C++ AllocaInst::getAllocatedType方法代码示例

bool CallAnalyzer::visitAlloca(AllocaInst &I) {
  // Check whether inlining will turn a dynamic alloca into a static
  // alloca, and handle that case.
  if (I.isArrayAllocation()) {
    if (Constant *Size = SimplifiedValues.lookup(I.getArraySize())) {
      ConstantInt *AllocSize = dyn_cast<ConstantInt>(Size);
      assert(AllocSize && "Allocation size not a constant int?");
      Type *Ty = I.getAllocatedType();
      AllocatedSize += Ty->getPrimitiveSizeInBits() * AllocSize->getZExtValue();
      return Base::visitAlloca(I);
    }
  }

  // Accumulate the allocated size.
  if (I.isStaticAlloca()) {
    Type *Ty = I.getAllocatedType();
    AllocatedSize += (DL ? DL->getTypeAllocSize(Ty) :
                      Ty->getPrimitiveSizeInBits());
  }

  // We will happily inline static alloca instructions.
  if (I.isStaticAlloca())
    return Base::visitAlloca(I);

  // FIXME: This is overly conservative. Dynamic allocas are inefficient for
  // a variety of reasons, and so we would like to not inline them into
  // functions which don't currently have a dynamic alloca. This simply
  // disables inlining altogether in the presence of a dynamic alloca.
  HasDynamicAlloca = true;
  return false;
}

SizeOffsetEvalType ObjectSizeOffsetEvaluator::visitAllocaInst(AllocaInst &I) {
  if (!I.getAllocatedType()->isSized())
    return unknown();

  // must be a VLA
  assert(I.isArrayAllocation());
  Value *ArraySize = I.getArraySize();
  Value *Size = ConstantInt::get(ArraySize->getType(),
                                 TD->getTypeAllocSize(I.getAllocatedType()));
  Size = Builder.CreateMul(Size, ArraySize);
  return std::make_pair(Size, Zero);
}

// -- handle unary instruction -- 
void UnsafeTypeCastingCheck::handleAllocInstruction (Instruction *inst) {
  AllocaInst *ainst = dyn_cast<AllocaInst>(inst); 
  if (ainst == NULL) 
    utccAbort("handleAllocInstruction cannot process with a non-alloca instruction"); 
  Type *atype = ainst->getAllocatedType(); 
  setPointedType(ainst, llvmT2utccT(atype, ainst));
}

void ArrayIndexChecker::visitAllocaInst(AllocaInst& I) {
  DEBUG(dbgs() << "ArrayIndexChecker: visiting alloca " << I << "\n");

  if (isa<StructType>(I.getAllocatedType()) || isa<ArrayType>(I.getAllocatedType())) {
    return;
  }

  auto pos = std::find(ptr_value_vec_.begin(), ptr_value_vec_.end(), &I);
  assert(pos != ptr_value_vec_.end());
  index_t varIdx = pos - ptr_value_vec_.begin();

  assert(idx2addr_.find(varIdx) != idx2addr_.end());
  if (addr2version_[idx2addr_[varIdx]] != 0)
    throw ArrayIndexIsNotConstant;

  DEBUG(dbgs() << "ArrayIndexChecker: visited alloca\n");
}

bool IRTranslator::translateStaticAlloca(const AllocaInst &AI) {
  assert(AI.isStaticAlloca() && "only handle static allocas now");
  MachineFunction &MF = MIRBuilder.getMF();
  unsigned ElementSize = DL->getTypeStoreSize(AI.getAllocatedType());
  unsigned Size =
      ElementSize * cast<ConstantInt>(AI.getArraySize())->getZExtValue();

