONMAP2(vqrshl_v, arm_neon_vqrshiftu, arm_neon_vqrshifts, Add1ArgType | UnsignedAlts),
  NEONMAP2(vqrshlq_v, arm_neon_vqrshiftu, arm_neon_vqrshifts, Add1ArgType | UnsignedAlts),
  NEONMAP2(vqshl_n_v, arm_neon_vqshiftu, arm_neon_vqshifts, UnsignedAlts),
  NEONMAP2(vqshl_v, arm_neon_vqshiftu, arm_neon_vqshifts, Add1ArgType | UnsignedAlts),
  NEONMAP2(vqshlq_n_v, arm_neon_vqshiftu, arm_neon_vqshifts, UnsignedAlts),
  NEONMAP2(vqshlq_v, arm_neon_vqshiftu, arm_neon_vqshifts, Add1ArgType | UnsignedAlts),
  NEONMAP1(vqshlu_n_v, arm_neon_vqshiftsu, 0),
  NEONMAP1(vqshluq_n_v, arm_neon_vqshiftsu, 0),
  NEONMAP2(vqsub_v, arm_neon_vqsubu, arm_neon_vqsubs, Add1ArgType | UnsignedAlts),
  NEONMAP2(vqsubq_v, arm_neon_vqsubu, arm_neon_vqsubs, Add1ArgType | UnsignedAlts),
  NEONMAP1(vraddhn_v, arm_neon_vraddhn, Add1ArgType),
  NEONMAP2(vrecpe_v, arm_neon_vrecpe, arm_neon_vrecpe, 0),
  NEONMAP2(vrecpeq_v, arm_neon_vrecpe, arm_neon_vrecpe, 0),
  NEONMAP1(vrecps_v, arm_neon_vrecps, Add1ArgType),
  NEONMAP1(vrecpsq_v, arm_neon_vrecps, Add1ArgType),
  NEONMAP2(vrhadd_v, arm_neon_vrhaddu, arm_neon_vrhadds, Add1ArgType | UnsignedAlts),
  NEONMAP2(vrhaddq_v, arm_neon_vrhaddu, arm_neon_vrhadds, Add1ArgType | UnsignedAlts),
  NEONMAP1(vrnd_v, arm_neon_vrintz, Add1ArgType),
  NEONMAP1(vrnda_v, arm_neon_vrinta, Add1ArgType),
  NEONMAP1(vrndaq_v, arm_neon_vrinta, Add1ArgType),
  NEONMAP0(vrndi_v),
  NEONMAP0(vrndiq_v),
  NEONMAP1(vrndm_v, arm_neon_vrintm, Add1ArgType),
  NEONMAP1(vrndmq_v, arm_neon_vrintm, Add1ArgType),
  NEONMAP1(vrndn_v, arm_neon_vrintn, Add1ArgType),
  NEONMAP1(vrndnq_v, arm_neon_vrintn, Add1ArgType),
  NEONMAP1(vrndp_v, arm_neon_vrintp, Add1ArgType),
  NEONMAP1(vrndpq_v, arm_neon_vrintp, Add1ArgType),
  NEONMAP1(vrndq_v, arm_neon_vrintz, Add1ArgType),
  NEONMAP1(vrndx_v, arm_neon_vrintx, Add1ArgType),
  NEONMAP1(vrndxq_v, arm_neon_vrintx, Add1ArgType),
  NEONMAP2(vrshl_v, arm_neon_vrshiftu, arm_neon_vrshifts, Add1ArgType | UnsignedAlts),
  NEONMAP2(vrshlq_v, arm_neon_vrshiftu, arm_neon_vrshifts, Add1ArgType | UnsignedAlts),
  NEONMAP2(vrshr_n_v, arm_neon_vrshiftu, arm_neon_vrshifts, UnsignedAlts),
  NEONMAP2(vrshrq_n_v, arm_neon_vrshiftu, arm_neon_vrshifts, UnsignedAlts),
  NEONMAP2(vrsqrte_v, arm_neon_vrsqrte, arm_neon_vrsqrte, 0),
  NEONMAP2(vrsqrteq_v, arm_neon_vrsqrte, arm_neon_vrsqrte, 0),
  NEONMAP1(vrsqrts_v, arm_neon_vrsqrts, Add1ArgType),
  NEONMAP1(vrsqrtsq_v, arm_neon_vrsqrts, Add1ArgType),
  NEONMAP1(vrsubhn_v, arm_neon_vrsubhn, Add1ArgType),
  NEONMAP1(vsha1su0q_v, arm_neon_sha1su0, 0),
  NEONMAP1(vsha1su1q_v, arm_neon_sha1su1, 0),
  NEONMAP1(vsha256h2q_v, arm_neon_sha256h2, 0),
  NEONMAP1(vsha256hq_v, arm_neon_sha256h, 0),
  NEONMAP1(vsha256su0q_v, arm_neon_sha256su0, 0),
  NEONMAP1(vsha256su1q_v, arm_neon_sha256su1, 0),
  NEONMAP0(vshl_n_v),
  NEONMAP2(vshl_v, arm_neon_vshiftu, arm_neon_vshifts, Add1ArgType | UnsignedAlts),
  NEONMAP0(vshll_n_v),
  NEONMAP0(vshlq_n_v),
  NEONMAP2(vshlq_v, arm_neon_vshiftu, arm_neon_vshifts, Add1ArgType | UnsignedAlts),
  NEONMAP0(vshr_n_v),
  NEONMAP0(vshrn_n_v),
  NEONMAP0(vshrq_n_v),
  NEONMAP1(vst1_v, arm_neon_vst1, 0),
  NEONMAP1(vst1_x2_v, arm_neon_vst1x2, 0),
  NEONMAP1(vst1_x3_v, arm_neon_vst1x3, 0),
  NEONMAP1(vst1_x4_v, arm_neon_vst1x4, 0),
  NEONMAP1(vst1q_v, arm_neon_vst1, 0),
  NEONMAP1(vst1q_x2_v, arm_neon_vst1x2, 0),
  NEONMAP1(vst1q_x3_v, arm_neon_vst1x3, 0),
  NEONMAP1(vst1q_x4_v, arm_neon_vst1x4, 0),
  NEONMAP1(vst2_lane_v, arm_neon_vst2lane, 0),
  NEONMAP1(vst2_v, arm_neon_vst2, 0),
  NEONMAP1(vst2q_lane_v, arm_neon_vst2lane, 0),
  NEONMAP1(vst2q_v, arm_neon_vst2, 0),
  NEONMAP1(vst3_lane_v, arm_neon_vst3lane, 0),
  NEONMAP1(vst3_v, arm_neon_vst3, 0),
  NEONMAP1(vst3q_lane_v, arm_neon_vst3lane, 0),
  NEONMAP1(vst3q_v, arm_neon_vst3, 0),
  NEONMAP1(vst4_lane_v, arm_neon_vst4lane, 0),
  NEONMAP1(vst4_v, arm_neon_vst4, 0),
  NEONMAP1(vst4q_lane_v, arm_neon_vst4lane, 0),
  NEONMAP1(vst4q_v, arm_neon_vst4, 0),
  NEONMAP0(vsubhn_v),
  NEONMAP0(vtrn_v),
  NEONMAP0(vtrnq_v),
  NEONMAP0(vtst_v),
  NEONMAP0(vtstq_v),
  NEONMAP0(vuzp_v),
  NEONMAP0(vuzpq_v),
  NEONMAP0(vzip_v),
  NEONMAP0(vzipq_v)
};

