yaxunl updated this revision to Diff 81309.
yaxunl added a comment.
Cast alloca to default address space.
https://reviews.llvm.org/D27627
Files:
include/clang/AST/ASTContext.h
include/clang/Basic/TargetInfo.h
lib/AST/ASTContext.cpp
lib/Basic/Targets.cpp
lib/CodeGen/CGCall.cpp
lib/CodeGen/CGDecl.cpp
lib/CodeGen/CGExpr.cpp
lib/CodeGen/CodeGenFunction.h
test/CodeGen/default-addr-space.cpp
Index: test/CodeGen/default-addr-space.cpp
===================================================================
--- /dev/null
+++ test/CodeGen/default-addr-space.cpp
@@ -0,0 +1,33 @@
+// RUN: %clang_cc1 -O0 -std=c++11 -emit-llvm -o - -triple amdgcn-- %s | FileCheck %s
+
+// CHECK: %struct.ATy = type { i32 addrspace(4)* }
+struct ATy {
+ int *p;
+};
+
+// CHECK-LABEL: @_Z1fPi(i32 addrspace(4)* %a)
+void f(int* a) {
+ // CHECK: %[[a_addr_0:.*]] = alloca i32 addrspace(4)*
+ // CHECK: %[[a_addr:.*]] = addrspacecast i32 addrspace(4)** %[[a_addr_0]] to i32 addrspace(4)* addrspace(4)*
+ // CHECK: %[[b0:.*]] = alloca i32
+ // CHECK: %[[b:.*]] = addrspacecast i32* %[[b0]] to i32 addrspace(4)*
+ // CHECK: %[[A0:.*]] = alloca %struct.ATy, align 4
+ // CHECK: %[[A:.*]] = addrspacecast %struct.ATy* %[[A0]] to %struct.ATy addrspace(4)*
+
+ // CHECK: store i32 addrspace(4)* %a, i32 addrspace(4)* addrspace(4)* %[[a_addr]]
+
+ // CHECK: store i32 1, i32 addrspace(4)* %[[b]]
+ int b = 1;
+
+ // CHECK: %[[p:.*]] = getelementptr inbounds %struct.ATy, %struct.ATy addrspace(4)* %[[A]], i32 0, i32 0
+ // CHECK: store i32 addrspace(4)* %[[b]], i32 addrspace(4)* addrspace(4)* %[[p]], align 4
+ ATy A{&b};
+
+ // CHECK: %[[r0:.*]] = load i32, i32 addrspace(4)* %b
+ // CHECK: %[[r1:.*]] = load i32 addrspace(4)*, i32 addrspace(4)* addrspace(4)* %[[a_addr]]
+ // CHECK: store i32 %[[r0]], i32 addrspace(4)* %[[r1]]
+ *a = b;
+
+ // CHECK: store i32 addrspace(4)* %[[b]], i32 addrspace(4)* addrspace(4)* %[[a_addr]], align 4
+ a = &b;
+}
Index: lib/CodeGen/CodeGenFunction.h
===================================================================
--- lib/CodeGen/CodeGenFunction.h
+++ lib/CodeGen/CodeGenFunction.h
@@ -344,7 +344,7 @@
};
/// i32s containing the indexes of the cleanup destinations.
- llvm::AllocaInst *NormalCleanupDest;
+ llvm::Instruction *NormalCleanupDest;
unsigned NextCleanupDestIndex;
@@ -359,8 +359,8 @@
llvm::Value *ExceptionSlot;
/// The selector slot. Under the MandatoryCleanup model, all landing pads
- /// write the current selector value into this alloca.
- llvm::AllocaInst *EHSelectorSlot;
+ /// write the current selector value into this instruction.
+ llvm::Instruction *EHSelectorSlot;
/// A stack of exception code slots. Entering an __except block pushes a slot
/// on the stack and leaving pops one. The __exception_code() intrinsic loads
@@ -395,11 +395,11 @@
/// An i1 variable indicating whether or not the @finally is
/// running for an exception.
- llvm::AllocaInst *ForEHVar;
+ llvm::Instruction *ForEHVar;
/// An i8* variable into which the exception pointer to rethrow
/// has been saved.
- llvm::AllocaInst *SavedExnVar;
+ llvm::Instruction *SavedExnVar;
public:
void enter(CodeGenFunction &CGF, const Stmt *Finally,
@@ -1788,14 +1788,23 @@
AlignmentSource *Source = nullptr);
LValue EmitLoadOfPointerLValue(Address Ptr, const PointerType *PtrTy);
+ /// Create an alloca instruction. If the default address space is not 0,
+ /// insert addrspacecast instruction which casts the alloca instruction
+ /// to the default address space.
+ llvm::Instruction *CreateAlloca(llvm::Type *Ty, const Twine &Name = "tmp",
+ llvm::Instruction *InsertPos = nullptr);
/// CreateTempAlloca - This creates a alloca and inserts it into the entry
/// block. The caller is responsible for setting an appropriate alignment on
- /// the alloca.
- llvm::AllocaInst *CreateTempAlloca(llvm::Type *Ty,
- const Twine &Name = "tmp");
+ /// the alloca. If the default address space is not 0, insert addrspacecast.
+ llvm::Instruction *CreateTempAlloca(llvm::Type *Ty,
+ const Twine &Name = "tmp");
Address CreateTempAlloca(llvm::Type *Ty, CharUnits align,
const Twine &Name = "tmp");
+ /// Get alloca instruction operand of an addrspacecast instruction.
+ /// If \p Inst is alloca instruction, returns \p Inst;
+ llvm::AllocaInst *getAddrSpaceCastedAlloca(llvm::Instruction *Inst) const;
+
/// CreateDefaultAlignedTempAlloca - This creates an alloca with the
/// default ABI alignment of the given LLVM type.
///
Index: lib/CodeGen/CGExpr.cpp
===================================================================
--- lib/CodeGen/CGExpr.cpp
+++ lib/CodeGen/CGExpr.cpp
@@ -59,16 +59,37 @@
/// block.
Address CodeGenFunction::CreateTempAlloca(llvm::Type *Ty, CharUnits Align,
const Twine &Name) {
- auto Alloca = CreateTempAlloca(Ty, Name);
+ auto CastedAlloca = CreateTempAlloca(Ty, Name);
+ auto *Alloca = getAddrSpaceCastedAlloca(CastedAlloca);
Alloca->setAlignment(Align.getQuantity());
- return Address(Alloca, Align);
+ return Address(CastedAlloca, Align);
}
/// CreateTempAlloca - This creates a alloca and inserts it into the entry
/// block.
-llvm::AllocaInst *CodeGenFunction::CreateTempAlloca(llvm::Type *Ty,
- const Twine &Name) {
- return new llvm::AllocaInst(Ty, nullptr, Name, AllocaInsertPt);
+llvm::Instruction *CodeGenFunction::CreateTempAlloca(llvm::Type *Ty,
+ const Twine &Name) {
+ return CreateAlloca(Ty, Name, AllocaInsertPt);
+}
+
+llvm::Instruction *CodeGenFunction::CreateAlloca(llvm::Type *Ty,
+ const Twine &Name,
+ llvm::Instruction *InsertPos) {
+ llvm::Instruction *V = new llvm::AllocaInst(Ty, nullptr, Name, InsertPos);
+ auto DefaultAddr = getTarget().getDefaultTargetAddressSpace(getLangOpts());
+ if (DefaultAddr != 0) {
+ auto *DestTy = llvm::PointerType::get(V->getType()->getPointerElementType(),
+ DefaultAddr);
+ V = new llvm::AddrSpaceCastInst(V, DestTy, "", InsertPos);
+ }
+ return V;
+}
+
+llvm::AllocaInst *
+CodeGenFunction::getAddrSpaceCastedAlloca(llvm::Instruction *V) const {
+ if (auto *Cast = dyn_cast<llvm::AddrSpaceCastInst>(V))
+ return cast<llvm::AllocaInst>(Cast->getOperand(0));
+ return cast<llvm::AllocaInst>(V);
}
/// CreateDefaultAlignTempAlloca - This creates an alloca with the
Index: lib/CodeGen/CGDecl.cpp
===================================================================
--- lib/CodeGen/CGDecl.cpp
+++ lib/CodeGen/CGDecl.cpp
@@ -1060,7 +1060,15 @@
llvm::AllocaInst *vla = Builder.CreateAlloca(llvmTy, elementCount, "vla");
vla->setAlignment(alignment.getQuantity());
- address = Address(vla, alignment);
+ llvm::Value *V = vla;
+ auto DefaultAddr = getTarget().getDefaultTargetAddressSpace(getLangOpts());
+ if (DefaultAddr != 0) {
+ auto *DestTy =
+ llvm::PointerType::get(vla->getType()->getElementType(), DefaultAddr);
+ V = Builder.CreateAddrSpaceCast(vla, DestTy);
+ }
+
+ address = Address(V, alignment);
}
setAddrOfLocalVar(&D, address);
Index: lib/CodeGen/CGCall.cpp
===================================================================
--- lib/CodeGen/CGCall.cpp
+++ lib/CodeGen/CGCall.cpp
@@ -3567,18 +3567,19 @@
if (llvm::StructType *ArgStruct = CallInfo.getArgStruct()) {
ArgMemoryLayout = CGM.getDataLayout().getStructLayout(ArgStruct);
llvm::Instruction *IP = CallArgs.getStackBase();
- llvm::AllocaInst *AI;
+ llvm::Instruction *CastedAI;
if (IP) {
IP = IP->getNextNode();
- AI = new llvm::AllocaInst(ArgStruct, "argmem", IP);
+ CastedAI = CreateAlloca(ArgStruct, "argmem", IP);
} else {
- AI = CreateTempAlloca(ArgStruct, "argmem");
+ CastedAI = CreateTempAlloca(ArgStruct, "argmem");
}
auto Align = CallInfo.getArgStructAlignment();
+ auto *AI = getAddrSpaceCastedAlloca(CastedAI);
AI->setAlignment(Align.getQuantity());
AI->setUsedWithInAlloca(true);
assert(AI->isUsedWithInAlloca() && !AI->isStaticAlloca());
- ArgMemory = Address(AI, Align);
+ ArgMemory = Address(CastedAI, Align);
}
// Helper function to drill into the inalloca allocation.
Index: lib/Basic/Targets.cpp
===================================================================
--- lib/Basic/Targets.cpp
+++ lib/Basic/Targets.cpp
@@ -1978,16 +1978,6 @@
}
};
-static const unsigned AMDGPUAddrSpaceMap[] = {
- 1, // opencl_global
- 3, // opencl_local
- 2, // opencl_constant
- 4, // opencl_generic
- 1, // cuda_device
- 2, // cuda_constant
- 3 // cuda_shared
-};
-
// If you edit the description strings, make sure you update
// getPointerWidthV().
@@ -2001,9 +1991,18 @@
"-v192:256-v256:256-v512:512-v1024:1024-v2048:2048-n32:64";
class AMDGPUTargetInfo final : public TargetInfo {
+ static const unsigned AddrSpaceMap_[7];
static const Builtin::Info BuiltinInfo[];
static const char * const GCCRegNames[];
+ enum AddrSpaceKind {
+ AS_Private = 0,
+ AS_Global = 1,
+ AS_Constant = 2,
+ AS_Local = 3,
+ AS_Generic = 4
+ };
+
/// \brief The GPU profiles supported by the AMDGPU target.
enum GPUKind {
GK_NONE,
@@ -2046,7 +2045,7 @@
resetDataLayout(getTriple().getArch() == llvm::Triple::amdgcn ?
DataLayoutStringSI : DataLayoutStringR600);
- AddrSpaceMap = &AMDGPUAddrSpaceMap;
+ AddrSpaceMap = &AddrSpaceMap_;
UseAddrSpaceMapMangling = true;
}
@@ -2232,6 +2231,15 @@
}
}
+ unsigned
+ getDefaultTargetAddressSpace(const LangOptions &Opts) const override {
+ // OpenCL sets address space explicitly in AST. The default case (type
+ // qualifier containing no address space) represents private address space.
+ if (Opts.OpenCL)
+ return AS_Private;
+ return AS_Generic;
+ }
+
LangAS::ID getOpenCLImageAddrSpace() const override {
return LangAS::opencl_constant;
}
@@ -2254,6 +2262,16 @@
}
};
+const unsigned AMDGPUTargetInfo::AddrSpaceMap_[] = {
+ AS_Global, // opencl_global
+ AS_Local, // opencl_local
+ AS_Constant, // opencl_constant
+ AS_Generic, // opencl_generic
+ AS_Global, // cuda_device
+ AS_Constant, // cuda_constant
+ AS_Local // cuda_shared
+};
+
const Builtin::Info AMDGPUTargetInfo::BuiltinInfo[] = {
#define BUILTIN(ID, TYPE, ATTRS) \
{ #ID, TYPE, ATTRS, nullptr, ALL_LANGUAGES, nullptr },
Index: lib/AST/ASTContext.cpp
===================================================================
--- lib/AST/ASTContext.cpp
+++ lib/AST/ASTContext.cpp
@@ -9436,6 +9436,12 @@
return getTargetInfo().getNullPointerValue(AS);
}
+unsigned ASTContext::getTargetAddressSpace(Qualifiers Q) const {
+ return Q.hasAddressSpace()
+ ? getTargetAddressSpace(Q.getAddressSpace())
+ : getTargetInfo().getDefaultTargetAddressSpace(LangOpts);
+}
+
// Explicitly instantiate this in case a Redeclarable<T> is used from a TU that
// doesn't include ASTContext.h
template
Index: include/clang/Basic/TargetInfo.h
===================================================================
--- include/clang/Basic/TargetInfo.h
+++ include/clang/Basic/TargetInfo.h
@@ -945,6 +945,10 @@
return *AddrSpaceMap;
}
+ virtual unsigned getDefaultTargetAddressSpace(const LangOptions &Opt) const {
+ return 0;
+ }
+
/// \brief Retrieve the name of the platform as it is used in the
/// availability attribute.
StringRef getPlatformName() const { return PlatformName; }
Index: include/clang/AST/ASTContext.h
===================================================================
--- include/clang/AST/ASTContext.h
+++ include/clang/AST/ASTContext.h
@@ -2284,9 +2284,7 @@
return getTargetAddressSpace(T.getQualifiers());
}
- unsigned getTargetAddressSpace(Qualifiers Q) const {
- return getTargetAddressSpace(Q.getAddressSpace());
- }
+ unsigned getTargetAddressSpace(Qualifiers Q) const;
unsigned getTargetAddressSpace(unsigned AS) const {
if (AS < LangAS::Offset || AS >= LangAS::Offset + LangAS::Count)
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