  // Always allocate at least one byte.
  Size = std::max(Size, 1u);

  unsigned Alignment = AI.getAlignment();
  if (!Alignment)
    Alignment = DL->getABITypeAlignment(AI.getAllocatedType());

  unsigned Res = getOrCreateVReg(AI);
  int FI = MF.getFrameInfo().CreateStackObject(Size, Alignment, false, &AI);
  MIRBuilder.buildFrameIndex(LLT::pointer(0), Res, FI);
  return true;
}

int IRTranslator::getOrCreateFrameIndex(const AllocaInst &AI) {
  if (FrameIndices.find(&AI) != FrameIndices.end())
    return FrameIndices[&AI];

  MachineFunction &MF = MIRBuilder.getMF();
  unsigned ElementSize = DL->getTypeStoreSize(AI.getAllocatedType());
  unsigned Size =
      ElementSize * cast<ConstantInt>(AI.getArraySize())->getZExtValue();

  // Always allocate at least one byte.
  Size = std::max(Size, 1u);

  unsigned Alignment = AI.getAlignment();
  if (!Alignment)
    Alignment = DL->getABITypeAlignment(AI.getAllocatedType());

  int &FI = FrameIndices[&AI];
  FI = MF.getFrameInfo().CreateStackObject(Size, Alignment, false, &AI);
  return FI;
}

SizeOffsetType ObjectSizeOffsetVisitor::visitAllocaInst(AllocaInst &I) {
  if (!I.getAllocatedType()->isSized())
    return unknown();

  APInt Size(IntTyBits, DL.getTypeAllocSize(I.getAllocatedType()));
  if (!I.isArrayAllocation())
    return std::make_pair(align(Size, I.getAlignment()), Zero);

  Value *ArraySize = I.getArraySize();
  if (const ConstantInt *C = dyn_cast<ConstantInt>(ArraySize)) {
    APInt NumElems = C->getValue();
    if (!CheckedZextOrTrunc(NumElems))
      return unknown();

    bool Overflow;
    Size = Size.umul_ov(NumElems, Overflow);
    return Overflow ? unknown() : std::make_pair(align(Size, I.getAlignment()),
                                                 Zero);
  }
  return unknown();
}

void LowerIntrinsics::InsertAutomaticGCRoots(Function &F, GCStrategy &S) {
  if (!S.usesAutomaticRoots())
    return;

  LLVMContext &C = F.getParent()->getContext();
  for (Function::iterator BBI = F.begin(), BBE = F.end(); BBI != BBE; ++BBI) {
    for (BasicBlock::iterator II = BBI->begin(),
                              IE = BBI->end(); II != IE; ++II) {
      AllocaInst *AI = dyn_cast<AllocaInst>(&*II);
      if (!AI)
        continue;

      Value *Index = ConstantInt::get(Type::getInt32Ty(C), 0);
      ArrayRef<Value *> Indices(&Index, 1);
      AutomaticallyRootValue(*AI, AI->getAllocatedType(), Indices);
    }
  }
}

//
// Method: visitAllocaInst()
//
// Description:
//  This method instruments an alloca instruction so that it is zero'ed out
//  before any data is loaded from it.
//
void
InitAllocas::visitAllocaInst (AllocaInst & AI) {
  //
  // Scan for a place to insert the instruction to initialize the
  // allocated memory.
  //
  Instruction * InsertPt = getInsertionPoint (AI);

  //
  // Zero the alloca with a memset.  If this is done more efficiently with stores
  // SelectionDAG will lower it appropriately based on target information.
  //
  TargetData & TD = getAnalysis<TargetData>();

  //
  // Get various types that we'll need.
  //
  Type * Int1Type    = IntegerType::getInt1Ty(AI.getContext());
  Type * Int8Type    = IntegerType::getInt8Ty(AI.getContext());
  Type * Int32Type   = IntegerType::getInt32Ty(AI.getContext());
  Type * VoidPtrType = getVoidPtrType (AI.getContext());
  Type * AllocType = AI.getAllocatedType();