static const NeonIntrinsicInfo AArch64SIMDIntrinsicMap[] = {
  NEONMAP1(vabs_v, aarch64_neon_abs, 0),
  NEONMAP1(vabsq_v, aarch64_neon_abs, 0),
  NEONMAP0(vaddhn_v),
  NEONMAP1(vaesdq_v, aarch64_crypto_aesd, 0),
  NEONMAP1(vaeseq_v, aarch64_crypto_aese, 0),
  NEONMAP1(vaesimcq_v, aarch64_crypto_aesimc, 0),
  NEONMAP1(vaesmcq_v, aarch64_crypto_aesmc, 0),
  NEONMAP1(vcage_v, aarch64_neon_facge, 0),
  NEONMAP1(vcageq_v, aarch64_neon_facge, 0),
  NEONMAP1(vcagt_v, aarch64_neon_facgt, 0),
  NEONMAP1(vcagtq_v, aarch64_neon_facgt, 0),
  NEONMAP1(vcale_v, aarch64_neon_facge, 0),
  NEONMAP1(vcaleq_v, aarch64_neon_facge, 0),
  NEONMAP1(vcalt_v, aarch64_neon_facgt, 0),
  NEONMAP1(vcaltq_v, aarch64_neon_facgt, 0),
  NEONMAP0(vceqz_v),
  NEONMAP0(vceqzq_v),
  NEONMAP0(vcgez_v),
  NEONMAP0(vcgezq_v),
  NEONMAP0(vcgtz_v),
  NEONMAP0(vcgtzq_v),
  NEONMAP0(vclez_v),
  NEONMAP0(vclezq_v),
  NEONMAP1(vcls_v, aarch64_neon_cls, Add1ArgType),
  NEONMAP1(vclsq_v, aarch64_neon_cls, Add1ArgType),
  NEONMAP0(vcltz_v),
  NEONMAP0(vcltzq_v),
  NEONMAP1(vclz_v, ctlz, Add1ArgType),
  NEONMAP1(vclzq_v, ctlz, Add1ArgType),
  NEONMAP1(vcnt_v, ctpop, Add1ArgType),
  NEONMAP1(vcntq_v, ctpop, Add1ArgType),
  NEONMAP1(vcvt_f16_f32, aarch64_neon_vcvtfp2hf, 0),
  NEONMAP0(vcvt_f16_v),
  NEONMAP1(vcvt_f32_f16, aarch64_neon_vcvthf2fp, 0),
  NEONMAP0(vcvt_f32_v),
  NEONMAP2(vcvt_n_f16_v, aarch64_neon_vcvtfxu2fp, aarch64_neon_vcvtfxs2fp, 0),
  NEONMAP2(vcvt_n_f32_v, aarch64_neon_vcvtfxu2fp, aarch64_neon_vcvtfxs2fp, 0),
  NEONMAP2(vcvt_n_f64_v, aarch64_neon_vcvtfxu2fp, aarch64_neon_vcvtfxs2fp, 0),
  NEONMAP1(vcvt_n_s16_v, aarch64_neon_vcvtfp2fxs, 0),
  NEONMAP1(vcvt_n_s32_v, aarch64_neon_vcvtfp2fxs, 0),
  NEONMAP1(vcvt_n_s64_v, aarch64_neon_vcvtfp2fxs, 0),
  NEONMAP1(vcvt_n_u16_v, aarch64_neon_vcvtfp2fxu, 0),
  NEONMAP1(vcvt_n_u32_v, aarch64_neon_vcvtfp2fxu, 0),
  NEONMAP1(vcvt_n_u64_v, aarch64_neon_vcvtfp2fxu, 0),
  NEONMAP0(vcvtq_f16_v),
  NEONMAP0(vcvtq_f32_v),
  NEONMAP2(vcvtq_n_f16_v, aarch64_neon_vcvtfxu2fp, aarch64_neon_vcvtfxs2fp, 0),
  NEONMAP2(vcvtq_n_f32_v, aarch64_neon_vcvtfxu2fp, aarch64_neon_vcvtfxs2fp, 0),
  NEONMAP2(vcvtq_n_f64_v, aarch64_neon_vcvtfxu2fp, aarch64_neon_vcvtfxs2fp, 0),
  NEONMAP1(vcvtq_n_s16_v, aarch64_neon_vcvtfp2fxs, 0),
  NEONMAP1(vcvtq_n_s32_v, aarch64_neon_vcvtfp2fxs, 0),
  NEONMAP1(vcvtq_n_s64_v, aarch64_neon_vcvtfp2fxs, 0),
  NEONMAP1(vcvtq_n_u16_v, aarch64_neon_vcvtfp2fxu, 0),
  NEONMAP1(vcvtq_n_u32_v, aarch64_neon_vcvtfp2fxu, 0),
  NEONMAP1(vcvtq_n_u64_v, aarch64_neon_vcvtfp2fxu, 0),
  NEONMAP1(vcvtx_f32_v, aarch64_neon_fcvtxn, AddRetType | Add1ArgType),
  NEONMAP2(vdot_v, aarch64_neon_udot, aarch64_neon_sdot, 0),
  NEONMAP2(vdotq_v, aarch64_neon_udot, aarch64_neon_sdot, 0),
  NEONMAP0(vext_v),
  NEONMAP0(vextq_v),
  NEONMAP0(vfma_v),
  NEONMAP0(vfmaq_v),
  NEONMAP2(vhadd_v, aarch64_neon_uhadd, aarch64_neon_shadd, Add1ArgType | UnsignedAlts),
  NEONMAP2(vhaddq_v, aarch64_neon_uhadd, aarch64_neon_shadd, Add1ArgType | UnsignedAlts),
  NEONMAP2(vhsub_v, aarch64_neon_uhsub, aarch64_neon_shsub, Add1ArgType | UnsignedAlts),
  NEONMAP2(vhsubq_v, aarch64_neon_uhsub, aarch64_neon_shsub, Add1ArgType | UnsignedAlts),
  NEONMAP1(vld1_x2_v, aarch64_neon_ld1x2, 0),
  NEONMAP1(vld1_x3_v, aarch64_neon_ld1x3, 0),
  NEONMAP1(vld1_x4_v, aarch64_neon_ld1x4, 0),
  NEONMAP1(vld1q_x2_v, aarch64_neon_ld1x2, 0),
  NEONMAP1(vld1q_x3_v, aarch64_neon_ld1x3, 0),
  NEONMAP1(vld1q_x4_v, aarch64_neon_ld1x4, 0),
  NEONMAP0(vmovl_v),
  NEONMAP0(vmovn_v),
  NEONMAP1(vmul_v, aarch64_neon_pmul, Add1ArgType),
  NEONMAP1(vmulq_v, aarch64_neon_pmul, Add1ArgType),
  NEONMAP1(vpadd_v, aarch64_neon_addp, Add1ArgType),
  NEONMAP2(vpaddl_v, aarch64_neon_uaddlp, aarch64_neon_saddlp, UnsignedAlts),
  NEONMAP2(vpaddlq_v, aarch64_neon_uaddlp, aarch64_neon_saddlp, UnsignedAlts),
  NEONMAP1(vpaddq_v, aarch64_neon_addp, Add1ArgType),
  NEONMAP1(vqabs_v, aarch64_neon_sqabs, Add1ArgType),
  NEONMAP1(vqabsq_v, aarch64_neon_sqabs, Add1ArgType),
  NEONMAP2(vqadd_v, aarch64_neon_uqadd, aarch64_neon_sqadd, Add1ArgType | UnsignedAlts),
  NEONMAP2(vqaddq_v, aarch64_neon_uqadd, aarch64_neon_sqadd, Add1ArgType | UnsignedAlts),
  NEONMAP2(vqdmlal_v, aarch64_neon_sqdmull, aarch64_neon_sqadd, 0),
  NEONMAP2(vqdmlsl_v, aarch64_neon_sqdmull, aarch64_neon_sqsub, 0),
  NEONMAP1(vqdmulh_v, aarch64_neon_sqdmulh, Add1ArgType),
  NEONMAP1(vqdmulhq_v, aarch64_neon_sqdmulh, Add1ArgType),
  NEONMAP1(vqdmull_v, aarch64_neon_sqdmull, Add1ArgType),
  NEONMAP2(vqmovn_v, aarch64_neon_uqxtn, aarch64_neon_sqxtn, Add1ArgType | UnsignedAlts),
  NEONMAP1(vqmovun_v, aarch64_neon_sqxtun, Add1ArgType),
  NEONMAP1(vqneg_v, aarch64_neon_sqneg, Add1ArgType),
  NEONMAP1(vqnegq_v, aarch64_neon_sqneg, Add1ArgType),
  NEONMAP1(vqrdmulh_v, aarch64_neon_sqrdmulh, Add1ArgType),
  NEONMAP1(vqrdmulhq_v, aarch64_neon_sqrdmulh, Add1ArgType),
  NEONMAP2(vqrshl_v, aarch64_neon_uqrshl, aarch64_neon_sqrshl, Add1ArgType | UnsignedAlts),
  NEONMAP2(vqrshlq_v, aarch64_neon_uqrshl, aarch64_neon_sqrshl, Add1ArgType | UnsignedAlts),
  NEONMAP2(vqshl_n_v, aarch64_neon_uqshl, aarch64_neon_sqshl, UnsignedAlts),
  NEONMAP2(vqshl_v, aarch64_neon_uqshl, aarch64_neon_sqshl, Add1ArgType | UnsignedAlts),
  NEONMAP2(vqshlq_n_v, aarch64_neon_uqshl, aarch64_neon_sqshl,UnsignedAlts),
  NEONMAP2(vqshlq_v, aarch64_neon_uqshl, aarch64_neon_sqshl, Add1ArgType | UnsignedAlts),
  NEONMAP1(vqshlu_n_v, aarch64_neon_sqshlu, 0),
  NEONMAP1(vqshluq_n_v, aarch64_neon_sqshlu, 0),
  NEONMAP2(vqsub_v, aarch64_neon_uqsub, aarch64_neon_sqsub, Add1ArgType | UnsignedAlts),
  NEONMAP2(vqsubq_v, aarch64_neon_uqsub, aarch64_neon_sqsub, Add1ArgType | UnsignedAlts),
  NEONMAP1(vraddhn_v, aarch64_neon_raddhn, Add1ArgType),
  NEONMAP2(vrecpe_v, aarch64_neon_frecpe, aarch64_neon_urecpe, 0),
  NEONMAP2(vrecpeq_v, aarch64_neon_frecpe, aarch64_neon_urecpe, 0),
  NEONMAP1(vrecps_v, aarch64_neon_frecps, Add1ArgType),
  NEONMAP1(vrecpsq_v, aarch64_neon_frecps, Add1ArgType),
  NEONMAP2(vrhadd_v, aarch64_neon_urhadd, aarch64_neon_srhadd, Add1ArgType | UnsignedAlts),
  NEONMAP2(vrhaddq_v, aarch64_neon_urhadd, aarch64_neon_srhadd, Add1ArgType | UnsignedAlts),
  NEONMAP0(vrndi_v),
  NEONMAP0(vrndiq_v),
  NEONMAP2(vrshl_v, aarch64_neon_urshl, aarch64_neon_srshl, Add1ArgType | UnsignedAlts),
  NEONMAP2(vrshlq_v, aarch64_neon_urshl, aarch64_neon_srshl, Add1ArgType | UnsignedAlts),
  NEONMAP2(vrshr_n_v, aarch64_neon_urshl, aarch64_neon_srshl, UnsignedAlts),
  NEONMAP2(vrshrq_n_v, aarch64_neon_urshl, aarch64_neon_srshl, UnsignedAlts),
  NEONMAP2(vrsqrte_v, aarch64_neon_frsqrte, aarch64_neon_ursqrte, 0),
  NEONMAP2(vrsqrteq_v, aarch64_neon_frsqrte, aarch64_neon_ursqrte, 0),
  NEONMAP1(vrsqrts_v, aarch64_neon_frsqrts, Add1ArgType),
  NEONMAP1(vrsqrtsq_v, aarch64_neon_frsqrts, Add1ArgType),
  NEONMAP1(vrsubhn_v, aarch64_neon_rsubhn, Add1ArgType),
  NEONMAP1(vsha1su0q_v, aarch64_crypto_sha1su0, 0),
  NEONMAP1(vsha1su1q_v, aarch64_crypto_sha1su1, 0),
  NEONMAP1(vsha256h2q_v, aarch64_crypto_sha256h2, 0),
  NEONMAP1(vsha256hq_v, aarch64_crypto_sha256h, 0),
  NEONMAP1(vsha256su0q_v, aarch64_crypto_sha256su0, 0),
  NEONMAP1(vsha256su1q_v, aarch64_crypto_sha256su1, 0),
  NEONMAP0(vshl_n_v),
  NEONMAP2(vshl_v, aarch64_neon_ushl, aarch64_neon_sshl, Add1ArgType | UnsignedAlts),
  NEONMAP0(vshll_n_v),
  NEONMAP0(vshlq_n_v),
  NEONMAP2(vshlq_v, aarch64_neon_ushl, aarch64_neon_sshl, Add1ArgType | UnsignedAlts),
  NEONMAP0(vshr_n_v),
  NEONMAP0(vshrn_n_v),
  NEONMAP0(vshrq_n_v),
  NEONMAP1(vst1_x2_v, aarch64_neon_st1x2, 0),
  NEONMAP1(vst1_x3_v, aarch64_neon_st1x3, 0),
  NEONMAP1(vst1_x4_v, aarch64_neon_st1x4, 0),
  NEONMAP1(vst1q_x2_v, aarch64_neon_st1x2, 0),
  NEONMAP1(vst1q_x3_v, aarch64_neon_st1x3, 0),
  NEONMAP1(vst1q_x4_v, aarch64_neon_st1x4, 0),
  NEONMAP0(vsubhn_v),
  NEONMAP0(vtst_v),
  NEONMAP0(vtstq_v),
};

static const NeonIntrinsicInfo AArch64SISDIntrinsicMap[] = {
  NEONMAP1(vabdd_f64, aarch64_sisd_fabd, Add1ArgType),
  NEONMAP1(vabds_f32, aarch64_sisd_fabd, Add1ArgType),
  NEONMAP1(vabsd_s64, aarch64_neon_abs, Add1ArgType),
  NEONMAP1(vaddlv_s32, aarch64_neon_saddlv, AddRetType | Add1ArgType),
  NEONMAP1(vaddlv_u32, aarch64_neon_uaddlv, AddRetType | Add1ArgType),
  NEONMAP1(vaddlvq_s32, aarch64_neon_saddlv, AddRetType | Add1ArgType),
  NEONMAP1(vaddlvq_u32, aarch64_neon_uaddlv, AddRetType | Add1ArgType),
  NEONMAP1(vaddv_f32, aarch64_neon_faddv, AddRetType | Add1ArgType),
  NEONMAP1(vaddv_s32, aarch64_neon_saddv, AddRetType | Add1ArgType),
  NEONMAP1(vaddv_u32, aarch64_neon_uaddv, AddRetType | Add1ArgType),
  NEONMAP1(vaddvq_f32, aarch64_neon_faddv, AddRetType | Add1ArgType),
  NEONMAP1(vaddvq_f64, aarch64_neon_faddv, AddRetType | Add1ArgType),
  NEONMAP1(vaddvq_s32, aarch64_neon_saddv, AddRetType | Add1ArgType),
  NEONMAP1(vaddvq_s64, aarch64_neon_saddv, AddRetType | Add1ArgType),
  NEONMAP1(vaddvq_u32, aarch64_neon_uaddv, AddRetType | Add1ArgType),
  NEONMAP1(vaddvq_u64, aarch64_neon_uaddv, AddRetType | Add1ArgType),
  NEONMAP1(vcaged_f64, aarch64_neon_facge, AddRetType | Add1ArgType),
  NEONMAP1(vcages_f32, aarch64_neon_facge, AddRetType | Add1ArgType),
  NEONMAP1(vcagtd_f64, aarch64_neon_facgt, AddRetType | Add1ArgType),
  NEONMAP1(vcagts_f32, aarch64_neon_facgt, AddRetType | Add1ArgType),
  NEONMAP1(vcaled_f64, aarch64_neon_facge, AddRetType | Add1ArgType),
  NEONMAP1(vcales_f32, aarch64_neon_facge, AddRetType | Add1ArgType),
  NEONMAP1(vcaltd_f64, aarch64_neon_facgt, AddRetType | Add1ArgType),
  NEONMAP1(vcalts_f32, aarch64_neon_facgt, AddRetType | Add1ArgType),
  NEONMAP1(vcvtad_s64_f64, aarch64_neon_fcvtas, AddRetType | Add1ArgType),
  NEONMAP1(vcvtad_u64_f64, aarch64_neon_fcvtau, AddRetType | Add1ArgType),
  NEONMAP1(vcvtas_s32_f32, aarch64_neon_fcvtas, AddRetType | Add1ArgType),
  NEONMAP1(vcvtas_u32_f32, aarch64_neon_fcvtau, AddRetType | Add1ArgType),
  NEONMAP1(vcvtd_n_f64_s64, aarch64_neon_vcvtfxs2fp, AddRetType | Add1ArgType),
  NEONMAP1(vcvtd_n_f64_u64, aarch64_neon_vcvtfxu2fp, AddRetType | Add1ArgType),
  NEONMAP1(vcvtd_n_s64_f64, aarch64_neon_vcvtfp2fxs, AddRetType | Add1ArgType),
  NEONMAP1(vcvtd_n_u64_f64, aarch64_neon_vcvtfp2fxu, AddRetType | Add1ArgType),
  NEONMAP1(vcvtmd_s64_f64, aarch64_neon_fcvtms, AddRetType | Add1ArgType),
  NEONMAP1(vcvtmd_u64_f64, aarch64_neon_fcvtmu, AddRetType | Add1ArgType),
  NEONMAP1(vcvtms_s32_f32, aarch64_neon_fcvtms, AddRetType | Add1ArgType),
  NEONMAP1(vcvtms_u32_f32, aarch64_neon_fcvtmu, AddRetType | Add1ArgType),
  NEONMAP1(vcvtnd_s64_f64, aarch64_neon_fcvtns, AddRetType | Add1ArgType),
  NEONMAP1(vcvtnd_u64_f64, aarch64_neon_fcvtnu, AddRetType | Add1ArgType),
  NEONMAP1(vcvtns_s32_f32, aarch64_neon_fcvtns, AddRetType | Add1ArgType),
  NEONMAP1(vcvtns_u32_f32, aarch64_neon_fcvtnu, AddRetType | Add1ArgType),
  NEONMAP1(vcvtpd_s64_f64, aarch64_neon_fcvtps, AddRetType | Add1ArgType),
  NEONMAP1(vcvtpd_u64_f64, aarch64_neon_fcvtpu, AddRetType | Add1ArgType),
  NEONMAP1(vcvtps_s32_f32, aarch64_neon_fcvtps, AddRetType | Add1ArgType),
  NEONMAP1(vcvtps_u32_f32, aarch64_neon_fcvtpu, AddRetType | Add1ArgType),
  NEONMAP1(vcvts_n_f32_s32, aarch64_neon_vcvtfxs2fp, AddRetType | Add1ArgType),
  NEONMAP1(vcvts_n_f32_u32, aarch64_neon_vcvtfxu2fp, AddRetType | Add1ArgType),
  NEONMAP1(vcvts_n_s32_f32, aarch64_neon_vcvtfp2fxs, AddRetType | Add1ArgType),
  NEONMAP1(vcvts_n_u32_f32, aarch64_neon_vcvtfp2fxu, AddRetType | Add1ArgType),
  NEONMAP1(vcvtxd_f32_f64, aarch64_sisd_fcvtxn, 0),
  NEONMAP1(vmaxnmv_f32, aarch64_neon_fmaxnmv, AddRetType | Add1ArgType),
  NEONMAP1(vmaxnmvq_f32, aarch64_neon_fmaxnmv, AddRetType | Add1ArgType),
  NEONMAP1(vmaxnmvq_f64, aarch64_neon_fmaxnmv, AddRetType | Add1ArgType),
  NEONMAP1(vmaxv_f32, aarch64_neon_fmaxv, AddRetType | Add1ArgType),
  NEONMAP1(vmaxv_s32, aarch64_neon_smaxv, AddRetType | Add1ArgType),
  NEONMAP1(vmaxv_u32, aarch64_neon_umaxv, AddRetType | Add1ArgType),
  NEONMAP1(vmaxvq_f32, aarch64_neon_fmaxv, AddRetType | Add1ArgType),
  NEONMAP1(vmaxvq_f64, aarch64_neon_fmaxv, AddRetType | Add1ArgType),
  NEONMAP1(vmaxvq_s32, aarch64_neon_smaxv, AddRetType | Add1ArgType),
  NEONMAP1(vmaxvq_u32, aarch64_neon_umaxv, AddRetType | Add1ArgType),
  NEONMAP1(vminnmv_f32, aarch64_neon_fminnmv, AddRetType | Add1ArgType),
  NEONMAP1(vminnmvq_f32, aarch64_neon_fminnmv, AddRetType | Add1ArgType),
  NEONMAP1(vminnmvq_f64, aarch64_neon_fminnmv, AddRetType | Add1ArgType),
  NEONMAP1(vminv_f32, aarch64_neon_fminv, AddRetType | Add1ArgType),
  NEONMAP1(vminv_s32, aarch64_neon_sminv, AddRetType | Add1ArgType),
  NEONMAP1(vminv_u32, aarch64_neon_uminv, AddRetType | Add1ArgType),
  NEONMAP1(vminvq_f32, aarch64_neon_fminv, AddRetType | Add1ArgType),
  NEONMAP1(vminvq_f64, aarch64_neon_fminv, AddRetType | Add1ArgType),
  NEONMAP1(vminvq_s32, aarch64_neon_sminv, AddRetType | Add1ArgType),
  NEONMAP1(vminvq_u32, aarch64_neon_uminv, AddRetType | Add1ArgType),
  NEONMAP1(vmull_p64, aarch64_neon_pmull64, 0),
  NEONMAP1(vmulxd_f64, aarch64_neon_fmulx, Add1ArgType),
  NEONMAP1(vmulxs_f32, aarch64_neon_fmulx, Add1ArgType),
  NEONMAP1(vpaddd_s64, aarch64_neon_uaddv, AddRetType | Add1ArgType),
  NEONMAP1(vpaddd_u64, aarch64_neon_uaddv, AddRetType | Add1ArgType),
  NEONMAP1(vpmaxnmqd_f64, aarch64_neon_fmaxnmv, AddRetType | Add1ArgType),
  NEONMAP1(vpmaxnms_f32, aarch64_neon_fmaxnmv, AddRetType | Add1ArgType),
  NEONMAP1(vpmaxqd_f64, aarch64_neon_fmaxv, AddRetType | Add1ArgType),
  NEONMAP1(vpmaxs_f32, aarch64_neon_fmaxv, AddRetType | Add1ArgType),
  NEONMAP1(vpminnmqd_f64, aarch64_neon_fminnmv, AddRetType | Add1ArgType),
  NEONMAP1(vpminnms_f32, aarch64_neon_fminnmv, AddRetType | Add1ArgType),
  NEONMAP1(vpminqd_f64, aarch64_neon_fminv, AddRetType | Add1ArgType),
  NEONMAP1(vpmins_f32, aarch64_neon_fminv, AddRetType | Add1ArgType),
  NEONMAP1(vqabsb_s8, aarch64_neon_sqabs, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqabsd_s64, aarch64_neon_sqabs, Add1ArgType),
  NEONMAP1(vqabsh_s16, aarch64_neon_sqabs, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqabss_s32, aarch64_neon_sqabs, Add1ArgType),
  NEONMAP1(vqaddb_s8, aarch64_neon_sqadd, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqaddb_u8, aarch64_neon_uqadd, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqaddd_s64, aarch64_neon_sqadd, Add1ArgType),
  NEONMAP1(vqaddd_u64, aarch64_neon_uqadd, Add1ArgType),
  NEONMAP1(vqaddh_s16, aarch64_neon_sqadd, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqaddh_u16, aarch64_neon_uqadd, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqadds_s32, aarch64_neon_sqadd, Add1ArgType),
  NEONMAP1(vqadds_u32, aarch64_neon_uqadd, Add1ArgType),
  NEONMAP1(vqdmulhh_s16, aarch64_neon_sqdmulh, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqdmulhs_s32, aarch64_neon_sqdmulh, Add1ArgType),
  NEONMAP1(vqdmullh_s16, aarch64_neon_sqdmull, VectorRet | Use128BitVectors),
  NEONMAP1(vqdmulls_s32, aarch64_neon_sqdmulls_scalar, 0),
  NEONMAP1(vqmovnd_s64, aarch64_neon_scalar_sqxtn, AddRetType | Add1ArgType),
  NEONMAP1(vqmovnd_u64, aarch64_neon_scalar_uqxtn, AddRetType | Add1ArgType),
  NEONMAP1(vqmovnh_s16, aarch64_neon_sqxtn, VectorRet | Use64BitVectors),
  NEONMAP1(vqmovnh_u16, aarch64_neon_uqxtn, VectorRet | Use64BitVectors),
  NEONMAP1(vqmovns_s32, aarch64_neon_sqxtn, VectorRet | Use64BitVectors),
  NEONMAP1(vqmovns_u32, aarch64_neon_uqxtn, VectorRet | Use64BitVectors),
  NEONMAP1(vqmovund_s64, aarch64_neon_scalar_sqxtun, AddRetType | Add1ArgType),
  NEONMAP1(vqmovunh_s16, aarch64_neon_sqxtun, VectorRet | Use64BitVectors),
  NEONMAP1(vqmovuns_s32, aarch64_neon_sqxtun, VectorRet | Use64BitVectors),
  NEONMAP1(vqnegb_s8, aarch64_neon_sqneg, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqnegd_s64, aarch64_neon_sqneg, Add1ArgType),