  //
  // Create a call to memset.
  //
  Module * M = AI.getParent()->getParent()->getParent();
  Function * Memset = cast<Function>(M->getFunction ("llvm.memset.p0i8.i32"));
  std::vector<Value *> args;
  args.push_back (castTo (&AI, VoidPtrType, AI.getName().str(), InsertPt));
  args.push_back (ConstantInt::get(Int8Type, 0));
  args.push_back (ConstantInt::get(Int32Type,TD.getTypeAllocSize(AllocType)));
  args.push_back (ConstantInt::get(Int32Type,
                                   TD.getABITypeAlignment(AllocType)));
  args.push_back (ConstantInt::get(Int1Type, 0));
  CallInst::Create (Memset, args, "", InsertPt);

  //
  // Update statistics.
  //
  ++InitedAllocas;
  return;
}

bool CallAnalyzer::visitAlloca(AllocaInst &I) {
  // FIXME: Check whether inlining will turn a dynamic alloca into a static
  // alloca, and handle that case.

  // Accumulate the allocated size.
  if (I.isStaticAlloca()) {
    Type *Ty = I.getAllocatedType();
    AllocatedSize += (TD ? TD->getTypeAllocSize(Ty) :
                      Ty->getPrimitiveSizeInBits());
  }

  // We will happily inline static alloca instructions.
  if (I.isStaticAlloca())
    return Base::visitAlloca(I);

  // FIXME: This is overly conservative. Dynamic allocas are inefficient for
  // a variety of reasons, and so we would like to not inline them into
  // functions which don't currently have a dynamic alloca. This simply
  // disables inlining altogether in the presence of a dynamic alloca.
  HasDynamicAlloca = true;
  return false;
}

/// InlineCallIfPossible - If it is possible to inline the specified call site,
/// do so and update the CallGraph for this operation.
///
/// This function also does some basic book-keeping to update the IR.  The
/// InlinedArrayAllocas map keeps track of any allocas that are already
/// available from other  functions inlined into the caller.  If we are able to
/// inline this call site we attempt to reuse already available allocas or add
/// any new allocas to the set if not possible.
static bool InlineCallIfPossible(CallSite CS, InlineFunctionInfo &IFI,
                                 InlinedArrayAllocasTy &InlinedArrayAllocas) {
  Function *Callee = CS.getCalledFunction();
  Function *Caller = CS.getCaller();

  // Try to inline the function.  Get the list of static allocas that were
  // inlined.
  if (!InlineFunction(CS, IFI))
    return false;

  // If the inlined function had a higher stack protection level than the
  // calling function, then bump up the caller's stack protection level.
  if (Callee->hasFnAttr(Attribute::StackProtectReq))
    Caller->addFnAttr(Attribute::StackProtectReq);
  else if (Callee->hasFnAttr(Attribute::StackProtect) &&
           !Caller->hasFnAttr(Attribute::StackProtectReq))
    Caller->addFnAttr(Attribute::StackProtect);

  
  // Look at all of the allocas that we inlined through this call site.  If we
  // have already inlined other allocas through other calls into this function,
  // then we know that they have disjoint lifetimes and that we can merge them.
  //
  // There are many heuristics possible for merging these allocas, and the
  // different options have different tradeoffs.  One thing that we *really*
  // don't want to hurt is SRoA: once inlining happens, often allocas are no
  // longer address taken and so they can be promoted.
  //
  // Our "solution" for that is to only merge allocas whose outermost type is an
  // array type.  These are usually not promoted because someone is using a
  // variable index into them.  These are also often the most important ones to
  // merge.
  //
  // A better solution would be to have real memory lifetime markers in the IR
  // and not have the inliner do any merging of allocas at all.  This would
  // allow the backend to do proper stack slot coloring of all allocas that
  // *actually make it to the backend*, which is really what we want.
  //
  // Because we don't have this information, we do this simple and useful hack.
  //
  SmallPtrSet<AllocaInst*, 16> UsedAllocas;
  