  NEONMAP1(vqnegh_s16, aarch64_neon_sqneg, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqnegs_s32, aarch64_neon_sqneg, Add1ArgType),
  NEONMAP1(vqrdmulhh_s16, aarch64_neon_sqrdmulh, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqrdmulhs_s32, aarch64_neon_sqrdmulh, Add1ArgType),
  NEONMAP1(vqrshlb_s8, aarch64_neon_sqrshl, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqrshlb_u8, aarch64_neon_uqrshl, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqrshld_s64, aarch64_neon_sqrshl, Add1ArgType),
  NEONMAP1(vqrshld_u64, aarch64_neon_uqrshl, Add1ArgType),
  NEONMAP1(vqrshlh_s16, aarch64_neon_sqrshl, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqrshlh_u16, aarch64_neon_uqrshl, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqrshls_s32, aarch64_neon_sqrshl, Add1ArgType),
  NEONMAP1(vqrshls_u32, aarch64_neon_uqrshl, Add1ArgType),
  NEONMAP1(vqrshrnd_n_s64, aarch64_neon_sqrshrn, AddRetType),
  NEONMAP1(vqrshrnd_n_u64, aarch64_neon_uqrshrn, AddRetType),
  NEONMAP1(vqrshrnh_n_s16, aarch64_neon_sqrshrn, VectorRet | Use64BitVectors),
  NEONMAP1(vqrshrnh_n_u16, aarch64_neon_uqrshrn, VectorRet | Use64BitVectors),
  NEONMAP1(vqrshrns_n_s32, aarch64_neon_sqrshrn, VectorRet | Use64BitVectors),
  NEONMAP1(vqrshrns_n_u32, aarch64_neon_uqrshrn, VectorRet | Use64BitVectors),
  NEONMAP1(vqrshrund_n_s64, aarch64_neon_sqrshrun, AddRetType),
  NEONMAP1(vqrshrunh_n_s16, aarch64_neon_sqrshrun, VectorRet | Use64BitVectors),
  NEONMAP1(vqrshruns_n_s32, aarch64_neon_sqrshrun, VectorRet | Use64BitVectors),
  NEONMAP1(vqshlb_n_s8, aarch64_neon_sqshl, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqshlb_n_u8, aarch64_neon_uqshl, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqshlb_s8, aarch64_neon_sqshl, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqshlb_u8, aarch64_neon_uqshl, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqshld_s64, aarch64_neon_sqshl, Add1ArgType),
  NEONMAP1(vqshld_u64, aarch64_neon_uqshl, Add1ArgType),
  NEONMAP1(vqshlh_n_s16, aarch64_neon_sqshl, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqshlh_n_u16, aarch64_neon_uqshl, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqshlh_s16, aarch64_neon_sqshl, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqshlh_u16, aarch64_neon_uqshl, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqshls_n_s32, aarch64_neon_sqshl, Add1ArgType),
  NEONMAP1(vqshls_n_u32, aarch64_neon_uqshl, Add1ArgType),
  NEONMAP1(vqshls_s32, aarch64_neon_sqshl, Add1ArgType),
  NEONMAP1(vqshls_u32, aarch64_neon_uqshl, Add1ArgType),
  NEONMAP1(vqshlub_n_s8, aarch64_neon_sqshlu, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqshluh_n_s16, aarch64_neon_sqshlu, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqshlus_n_s32, aarch64_neon_sqshlu, Add1ArgType),
  NEONMAP1(vqshrnd_n_s64, aarch64_neon_sqshrn, AddRetType),
  NEONMAP1(vqshrnd_n_u64, aarch64_neon_uqshrn, AddRetType),
  NEONMAP1(vqshrnh_n_s16, aarch64_neon_sqshrn, VectorRet | Use64BitVectors),
  NEONMAP1(vqshrnh_n_u16, aarch64_neon_uqshrn, VectorRet | Use64BitVectors),
  NEONMAP1(vqshrns_n_s32, aarch64_neon_sqshrn, VectorRet | Use64BitVectors),
  NEONMAP1(vqshrns_n_u32, aarch64_neon_uqshrn, VectorRet | Use64BitVectors),
  NEONMAP1(vqshrund_n_s64, aarch64_neon_sqshrun, AddRetType),
  NEONMAP1(vqshrunh_n_s16, aarch64_neon_sqshrun, VectorRet | Use64BitVectors),
  NEONMAP1(vqshruns_n_s32, aarch64_neon_sqshrun, VectorRet | Use64BitVectors),
  NEONMAP1(vqsubb_s8, aarch64_neon_sqsub, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqsubb_u8, aarch64_neon_uqsub, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqsubd_s64, aarch64_neon_sqsub, Add1ArgType),
  NEONMAP1(vqsubd_u64, aarch64_neon_uqsub, Add1ArgType),
  NEONMAP1(vqsubh_s16, aarch64_neon_sqsub, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqsubh_u16, aarch64_neon_uqsub, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vqsubs_s32, aarch64_neon_sqsub, Add1ArgType),
  NEONMAP1(vqsubs_u32, aarch64_neon_uqsub, Add1ArgType),
  NEONMAP1(vrecped_f64, aarch64_neon_frecpe, Add1ArgType),
  NEONMAP1(vrecpes_f32, aarch64_neon_frecpe, Add1ArgType),
  NEONMAP1(vrecpxd_f64, aarch64_neon_frecpx, Add1ArgType),
  NEONMAP1(vrecpxs_f32, aarch64_neon_frecpx, Add1ArgType),
  NEONMAP1(vrshld_s64, aarch64_neon_srshl, Add1ArgType),
  NEONMAP1(vrshld_u64, aarch64_neon_urshl, Add1ArgType),
  NEONMAP1(vrsqrted_f64, aarch64_neon_frsqrte, Add1ArgType),
  NEONMAP1(vrsqrtes_f32, aarch64_neon_frsqrte, Add1ArgType),
  NEONMAP1(vrsqrtsd_f64, aarch64_neon_frsqrts, Add1ArgType),
  NEONMAP1(vrsqrtss_f32, aarch64_neon_frsqrts, Add1ArgType),
  NEONMAP1(vsha1cq_u32, aarch64_crypto_sha1c, 0),
  NEONMAP1(vsha1h_u32, aarch64_crypto_sha1h, 0),
  NEONMAP1(vsha1mq_u32, aarch64_crypto_sha1m, 0),
  NEONMAP1(vsha1pq_u32, aarch64_crypto_sha1p, 0),
  NEONMAP1(vshld_s64, aarch64_neon_sshl, Add1ArgType),
  NEONMAP1(vshld_u64, aarch64_neon_ushl, Add1ArgType),
  NEONMAP1(vslid_n_s64, aarch64_neon_vsli, Vectorize1ArgType),
  NEONMAP1(vslid_n_u64, aarch64_neon_vsli, Vectorize1ArgType),
  NEONMAP1(vsqaddb_u8, aarch64_neon_usqadd, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vsqaddd_u64, aarch64_neon_usqadd, Add1ArgType),
  NEONMAP1(vsqaddh_u16, aarch64_neon_usqadd, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vsqadds_u32, aarch64_neon_usqadd, Add1ArgType),
  NEONMAP1(vsrid_n_s64, aarch64_neon_vsri, Vectorize1ArgType),
  NEONMAP1(vsrid_n_u64, aarch64_neon_vsri, Vectorize1ArgType),
  NEONMAP1(vuqaddb_s8, aarch64_neon_suqadd, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vuqaddd_s64, aarch64_neon_suqadd, Add1ArgType),
  NEONMAP1(vuqaddh_s16, aarch64_neon_suqadd, Vectorize1ArgType | Use64BitVectors),
  NEONMAP1(vuqadds_s32, aarch64_neon_suqadd, Add1ArgType),
  // FP16 scalar intrinisics go here.
  NEONMAP1(vabdh_f16, aarch64_sisd_fabd, Add1ArgType),
  NEONMAP1(vcvtah_s32_f16, aarch64_neon_fcvtas, AddRetType | Add1ArgType),
  NEONMAP1(vcvtah_s64_f16, aarch64_neon_fcvtas, AddRetType | Add1ArgType),
  NEONMAP1(vcvtah_u32_f16, aarch64_neon_fcvtau, AddRetType | Add1ArgType),
  NEONMAP1(vcvtah_u64_f16, aarch64_neon_fcvtau, AddRetType | Add1ArgType),
  NEONMAP1(vcvth_n_f16_s32, aarch64_neon_vcvtfxs2fp, AddRetType | Add1ArgType),
  NEONMAP1(vcvth_n_f16_s64, aarch64_neon_vcvtfxs2fp, AddRetType | Add1ArgType),
  NEONMAP1(vcvth_n_f16_u32, aarch64_neon_vcvtfxu2fp, AddRetType | Add1ArgType),
  NEONMAP1(vcvth_n_f16_u64, aarch64_neon_vcvtfxu2fp, AddRetType | Add1ArgType),
  NEONMAP1(vcvth_n_s32_f16, aarch64_neon_vcvtfp2fxs, AddRetType | Add1ArgType),
  NEONMAP1(vcvth_n_s64_f16, aarch64_neon_vcvtfp2fxs, AddRetType | Add1ArgType),
  NEONMAP1(vcvth_n_u32_f16, aarch64_neon_vcvtfp2fxu, AddRetType | Add1ArgType),
  NEONMAP1(vcvth_n_u64_f16, aarch64_neon_vcvtfp2fxu, AddRetType | Add1ArgType),
  NEONMAP1(vcvtmh_s32_f16, aarch64_neon_fcvtms, AddRetType | Add1ArgType),
  NEONMAP1(vcvtmh_s64_f16, aarch64_neon_fcvtms, AddRetType | Add1ArgType),
  NEONMAP1(vcvtmh_u32_f16, aarch64_neon_fcvtmu, AddRetType | Add1ArgType),
  NEONMAP1(vcvtmh_u64_f16, aarch64_neon_fcvtmu, AddRetType | Add1ArgType),
  NEONMAP1(vcvtnh_s32_f16, aarch64_neon_fcvtns, AddRetType | Add1ArgType),
  NEONMAP1(vcvtnh_s64_f16, aarch64_neon_fcvtns, AddRetType | Add1ArgType),
  NEONMAP1(vcvtnh_u32_f16, aarch64_neon_fcvtnu, AddRetType | Add1ArgType),
  NEONMAP1(vcvtnh_u64_f16, aarch64_neon_fcvtnu, AddRetType | Add1ArgType),
  NEONMAP1(vcvtph_s32_f16, aarch64_neon_fcvtps, AddRetType | Add1ArgType),
  NEONMAP1(vcvtph_s64_f16, aarch64_neon_fcvtps, AddRetType | Add1ArgType),
  NEONMAP1(vcvtph_u32_f16, aarch64_neon_fcvtpu, AddRetType | Add1ArgType),
  NEONMAP1(vcvtph_u64_f16, aarch64_neon_fcvtpu, AddRetType | Add1ArgType),
  NEONMAP1(vmulxh_f16, aarch64_neon_fmulx, Add1ArgType),
  NEONMAP1(vrecpeh_f16, aarch64_neon_frecpe, Add1ArgType),
  NEONMAP1(vrecpxh_f16, aarch64_neon_frecpx, Add1ArgType),
  NEONMAP1(vrsqrteh_f16, aarch64_neon_frsqrte, Add1ArgType),
  NEONMAP1(vrsqrtsh_f16, aarch64_neon_frsqrts, Add1ArgType),
};

#undef NEONMAP0
#undef NEONMAP1
#undef NEONMAP2

static bool NEONSIMDIntrinsicsProvenSorted = false;

static bool AArch64SIMDIntrinsicsProvenSorted = false;
static bool AArch64SISDIntrinsicsProvenSorted = false;


static const NeonIntrinsicInfo *
findNeonIntrinsicInMap(ArrayRef<NeonIntrinsicInfo> IntrinsicMap,
                       unsigned BuiltinID, bool &MapProvenSorted) {

#ifndef NDEBUG
  if (!MapProvenSorted) {
    assert(std::is_sorted(std::begin(IntrinsicMap), std::end(IntrinsicMap)));
    MapProvenSorted = true;
  }
#endif

  const NeonIntrinsicInfo *Builtin =
      std::lower_bound(IntrinsicMap.begin(), IntrinsicMap.end(), BuiltinID);

  if (Builtin != IntrinsicMap.end() && Builtin->BuiltinID == BuiltinID)
    return Builtin;

  return nullptr;
}

Function *CodeGenFunction::LookupNeonLLVMIntrinsic(unsigned IntrinsicID,
                                                   unsigned Modifier,
                                                   llvm::Type *ArgType,
                                                   const CallExpr *E) {
  int VectorSize = 0;
  if (Modifier & Use64BitVectors)
    VectorSize = 64;
  else if (Modifier & Use128BitVectors)
    VectorSize = 128;

  // Return type.
  SmallVector<llvm::Type *, 3> Tys;
  if (Modifier & AddRetType) {
    llvm::Type *Ty = ConvertType(E->getCallReturnType(getContext()));
    if (Modifier & VectorizeRetType)
      Ty = llvm::VectorType::get(
          Ty, VectorSize ? VectorSize / Ty->getPrimitiveSizeInBits() : 1);