  // Loop over all the allocas we have so far and see if they can be merged with
  // a previously inlined alloca.  If not, remember that we had it.
  for (unsigned AllocaNo = 0, e = IFI.StaticAllocas.size();
       AllocaNo != e; ++AllocaNo) {
    AllocaInst *AI = IFI.StaticAllocas[AllocaNo];
    
    // Don't bother trying to merge array allocations (they will usually be
    // canonicalized to be an allocation *of* an array), or allocations whose
    // type is not itself an array (because we're afraid of pessimizing SRoA).
    const ArrayType *ATy = dyn_cast<ArrayType>(AI->getAllocatedType());
    if (ATy == 0 || AI->isArrayAllocation())
      continue;
    
    // Get the list of all available allocas for this array type.
    std::vector<AllocaInst*> &AllocasForType = InlinedArrayAllocas[ATy];
    
    // Loop over the allocas in AllocasForType to see if we can reuse one.  Note
    // that we have to be careful not to reuse the same "available" alloca for
    // multiple different allocas that we just inlined, we use the 'UsedAllocas'
    // set to keep track of which "available" allocas are being used by this
    // function.  Also, AllocasForType can be empty of course!
    bool MergedAwayAlloca = false;
    for (unsigned i = 0, e = AllocasForType.size(); i != e; ++i) {
      AllocaInst *AvailableAlloca = AllocasForType[i];
      
      // The available alloca has to be in the right function, not in some other
      // function in this SCC.
      if (AvailableAlloca->getParent() != AI->getParent())
        continue;
      
      // If the inlined function already uses this alloca then we can't reuse
      // it.
      if (!UsedAllocas.insert(AvailableAlloca))
        continue;
      
      // Otherwise, we *can* reuse it, RAUW AI into AvailableAlloca and declare
      // success!
      DEBUG(dbgs() << "    ***MERGED ALLOCA: " << *AI);
      
      AI->replaceAllUsesWith(AvailableAlloca);
      AI->eraseFromParent();
      MergedAwayAlloca = true;
      ++NumMergedAllocas;
      break;
    }

    // If we already nuked the alloca, we're done with it.
    if (MergedAwayAlloca)
      continue;

//.........这里部分代码省略.........

void AMDGPUPromoteAlloca::handleAlloca(AllocaInst &I) {
  // Array allocations are probably not worth handling, since an allocation of
  // the array type is the canonical form.
  if (!I.isStaticAlloca() || I.isArrayAllocation())
    return;

  IRBuilder<> Builder(&I);

  // First try to replace the alloca with a vector
  Type *AllocaTy = I.getAllocatedType();

  DEBUG(dbgs() << "Trying to promote " << I << '\n');

  if (tryPromoteAllocaToVector(&I))
    return;

  DEBUG(dbgs() << " alloca is not a candidate for vectorization.\n");

  const Function &ContainingFunction = *I.getParent()->getParent();

  // FIXME: We should also try to get this value from the reqd_work_group_size
  // function attribute if it is available.
  unsigned WorkGroupSize = AMDGPU::getMaximumWorkGroupSize(ContainingFunction);

  int AllocaSize =
      WorkGroupSize * Mod->getDataLayout().getTypeAllocSize(AllocaTy);

  if (AllocaSize > LocalMemAvailable) {
    DEBUG(dbgs() << " Not enough local memory to promote alloca.\n");
    return;
  }

  std::vector<Value*> WorkList;

  if (!collectUsesWithPtrTypes(&I, WorkList)) {
    DEBUG(dbgs() << " Do not know how to convert all uses\n");
    return;
  }

  DEBUG(dbgs() << "Promoting alloca to local memory\n");
  LocalMemAvailable -= AllocaSize;

  Function *F = I.getParent()->getParent();

  Type *GVTy = ArrayType::get(I.getAllocatedType(), WorkGroupSize);
  GlobalVariable *GV = new GlobalVariable(
      *Mod, GVTy, false, GlobalValue::InternalLinkage,
      UndefValue::get(GVTy),
      Twine(F->getName()) + Twine('.') + I.getName(),
      nullptr,
      GlobalVariable::NotThreadLocal,
      AMDGPUAS::LOCAL_ADDRESS);
  GV->setUnnamedAddr(true);
  GV->setAlignment(I.getAlignment());

  Value *TCntY, *TCntZ;

  std::tie(TCntY, TCntZ) = getLocalSizeYZ(Builder);
  Value *TIdX = getWorkitemID(Builder, 0);
  Value *TIdY = getWorkitemID(Builder, 1);
  Value *TIdZ = getWorkitemID(Builder, 2);

  Value *Tmp0 = Builder.CreateMul(TCntY, TCntZ, "", true, true);
  Tmp0 = Builder.CreateMul(Tmp0, TIdX);
  Value *Tmp1 = Builder.CreateMul(TIdY, TCntZ, "", true, true);
  Value *TID = Builder.CreateAdd(Tmp0, Tmp1);
  TID = Builder.CreateAdd(TID, TIdZ);

  Value *Indices[] = {
    Constant::getNullValue(Type::getInt32Ty(Mod->getContext())),
    TID
  };

  Value *Offset = Builder.CreateInBoundsGEP(GVTy, GV, Indices);
  I.mutateType(Offset->getType());
  I.replaceAllUsesWith(Offset);
  I.eraseFromParent();

  for (Value *V : WorkList) {
    CallInst *Call = dyn_cast<CallInst>(V);
    if (!Call) {
      Type *EltTy = V->getType()->getPointerElementType();
      PointerType *NewTy = PointerType::get(EltTy, AMDGPUAS::LOCAL_ADDRESS);

      // The operand's value should be corrected on its own.
      if (isa<AddrSpaceCastInst>(V))
        continue;

      // FIXME: It doesn't really make sense to try to do this for all
      // instructions.
      V->mutateType(NewTy);
      continue;
    }

    IntrinsicInst *Intr = dyn_cast<IntrinsicInst>(Call);
    if (!Intr) {
      // FIXME: What is this for? It doesn't make sense to promote arbitrary
      // function calls. If the call is to a defined function that can also be
      // promoted, we should be able to do this once that function is also
      // rewritten.
//.........这里部分代码省略.........

Instruction *InstCombiner::visitAllocaInst(AllocaInst &AI) {
  if (auto *I = simplifyAllocaArraySize(*this, AI))
    return I;

  if (AI.getAllocatedType()->isSized()) {
    // If the alignment is 0 (unspecified), assign it the preferred alignment.
    if (AI.getAlignment() == 0)
      AI.setAlignment(DL.getPrefTypeAlignment(AI.getAllocatedType()));

    // Move all alloca's of zero byte objects to the entry block and merge them
    // together.  Note that we only do this for alloca's, because malloc should
    // allocate and return a unique pointer, even for a zero byte allocation.
    if (DL.getTypeAllocSize(AI.getAllocatedType()) == 0) {
      // For a zero sized alloca there is no point in doing an array allocation.
      // This is helpful if the array size is a complicated expression not used
      // elsewhere.
      if (AI.isArrayAllocation()) {
        AI.setOperand(0, ConstantInt::get(AI.getArraySize()->getType(), 1));
        return &AI;
      }

      // Get the first instruction in the entry block.
      BasicBlock &EntryBlock = AI.getParent()->getParent()->getEntryBlock();
      Instruction *FirstInst = EntryBlock.getFirstNonPHIOrDbg();
      if (FirstInst != &AI) {
        // If the entry block doesn't start with a zero-size alloca then move
        // this one to the start of the entry block.  There is no problem with
        // dominance as the array size was forced to a constant earlier already.
        AllocaInst *EntryAI = dyn_cast<AllocaInst>(FirstInst);
        if (!EntryAI || !EntryAI->getAllocatedType()->isSized() ||
            DL.getTypeAllocSize(EntryAI->getAllocatedType()) != 0) {
          AI.moveBefore(FirstInst);
          return &AI;
        }

        // If the alignment of the entry block alloca is 0 (unspecified),
        // assign it the preferred alignment.
        if (EntryAI->getAlignment() == 0)
          EntryAI->setAlignment(
              DL.getPrefTypeAlignment(EntryAI->getAllocatedType()));
        // Replace this zero-sized alloca with the one at the start of the entry
        // block after ensuring that the address will be aligned enough for both
        // types.
        unsigned MaxAlign = std::max(EntryAI->getAlignment(),
                                     AI.getAlignment());
        EntryAI->setAlignment(MaxAlign);
        if (AI.getType() != EntryAI->getType())
          return new BitCastInst(EntryAI, AI.getType());
        return ReplaceInstUsesWith(AI, EntryAI);
      }
    }
  }

  if (AI.getAlignment()) {
    // Check to see if this allocation is only modified by a memcpy/memmove from
    // a constant global whose alignment is equal to or exceeds that of the
    // allocation.  If this is the case, we can change all users to use
    // the constant global instead.  This is commonly produced by the CFE by
    // constructs like "void foo() { int A[] = {1,2,3,4,5,6,7,8,9...}; }" if 'A'
    // is only subsequently read.
    SmallVector<Instruction *, 4> ToDelete;
    if (MemTransferInst *Copy = isOnlyCopiedFromConstantGlobal(&AI, ToDelete)) {
      unsigned SourceAlign = getOrEnforceKnownAlignment(
          Copy->getSource(), AI.getAlignment(), DL, &AI, AC, DT);
      if (AI.getAlignment() <= SourceAlign) {
        DEBUG(dbgs() << "Found alloca equal to global: " << AI << '\n');
        DEBUG(dbgs() << "  memcpy = " << *Copy << '\n');
        for (unsigned i = 0, e = ToDelete.size(); i != e; ++i)
          EraseInstFromFunction(*ToDelete[i]);
        Constant *TheSrc = cast<Constant>(Copy->getSource());
        Constant *Cast
          = ConstantExpr::getPointerBitCastOrAddrSpaceCast(TheSrc, AI.getType());
        Instruction *NewI = ReplaceInstUsesWith(AI, Cast);
        EraseInstFromFunction(*Copy);
        ++NumGlobalCopies;
        return NewI;
      }
    }
  }

  // At last, use the generic allocation site handler to aggressively remove
  // unused allocas.
  return visitAllocSite(AI);
}

// emit code to unpack a raw value from a box into registers or a stack slot
static Value *emit_unbox(jl_codectx_t &ctx, Type *to, const jl_cgval_t &x, jl_value_t *jt, Value *dest, bool volatile_store)
{
    assert(to != T_void);
    // TODO: fully validate that x.typ == jt?
    if (x.isghost) {
        // this can happen when a branch yielding a different type ends
        // up being dead code, and type inference knows that the other
        // branch's type is the only one that matters.
        if (type_is_ghost(to)) {
            return NULL;
        }
        //emit_error(ctx, "emit_unbox: a type mismatch error in occurred during codegen");
        return UndefValue::get(to); // type mismatch error
    }

    Constant *c = x.constant ? julia_const_to_llvm(x.constant) : NULL;
    if (!x.ispointer() || c) { // already unboxed, but sometimes need conversion
        Value *unboxed = emit_unboxed_coercion(ctx, to, c ? c : x.V);
        if (!dest)
            return unboxed;
        Type *dest_ty = unboxed->getType()->getPointerTo();
        if (dest->getType() != dest_ty)
            dest = emit_bitcast(ctx, dest, dest_ty);
        ctx.builder.CreateStore(unboxed, dest, volatile_store);
        return NULL;
    }

    // bools stored as int8, so an extra Trunc is needed to get an int1
    Value *p = x.constant ? literal_pointer_val(ctx, x.constant) : x.V;
    Type *ptype = (to == T_int1 ? T_pint8 : to->getPointerTo());

    Value *unboxed = NULL;
    if (to == T_int1)
        unboxed = ctx.builder.CreateTrunc(tbaa_decorate(x.tbaa, ctx.builder.CreateLoad(maybe_bitcast(ctx, p, ptype))), T_int1);
    else if (jt == (jl_value_t*)jl_bool_type)
        unboxed = ctx.builder.CreateZExt(ctx.builder.CreateTrunc(tbaa_decorate(x.tbaa, ctx.builder.CreateLoad(maybe_bitcast(ctx, p, ptype))), T_int1), to);
    if (unboxed) {
        if (!dest)
            return unboxed;
        ctx.builder.CreateStore(unboxed, dest);
        return NULL;
    }

    unsigned alignment = julia_alignment(jt, 0);
    if (dest) {
        MDNode *tbaa = x.tbaa;
        // the memcpy intrinsic does not allow to specify different alias tags
        // for the load part (x.tbaa) and the store part (tbaa_stack).
        // since the tbaa lattice has to be a tree we have unfortunately
        // x.tbaa ∪ tbaa_stack = tbaa_root if x.tbaa != tbaa_stack
        if (tbaa != tbaa_stack)
            tbaa = NULL;
        emit_memcpy(ctx, dest, p, jl_datatype_size(jt), alignment, volatile_store, tbaa);
        return NULL;
    }
    else {
        if (p->getType() != ptype && isa<AllocaInst>(p)) {
            // LLVM's mem2reg can't handle coercion if the load/store type does
            // not match the type of the alloca. As such, it is better to
            // perform the load using the alloca's type and then perform the
            // appropriate coercion manually.
            AllocaInst *AI = cast<AllocaInst>(p);
            Type *AllocType = AI->getAllocatedType();
#if JL_LLVM_VERSION >= 40000
            const DataLayout &DL = jl_data_layout;
#else
            const DataLayout &DL = jl_ExecutionEngine->getDataLayout();
#endif
            if (!AI->isArrayAllocation() &&
                    (AllocType->isFloatingPointTy() || AllocType->isIntegerTy() || AllocType->isPointerTy()) &&
                    (to->isFloatingPointTy() || to->isIntegerTy() || to->isPointerTy()) &&
                    DL.getTypeSizeInBits(AllocType) == DL.getTypeSizeInBits(to)) {
                Instruction *load = ctx.builder.CreateAlignedLoad(p, alignment);
                return emit_unboxed_coercion(ctx, to, tbaa_decorate(x.tbaa, load));
            }
        }
        p = maybe_bitcast(ctx, p, ptype);
        Instruction *load = ctx.builder.CreateAlignedLoad(p, alignment);
        return tbaa_decorate(x.tbaa, load);
    }
}

/// If it is possible to inline the specified call site,
/// do so and update the CallGraph for this operation.
///
/// This function also does some basic book-keeping to update the IR.  The
/// InlinedArrayAllocas map keeps track of any allocas that are already
/// available from other functions inlined into the caller.  If we are able to
/// inline this call site we attempt to reuse already available allocas or add
/// any new allocas to the set if not possible.
static bool InlineCallIfPossible(CallSite CS, InlineFunctionInfo &IFI,
                                 InlinedArrayAllocasTy &InlinedArrayAllocas,
                                 int InlineHistory, bool InsertLifetime) {
  Function *Callee = CS.getCalledFunction();
  Function *Caller = CS.getCaller();

  // Try to inline the function.  Get the list of static allocas that were
  // inlined.
  if (!InlineFunction(CS, IFI, InsertLifetime))
    return false;

  AdjustCallerSSPLevel(Caller, Callee);

  // Look at all of the allocas that we inlined through this call site.  If we
  // have already inlined other allocas through other calls into this function,
  // then we know that they have disjoint lifetimes and that we can merge them.
  //
  // There are many heuristics possible for merging these allocas, and the
  // different options have different tradeoffs.  One thing that we *really*
  // don't want to hurt is SRoA: once inlining happens, often allocas are no
  // longer address taken and so they can be promoted.
  //
  // Our "solution" for that is to only merge allocas whose outermost type is an
  // array type.  These are usually not promoted because someone is using a
  // variable index into them.  These are also often the most important ones to
  // merge.
  //
  // A better solution would be to have real memory lifetime markers in the IR
  // and not have the inliner do any merging of allocas at all.  This would
  // allow the backend to do proper stack slot coloring of all allocas that
  // *actually make it to the backend*, which is really what we want.
  //
  // Because we don't have this information, we do this simple and useful hack.
  //
  SmallPtrSet<AllocaInst*, 16> UsedAllocas;
  
  // When processing our SCC, check to see if CS was inlined from some other
  // call site.  For example, if we're processing "A" in this code:
  //   A() { B() }
  //   B() { x = alloca ... C() }
  //   C() { y = alloca ... }
  // Assume that C was not inlined into B initially, and so we're processing A
  // and decide to inline B into A.  Doing this makes an alloca available for
  // reuse and makes a callsite (C) available for inlining.  When we process
  // the C call site we don't want to do any alloca merging between X and Y
  // because their scopes are not disjoint.  We could make this smarter by
  // keeping track of the inline history for each alloca in the
  // InlinedArrayAllocas but this isn't likely to be a significant win.
  if (InlineHistory != -1)  // Only do merging for top-level call sites in SCC.
    return true;
  
  // Loop over all the allocas we have so far and see if they can be merged with
  // a previously inlined alloca.  If not, remember that we had it.
  for (unsigned AllocaNo = 0, e = IFI.StaticAllocas.size();
       AllocaNo != e; ++AllocaNo) {
    AllocaInst *AI = IFI.StaticAllocas[AllocaNo];
    
    // Don't bother trying to merge array allocations (they will usually be
    // canonicalized to be an allocation *of* an array), or allocations whose
    // type is not itself an array (because we're afraid of pessimizing SRoA).
    ArrayType *ATy = dyn_cast<ArrayType>(AI->getAllocatedType());
    if (!ATy || AI->isArrayAllocation())
      continue;
    
    // Get the list of all available allocas for this array type.
    std::vector<AllocaInst*> &AllocasForType = InlinedArrayAllocas[ATy];
    
    // Loop over the allocas in AllocasForType to see if we can reuse one.  Note
    // that we have to be careful not to reuse the same "available" alloca for
    // multiple different allocas that we just inlined, we use the 'UsedAllocas'
    // set to keep track of which "available" allocas are being used by this
    // function.  Also, AllocasForType can be empty of course!
    bool MergedAwayAlloca = false;
    for (unsigned i = 0, e = AllocasForType.size(); i != e; ++i) {
      AllocaInst *AvailableAlloca = AllocasForType[i];

      unsigned Align1 = AI->getAlignment(),
               Align2 = AvailableAlloca->getAlignment();
      
      // The available alloca has to be in the right function, not in some other
      // function in this SCC.
      if (AvailableAlloca->getParent() != AI->getParent())
        continue;
      
      // If the inlined function already uses this alloca then we can't reuse
      // it.
      if (!UsedAllocas.insert(AvailableAlloca).second)
        continue;
      
      // Otherwise, we *can* reuse it, RAUW AI into AvailableAlloca and declare
      // success!
      DEBUG(dbgs() << "    ***MERGED ALLOCA: " << *AI << "\n\t\tINTO: "
//.........这里部分代码省略.........