    Tys.push_back(Ty);
  }

  // Arguments.
  if (Modifier & VectorizeArgTypes) {
    int Elts = VectorSize ? VectorSize / ArgType->getPrimitiveSizeInBits() : 1;
    ArgType = llvm::VectorType::get(ArgType, Elts);
  }

  if (Modifier & (Add1ArgType | Add2ArgTypes))
    Tys.push_back(ArgType);

  if (Modifier & Add2ArgTypes)
    Tys.push_back(ArgType);

  if (Modifier & InventFloatType)
    Tys.push_back(FloatTy);

  return CGM.getIntrinsic(IntrinsicID, Tys);
}

static Value *EmitCommonNeonSISDBuiltinExpr(CodeGenFunction &CGF,
                                            const NeonIntrinsicInfo &SISDInfo,
                                            SmallVectorImpl<Value *> &Ops,
                                            const CallExpr *E) {
  unsigned BuiltinID = SISDInfo.BuiltinID;
  unsigned int Int = SISDInfo.LLVMIntrinsic;
  unsigned Modifier = SISDInfo.TypeModifier;
  const char *s = SISDInfo.NameHint;

  switch (BuiltinID) {
  case NEON::BI__builtin_neon_vcled_s64:
  case NEON::BI__builtin_neon_vcled_u64:
  case NEON::BI__builtin_neon_vcles_f32:
  case NEON::BI__builtin_neon_vcled_f64:
  case NEON::BI__builtin_neon_vcltd_s64:
  case NEON::BI__builtin_neon_vcltd_u64:
  case NEON::BI__builtin_neon_vclts_f32:
  case NEON::BI__builtin_neon_vcltd_f64:
  case NEON::BI__builtin_neon_vcales_f32:
  case NEON::BI__builtin_neon_vcaled_f64:
  case NEON::BI__builtin_neon_vcalts_f32:
  case NEON::BI__builtin_neon_vcaltd_f64:
    // Only one direction of comparisons actually exist, cmle is actually a cmge
    // with swapped operands. The table gives us the right intrinsic but we
    // still need to do the swap.
    std::swap(Ops[0], Ops[1]);
    break;
  }

  assert(Int && "Generic code assumes a valid intrinsic");

  // Determine the type(s) of this overloaded AArch64 intrinsic.
  const Expr *Arg = E->getArg(0);
  llvm::Type *ArgTy = CGF.ConvertType(Arg->getType());
  Function *F = CGF.LookupNeonLLVMIntrinsic(Int, Modifier, ArgTy, E);

  int j = 0;
  ConstantInt *C0 = ConstantInt::get(CGF.SizeTy, 0);
  for (Function::const_arg_iterator ai = F->arg_begin(), ae = F->arg_end();
       ai != ae; ++ai, ++j) {
    llvm::Type *ArgTy = ai->getType();
    if (Ops[j]->getType()->getPrimitiveSizeInBits() ==
             ArgTy->getPrimitiveSizeInBits())
      continue;

    assert(ArgTy->isVectorTy() && !Ops[j]->getType()->isVectorTy());
    // The constant argument to an _n_ intrinsic always has Int32Ty, so truncate
    // it before inserting.
    Ops[j] =
        CGF.Builder.CreateTruncOrBitCast(Ops[j], ArgTy->getVectorElementType());
    Ops[j] =
        CGF.Builder.CreateInsertElement(UndefValue::get(ArgTy), Ops[j], C0);
  }

  Value *Result = CGF.EmitNeonCall(F, Ops, s);
  llvm::Type *ResultType = CGF.ConvertType(E->getType());
  if (ResultType->getPrimitiveSizeInBits() <
      Result->getType()->getPrimitiveSizeInBits())
    return CGF.Builder.CreateExtractElement(Result, C0);

  return CGF.Builder.CreateBitCast(Result, ResultType, s);
}

Value *CodeGenFunction::EmitCommonNeonBuiltinExpr(
    unsigned BuiltinID, unsigned LLVMIntrinsic, unsigned AltLLVMIntrinsic,
    const char *NameHint, unsigned Modifier, const CallExpr *E,
    SmallVectorImpl<llvm::Value *> &Ops, Address PtrOp0, Address PtrOp1,
    llvm::Triple::ArchType Arch) {
  // Get the last argument, which specifies the vector type.
  llvm::APSInt NeonTypeConst;
  const Expr *Arg = E->getArg(E->getNumArgs() - 1);
  if (!Arg->isIntegerConstantExpr(NeonTypeConst, getContext()))
    return nullptr;

  // Determine the type of this overloaded NEON intrinsic.
  NeonTypeFlags Type(NeonTypeConst.getZExtValue());
  bool Usgn = Type.isUnsigned();
  bool Quad = Type.isQuad();
  const bool HasLegalHalfType = getTarget().hasLegalHalfType();

  llvm::VectorType *VTy = GetNeonType(this, Type, HasLegalHalfType);
  llvm::Type *Ty = VTy;
  if (!Ty)
    return nullptr;

  auto getAlignmentValue32 = [&](Address addr) -> Value* {
    return Builder.getInt32(addr.getAlignment().getQuantity());
  };

  unsigned Int = LLVMIntrinsic;
  if ((Modifier & UnsignedAlts) && !Usgn)
    Int = AltLLVMIntrinsic;

  switch (BuiltinID) {
  default: break;
  case NEON::BI__builtin_neon_vabs_v:
  case NEON::BI__builtin_neon_vabsq_v:
    if (VTy->getElementType()->isFloatingPointTy())
      return EmitNeonCall(CGM.getIntrinsic(Intrinsic::fabs, Ty), Ops, "vabs");
    return EmitNeonCall(CGM.getIntrinsic(LLVMIntrinsic, Ty), Ops, "vabs");
  case NEON::BI__builtin_neon_vaddhn_v: {
    llvm::VectorType *SrcTy =
        llvm::VectorType::getExtendedElementVectorType(VTy);

    // %sum = add <4 x i32> %lhs, %rhs
    Ops[0] = Builder.CreateBitCast(Ops[0], SrcTy);
    Ops[1] = Builder.CreateBitCast(Ops[1], SrcTy);
    Ops[0] = Builder.CreateAdd(Ops[0], Ops[1], "vaddhn");

    // %high = lshr <4 x i32> %sum, <i32 16, i32 16, i32 16, i32 16>
    Constant *ShiftAmt =
        ConstantInt::get(SrcTy, SrcTy->getScalarSizeInBits() / 2);
    Ops[0] = Builder.CreateLShr(Ops[0], ShiftAmt, "vaddhn");

    // %res = trunc <4 x i32> %high to <4 x i16>
    return Builder.CreateTrunc(Ops[0], VTy, "vaddhn");
  }
  case NEON::BI__builtin_neon_vcale_v:
  case NEON::BI__builtin_neon_vcaleq_v:
  case NEON::BI__builtin_neon_vcalt_v:
  case NEON::BI__builtin_neon_vcaltq_v:
    std::swap(Ops[0], Ops[1]);
    LLVM_FALLTHROUGH;
  case NEON::BI__builtin_neon_vcage_v:
  case NEON::BI__builtin_neon_vcageq_v:
  case NEON::BI__builtin_neon_vcagt_v:
  case NEON::BI__builtin_neon_vcagtq_v: {
    llvm::Type *Ty;
    switch (VTy->getScalarSizeInBits()) {
    default: llvm_unreachable("unexpected type");
    case 32:
      Ty = FloatTy;
      break;
    case 64:
      Ty = DoubleTy;
      break;
    case 16:
      Ty = HalfTy;
      break;
    }
    llvm::Type *VecFlt = llvm::VectorType::get(Ty, VTy->getNumElements());
    llvm::Type *Tys[] = { VTy, VecFlt };
    Function *F = CGM.getIntrinsic(LLVMIntrinsic, Tys);
    return EmitNeonCall(F, Ops, NameHint);
  }
  case NEON::BI__builtin_neon_vceqz_v:
  case NEON::BI__builtin_neon_vceqzq_v:
    return EmitAArch64CompareBuiltinExpr(Ops[0], Ty, ICmpInst::FCMP_OEQ,
                                         ICmpInst::ICMP_EQ, "vceqz");
  case NEON::BI__builtin_neon_vcgez_v:
  case NEON::BI__builtin_neon_vcgezq_v:
    return EmitAArch64CompareBuiltinExpr(Ops[0], Ty, ICmpInst::FCMP_OGE,
                                         ICmpInst::ICMP_SGE, "vcgez");
  case NEON::BI__builtin_neon_vclez_v:
  case NEON::BI__builtin_neon_vclezq_v:
    return EmitAArch64CompareBuiltinExpr(Ops[0], Ty, ICmpInst::FCMP_OLE,
                                         ICmpInst::ICMP_SLE, "vclez");
  case NEON::BI__builtin_neon_vcgtz_v:
  case NEON::BI__builtin_neon_vcgtzq_v:
    return EmitAArch64CompareBuiltinExpr(Ops[0], Ty, ICmpInst::FCMP_OGT,
                                         ICmpInst::ICMP_SGT, "vcgtz");
  case NEON::BI__builtin_neon_vcltz_v:
  case NEON::BI__builtin_neon_vcltzq_v:
    return EmitAArch64CompareBuiltinExpr(Ops[0], Ty, ICmpInst::FCMP_OLT,
                                         ICmpInst::ICMP_SLT, "vcltz");
  case NEON::BI__builtin_neon_vclz_v:
  case NEON::BI__builtin_neon_vclzq_v:
    // We generate target-independent intrinsic, which needs a second argument
    // for whether or not clz of zero is undefined; on ARM it isn't.
    Ops.push_back(Builder.getInt1(getTarget().isCLZForZeroUndef()));
    break;
  case NEON::BI__builtin_neon_vcvt_f32_v:
  case NEON::BI__builtin_neon_vcvtq_f32_v:
    Ops[0] = Builder.CreateBitCast(Ops[0], Ty);
    Ty = GetNeonType(this, NeonTypeFlags(NeonTypeFlags::Float32, false, Quad),
                     HasLegalHalfType);
    return Usgn ? Builder.CreateUIToFP(Ops[0], Ty, "vcvt")
                : Builder.CreateSIToFP(Ops[0], Ty, "vcvt");
  case NEON::BI__builtin_neon_vcvt_f16_v:
  case NEON::BI__builtin_neon_vcvtq_f16_v:
    Ops[0] = Builder.CreateBitCast(Ops[0], Ty);
    Ty = GetNeonType(this, NeonTypeFlags(NeonTypeFlags::Float16, false, Quad),
                     HasLegalHalfType);
    return Usgn ? Builder.CreateUIToFP(Ops[0], Ty, "vcvt")
                : Builder.CreateSIToFP(Ops[0], Ty, "vcvt");
  case NEON::BI__builtin_neon_vcvt_n_f16_v:
  case NEON::BI__builtin_neon_vcvt_n_f32_v:
  case NEON::BI__builtin_neon_vcvt_n_f64_v:
  case NEON::BI__builtin_neon_vcvtq_n_f16_v:
  case NEON::BI__builtin_neon_vcvtq_n_f32_v:
  case NEON::BI__builtin_neon_vcvtq_n_f64_v: {
    llvm::Type *Tys[2] = { GetFloatNeonType(this, Type), Ty };
    Int = Usgn ? LLVMIntrinsic : AltLLVMIntrinsic;
    Function *F = CGM.getIntrinsic(Int, Tys);
    return EmitNeonCall(F, Ops, "vcvt_n");
  }
  case NEON::BI__builtin_neon_vcvt_n_s16_v:
  case NEON::BI__builtin_neon_vcvt_n_s32_v:
  case NEON::BI__builtin_neon_vcvt_n_u16_v:
  case NEON::BI__builtin_neon_vcvt_n_u32_v:
  case NEON::BI__builtin_neon_vcvt_n_s64_v:
  case NEON::BI__builtin_neon_vcvt_n_u64_v:
  case NEON::BI__builtin_neon_vcvtq_n_s16_v:
  case NEON::BI__builtin_neon_vcvtq_n_s32_v:
  case NEON::BI__builtin_neon_vcvtq_n_u16_v:
  case NEON::BI__builtin_neon_vcvtq_n_u32_v:
  case NEON::BI__builtin_neon_vcvtq_n_s64_v:
  case NEON::BI__builtin_neon_vcvtq_n_u64_v: {
    llvm::Type *Tys[2] = { Ty, GetFloatNeonType(this, Type) };
    Function *F = CGM.getIntrinsic(LLVMIntrinsic, Tys);
    return EmitNeonCall(F, Ops, "vcvt_n");
  }
  case NEON::BI__builtin_neon_vcvt_s32_v:
  case NEON::BI__builtin_neon_vcvt_u32_v:
  case NEON::BI__builtin_neon_vcvt_s64_v:
  case NEON::BI__builtin_neon_vcvt_u64_v:
  case NEON::BI__builtin_neon_vcvt_s16_v:
  case NEON::BI__builtin_neon_vcvt_u16_v:
  case NEON::BI__builtin_neon_vcvtq_s32_v:
  case NEON::BI__builtin_neon_vcvtq_u32_v:
  case NEON::BI__builtin_neon_vcvtq_s64_v:
  case NEON::BI__builtin_neon_vcvtq_u64_v:
  case NEON::BI__builtin_neon_vcvtq_s16_v:
  case NEON::BI__builtin_neon_vcvtq_u16_v: {
    Ops[0] = Builder.CreateBitCast(Ops[0], GetFloatNeonType(this, Type));
    return Usgn ? Builder.CreateFPToUI(Ops[0], Ty, "vcvt")
                : Builder.CreateFPToSI(Ops[0], Ty, "vcvt");
  }
  case NEON::BI__builtin_neon_vcvta_s16_v:
  case NEON::BI__builtin_neon_vcvta_s32_v:
  case NEON::BI__builtin_neon_vcvta_s64_v:
  case NEON::BI__builtin_neon_vcvta_u16_v:
  case NEON::BI__builtin_neon_vcvta_u32_v:
  case NEON::BI__builtin_neon_vcvta_u64_v:
  case NEON::BI__builtin_neon_vcvtaq_s16_v:
  case NEON::BI__builtin_neon_vcvtaq_s32_v:
  case NEON::BI__builtin_neon_vcvtaq_s64_v:
  case NEON::BI__builtin_neon_vcvtaq_u16_v:
  case NEON::BI__builtin_neon_vcvtaq_u32_v:
  case NEON::BI__builtin_neon_vcvtaq_u64_v:
  case NEON::BI__builtin_neon_vcvtn_s16_v:
  case NEON::BI__builtin_neon_vcvtn_s32_v:
  case NEON::BI__builtin_neon_vcvtn_s64_v:
  case NEON::BI__builtin_neon_vcvtn_u16_v:
  case NEON::BI__builtin_neon_vcvtn_u32_v:
  case NEON::BI__builtin_neon_vcvtn_u64_v:
  case NEON::BI__builtin_neon_vcvtnq_s16_v:
  case NEON::BI__builtin_neon_vcvtnq_s32_v:
  case NEON::BI__builtin_neon_vcvtnq_s64_v:
  case NEON::BI__builtin_neon_vcvtnq_u16_v:
  case NEON::BI__builtin_neon_vcvtnq_u32_v:
  case NEON::BI__builtin_neon_vcvtnq_u64_v:
  case NEON::BI__builtin_neon_vcvtp_s16_v:
  case NEON::BI__builtin_neon_vcvtp_s32_v:
  case NEON::BI__builtin_neon_vcvtp_s64_v:
  case NEON::BI__builtin_neon_vcvtp_u16_v:
  case NEON::BI__builtin_neon_vcvtp_u32_v:
  case NEON::BI__builtin_neon_vcvtp_u64_v:
  case NEON::BI__builtin_neon_vcvtpq_s16_v:
  case NEON::BI__builtin_neon_vcvtpq_s32_v:
  case NEON::BI__builtin_neon_vcvtpq_s64_v:
  case NEON::BI__builtin_neon_vcvtpq_u16_v:
  case NEON::BI__builtin_neon_vcvtpq_u32_v:
  case NEON::BI__builtin_neon_vcvtpq_u64_v:
  case NEON::BI__builtin_neon_vcvtm_s16_v:
  case NEON::BI__builtin_neon_vcvtm_s32_v:
  case NEON::BI__builtin_neon_vcvtm_s64_v:
  case NEON::BI__builtin_neon_vcvtm_u16_v:
  case NEON::BI__builtin_neon_vcvtm_u32_v:
  case NEON::BI__builtin_neon_vcvtm_u64_v:
  case NEON::BI__builtin_neon_vcvtmq_s16_v:
  case NEON::BI__builtin_neon_vcvtmq_s32_v:
  case NEON::BI__builtin_neon_vcvtmq_s64_v:
  case NEON::BI__builtin_neon_vcvtmq_u16_v:
  case NEON::BI__builtin_neon_vcvtmq_u32_v:
  case NEON::BI__builtin_neon_vcvtmq_u64_v: {
    llvm::Type *Tys[2] = { Ty, GetFloatNeonType(this, Type) };
    return EmitNeonCall(CGM.getIntrinsic(LLVMIntrinsic, Tys), Ops, NameHint);
  }
  case NEON::BI__builtin_neon_vext_v:
  case NEON::BI__builtin_neon_vextq_v: {
    int CV = cast<ConstantInt>(Ops[2])->getSExtValue();
    SmallVector<uint32_t, 16> Indices;
    for (unsigned i = 0, e = VTy->getNumElements(); i != e; ++i)
      Indices.push_back(i+CV);

    Ops[0] = Builder.CreateBitCast(Ops[0], Ty);
    Ops[1] = Builder.CreateBitCast(Ops[1], Ty);
    return Builder.CreateShuffleVector(Ops[0], Ops[1], Indices, "vext");
  }
  case NEON::BI__builtin_neon_vfma_v:
  case NEON::BI__builtin_neon_vfmaq_v: {
    Value *F = CGM.getIntrinsic(Intrinsic::fma, Ty);
    Ops[0] = Builder.CreateBitCast(Ops[0], Ty);
    Ops[1] = Builder.CreateBitCast(Ops[1], Ty);
    Ops[2] = Builder.CreateBitCast(Ops[2], Ty);

    // NEON intrinsic puts accumulator first, unlike the LLVM fma.
    return Builder.CreateCall(F, {Ops[1], Ops[2], Ops[0]});
  }
  case NEON::BI__builtin_neon_vld1_v:
  case NEON::BI__builtin_neon_vld1q_v: {
    llvm::Type *Tys[] = {Ty, Int8PtrTy};
    Ops.push_back(getAlignmentValue32(PtrOp0));
    return EmitNeonCall(CGM.getIntrinsic(LLVMIntrinsic, Tys), Ops, "vld1");
  }
  case NEON::BI__builtin_neon_vld1_x2_v:
  case NEON::BI__builtin_neon_vld1q_x2_v:
  case NEON::BI__builtin_neon_vld1_x3_v:
  case NEON::BI__builtin_neon_vld1q_x3_v:
  case NEON::BI__builtin_neon_vld1_x4_v:
  case NEON::BI__builtin_neon_vld1q_x4_v: {
    llvm::Type *PTy = llvm::PointerType::getUnqual(VTy->getVectorElementType());
    Ops[1] = Builder.CreateBitCast(Ops[1], PTy);
    llvm::Type *Tys[2] = { VTy, PTy };
    Function *F = CGM.getIntrinsic(LLVMIntrinsic, Tys);
    Ops[1] = Builder.CreateCall(F, Ops[1], "vld1xN");
    Ty = llvm::PointerType::getUnqual(Ops[1]->getType());
    Ops[0] = Builder.CreateBitCast(Ops[0], Ty);
    return Builder.CreateDefaultAlignedStore(Ops[1], Ops[0]);
  }
  case NEON::BI__builtin_neon_vld2_v:
  case NEON::BI__builtin_neon_vld2q_v:
  case NEON::BI__builtin_neon_vld3_v:
  case NEON::BI__builtin_neon_vld3q_v:
  case NEON::BI__builtin_neon_vld4_v:
  case NEON::BI__builtin_neon_vld4q_v:
  case NEON::BI__builtin_neon_vld2_dup_v:
  case NEON::BI__builtin_neon_vld2q_dup_v:
  case NEON::BI__builtin_neon_vld3_dup_v:
  case NEON::BI__builtin_neon_vld3q_dup_v:
  case NEON::BI__builtin_neon_vld4_dup_v:
  case NEON::BI__builtin_neon_vld4q_dup_v: {
    llvm::Type *Tys[] = {Ty, Int8PtrTy};
    Function *F = CGM.getIntrinsic(LLVMIntrinsic, Tys);
    Value *Align = getAlignmentValue32(PtrOp1);
    Ops[1] = Builder.CreateCall(F, {Ops[1], Align}, NameHint);
    Ty = llvm::PointerType::getUnqual(Ops[1]->getType());
    Ops[0] = Builder.CreateBitCast(Ops[0], Ty);
    return Builder.CreateDefaultAlignedStore(Ops[1], Ops[0]);
  }
  case NEON::BI__builtin_neon_vld1_dup_v:
  case NEON::BI__builtin_neon_vld1q_dup_v: {
    Value *V = UndefValue::get(Ty);
    Ty = llvm::PointerType::getUnqual(VTy->getElementType());
    PtrOp0 = Builder.CreateBitCast(PtrOp0, Ty);
    LoadInst *Ld = Builder.CreateLoad(PtrOp0);
    llvm::Constant *CI = ConstantInt::get(SizeTy, 0);
    Ops[0] = Builder.CreateInsertElement(V, Ld, CI);
    return EmitNeonSplat(Ops[0], CI);
  }
  case NEON::BI__builtin_neon_vld2_lane_v:
  case NEON::BI__builtin_neon_vld2q_lane_v:
  case NEON::BI__builtin_neon_vld3_lane_v:
  case NEON::BI__builtin_neon_vld3q_lane_v:
  case NEON::BI__builtin_neon_vld4_lane_v:
  case NEON::BI__builtin_neon_vld4q_lane_v: {
    llvm::Type *Tys[] = {Ty, Int8PtrTy};
    Function *F = CGM.getIntrinsic(LLVMIntrinsic, Tys);
    for (unsigned I = 2; I < Ops.size() - 1; ++I)
      Ops[I] = Builder.CreateBitCast(Ops[I], Ty);
    Ops.push_back(getAlignmentValue32(PtrOp1));
    Ops[1] = Builder.CreateCall(F, makeArrayRef(Ops).slice(1), NameHint);
    Ty = llvm::PointerType::getUnqual(Ops[1]->getType());
    Ops[0] = Builder.CreateBitCast(Ops[0], Ty);
    return Builder.CreateDefaultAlignedStore(Ops[1], Ops[0]);
  }
  case NEON::BI__builtin_neon_vmovl_v: {
    llvm::Type *DTy =llvm::VectorType::getTruncatedElementVectorType(VTy);
    Ops[0] = Builder.CreateBitCast(Ops[0], DTy);
    if (Usgn)
      return Builder.CreateZExt(Ops[0], Ty, "vmovl");
    return Builder.CreateSExt(Ops[0], Ty, "vmovl");
  }
  case NEON::BI__builtin_neon_vmovn_v: {
    llvm::Type *QTy = llvm::VectorType::getExtendedElementVectorType(VTy);
    Ops[0] = Builder.CreateBitCast(Ops[0], QTy);
    return Builder.CreateTrunc(Ops[0], Ty, "vmovn");
  }
  case NEON::BI__builtin_neon_vmull_v:
    // FIXME: the integer vmull operations could be emitted in terms of pure
    // LLVM IR (2 exts followed by a mul). Unfortunately LLVM has a habit of
    // hoisting the exts outside loops. Until global ISel comes along that can
    // see through such movement this leads to bad CodeGen. So we need an
    // intrinsic for now.
    Int = Usgn ? Intrinsic::arm_neon_vmullu : Intrinsic::arm_neon_vmulls;
    Int = Type.isPoly() ? (unsigned)Intrinsic::arm_neon_vmullp : Int;
    return EmitNeonCall(CGM.getIntrinsic(Int, Ty), Ops, "vmull");
  case NEON::BI__builtin_neon_vpadal_v:
  case NEON::BI__builtin_neon_vpadalq_v: {
    // The source operand type has twice as many elements of half the size.
    unsigned EltBits = VTy->getElementType()->getPrimitiveSizeInBits();
    llvm::Type *EltTy =
      llvm::IntegerType::get(getLLVMContext(), EltBits / 2);
    llvm::Type *NarrowTy =
      llvm::VectorType::get(EltTy, VTy->getNumElements() * 2);
    llvm::Type *Tys[2] = { Ty, NarrowTy };
    return EmitNeonCall(CGM.getIntrinsic(Int, Tys), Ops, NameHint);
  }
  case NEON::BI__builtin_neon_vpaddl_v:
  case NEON::BI__builtin_neon_vpaddlq_v: {
    // The source operand type has twice as many elements of half the size.
    unsigned EltBits = VTy->getElementType()->getPrimitiveSizeInBits();
    llvm::Type *EltTy = llvm::IntegerType::get(getLLVMContext(), EltBits / 2);
    llvm::Type *NarrowTy =
      llvm::VectorType::get(EltTy, VTy->getNumElements() * 2);
    llvm::Type *Tys[2] = { Ty, NarrowTy };
    return EmitNeonCall(CGM.getIntrinsic(Int, Tys), Ops, "vpaddl");
  }
  case NEON::BI__builtin_neon_vqdmlal_v:
  case NEON::BI__builtin_neon_vqdmlsl_v: {
    SmallVector<Value *, 2> MulOps(Ops.begin() + 1, Ops.end());
    Ops[1] =
        EmitNeonCall(CGM.getIntrinsic(LLVMIntrinsic, Ty), MulOps, "vqdmlal");
    Ops.resize(2);
    return EmitNeonCall(CGM.getIntrinsic(AltLLVMIntrinsic, Ty), Ops, NameHint);
  }
  case NEON::BI__builtin_neon_vqshl_n_v:
  case NEON::BI__builtin_neon_vqshlq_n_v:
    return EmitNeonCall(CGM.getIntrinsic(Int, Ty), Ops, "vqshl_n",
                        1, false);
  case NEON::BI__builtin_neon_vqshlu_n_v:
  case NEON::BI__builtin_neon_vqshluq_n_v:
    return EmitNeonCall(CGM.getIntrinsic(Int, Ty), Ops, "vqshlu_n",
                        1, false);
  case NEON::BI__builtin_neon_vrecpe_v:
  case NEON::BI__builtin_neon_vrecpeq_v:
  case NEON::BI__builtin_neon_vrsqrte_v:
  case NEON::BI__builtin_neon_vrsqrteq_v:
    Int = Ty->isFPOrFPVectorTy() ? LLVMIntrinsic : AltLLVMIntrinsic;
    return EmitNeonCall(CGM.getIntrinsic(Int, Ty), Ops, NameHint);
  case NEON::BI__builtin_neon_vrndi_v:
  case NEON::BI__builtin_neon_vrndiq_v:
    Int = Intrinsic::nearbyint;
    return EmitNeonCall(CGM.getIntrinsic(Int, Ty), Ops, NameHint);
  case NEON::BI__builtin_neon_vrshr_n_v:
  case NEON::BI__builtin_neon_vrshrq_n_v:
    return EmitNeonCall(CGM.getIntrinsic(Int, Ty), Ops, "vrshr_n",
                        1, true);
  case NEON::BI__builtin_neon_vshl_n_v:
  case NEON::BI__builtin_neon_vshlq_n_v:
    Ops[1] = EmitNeonShiftVector(Ops[1], Ty, false);
    return Builder.CreateShl(Builder.CreateBitCast(Ops[0],Ty), Ops[1],
                             "vshl_n");
  case NEON::BI__builtin_neon_vshll_n_v: {
    llvm::Type *SrcTy = llvm::VectorType::getTruncatedElementVectorType(VTy);
    Ops[0] = Builder.CreateBitCast(Ops[0], SrcTy);
    if (Usgn)
      Ops[0] = Builder.CreateZExt(Ops[0], VTy);
    else
      Ops[0] = Builder.CreateSExt(Ops[0], VTy);
    Ops[1] = EmitNeonShiftVector(Ops[1], VTy, false);
    return Builder.CreateShl(Ops[0], Ops[1], "vshll_n");
  }
  case NEON::BI__builtin_neon_vshrn_n_v: {
    llvm::Type *SrcTy = llvm::VectorType::getExtendedElementVectorType(VTy);
    Ops[0] = Builder.CreateBitCast(Ops[0], SrcTy);
    Ops[1] = EmitNeonShiftVector(Ops[1], SrcTy, false);
    if (Usgn)
      Ops[0] = Builder.CreateLShr(Ops[0], Ops[1]);
    else
      Ops[0] = Builder.CreateAShr(Ops[0], Ops[1]);
    return Builder.CreateTrunc(Ops[0], Ty, "vshrn_n");
  }
  case NEON::BI__builtin_neon_vshr_n_v: