r315453 - [OpenCL] Allow function declaration with empty argument list.
Author: bader
Date: Wed Oct 11 04:16:31 2017
New Revision: 315453
URL: http://llvm.org/viewvc/llvm-project?rev=315453&view=rev
Log:
[OpenCL] Allow function declaration with empty argument list.
Treat 'f()' as 'f(void)' rather than a function w/o a prototype.
Reviewers: Anastasia, yaxunl
Reviewed By: Anastasia, yaxunl
Subscribers: cfe-commits, echuraev, chapuni
Differential Revision: https://reviews.llvm.org/D33681
Re-apply revision 306653.
Modified:
cfe/trunk/lib/Parse/ParseDecl.cpp
cfe/trunk/lib/Sema/SemaType.cpp
cfe/trunk/test/SemaOpenCL/func.cl
cfe/trunk/test/SemaOpenCL/invalid-pipes-cl2.0.cl
Modified: cfe/trunk/lib/Parse/ParseDecl.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Parse/ParseDecl.cpp?rev=315453&r1=315452&r2=315453&view=diff
==
--- cfe/trunk/lib/Parse/ParseDecl.cpp (original)
+++ cfe/trunk/lib/Parse/ParseDecl.cpp Wed Oct 11 04:16:31 2017
@@ -5989,7 +5989,8 @@ void Parser::ParseFunctionDeclarator(Dec
else if (RequiresArg)
Diag(Tok, diag::err_argument_required_after_attribute);
-HasProto = ParamInfo.size() || getLangOpts().CPlusPlus;
+HasProto = ParamInfo.size() || getLangOpts().CPlusPlus
+|| getLangOpts().OpenCL;
// If we have the closing ')', eat it.
Tracker.consumeClose();
Modified: cfe/trunk/lib/Sema/SemaType.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Sema/SemaType.cpp?rev=315453&r1=315452&r2=315453&view=diff
==
--- cfe/trunk/lib/Sema/SemaType.cpp (original)
+++ cfe/trunk/lib/Sema/SemaType.cpp Wed Oct 11 04:16:31 2017
@@ -4460,7 +4460,8 @@ static TypeSourceInfo *GetFullTypeForDec
FunctionType::ExtInfo EI(getCCForDeclaratorChunk(S, D, FTI, chunkIndex));
- if (!FTI.NumParams && !FTI.isVariadic && !LangOpts.CPlusPlus) {
+ if (!FTI.NumParams && !FTI.isVariadic && !LangOpts.CPlusPlus
+&& !LangOpts.OpenCL) {
// Simple void foo(), where the incoming T is the result type.
T = Context.getFunctionNoProtoType(T, EI);
} else {
Modified: cfe/trunk/test/SemaOpenCL/func.cl
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/SemaOpenCL/func.cl?rev=315453&r1=315452&r2=315453&view=diff
==
--- cfe/trunk/test/SemaOpenCL/func.cl (original)
+++ cfe/trunk/test/SemaOpenCL/func.cl Wed Oct 11 04:16:31 2017
@@ -1,4 +1,4 @@
-// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only
+// RUN: %clang_cc1 %s -verify -pedantic -fsyntax-only -triple
spir-unknown-unknown
// Variadic functions
void vararg_f(int, ...);// expected-error {{invalid
prototype, variadic arguments are not allowed in OpenCL}}
@@ -16,6 +16,9 @@ typedef struct s
//Function pointer
void foo(void*);
+// Expect no diagnostics for an empty parameter list.
+void bar();
+
void bar()
{
// declaring a function pointer is an error
Modified: cfe/trunk/test/SemaOpenCL/invalid-pipes-cl2.0.cl
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/SemaOpenCL/invalid-pipes-cl2.0.cl?rev=315453&r1=315452&r2=315453&view=diff
==
--- cfe/trunk/test/SemaOpenCL/invalid-pipes-cl2.0.cl (original)
+++ cfe/trunk/test/SemaOpenCL/invalid-pipes-cl2.0.cl Wed Oct 11 04:16:31 2017
@@ -3,7 +3,7 @@
global pipe int gp;// expected-error {{type '__global read_only
pipe int' can only be used as a function parameter in OpenCL}}
global reserve_id_t rid; // expected-error {{the '__global
reserve_id_t' type cannot be used to declare a program scope variable}}
-extern pipe write_only int get_pipe(); // expected-error {{type '__global
write_only pipe int ()' can only be used as a function parameter in OpenCL}}
+extern pipe write_only int get_pipe(); // expected-error {{type '__global
write_only pipe int (void)' can only be used as a function parameter in OpenCL}}
kernel void test_invalid_reserved_id(reserve_id_t ID) { // expected-error
{{'reserve_id_t' cannot be used as the type of a kernel parameter}}
}
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
r304575 - [OpenCL] Harden function pointer diagnostics.
Author: bader
Date: Fri Jun 2 13:08:58 2017
New Revision: 304575
URL: http://llvm.org/viewvc/llvm-project?rev=304575&view=rev
Log:
[OpenCL] Harden function pointer diagnostics.
Summary: Improve OpenCL type checking by rejecting function pointer types.
Reviewers: Anastasia, yaxunl
Reviewed By: Anastasia
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D33821
Modified:
cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td
cfe/trunk/lib/Sema/SemaDecl.cpp
cfe/trunk/lib/Sema/SemaType.cpp
cfe/trunk/test/SemaOpenCL/func.cl
Modified: cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td?rev=304575&r1=304574&r2=304575&view=diff
==
--- cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td (original)
+++ cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td Fri Jun 2 13:08:58
2017
@@ -7268,7 +7268,7 @@ def err_invalid_conversion_between_vecto
"invalid conversion between vector type %0 and integer type %1 "
"of different size">;
-def err_opencl_function_pointer_variable : Error<
+def err_opencl_function_pointer : Error<
"pointers to functions are not allowed">;
def err_opencl_taking_function_address : Error<
Modified: cfe/trunk/lib/Sema/SemaDecl.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Sema/SemaDecl.cpp?rev=304575&r1=304574&r2=304575&view=diff
==
--- cfe/trunk/lib/Sema/SemaDecl.cpp (original)
+++ cfe/trunk/lib/Sema/SemaDecl.cpp Fri Jun 2 13:08:58 2017
@@ -6160,7 +6160,7 @@ NamedDecl *Sema::ActOnVariableDeclarator
QualType NR = R;
while (NR->isPointerType()) {
if (NR->isFunctionPointerType()) {
-Diag(D.getIdentifierLoc(), diag::err_opencl_function_pointer_variable);
+Diag(D.getIdentifierLoc(), diag::err_opencl_function_pointer);
D.setInvalidType();
break;
}
Modified: cfe/trunk/lib/Sema/SemaType.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Sema/SemaType.cpp?rev=304575&r1=304574&r2=304575&view=diff
==
--- cfe/trunk/lib/Sema/SemaType.cpp (original)
+++ cfe/trunk/lib/Sema/SemaType.cpp Fri Jun 2 13:08:58 2017
@@ -1881,6 +1881,11 @@ QualType Sema::BuildPointerType(QualType
return QualType();
}
+ if (T->isFunctionType() && getLangOpts().OpenCL) {
+Diag(Loc, diag::err_opencl_function_pointer);
+return QualType();
+ }
+
if (checkQualifiedFunction(*this, T, Loc, QFK_Pointer))
return QualType();
Modified: cfe/trunk/test/SemaOpenCL/func.cl
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/SemaOpenCL/func.cl?rev=304575&r1=304574&r2=304575&view=diff
==
--- cfe/trunk/test/SemaOpenCL/func.cl (original)
+++ cfe/trunk/test/SemaOpenCL/func.cl Fri Jun 2 13:08:58 2017
@@ -4,8 +4,15 @@
void vararg_f(int, ...);// expected-error {{invalid
prototype, variadic arguments are not allowed in OpenCL}}
void __vararg_f(int, ...);
typedef void (*vararg_fptr_t)(int, ...);// expected-error {{invalid
prototype, variadic arguments are not allowed in OpenCL}}
+// expected-error@-1{{pointers to
functions are not allowed}}
int printf(__constant const char *st, ...); // expected-error {{invalid
prototype, variadic arguments are not allowed in OpenCL}}
+// Struct type with function pointer field
+typedef struct s
+{
+ void (*f)(struct s *self, int *i); // expected-error{{pointers to
functions are not allowed}}
+} s_t;
+
//Function pointer
void foo(void*);
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
r305796 - [OpenCL] Fix OpenCL and SPIR version metadata generation.
Author: bader
Date: Tue Jun 20 09:30:18 2017
New Revision: 305796
URL: http://llvm.org/viewvc/llvm-project?rev=305796&view=rev
Log:
[OpenCL] Fix OpenCL and SPIR version metadata generation.
Summary: OpenCL and SPIR version metadata must be generated once per module
instead of once per mangled global value.
Reviewers: Anastasia, yaxunl
Reviewed By: Anastasia
Subscribers: ahatanak, cfe-commits
Differential Revision: https://reviews.llvm.org/D34235
Modified:
cfe/trunk/lib/CodeGen/CodeGenModule.cpp
cfe/trunk/lib/CodeGen/CodeGenModule.h
cfe/trunk/lib/CodeGen/TargetInfo.cpp
cfe/trunk/test/CodeGenOpenCL/spir_version.cl
Modified: cfe/trunk/lib/CodeGen/CodeGenModule.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/CodeGen/CodeGenModule.cpp?rev=305796&r1=305795&r2=305796&view=diff
==
--- cfe/trunk/lib/CodeGen/CodeGenModule.cpp (original)
+++ cfe/trunk/lib/CodeGen/CodeGenModule.cpp Tue Jun 20 09:30:18 2017
@@ -506,6 +506,26 @@ void CodeGenModule::Release() {
LangOpts.CUDADeviceFlushDenormalsToZero ? 1 : 0);
}
+ // Emit OpenCL specific module metadata: OpenCL/SPIR version.
+ if (LangOpts.OpenCL) {
+EmitOpenCLMetadata();
+// Emit SPIR version.
+if (getTriple().getArch() == llvm::Triple::spir ||
+getTriple().getArch() == llvm::Triple::spir64) {
+ // SPIR v2.0 s2.12 - The SPIR version used by the module is stored in the
+ // opencl.spir.version named metadata.
+ llvm::Metadata *SPIRVerElts[] = {
+ llvm::ConstantAsMetadata::get(llvm::ConstantInt::get(
+ Int32Ty, LangOpts.OpenCLVersion / 100)),
+ llvm::ConstantAsMetadata::get(llvm::ConstantInt::get(
+ Int32Ty, (LangOpts.OpenCLVersion / 100 > 1) ? 0 : 2))};
+ llvm::NamedMDNode *SPIRVerMD =
+ TheModule.getOrInsertNamedMetadata("opencl.spir.version");
+ llvm::LLVMContext &Ctx = TheModule.getContext();
+ SPIRVerMD->addOperand(llvm::MDNode::get(Ctx, SPIRVerElts));
+}
+ }
+
if (uint32_t PLevel = Context.getLangOpts().PICLevel) {
assert(PLevel < 3 && "Invalid PIC Level");
getModule().setPICLevel(static_cast(PLevel));
@@ -529,6 +549,20 @@ void CodeGenModule::Release() {
EmitTargetMetadata();
}
+void CodeGenModule::EmitOpenCLMetadata() {
+ // SPIR v2.0 s2.13 - The OpenCL version used by the module is stored in the
+ // opencl.ocl.version named metadata node.
+ llvm::Metadata *OCLVerElts[] = {
+ llvm::ConstantAsMetadata::get(llvm::ConstantInt::get(
+ Int32Ty, LangOpts.OpenCLVersion / 100)),
+ llvm::ConstantAsMetadata::get(llvm::ConstantInt::get(
+ Int32Ty, (LangOpts.OpenCLVersion % 100) / 10))};
+ llvm::NamedMDNode *OCLVerMD =
+ TheModule.getOrInsertNamedMetadata("opencl.ocl.version");
+ llvm::LLVMContext &Ctx = TheModule.getContext();
+ OCLVerMD->addOperand(llvm::MDNode::get(Ctx, OCLVerElts));
+}
+
void CodeGenModule::UpdateCompletedType(const TagDecl *TD) {
// Make sure that this type is translated.
Types.UpdateCompletedType(TD);
Modified: cfe/trunk/lib/CodeGen/CodeGenModule.h
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/CodeGen/CodeGenModule.h?rev=305796&r1=305795&r2=305796&view=diff
==
--- cfe/trunk/lib/CodeGen/CodeGenModule.h (original)
+++ cfe/trunk/lib/CodeGen/CodeGenModule.h Tue Jun 20 09:30:18 2017
@@ -1321,6 +1321,9 @@ private:
/// Emits target specific Metadata for global declarations.
void EmitTargetMetadata();
+ /// Emits OpenCL specific Metadata e.g. OpenCL version.
+ void EmitOpenCLMetadata();
+
/// Emit the llvm.gcov metadata used to tell LLVM where to emit the .gcno and
/// .gcda files in a way that persists in .bc files.
void EmitCoverageFile();
Modified: cfe/trunk/lib/CodeGen/TargetInfo.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/CodeGen/TargetInfo.cpp?rev=305796&r1=305795&r2=305796&view=diff
==
--- cfe/trunk/lib/CodeGen/TargetInfo.cpp (original)
+++ cfe/trunk/lib/CodeGen/TargetInfo.cpp Tue Jun 20 09:30:18 2017
@@ -7344,8 +7344,6 @@ public:
};
}
-static void appendOpenCLVersionMD (CodeGen::CodeGenModule &CGM);
-
void AMDGPUTargetCodeGenInfo::setTargetAttributes(
const Decl *D,
llvm::GlobalValue *GV,
@@ -7402,8 +7400,6 @@ void AMDGPUTargetCodeGenInfo::setTargetA
if (NumVGPR != 0)
F->addFnAttr("amdgpu-num-vgpr", llvm::utostr(NumVGPR));
}
-
- appendOpenCLVersionMD(M);
}
unsigned AMDGPUTargetCodeGenInfo::getOpenCLKernelCallingConv() const {
@@ -8074,8 +8070,6 @@ class SPIRTargetCodeGenInfo : public Tar
public:
SPIRTargetCodeGenInfo(CodeGen::CodeGenTypes &CGT)
: TargetCodeGenInfo(new DefaultABIInfo(CGT)) {}
- void emitTargetMD(const Decl *D, llvm::GlobalValue *GV,
-CodeGen::CodeGenModule &M)
r335103 - [Sema] Allow creating types with multiple of the same addrspace.
Author: bader
Date: Wed Jun 20 01:31:24 2018
New Revision: 335103
URL: http://llvm.org/viewvc/llvm-project?rev=335103&view=rev
Log:
[Sema] Allow creating types with multiple of the same addrspace.
Summary:
The comment with the OpenCL clause about this clearly
says: "No type shall be qualified by qualifiers for
two or more different address spaces."
This must mean that two or more qualifiers for the
_same_ address space is allowed. However, it is
likely unintended by the programmer, so emit a
warning.
For dependent address space types, reject them like
before since we cannot know what the address space
will be.
Patch by Bevin Hansson (ebevhan).
Reviewers: Anastasia
Reviewed By: Anastasia
Subscribers: bader, cfe-commits
Differential Revision: https://reviews.llvm.org/D47630
Modified:
cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td
cfe/trunk/lib/Sema/SemaType.cpp
cfe/trunk/test/Sema/address_spaces.c
cfe/trunk/test/SemaOpenCL/address-spaces.cl
Modified: cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td?rev=335103&r1=335102&r2=335103&view=diff
==
--- cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td (original)
+++ cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td Wed Jun 20 01:31:24
2018
@@ -2576,6 +2576,9 @@ def err_attribute_address_space_too_high
"address space is larger than the maximum supported (%0)">;
def err_attribute_address_multiple_qualifiers : Error<
"multiple address spaces specified for type">;
+def warn_attribute_address_multiple_identical_qualifiers : Warning<
+ "multiple identical address spaces specified for type">,
+ InGroup;
def err_attribute_address_function_type : Error<
"function type may not be qualified with an address space">;
def err_as_qualified_auto_decl : Error<
Modified: cfe/trunk/lib/Sema/SemaType.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Sema/SemaType.cpp?rev=335103&r1=335102&r2=335103&view=diff
==
--- cfe/trunk/lib/Sema/SemaType.cpp (original)
+++ cfe/trunk/lib/Sema/SemaType.cpp Wed Jun 20 01:31:24 2018
@@ -5758,14 +5758,6 @@ QualType Sema::BuildAddressSpaceAttr(Qua
SourceLocation AttrLoc) {
if (!AddrSpace->isValueDependent()) {
-// If this type is already address space qualified, reject it.
-// ISO/IEC TR 18037 S5.3 (amending C99 6.7.3): "No type shall be qualified
-// by qualifiers for two or more different address spaces."
-if (T.getAddressSpace() != LangAS::Default) {
- Diag(AttrLoc, diag::err_attribute_address_multiple_qualifiers);
- return QualType();
-}
-
llvm::APSInt addrSpace(32);
if (!AddrSpace->isIntegerConstantExpr(addrSpace, Context)) {
Diag(AttrLoc, diag::err_attribute_argument_type)
@@ -5796,6 +5788,20 @@ QualType Sema::BuildAddressSpaceAttr(Qua
LangAS ASIdx =
getLangASFromTargetAS(static_cast(addrSpace.getZExtValue()));
+// If this type is already address space qualified with a different
+// address space, reject it.
+// ISO/IEC TR 18037 S5.3 (amending C99 6.7.3): "No type shall be qualified
+// by qualifiers for two or more different address spaces."
+if (T.getAddressSpace() != LangAS::Default) {
+ if (T.getAddressSpace() != ASIdx) {
+Diag(AttrLoc, diag::err_attribute_address_multiple_qualifiers);
+return QualType();
+ } else
+// Emit a warning if they are identical; it's likely unintended.
+Diag(AttrLoc,
+ diag::warn_attribute_address_multiple_identical_qualifiers);
+}
+
return Context.getAddrSpaceQualType(T, ASIdx);
}
@@ -5817,15 +5823,6 @@ QualType Sema::BuildAddressSpaceAttr(Qua
/// space for the type.
static void HandleAddressSpaceTypeAttribute(QualType &Type,
const AttributeList &Attr, Sema
&S){
- // If this type is already address space qualified, reject it.
- // ISO/IEC TR 18037 S5.3 (amending C99 6.7.3): "No type shall be qualified by
- // qualifiers for two or more different address spaces."
- if (Type.getAddressSpace() != LangAS::Default) {
-S.Diag(Attr.getLoc(), diag::err_attribute_address_multiple_qualifiers);
-Attr.setInvalid();
-return;
- }
-
// ISO/IEC TR 18037 S5.3 (amending C99 6.7.3): "A function type shall not be
// qualified by an address-space qualifier."
if (Type->isFunctionType()) {
@@ -5888,6 +5885,21 @@ static void HandleAddressSpaceTypeAttrib
llvm_unreachable("Invalid address space");
}
+// If this type is already address space qualified with a different
+// address space, reject it.
+// ISO/IEC TR 18037 S5.3 (amending C99 6.7.3): "No type shall be qualified
by
+// qualifiers for two or more different address spaces."
+i
r354337 - [OpenCL] Change type of block pointer for OpenCL
Author: bader
Date: Tue Feb 19 07:19:06 2019
New Revision: 354337
URL: http://llvm.org/viewvc/llvm-project?rev=354337&view=rev
Log:
[OpenCL] Change type of block pointer for OpenCL
Summary:
For some reason OpenCL blocks in LLVM IR are represented as function pointers.
These pointers do not point to any real function and never get called. Actually
they point to some structure, which in turn contains pointer to the real block
invoke function.
This patch changes represntation of OpenCL blocks in LLVM IR from function
pointers to pointers to `%struct.__block_literal_generic`.
Such representation allows to avoid unnecessary bitcasts and simplifies
further processing (e.g. translation to SPIR-V ) of the module for targets
which do not support function pointers.
Patch by: Alexey Sotkin.
Reviewers: Anastasia, yaxunl, svenvh
Reviewed By: Anastasia
Subscribers: alexbatashev, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D58277
Modified:
cfe/trunk/lib/CodeGen/CodeGenTypes.cpp
cfe/trunk/test/CodeGenOpenCL/blocks.cl
cfe/trunk/test/CodeGenOpenCL/cl20-device-side-enqueue.cl
Modified: cfe/trunk/lib/CodeGen/CodeGenTypes.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/CodeGen/CodeGenTypes.cpp?rev=354337&r1=354336&r2=354337&view=diff
==
--- cfe/trunk/lib/CodeGen/CodeGenTypes.cpp (original)
+++ cfe/trunk/lib/CodeGen/CodeGenTypes.cpp Tue Feb 19 07:19:06 2019
@@ -635,7 +635,9 @@ llvm::Type *CodeGenTypes::ConvertType(Qu
case Type::BlockPointer: {
const QualType FTy = cast(Ty)->getPointeeType();
-llvm::Type *PointeeType = ConvertTypeForMem(FTy);
+llvm::Type *PointeeType = CGM.getLangOpts().OpenCL
+ ? CGM.getGenericBlockLiteralType()
+ : ConvertTypeForMem(FTy);
unsigned AS = Context.getTargetAddressSpace(FTy);
ResultType = llvm::PointerType::get(PointeeType, AS);
break;
Modified: cfe/trunk/test/CodeGenOpenCL/blocks.cl
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/CodeGenOpenCL/blocks.cl?rev=354337&r1=354336&r2=354337&view=diff
==
--- cfe/trunk/test/CodeGenOpenCL/blocks.cl (original)
+++ cfe/trunk/test/CodeGenOpenCL/blocks.cl Tue Feb 19 07:19:06 2019
@@ -35,11 +35,10 @@ void foo(){
// SPIR: %[[block_captured:.*]] = getelementptr inbounds <{ i32, i32, i8
addrspace(4)*, i32 }>, <{ i32, i32, i8 addrspace(4)*, i32 }>* %[[block]], i32
0, i32 3
// SPIR: %[[i_value:.*]] = load i32, i32* %i
// SPIR: store i32 %[[i_value]], i32* %[[block_captured]],
- // SPIR: %[[blk_ptr:.*]] = bitcast <{ i32, i32, i8 addrspace(4)*, i32 }>*
%[[block]] to i32 ()*
- // SPIR: %[[blk_gen_ptr:.*]] = addrspacecast i32 ()* %[[blk_ptr]] to i32 ()
addrspace(4)*
- // SPIR: store i32 () addrspace(4)* %[[blk_gen_ptr]], i32 () addrspace(4)**
%[[block_B:.*]],
- // SPIR: %[[blk_gen_ptr:.*]] = load i32 () addrspace(4)*, i32 ()
addrspace(4)** %[[block_B]]
- // SPIR: %[[block_literal:.*]] = bitcast i32 () addrspace(4)*
%[[blk_gen_ptr]] to %struct.__opencl_block_literal_generic addrspace(4)*
+ // SPIR: %[[blk_ptr:.*]] = bitcast <{ i32, i32, i8 addrspace(4)*, i32 }>*
%[[block]] to %struct.__opencl_block_literal_generic*
+ // SPIR: %[[blk_gen_ptr:.*]] = addrspacecast
%struct.__opencl_block_literal_generic* %[[blk_ptr]] to
%struct.__opencl_block_literal_generic addrspace(4)*
+ // SPIR: store %struct.__opencl_block_literal_generic addrspace(4)*
%[[blk_gen_ptr]], %struct.__opencl_block_literal_generic addrspace(4)**
%[[block_B:.*]],
+ // SPIR: %[[block_literal:.*]] = load %struct.__opencl_block_literal_generic
addrspace(4)*, %struct.__opencl_block_literal_generic addrspace(4)**
%[[block_B]]
// SPIR: %[[invoke_addr:.*]] = getelementptr inbounds
%struct.__opencl_block_literal_generic, %struct.__opencl_block_literal_generic
addrspace(4)* %[[block_literal]], i32 0, i32 2
// SPIR: %[[blk_gen_ptr:.*]] = bitcast
%struct.__opencl_block_literal_generic addrspace(4)* %[[block_literal]] to i8
addrspace(4)*
// SPIR: %[[invoke_func_ptr:.*]] = load i8 addrspace(4)*, i8 addrspace(4)*
addrspace(4)* %[[invoke_addr]]
@@ -50,11 +49,10 @@ void foo(){
// AMDGCN: %[[block_captured:.*]] = getelementptr inbounds <{ i32, i32, i8*,
i32 }>, <{ i32, i32, i8*, i32 }> addrspace(5)* %[[block]], i32 0, i32 3
// AMDGCN: %[[i_value:.*]] = load i32, i32 addrspace(5)* %i
// AMDGCN: store i32 %[[i_value]], i32 addrspace(5)* %[[block_captured]],
- // AMDGCN: %[[blk_ptr:.*]] = bitcast <{ i32, i32, i8*, i32 }> addrspace(5)*
%[[block]] to i32 () addrspace(5)*
- // AMDGCN: %[[blk_gen_ptr:.*]] = addrspacecast i32 () addrspace(5)*
%[[blk_ptr]] to i32 ()*
- // AMDGCN: store i32 ()* %[[blk_gen_ptr]], i32 ()* addrspace(5)*
%[[block_B:.*]],
- // AMDGCN: %[[blk_gen_ptr:.*]] = load i32 ()*, i32 ()* addrspace(5)*
%[[block_B]]
- // AMDGCN: %[[block_
r354773 - [SYCL] Add clang front-end option to enable SYCL device compilation flow.
Author: bader
Date: Mon Feb 25 03:48:48 2019
New Revision: 354773
URL: http://llvm.org/viewvc/llvm-project?rev=354773&view=rev
Log:
[SYCL] Add clang front-end option to enable SYCL device compilation flow.
Patch by Mariya Podchishchaeva
Added:
cfe/trunk/test/Preprocessor/sycl-macro.cpp (with props)
Modified:
cfe/trunk/include/clang/Basic/LangOptions.def
cfe/trunk/include/clang/Driver/CC1Options.td
cfe/trunk/lib/Frontend/CompilerInvocation.cpp
cfe/trunk/lib/Frontend/InitPreprocessor.cpp
Modified: cfe/trunk/include/clang/Basic/LangOptions.def
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/include/clang/Basic/LangOptions.def?rev=354773&r1=354772&r2=354773&view=diff
==
--- cfe/trunk/include/clang/Basic/LangOptions.def (original)
+++ cfe/trunk/include/clang/Basic/LangOptions.def Mon Feb 25 03:48:48 2019
@@ -217,6 +217,8 @@ LANGOPT(CUDAHostDeviceConstexpr, 1, 1, "
LANGOPT(CUDADeviceApproxTranscendentals, 1, 0, "using approximate
transcendental functions")
LANGOPT(GPURelocatableDeviceCode, 1, 0, "generate relocatable device code")
+LANGOPT(SYCLIsDevice , 1, 0, "Generate code for SYCL device")
+
LANGOPT(SizedDeallocation , 1, 0, "sized deallocation")
LANGOPT(AlignedAllocation , 1, 0, "aligned allocation")
LANGOPT(AlignedAllocationUnavailable, 1, 0, "aligned allocation functions are
unavailable")
Modified: cfe/trunk/include/clang/Driver/CC1Options.td
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/include/clang/Driver/CC1Options.td?rev=354773&r1=354772&r2=354773&view=diff
==
--- cfe/trunk/include/clang/Driver/CC1Options.td (original)
+++ cfe/trunk/include/clang/Driver/CC1Options.td Mon Feb 25 03:48:48 2019
@@ -836,8 +836,14 @@ def fopenmp_is_device : Flag<["-"], "fop
def fopenmp_host_ir_file_path : Separate<["-"], "fopenmp-host-ir-file-path">,
HelpText<"Path to the IR file produced by the frontend for the host.">;
-} // let Flags = [CC1Option]
+//===--===//
+// SYCL Options
+//===--===//
+def fsycl_is_device : Flag<["-"], "fsycl-is-device">,
+ HelpText<"Generate code for SYCL device.">;
+
+} // let Flags = [CC1Option]
//===--===//
// cc1as-only Options
Modified: cfe/trunk/lib/Frontend/CompilerInvocation.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Frontend/CompilerInvocation.cpp?rev=354773&r1=354772&r2=354773&view=diff
==
--- cfe/trunk/lib/Frontend/CompilerInvocation.cpp (original)
+++ cfe/trunk/lib/Frontend/CompilerInvocation.cpp Mon Feb 25 03:48:48 2019
@@ -2879,6 +2879,8 @@ static void ParseLangArgs(LangOptions &O
<< Opts.OMPHostIRFile;
}
+ Opts.SYCLIsDevice = Args.hasArg(options::OPT_fsycl_is_device);
+
// Set CUDA mode for OpenMP target NVPTX if specified in options
Opts.OpenMPCUDAMode = Opts.OpenMPIsDevice && T.isNVPTX() &&
Args.hasArg(options::OPT_fopenmp_cuda_mode);
Modified: cfe/trunk/lib/Frontend/InitPreprocessor.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Frontend/InitPreprocessor.cpp?rev=354773&r1=354772&r2=354773&view=diff
==
--- cfe/trunk/lib/Frontend/InitPreprocessor.cpp (original)
+++ cfe/trunk/lib/Frontend/InitPreprocessor.cpp Mon Feb 25 03:48:48 2019
@@ -1057,6 +1057,12 @@ static void InitializePredefinedMacros(c
Builder.defineMacro("__CLANG_CUDA_APPROX_TRANSCENDENTALS__");
}
+ // Define a macro indicating that the source file is being compiled with a
+ // SYCL device compiler which doesn't produce host binary.
+ if (LangOpts.SYCLIsDevice) {
+Builder.defineMacro("__SYCL_DEVICE_ONLY__", "1");
+ }
+
// OpenCL definitions.
if (LangOpts.OpenCL) {
#define OPENCLEXT(Ext)
\
Added: cfe/trunk/test/Preprocessor/sycl-macro.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Preprocessor/sycl-macro.cpp?rev=354773&view=auto
==
--- cfe/trunk/test/Preprocessor/sycl-macro.cpp (added)
+++ cfe/trunk/test/Preprocessor/sycl-macro.cpp Mon Feb 25 03:48:48 2019
@@ -0,0 +1,5 @@
+// RUN: %clang_cc1 %s -E -dM | FileCheck %s
+// RUN: %clang_cc1 %s -fsycl-is-device -E -dM | FileCheck
--check-prefix=CHECK-SYCL %s
+
+// CHECK-NOT:#define __SYCL_DEVICE_ONLY__ 1
+// CHECK-SYCL:#define __SYCL_DEVICE_ONLY__ 1
Propchange: cfe/trunk/test/Preprocessor/sycl-macro.cpp
--
svn:eol-style = native
Propchange: cfe/trunk/test/Preprocessor/sycl-macro.cpp
[clang] e95ee30 - [SYCL] Prohibit arithmetic operations for incompatible pointers
Author: Alexey Bader
Date: 2020-05-22T13:43:24+03:00
New Revision: e95ee300c0530158d86430fd82ffabd36262e862
URL:
https://github.com/llvm/llvm-project/commit/e95ee300c0530158d86430fd82ffabd36262e862
DIFF:
https://github.com/llvm/llvm-project/commit/e95ee300c0530158d86430fd82ffabd36262e862.diff
LOG: [SYCL] Prohibit arithmetic operations for incompatible pointers
Summary:
This change enables OpenCL diagnostics for the pointers annotated with
address space attribute SYCL mode.
Move `isAddressSpaceOverlapping` method from PointerType to QualType.
Reviewers: Anastasia, rjmccall
Reviewed By: rjmccall
Subscribers: rjmccall, jeroen.dobbelaere, Fznamznon, yaxunl, ebevhan,
cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D80317
Added:
clang/test/SemaCXX/address-space-arithmetic.cpp
Modified:
clang/include/clang/AST/Type.h
clang/lib/Sema/SemaCast.cpp
clang/lib/Sema/SemaExpr.cpp
clang/test/Sema/address_spaces.c
Removed:
diff --git a/clang/include/clang/AST/Type.h b/clang/include/clang/AST/Type.h
index 21b14de997cb..ed31dea925f3 100644
--- a/clang/include/clang/AST/Type.h
+++ b/clang/include/clang/AST/Type.h
@@ -1064,6 +1064,21 @@ class QualType {
/// Return the address space of this type.
inline LangAS getAddressSpace() const;
+ /// Returns true if address space qualifiers overlap with T address space
+ /// qualifiers.
+ /// OpenCL C defines conversion rules for pointers to
diff erent address spaces
+ /// and notion of overlapping address spaces.
+ /// CL1.1 or CL1.2:
+ /// address spaces overlap iff they are they same.
+ /// OpenCL C v2.0 s6.5.5 adds:
+ /// __generic overlaps with any address space except for __constant.
+ bool isAddressSpaceOverlapping(QualType T) const {
+Qualifiers Q = getQualifiers();
+Qualifiers TQ = T.getQualifiers();
+// Address spaces overlap if at least one of them is a superset of another
+return Q.isAddressSpaceSupersetOf(TQ) || TQ.isAddressSpaceSupersetOf(Q);
+ }
+
/// Returns gc attribute of this type.
inline Qualifiers::GC getObjCGCAttr() const;
@@ -2631,22 +2646,6 @@ class PointerType : public Type, public
llvm::FoldingSetNode {
public:
QualType getPointeeType() const { return PointeeType; }
- /// Returns true if address spaces of pointers overlap.
- /// OpenCL v2.0 defines conversion rules for pointers to
diff erent
- /// address spaces (OpenCLC v2.0 s6.5.5) and notion of overlapping
- /// address spaces.
- /// CL1.1 or CL1.2:
- /// address spaces overlap iff they are they same.
- /// CL2.0 adds:
- /// __generic overlaps with any address space except for __constant.
- bool isAddressSpaceOverlapping(const PointerType &other) const {
-Qualifiers thisQuals = PointeeType.getQualifiers();
-Qualifiers otherQuals = other.getPointeeType().getQualifiers();
-// Address spaces overlap if at least one of them is a superset of another
-return thisQuals.isAddressSpaceSupersetOf(otherQuals) ||
- otherQuals.isAddressSpaceSupersetOf(thisQuals);
- }
-
bool isSugared() const { return false; }
QualType desugar() const { return QualType(this, 0); }
diff --git a/clang/lib/Sema/SemaCast.cpp b/clang/lib/Sema/SemaCast.cpp
index a4fe90f79eb9..fe4fcdd01301 100644
--- a/clang/lib/Sema/SemaCast.cpp
+++ b/clang/lib/Sema/SemaCast.cpp
@@ -2391,7 +2391,7 @@ static TryCastResult TryAddressSpaceCast(Sema &Self,
ExprResult &SrcExpr,
return TC_NotApplicable;
auto SrcPointeeType = SrcPtrType->getPointeeType();
auto DestPointeeType = DestPtrType->getPointeeType();
- if (!DestPtrType->isAddressSpaceOverlapping(*SrcPtrType)) {
+ if (!DestPointeeType.isAddressSpaceOverlapping(SrcPointeeType)) {
msg = diag::err_bad_cxx_cast_addr_space_mismatch;
return TC_Failed;
}
@@ -2434,9 +2434,9 @@ void CastOperation::checkAddressSpaceCast(QualType
SrcType, QualType DestType) {
const PointerType *SrcPPtr = cast(SrcPtr);
QualType DestPPointee = DestPPtr->getPointeeType();
QualType SrcPPointee = SrcPPtr->getPointeeType();
- if (Nested ? DestPPointee.getAddressSpace() !=
- SrcPPointee.getAddressSpace()
- : !DestPPtr->isAddressSpaceOverlapping(*SrcPPtr)) {
+ if (Nested
+ ? DestPPointee.getAddressSpace() != SrcPPointee.getAddressSpace()
+ : !DestPPointee.isAddressSpaceOverlapping(SrcPPointee)) {
Self.Diag(OpRange.getBegin(), DiagID)
<< SrcType << DestType << Sema::AA_Casting
<< SrcExpr.get()->getSourceRange();
diff --git a/clang/lib/Sema/SemaExpr.cpp b/clang/lib/Sema/SemaExpr.cpp
index 93e67ad2940c..261e69b44052 100644
--- a/clang/lib/Sema/SemaExpr.cpp
+++ b/clang/lib/Sema/SemaExpr.cpp
@@ -10087,10 +10087,8 @@ static bool checkArithmeticBinOpPointerOperands(Sema
&S, SourceLocation Loc,
if (isRHSPointer) RHSPointeeTy = RHSExpr->getType()
[clang] cf6cc66 - [OpenMP][SYCL] Improve diagnosing of unsupported types usage
Author: Mariya Podchishchaeva
Date: 2020-05-29T18:00:48+03:00
New Revision: cf6cc6622b1416430f517850be9032788e39
URL:
https://github.com/llvm/llvm-project/commit/cf6cc6622b1416430f517850be9032788e39
DIFF:
https://github.com/llvm/llvm-project/commit/cf6cc6622b1416430f517850be9032788e39.diff
LOG: [OpenMP][SYCL] Improve diagnosing of unsupported types usage
Summary:
Diagnostic is emitted if some declaration of unsupported type
declaration is used inside device code.
Memcpy operations for structs containing member with unsupported type
are allowed. Fixed crash on attempt to emit diagnostic outside of the
functions.
The approach is generalized between SYCL and OpenMP.
CUDA/OMP deferred diagnostic interface is going to be used for SYCL device.
Reviewers: rsmith, rjmccall, ABataev, erichkeane, bader, jdoerfert,
aaron.ballman
Reviewed By: jdoerfert
Subscribers: guansong, sstefan1, yaxunl, mgorny, bader, ebevhan, Anastasia,
cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D74387
Added:
clang/lib/Sema/SemaSYCL.cpp
clang/test/SemaSYCL/float128.cpp
Modified:
clang/include/clang/Basic/DiagnosticSemaKinds.td
clang/include/clang/Sema/Sema.h
clang/lib/Sema/CMakeLists.txt
clang/lib/Sema/Sema.cpp
clang/lib/Sema/SemaDecl.cpp
clang/lib/Sema/SemaDeclCXX.cpp
clang/lib/Sema/SemaExpr.cpp
clang/lib/Sema/SemaOpenMP.cpp
clang/lib/Sema/SemaType.cpp
clang/test/Headers/nvptx_device_math_sin.c
clang/test/Headers/nvptx_device_math_sin.cpp
clang/test/OpenMP/nvptx_unsupported_type_codegen.cpp
clang/test/OpenMP/nvptx_unsupported_type_messages.cpp
Removed:
diff --git a/clang/include/clang/Basic/DiagnosticSemaKinds.td
b/clang/include/clang/Basic/DiagnosticSemaKinds.td
index 845e329033c3..63af9f42dfd3 100644
--- a/clang/include/clang/Basic/DiagnosticSemaKinds.td
+++ b/clang/include/clang/Basic/DiagnosticSemaKinds.td
@@ -10204,8 +10204,8 @@ def err_omp_invariant_or_linear_dependency : Error<
"expected loop invariant expression or ' * %0 + '
kind of expression">;
def err_omp_wrong_dependency_iterator_type : Error<
"expected an integer or a pointer type of the outer loop counter '%0' for
non-rectangular nests">;
-def err_omp_unsupported_type : Error <
- "host requires %0 bit size %1 type support, but device '%2' does not support
it">;
+def err_device_unsupported_type : Error <
+ "%0 requires %1 bit size %2 type support, but device '%3' does not support
it">;
def err_omp_lambda_capture_in_declare_target_not_to : Error<
"variable captured in declare target region must appear in a to clause">;
def err_omp_device_type_mismatch : Error<
diff --git a/clang/include/clang/Sema/Sema.h b/clang/include/clang/Sema/Sema.h
index dc7ee2ddd0b8..594c6e03aa38 100644
--- a/clang/include/clang/Sema/Sema.h
+++ b/clang/include/clang/Sema/Sema.h
@@ -9868,10 +9868,6 @@ class Sema final {
/// Pop OpenMP function region for non-capturing function.
void popOpenMPFunctionRegion(const sema::FunctionScopeInfo *OldFSI);
- /// Check if the expression is allowed to be used in expressions for the
- /// OpenMP devices.
- void checkOpenMPDeviceExpr(const Expr *E);
-
/// Checks if a type or a declaration is disabled due to the owning extension
/// being disabled, and emits diagnostic messages if it is disabled.
/// \param D type or declaration to be checked.
@@ -11654,6 +11650,10 @@ class Sema final {
DeviceDiagBuilder targetDiag(SourceLocation Loc, unsigned DiagID);
+ /// Check if the expression is allowed to be used in expressions for the
+ /// offloading devices.
+ void checkDeviceDecl(const ValueDecl *D, SourceLocation Loc);
+
enum CUDAFunctionTarget {
CFT_Device,
CFT_Global,
@@ -12396,6 +12396,40 @@ class Sema final {
ConstructorDestructor,
BuiltinFunction
};
+ /// Creates a DeviceDiagBuilder that emits the diagnostic if the current
+ /// context is "used as device code".
+ ///
+ /// - If CurLexicalContext is a kernel function or it is known that the
+ /// function will be emitted for the device, emits the diagnostics
+ /// immediately.
+ /// - If CurLexicalContext is a function and we are compiling
+ /// for the device, but we don't know that this function will be codegen'ed
+ /// for devive yet, creates a diagnostic which is emitted if and when we
+ /// realize that the function will be codegen'ed.
+ ///
+ /// Example usage:
+ ///
+ /// Diagnose __float128 type usage only from SYCL device code if the current
+ /// target doesn't support it
+ /// if (!S.Context.getTargetInfo().hasFloat128Type() &&
+ /// S.getLangOpts().SYCLIsDevice)
+ /// SYCLDiagIfDeviceCode(Loc, diag::err_type_unsupported) << "__float128";
+ DeviceDiagBuilder SYCLDiagIfDeviceCode(SourceLocation Loc, unsigned DiagID);
+
+ /// Check whether we're allowed to call Callee from the current context.
+
[clang] bd85b7d - [OpenMP][SYCL] Do not crash on attempt to diagnose unsupported type use
Author: Mariya Podchishchaeva
Date: 2020-05-30T12:27:58+03:00
New Revision: bd85b7d6688725e854a694f9f3e8baa6a3077a4a
URL:
https://github.com/llvm/llvm-project/commit/bd85b7d6688725e854a694f9f3e8baa6a3077a4a
DIFF:
https://github.com/llvm/llvm-project/commit/bd85b7d6688725e854a694f9f3e8baa6a3077a4a.diff
LOG: [OpenMP][SYCL] Do not crash on attempt to diagnose unsupported type use
Summary:
Do not ask size of type if it is dependent. ASTContext doesn't seem expecting
this.
Reviewers: jdoerfert, ABataev, bader
Reviewed By: ABataev
Subscribers: yaxunl, guansong, ebevhan, Anastasia, sstefan1, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D80829
Added:
Modified:
clang/lib/Sema/Sema.cpp
clang/test/OpenMP/nvptx_unsupported_type_messages.cpp
Removed:
diff --git a/clang/lib/Sema/Sema.cpp b/clang/lib/Sema/Sema.cpp
index 8c11a1a59e9c..ffe2e4d4d56a 100644
--- a/clang/lib/Sema/Sema.cpp
+++ b/clang/lib/Sema/Sema.cpp
@@ -1725,6 +1725,9 @@ void Sema::checkDeviceDecl(const ValueDecl *D,
SourceLocation Loc) {
}
auto CheckType = [&](QualType Ty) {
+if (Ty->isDependentType())
+ return;
+
if ((Ty->isFloat16Type() && !Context.getTargetInfo().hasFloat16Type()) ||
((Ty->isFloat128Type() ||
(Ty->isRealFloatingType() && Context.getTypeSize(Ty) == 128)) &&
diff --git a/clang/test/OpenMP/nvptx_unsupported_type_messages.cpp
b/clang/test/OpenMP/nvptx_unsupported_type_messages.cpp
index 22ce8175fd05..e56105adeb83 100644
--- a/clang/test/OpenMP/nvptx_unsupported_type_messages.cpp
+++ b/clang/test/OpenMP/nvptx_unsupported_type_messages.cpp
@@ -120,3 +120,14 @@ void hostFoo() {
long double qa, qb;
decltype(qa + qb) qc;
double qd[sizeof(-(-(qc * 2)))];
+
+struct A { };
+
+template
+struct A_type { typedef A type; };
+
+template
+struct B {
+ enum { value = bool(Sp::value) || bool(Tp::value) };
+ typedef typename A_type::type type;
+};
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] 8d27be8 - [OpenCL] Add global_device and global_host address spaces
Author: Alexey Bader
Date: 2020-07-29T17:24:53+03:00
New Revision: 8d27be8dbaffce0519ac41173d51923fc2524b1b
URL:
https://github.com/llvm/llvm-project/commit/8d27be8dbaffce0519ac41173d51923fc2524b1b
DIFF:
https://github.com/llvm/llvm-project/commit/8d27be8dbaffce0519ac41173d51923fc2524b1b.diff
LOG: [OpenCL] Add global_device and global_host address spaces
This patch introduces 2 new address spaces in OpenCL: global_device and
global_host
which are a subset of a global address space, so the address space scheme will
be
looking like:
```
generic->global->host
->device
->private
->local
constant
```
Justification: USM allocations may be associated with both host and device
memory. We
want to give users a way to tell the compiler the allocation type of a USM
pointer for
optimization purposes. (Link to the Unified Shared Memory extension:
https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/cl_intel_unified_shared_memory.asciidoc)
Before this patch USM pointer could be only in opencl_global
address space, hence a device backend can't tell if a particular pointer
points to host or device memory. On FPGAs at least we can generate more
efficient hardware code if the user tells us where the pointer can point -
being able to distinguish between these types of pointers at compile time
allows us to instantiate simpler load-store units to perform memory
transactions.
Patch by Dmitry Sidorov.
Reviewed By: Anastasia
Differential Revision: https://reviews.llvm.org/D82174
Added:
clang/test/SemaOpenCL/usm-address-spaces-conversions.cl
Modified:
clang/include/clang/AST/Type.h
clang/include/clang/Basic/AddressSpaces.h
clang/include/clang/Basic/Attr.td
clang/include/clang/Basic/AttrDocs.td
clang/include/clang/Sema/ParsedAttr.h
clang/lib/AST/ASTContext.cpp
clang/lib/AST/ItaniumMangle.cpp
clang/lib/AST/MicrosoftMangle.cpp
clang/lib/AST/TypePrinter.cpp
clang/lib/Basic/Targets/AMDGPU.cpp
clang/lib/Basic/Targets/NVPTX.h
clang/lib/Basic/Targets/SPIR.h
clang/lib/Basic/Targets/TCE.h
clang/lib/Basic/Targets/X86.h
clang/lib/CodeGen/CodeGenModule.cpp
clang/lib/Sema/SemaType.cpp
clang/test/AST/language_address_space_attribute.cpp
clang/test/CodeGenCXX/mangle-address-space.cpp
clang/test/CodeGenOpenCL/address-spaces-conversions.cl
clang/test/CodeGenOpenCL/address-spaces.cl
clang/test/SemaTemplate/address_space-dependent.cpp
Removed:
diff --git a/clang/include/clang/AST/Type.h b/clang/include/clang/AST/Type.h
index 7fe652492b0e..7d943ebc78c0 100644
--- a/clang/include/clang/AST/Type.h
+++ b/clang/include/clang/AST/Type.h
@@ -480,6 +480,11 @@ class Qualifiers {
// Otherwise in OpenCLC v2.0 s6.5.5: every address space except
// for __constant can be used as __generic.
(A == LangAS::opencl_generic && B != LangAS::opencl_constant) ||
+ // We also define global_device and global_host address spaces,
+ // to distinguish global pointers allocated on host from pointers
+ // allocated on device, which are a subset of __global.
+ (A == LangAS::opencl_global && (B == LangAS::opencl_global_device ||
+ B == LangAS::opencl_global_host)) ||
// Consider pointer size address spaces to be equivalent to default.
((isPtrSizeAddressSpace(A) || A == LangAS::Default) &&
(isPtrSizeAddressSpace(B) || B == LangAS::Default));
diff --git a/clang/include/clang/Basic/AddressSpaces.h
b/clang/include/clang/Basic/AddressSpaces.h
index faf7f303aa2d..a9db52dfcc9c 100644
--- a/clang/include/clang/Basic/AddressSpaces.h
+++ b/clang/include/clang/Basic/AddressSpaces.h
@@ -36,6 +36,8 @@ enum class LangAS : unsigned {
opencl_constant,
opencl_private,
opencl_generic,
+ opencl_global_device,
+ opencl_global_host,
// CUDA specific address spaces.
cuda_device,
diff --git a/clang/include/clang/Basic/Attr.td
b/clang/include/clang/Basic/Attr.td
index 0ee3c5188563..f9bf3f0acd55 100644
--- a/clang/include/clang/Basic/Attr.td
+++ b/clang/include/clang/Basic/Attr.td
@@ -1178,6 +1178,16 @@ def OpenCLGlobalAddressSpace : TypeAttr {
let Documentation = [OpenCLAddressSpaceGlobalDocs];
}
+def OpenCLGlobalDeviceAddressSpace : TypeAttr {
+ let Spellings = [Clang<"opencl_global_device">];
+ let Documentation = [OpenCLAddressSpaceGlobalExtDocs];
+}
+
+def OpenCLGlobalHostAddressSpace : TypeAttr {
+ let Spellings = [Clang<"opencl_global_host">];
+ let Documentation = [OpenCLAddressSpaceGlobalExtDocs];
+}
+
def OpenCLLocalAddressSpace : TypeAttr {
let Spellings = [Keyword<"__local">, Keyword<"local">,
Clang<"opencl_local">];
let Documentation = [OpenCLAddressSpaceLocalDocs];
diff --git a/clang/include/clang/Basic/AttrDocs.td
b/clang/include/clang/Ba
[clang] 93cd411 - [NFC] Run clang-format on clang/test/OpenMP/nvptx_target_codegen.cpp
Author: Alexey Bader
Date: 2020-06-17T13:04:01+03:00
New Revision: 93cd4115799cefa698833ca7a2f1899243d94c77
URL:
https://github.com/llvm/llvm-project/commit/93cd4115799cefa698833ca7a2f1899243d94c77
DIFF:
https://github.com/llvm/llvm-project/commit/93cd4115799cefa698833ca7a2f1899243d94c77.diff
LOG: [NFC] Run clang-format on clang/test/OpenMP/nvptx_target_codegen.cpp
Added:
Modified:
clang/test/OpenMP/nvptx_target_codegen.cpp
Removed:
diff --git a/clang/test/OpenMP/nvptx_target_codegen.cpp
b/clang/test/OpenMP/nvptx_target_codegen.cpp
index 20415c0dc1b6..d615b8536c48 100644
--- a/clang/test/OpenMP/nvptx_target_codegen.cpp
+++ b/clang/test/OpenMP/nvptx_target_codegen.cpp
@@ -16,16 +16,16 @@
// CHECK-DAG: {{@__omp_offloading_.+l123}}_exec_mode = weak constant i8 1
// CHECK-DAG: {{@__omp_offloading_.+l200}}_exec_mode = weak constant i8 1
// CHECK-DAG: {{@__omp_offloading_.+l310}}_exec_mode = weak constant i8 1
-// CHECK-DAG: {{@__omp_offloading_.+l348}}_exec_mode = weak constant i8 1
-// CHECK-DAG: {{@__omp_offloading_.+l366}}_exec_mode = weak constant i8 1
+// CHECK-DAG: {{@__omp_offloading_.+l347}}_exec_mode = weak constant i8 1
+// CHECK-DAG: {{@__omp_offloading_.+l365}}_exec_mode = weak constant i8 1
// CHECK-DAG: {{@__omp_offloading_.+l331}}_exec_mode = weak constant i8 1
__thread int id;
int baz(int f, double &a);
-template
-struct TT{
+template
+struct TT {
tx X;
ty Y;
tx &operator[](int i) { return X; }
@@ -56,258 +56,258 @@ int foo(int n) {
double cn[5][n];
TT d;
- // CHECK-LABEL: define {{.*}}void {{@__omp_offloading_.+foo.+l123}}_worker()
- // CHECK-DAG: [[OMP_EXEC_STATUS:%.+]] = alloca i8,
- // CHECK-DAG: [[OMP_WORK_FN:%.+]] = alloca i8*,
- // CHECK: store i8* null, i8** [[OMP_WORK_FN]],
- // CHECK: store i8 0, i8* [[OMP_EXEC_STATUS]],
- // CHECK: br label {{%?}}[[AWAIT_WORK:.+]]
- //
- // CHECK: [[AWAIT_WORK]]
- // CHECK: call void @__kmpc_barrier_simple_spmd(%struct.ident_t* null, i32 0)
- // CHECK: [[WORK:%.+]] = load i8*, i8** [[OMP_WORK_FN]],
- // CHECK: [[SHOULD_EXIT:%.+]] = icmp eq i8* [[WORK]], null
- // CHECK: br i1 [[SHOULD_EXIT]], label {{%?}}[[EXIT:.+]], label
{{%?}}[[SEL_WORKERS:.+]]
- //
- // CHECK: [[SEL_WORKERS]]
- // CHECK: [[ST:%.+]] = load i8, i8* [[OMP_EXEC_STATUS]],
- // CHECK: [[IS_ACTIVE:%.+]] = icmp ne i8 [[ST]], 0
- // CHECK: br i1 [[IS_ACTIVE]], label {{%?}}[[EXEC_PARALLEL:.+]], label
{{%?}}[[BAR_PARALLEL:.+]]
- //
- // CHECK: [[EXEC_PARALLEL]]
- // CHECK: br label {{%?}}[[TERM_PARALLEL:.+]]
- //
- // CHECK: [[TERM_PARALLEL]]
- // CHECK: br label {{%?}}[[BAR_PARALLEL]]
- //
- // CHECK: [[BAR_PARALLEL]]
- // CHECK: call void @__kmpc_barrier_simple_spmd(%struct.ident_t* null, i32 0)
- // CHECK: br label {{%?}}[[AWAIT_WORK]]
- //
- // CHECK: [[EXIT]]
- // CHECK: ret void
-
- // CHECK: define {{.*}}void [[T1:@__omp_offloading_.+foo.+l123]]()
- // CHECK-DAG: [[TID:%.+]] = call i32 @llvm.nvvm.read.ptx.sreg.tid.x()
- // CHECK-DAG: [[NTH:%.+]] = call i32 @llvm.nvvm.read.ptx.sreg.ntid.x()
- // CHECK-DAG: [[WS:%.+]] = call i32 @llvm.nvvm.read.ptx.sreg.warpsize()
- // CHECK-DAG: [[TH_LIMIT:%.+]] = sub nuw i32 [[NTH]], [[WS]]
- // CHECK: [[IS_WORKER:%.+]] = icmp ult i32 [[TID]], [[TH_LIMIT]]
- // CHECK: br i1 [[IS_WORKER]], label {{%?}}[[WORKER:.+]], label
{{%?}}[[CHECK_MASTER:.+]]
- //
- // CHECK: [[WORKER]]
- // CHECK: {{call|invoke}} void [[T1]]_worker()
- // CHECK: br label {{%?}}[[EXIT:.+]]
- //
- // CHECK: [[CHECK_MASTER]]
- // CHECK-DAG: [[CMTID:%.+]] = call i32 @llvm.nvvm.read.ptx.sreg.tid.x()
- // CHECK-DAG: [[CMNTH:%.+]] = call i32 @llvm.nvvm.read.ptx.sreg.ntid.x()
- // CHECK-DAG: [[CMWS:%.+]] = call i32 @llvm.nvvm.read.ptx.sreg.warpsize()
- // CHECK: [[IS_MASTER:%.+]] = icmp eq i32 [[CMTID]],
- // CHECK: br i1 [[IS_MASTER]], label {{%?}}[[MASTER:.+]], label
{{%?}}[[EXIT]]
- //
- // CHECK: [[MASTER]]
- // CHECK-DAG: [[MNTH:%.+]] = call i32 @llvm.nvvm.read.ptx.sreg.ntid.x()
- // CHECK-DAG: [[MWS:%.+]] = call i32 @llvm.nvvm.read.ptx.sreg.warpsize()
- // CHECK: [[MTMP1:%.+]] = sub nuw i32 [[MNTH]], [[MWS]]
- // CHECK: call void @__kmpc_kernel_init(i32 [[MTMP1]]
- // CHECK: br label {{%?}}[[TERMINATE:.+]]
- //
- // CHECK: [[TERMINATE]]
- // CHECK: call void @__kmpc_kernel_deinit(
- // CHECK: call void @__kmpc_barrier_simple_spmd(%struct.ident_t* null, i32 0)
- // CHECK: br label {{%?}}[[EXIT]]
- //
- // CHECK: [[EXIT]]
- // CHECK: ret void
- #pragma omp target
+// CHECK-LABEL: define {{.*}}void {{@__omp_offloading_.+foo.+l123}}_worker()
+// CHECK-DAG: [[OMP_EXEC_STATUS:%.+]] = alloca i8,
+// CHECK-DAG: [[OMP_WORK_FN:%.+]] = alloca i8*,
+// CHECK: store i8* null, i8** [[OMP_WORK_FN]],
+// CHECK: store i8 0, i8* [[OMP_EXEC_STATUS]],
+// CHECK: br label {{%?}}[[AWAIT_WORK:.+]]
+//
+// CHECK: [[AWAIT_WORK]]
+// CHECK: call void @__kmpc_barrier_simple_spmd(%struct.ident_t* null, i3
[clang] 0bdcd95 - [SYCL][OpenMP] Implement thread-local storage restriction
Author: Mariya Podchishchaeva
Date: 2020-06-17T14:36:00+03:00
New Revision: 0bdcd95bf20f159a2512aff1ef032bec52039bf6
URL:
https://github.com/llvm/llvm-project/commit/0bdcd95bf20f159a2512aff1ef032bec52039bf6
DIFF:
https://github.com/llvm/llvm-project/commit/0bdcd95bf20f159a2512aff1ef032bec52039bf6.diff
LOG: [SYCL][OpenMP] Implement thread-local storage restriction
Summary:
SYCL and OpenMP prohibits thread local storage in device code,
so this commit ensures that error is emitted for device code and not
emitted for host code when host target supports it.
Reviewers: jdoerfert, erichkeane, bader
Reviewed By: jdoerfert, erichkeane
Subscribers: guansong, riccibruno, ABataev, yaxunl, ebevhan, Anastasia,
sstefan1, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D81641
Added:
clang/test/OpenMP/nvptx_prohibit_thread_local.cpp
clang/test/SemaSYCL/prohibit-thread-local.cpp
Modified:
clang/lib/Sema/SemaDecl.cpp
clang/lib/Sema/SemaExpr.cpp
clang/test/OpenMP/nvptx_target_codegen.cpp
Removed:
diff --git a/clang/lib/Sema/SemaDecl.cpp b/clang/lib/Sema/SemaDecl.cpp
index 2bf16d138d5a..80469e3bedbe 100644
--- a/clang/lib/Sema/SemaDecl.cpp
+++ b/clang/lib/Sema/SemaDecl.cpp
@@ -7077,7 +7077,8 @@ NamedDecl *Sema::ActOnVariableDeclarator(
diag::err_thread_non_global)
<< DeclSpec::getSpecifierName(TSCS);
else if (!Context.getTargetInfo().isTLSSupported()) {
- if (getLangOpts().CUDA || getLangOpts().OpenMPIsDevice) {
+ if (getLangOpts().CUDA || getLangOpts().OpenMPIsDevice ||
+ getLangOpts().SYCLIsDevice) {
// Postpone error emission until we've collected attributes required to
// figure out whether it's a host or device variable and whether the
// error should be ignored.
@@ -7179,13 +7180,18 @@ NamedDecl *Sema::ActOnVariableDeclarator(
// Handle attributes prior to checking for duplicates in MergeVarDecl
ProcessDeclAttributes(S, NewVD, D);
- if (getLangOpts().CUDA || getLangOpts().OpenMPIsDevice) {
+ if (getLangOpts().CUDA || getLangOpts().OpenMPIsDevice ||
+ getLangOpts().SYCLIsDevice) {
if (EmitTLSUnsupportedError &&
((getLangOpts().CUDA && DeclAttrsMatchCUDAMode(getLangOpts(), NewVD))
||
(getLangOpts().OpenMPIsDevice &&
OMPDeclareTargetDeclAttr::isDeclareTargetDeclaration(NewVD
Diag(D.getDeclSpec().getThreadStorageClassSpecLoc(),
diag::err_thread_unsupported);
+
+if (EmitTLSUnsupportedError &&
+(LangOpts.SYCLIsDevice || (LangOpts.OpenMP &&
LangOpts.OpenMPIsDevice)))
+ targetDiag(D.getIdentifierLoc(), diag::err_thread_unsupported);
// CUDA B.2.5: "__shared__ and __constant__ variables have implied static
// storage [duration]."
if (SC == SC_None && S->getFnParent() != nullptr &&
diff --git a/clang/lib/Sema/SemaExpr.cpp b/clang/lib/Sema/SemaExpr.cpp
index 66a2ec1fe9dc..ffc72140dcf4 100644
--- a/clang/lib/Sema/SemaExpr.cpp
+++ b/clang/lib/Sema/SemaExpr.cpp
@@ -355,10 +355,16 @@ bool Sema::DiagnoseUseOfDecl(NamedDecl *D,
ArrayRef Locs,
diagnoseUseOfInternalDeclInInlineFunction(*this, D, Loc);
- if (LangOpts.SYCLIsDevice || (LangOpts.OpenMP && LangOpts.OpenMPIsDevice))
+ if (LangOpts.SYCLIsDevice || (LangOpts.OpenMP && LangOpts.OpenMPIsDevice)) {
if (const auto *VD = dyn_cast(D))
checkDeviceDecl(VD, Loc);
+if (!Context.getTargetInfo().isTLSSupported())
+ if (const auto *VD = dyn_cast(D))
+if (VD->getTLSKind() != VarDecl::TLS_None)
+ targetDiag(*Locs.begin(), diag::err_thread_unsupported);
+ }
+
if (isa(D) && isa(D->getDeclContext()) &&
!isUnevaluatedContext()) {
// C++ [expr.prim.req.nested] p3
diff --git a/clang/test/OpenMP/nvptx_prohibit_thread_local.cpp
b/clang/test/OpenMP/nvptx_prohibit_thread_local.cpp
new file mode 100644
index ..b84918e528cb
--- /dev/null
+++ b/clang/test/OpenMP/nvptx_prohibit_thread_local.cpp
@@ -0,0 +1,44 @@
+// RUN: %clang_cc1 -fopenmp -x c++ -triple x86_64-unknown-linux
-fopenmp-targets=nvptx64-nvidia-cuda -emit-llvm-bc %s -o %t-host.bc
+// RUN: %clang_cc1 -verify -fopenmp -x c++ -triple nvptx64-unknown-unknown
-aux-triple x86_64-unknown-linux -fopenmp-targets=nvptx64-nvidia-cuda %s
-fopenmp-is-device -fopenmp-host-ir-file-path %t-host.bc -fsyntax-only
+
+thread_local const int prohobit_ns_scope = 0;
+thread_local int prohobit_ns_scope2 = 0;
+thread_local const int allow_ns_scope = 0;
+
+struct S {
+ static const thread_local int prohibit_static_member;
+ static thread_local int prohibit_static_member2;
+};
+
+struct T {
+ static const thread_local int allow_static_member;
+};
+
+void foo() {
+ // expected-error@+1{{thread-local storage is not supported for the current
target}}
+ thread_local const int prohibit_local = 0;
+ // expected-error@+1{{thread-local storage is not supported for t
Re: [PATCH] D24666: [OpenCL] Allow half type kernel argument when cl_khr_fp16 is enabled
bader added inline comments. Comment at: lib/Sema/SemaDecl.cpp:7599-7602 @@ -7595,3 +7598,6 @@ // of event_t type. -S.Diag(Param->getLocation(), diag::err_bad_kernel_param_type) << PT; +// Do not diagnose half type since it is diagnosed as invalid argument +// type for any function eleswhere. +if (!PT->isHalfType()) + S.Diag(Param->getLocation(), diag::err_bad_kernel_param_type) << PT; D.setInvalidType(); yaxunl wrote: > bader wrote: > > It looks strange to me. First we check if parameter type is half - we set > > InvalidKernelParam status, later we check again and do not report an error. > > I think it would be simpler just return ValidKernelParam for half data type > > from getOpenCLKernelParameterType, > > > > I think the whole patch should two deleted lines from > > getOpenCLKernelParameterType function + test case. > getOpenCLKernelParameterType should report half type as InvalidKernelParam > since it really is an invalid kernel argument type. We do not emit diagnostic > msg because the msg is redundant, not because half type is a valid kernel > argument type. > > getOpenCLKernelParameterType may be used for other purpose. Reporting half > type as a valid kernel argument violates the semantics of > getOpenCLKernelParameterType and can cause confusion and potential error. Maybe we should the other way. Leave half parameter check here only and remove duplicate check that reports "declaring function parameter of type 'half' is not allowed; did you forget * ?" message. https://reviews.llvm.org/D24666 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
r282252 - [OpenCL] Augment pipe built-ins with pipe packet size and alignment.
Author: bader
Date: Fri Sep 23 09:20:00 2016
New Revision: 282252
URL: http://llvm.org/viewvc/llvm-project?rev=282252&view=rev
Log:
[OpenCL] Augment pipe built-ins with pipe packet size and alignment.
Reviewers: Anastasia, vpykhtin
Subscribers: dmitry, cfe-commits
Differential Revision: https://reviews.llvm.org/D23992
Modified:
cfe/trunk/lib/CodeGen/CGBuiltin.cpp
cfe/trunk/lib/CodeGen/CGOpenCLRuntime.cpp
cfe/trunk/lib/CodeGen/CGOpenCLRuntime.h
cfe/trunk/test/CodeGenOpenCL/pipe_builtin.cl
cfe/trunk/test/CodeGenOpenCL/pipe_types.cl
Modified: cfe/trunk/lib/CodeGen/CGBuiltin.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/CodeGen/CGBuiltin.cpp?rev=282252&r1=282251&r2=282252&view=diff
==
--- cfe/trunk/lib/CodeGen/CGBuiltin.cpp (original)
+++ cfe/trunk/lib/CodeGen/CGBuiltin.cpp Fri Sep 23 09:20:00 2016
@@ -14,6 +14,7 @@
#include "CodeGenFunction.h"
#include "CGCXXABI.h"
#include "CGObjCRuntime.h"
+#include "CGOpenCLRuntime.h"
#include "CodeGenModule.h"
#include "TargetInfo.h"
#include "clang/AST/ASTContext.h"
@@ -2139,6 +2140,9 @@ RValue CodeGenFunction::EmitBuiltinExpr(
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 =
@@ -2152,19 +2156,21 @@ RValue CodeGenFunction::EmitBuiltinExpr(
:
"__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};
+ llvm::Type *ArgTys[] = {Arg0->getType(), I8PTy, Int32Ty, Int32Ty};
llvm::FunctionType *FTy = llvm::FunctionType::get(
Int32Ty, llvm::ArrayRef(ArgTys), false);
Value *BCast = Builder.CreatePointerCast(Arg1, I8PTy);
- return RValue::get(Builder.CreateCall(
- CGM.CreateRuntimeFunction(FTy, Name), {Arg0, BCast}));
+ 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};
+ 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(
@@ -2175,7 +2181,8 @@ RValue CodeGenFunction::EmitBuiltinExpr(
if (Arg2->getType() != Int32Ty)
Arg2 = Builder.CreateZExtOrTrunc(Arg2, Int32Ty);
return RValue::get(Builder.CreateCall(
- CGM.CreateRuntimeFunction(FTy, Name), {Arg0, Arg1, Arg2, BCast}));
+ 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
@@ -2204,9 +2211,12 @@ RValue CodeGenFunction::EmitBuiltinExpr(
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};
+llvm::Type *ArgTys[] = {Arg0->getType(), Int32Ty, Int32Ty, Int32Ty};
llvm::FunctionType *FTy = llvm::FunctionType::get(
ReservedIDTy, llvm::ArrayRef(ArgTys), false);
// We know the second argument is an integer type, but we may need to cast
@@ -2214,7 +2224,8 @@ RValue CodeGenFunction::EmitBuiltinExpr(
if (Arg1->getType() != Int32Ty)
Arg1 = Builder.CreateZExtOrTrunc(Arg1, Int32Ty);
return RValue::get(
-Builder.CreateCall(CGM.CreateRuntimeFunction(FTy, Name), {Arg0,
Arg1}));
+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
@@ -2240,15 +2251,19 @@ RValue CodeGenFunction::EmitBuiltinExpr(
Value *Arg0 = EmitScalarExpr(E->getArg(0)),
*Arg1 = EmitScalarExpr(E->getArg(1));
+
Re: [PATCH] D23992: [OpenCL] Augment pipe built-ins with pipe packet size and alignment.
This revision was automatically updated to reflect the committed changes.
Closed by commit rL282252: [OpenCL] Augment pipe built-ins with pipe packet
size and alignment. (authored by bader).
Changed prior to commit:
https://reviews.llvm.org/D23992?vs=70366&id=72276#toc
Repository:
rL LLVM
https://reviews.llvm.org/D23992
Files:
cfe/trunk/lib/CodeGen/CGBuiltin.cpp
cfe/trunk/lib/CodeGen/CGOpenCLRuntime.cpp
cfe/trunk/lib/CodeGen/CGOpenCLRuntime.h
cfe/trunk/test/CodeGenOpenCL/pipe_builtin.cl
cfe/trunk/test/CodeGenOpenCL/pipe_types.cl
Index: cfe/trunk/test/CodeGenOpenCL/pipe_builtin.cl
===
--- cfe/trunk/test/CodeGenOpenCL/pipe_builtin.cl
+++ cfe/trunk/test/CodeGenOpenCL/pipe_builtin.cl
@@ -4,59 +4,59 @@
// CHECK: %opencl.reserve_id_t = type opaque
void test1(read_only pipe int p, global int *ptr) {
- // CHECK: call i32 @__read_pipe_2(%opencl.pipe_t* %{{.*}}, i8* %{{.*}})
+ // CHECK: call i32 @__read_pipe_2(%opencl.pipe_t* %{{.*}}, i8* %{{.*}}, i32 4, i32 4)
read_pipe(p, ptr);
- // CHECK: call %opencl.reserve_id_t* @__reserve_read_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}})
+ // CHECK: call %opencl.reserve_id_t* @__reserve_read_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}}, i32 4, i32 4)
reserve_id_t rid = reserve_read_pipe(p, 2);
- // CHECK: call i32 @__read_pipe_4(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 {{.*}}, i8* %{{.*}})
+ // CHECK: call i32 @__read_pipe_4(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 {{.*}}, i8* %{{.*}}, i32 4, i32 4)
read_pipe(p, rid, 2, ptr);
- // CHECK: call void @__commit_read_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}})
+ // CHECK: call void @__commit_read_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 4, i32 4)
commit_read_pipe(p, rid);
}
void test2(write_only pipe int p, global int *ptr) {
- // CHECK: call i32 @__write_pipe_2(%opencl.pipe_t* %{{.*}}, i8* %{{.*}})
+ // CHECK: call i32 @__write_pipe_2(%opencl.pipe_t* %{{.*}}, i8* %{{.*}}, i32 4, i32 4)
write_pipe(p, ptr);
- // CHECK: call %opencl.reserve_id_t* @__reserve_write_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}})
+ // CHECK: call %opencl.reserve_id_t* @__reserve_write_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}}, i32 4, i32 4)
reserve_id_t rid = reserve_write_pipe(p, 2);
- // CHECK: call i32 @__write_pipe_4(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 {{.*}}, i8* %{{.*}})
+ // CHECK: call i32 @__write_pipe_4(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 {{.*}}, i8* %{{.*}}, i32 4, i32 4)
write_pipe(p, rid, 2, ptr);
- // CHECK: call void @__commit_write_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}})
+ // CHECK: call void @__commit_write_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 4, i32 4)
commit_write_pipe(p, rid);
}
void test3(read_only pipe int p, global int *ptr) {
- // CHECK: call %opencl.reserve_id_t* @__work_group_reserve_read_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}})
+ // CHECK: call %opencl.reserve_id_t* @__work_group_reserve_read_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}}, i32 4, i32 4)
reserve_id_t rid = work_group_reserve_read_pipe(p, 2);
- // CHECK: call void @__work_group_commit_read_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}})
+ // CHECK: call void @__work_group_commit_read_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 4, i32 4)
work_group_commit_read_pipe(p, rid);
}
void test4(write_only pipe int p, global int *ptr) {
- // CHECK: call %opencl.reserve_id_t* @__work_group_reserve_write_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}})
+ // CHECK: call %opencl.reserve_id_t* @__work_group_reserve_write_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}}, i32 4, i32 4)
reserve_id_t rid = work_group_reserve_write_pipe(p, 2);
- // CHECK: call void @__work_group_commit_write_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}})
+ // CHECK: call void @__work_group_commit_write_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 4, i32 4)
work_group_commit_write_pipe(p, rid);
}
void test5(read_only pipe int p, global int *ptr) {
- // CHECK: call %opencl.reserve_id_t* @__sub_group_reserve_read_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}})
+ // CHECK: call %opencl.reserve_id_t* @__sub_group_reserve_read_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}}, i32 4, i32 4)
reserve_id_t rid = sub_group_reserve_read_pipe(p, 2);
- // CHECK: call void @__sub_group_commit_read_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}})
+ // CHECK: call void @__sub_group_commit_read_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 4, i32 4)
sub_group_commit_read_pipe(p, rid);
}
void test6(write_only pipe int p, global int *ptr) {
- // CHECK: call %opencl.reserve_id_t* @__sub_group_reserve_write_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}})
+ // CHECK: call %opencl.reserve_id_t* @__sub_group_reserve_write_pipe(%
[clang] 923b56e - [NFC] Add a TODO comment to apply nounwind attribute in all GPU modes.
Author: Alexey Bader
Date: 2022-07-06T06:20:09-07:00
New Revision: 923b56e7ca96e03cedcb0e3a5df5c05e8e975a38
URL:
https://github.com/llvm/llvm-project/commit/923b56e7ca96e03cedcb0e3a5df5c05e8e975a38
DIFF:
https://github.com/llvm/llvm-project/commit/923b56e7ca96e03cedcb0e3a5df5c05e8e975a38.diff
LOG: [NFC] Add a TODO comment to apply nounwind attribute in all GPU modes.
Added:
Modified:
clang/lib/CodeGen/CGCall.cpp
Removed:
diff --git a/clang/lib/CodeGen/CGCall.cpp b/clang/lib/CodeGen/CGCall.cpp
index 4e26c35c6342..104a30dd6b25 100644
--- a/clang/lib/CodeGen/CGCall.cpp
+++ b/clang/lib/CodeGen/CGCall.cpp
@@ -1931,6 +1931,9 @@ void
CodeGenModule::getDefaultFunctionAttributes(StringRef Name,
FuncAttrs.addAttribute(llvm::Attribute::Convergent);
}
+ // TODO: NoUnwind attribute should be added for other GPU modes OpenCL, HIP,
+ // SYCL, OpenMP offload. AFAIK, none of them support exceptions in device
+ // code.
if (getLangOpts().CUDA && getLangOpts().CUDAIsDevice) {
// Exceptions aren't supported in CUDA device code.
FuncAttrs.addAttribute(llvm::Attribute::NoUnwind);
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] 87b28f5 - [clang][NFC] Extract the EmitAssemblyHelper::TargetTriple member
Author: Pavel Samolysov
Date: 2022-04-04T12:16:39+03:00
New Revision: 87b28f5092f2f92fc380f18e8578746bdd2a54b2
URL:
https://github.com/llvm/llvm-project/commit/87b28f5092f2f92fc380f18e8578746bdd2a54b2
DIFF:
https://github.com/llvm/llvm-project/commit/87b28f5092f2f92fc380f18e8578746bdd2a54b2.diff
LOG: [clang][NFC] Extract the EmitAssemblyHelper::TargetTriple member
Few times in different methods of the EmitAssemblyHelper class the following
code snippet is used to get the TargetTriple and then use it's single method
to check some conditions:
TargetTriple(TheModule->getTargetTriple())
The parsing of a target triple string is not a trivial operation and it takes
time to repeat the parsing many times in different methods of the class and
even numerous times in one method just to call a getter
(llvm::Triple(TheModule->getTargetTriple()).getVendor()), for example.
The patch extracts the TargetTriple member of the EmitAssemblyHelper class to
parse the triple only once in the class' constructor.
Reviewed By: tejohnson
Differential Revision: https://reviews.llvm.org/D122587
Added:
Modified:
clang/lib/CodeGen/BackendUtil.cpp
Removed:
diff --git a/clang/lib/CodeGen/BackendUtil.cpp
b/clang/lib/CodeGen/BackendUtil.cpp
index 61677d368029e..36776d39b952a 100644
--- a/clang/lib/CodeGen/BackendUtil.cpp
+++ b/clang/lib/CodeGen/BackendUtil.cpp
@@ -118,6 +118,8 @@ class EmitAssemblyHelper {
std::unique_ptr OS;
+ Triple TargetTriple;
+
TargetIRAnalysis getTargetIRAnalysis() const {
if (TM)
return TM->getTargetIRAnalysis();
@@ -170,7 +172,8 @@ class EmitAssemblyHelper {
const LangOptions &LOpts, Module *M)
: Diags(_Diags), HSOpts(HeaderSearchOpts), CodeGenOpts(CGOpts),
TargetOpts(TOpts), LangOpts(LOpts), TheModule(M),
-CodeGenerationTime("codegen", "Code Generation Time") {}
+CodeGenerationTime("codegen", "Code Generation Time"),
+TargetTriple(TheModule->getTargetTriple()) {}
~EmitAssemblyHelper() {
if (CodeGenOpts.DisableFree)
@@ -695,7 +698,6 @@ void EmitAssemblyHelper::CreatePasses(legacy::PassManager
&MPM,
// manually (and not via PMBuilder), since some passes (eg. InstrProfiling)
// are inserted before PMBuilder ones - they'd get the default-constructed
// TLI with an unknown target otherwise.
- Triple TargetTriple(TheModule->getTargetTriple());
std::unique_ptr TLII(
createTLII(TargetTriple, CodeGenOpts));
@@ -965,7 +967,6 @@ bool EmitAssemblyHelper::AddEmitPasses(legacy::PassManager
&CodeGenPasses,
raw_pwrite_stream &OS,
raw_pwrite_stream *DwoOS) {
// Add LibraryInfo.
- llvm::Triple TargetTriple(TheModule->getTargetTriple());
std::unique_ptr TLII(
createTLII(TargetTriple, CodeGenOpts));
CodeGenPasses.add(new TargetLibraryInfoWrapperPass(*TLII));
@@ -1054,10 +1055,8 @@ void
EmitAssemblyHelper::EmitAssemblyWithLegacyPassManager(
// Emit a module summary by default for Regular LTO except for ld64
// targets
bool EmitLTOSummary =
- (CodeGenOpts.PrepareForLTO &&
- !CodeGenOpts.DisableLLVMPasses &&
- llvm::Triple(TheModule->getTargetTriple()).getVendor() !=
- llvm::Triple::Apple);
+ (CodeGenOpts.PrepareForLTO && !CodeGenOpts.DisableLLVMPasses &&
+ TargetTriple.getVendor() != llvm::Triple::Apple);
if (EmitLTOSummary) {
if (!TheModule->getModuleFlag("ThinLTO"))
TheModule->addModuleFlag(Module::Error, "ThinLTO", uint32_t(0));
@@ -1338,7 +1337,6 @@ void EmitAssemblyHelper::RunOptimizationPipeline(
// Register the target library analysis directly and give it a customized
// preset TLI.
- Triple TargetTriple(TheModule->getTargetTriple());
std::unique_ptr TLII(
createTLII(TargetTriple, CodeGenOpts));
FAM.registerPass([&] { return TargetLibraryAnalysis(*TLII); });
@@ -1474,8 +1472,7 @@ void EmitAssemblyHelper::RunOptimizationPipeline(
// targets
bool EmitLTOSummary =
(CodeGenOpts.PrepareForLTO && !CodeGenOpts.DisableLLVMPasses &&
- llvm::Triple(TheModule->getTargetTriple()).getVendor() !=
- llvm::Triple::Apple);
+ TargetTriple.getVendor() != llvm::Triple::Apple);
if (EmitLTOSummary) {
if (!TheModule->getModuleFlag("ThinLTO"))
TheModule->addModuleFlag(Module::Error, "ThinLTO", uint32_t(0));
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [AMDGPU] Adding the amdgpu-num-work-groups function attribute (PR #75647)
bader wrote: > How does this attribute relate to `reqd_work_group_size` and related existing > attributes? They seems to be different/"unrelated". Based on the description of the `amdgpu-num-work-groups` attribute it provides "number of work-groups", whereas `reqd_work_group_size` provides "number of work-items in a work-group". I think this attributed can be useful for other targets. The optimization ideas described in https://github.com/llvm/llvm-project/pull/75647#issuecomment-1863352459 seems to be generic. There is an RFC to unify some existing functionality exposing "grid" information: https://discourse.llvm.org/t/proposing-llvm-gpu-intrinsics/75374. This might fall into similar category. https://github.com/llvm/llvm-project/pull/75647 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] 128f39d - Fix crash in getFullyQualifiedName for inline namespace
Author: Alexey Bader
Date: 2019-12-28T16:35:51+03:00
New Revision: 128f39da932be50cb49646084820119e6e0d1e22
URL:
https://github.com/llvm/llvm-project/commit/128f39da932be50cb49646084820119e6e0d1e22
DIFF:
https://github.com/llvm/llvm-project/commit/128f39da932be50cb49646084820119e6e0d1e22.diff
LOG: Fix crash in getFullyQualifiedName for inline namespace
Summary: The ICE happens when the most outer namespace is an inline namespace.
Reviewers: bkramer, ilya-biryukov
Reviewed By: ilya-biryukov
Subscribers: ebevhan, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D71962
Added:
Modified:
clang/lib/AST/QualTypeNames.cpp
clang/unittests/Tooling/QualTypeNamesTest.cpp
Removed:
diff --git a/clang/lib/AST/QualTypeNames.cpp b/clang/lib/AST/QualTypeNames.cpp
index f28f00171cce..73a33a208233 100644
--- a/clang/lib/AST/QualTypeNames.cpp
+++ b/clang/lib/AST/QualTypeNames.cpp
@@ -192,7 +192,7 @@ static NestedNameSpecifier *createOuterNNS(const ASTContext
&Ctx, const Decl *D,
// Ignore inline namespace;
NS = dyn_cast(NS->getDeclContext());
}
-if (NS->getDeclName()) {
+if (NS && NS->getDeclName()) {
return createNestedNameSpecifier(Ctx, NS, WithGlobalNsPrefix);
}
return nullptr; // no starting '::', no anonymous
diff --git a/clang/unittests/Tooling/QualTypeNamesTest.cpp
b/clang/unittests/Tooling/QualTypeNamesTest.cpp
index b6c302977876..ff6b78c2666d 100644
--- a/clang/unittests/Tooling/QualTypeNamesTest.cpp
+++ b/clang/unittests/Tooling/QualTypeNamesTest.cpp
@@ -223,6 +223,17 @@ TEST(QualTypeNameTest, getFullyQualifiedName) {
"}\n"
);
+ TypeNameVisitor InlineNamespace;
+ InlineNamespace.ExpectedQualTypeNames["c"] = "B::C";
+ InlineNamespace.runOver("inline namespace A {\n"
+ " namespace B {\n"
+ "class C {};\n"
+ " }\n"
+ "}\n"
+ "using namespace A::B;\n"
+ "C c;\n",
+ TypeNameVisitor::Lang_CXX11);
+
TypeNameVisitor AnonStrucs;
AnonStrucs.ExpectedQualTypeNames["a"] = "short";
AnonStrucs.ExpectedQualTypeNames["un_in_st_1"] =
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] a500a43 - [CodeGen][AMDGPU] Fix ICE for static initializer IR generation
Author: Alexey Bader
Date: 2020-12-12T23:26:54+03:00
New Revision: a500a4358789d1794bc672421c55900ea2bbc938
URL:
https://github.com/llvm/llvm-project/commit/a500a4358789d1794bc672421c55900ea2bbc938
DIFF:
https://github.com/llvm/llvm-project/commit/a500a4358789d1794bc672421c55900ea2bbc938.diff
LOG: [CodeGen][AMDGPU] Fix ICE for static initializer IR generation
Differential Revision: https://reviews.llvm.org/D92782
Added:
Modified:
clang/lib/CodeGen/CGDecl.cpp
clang/test/CodeGen/address-space.c
Removed:
diff --git a/clang/lib/CodeGen/CGDecl.cpp b/clang/lib/CodeGen/CGDecl.cpp
index 478821665e45..a01638f0b67b 100644
--- a/clang/lib/CodeGen/CGDecl.cpp
+++ b/clang/lib/CodeGen/CGDecl.cpp
@@ -353,12 +353,11 @@ CodeGenFunction::AddInitializerToStaticVarDecl(const
VarDecl &D,
if (GV->getValueType() != Init->getType()) {
llvm::GlobalVariable *OldGV = GV;
-GV = new llvm::GlobalVariable(CGM.getModule(), Init->getType(),
- OldGV->isConstant(),
- OldGV->getLinkage(), Init, "",
- /*InsertBefore*/ OldGV,
- OldGV->getThreadLocalMode(),
-
CGM.getContext().getTargetAddressSpace(D.getType()));
+GV = new llvm::GlobalVariable(
+CGM.getModule(), Init->getType(), OldGV->isConstant(),
+OldGV->getLinkage(), Init, "",
+/*InsertBefore*/ OldGV, OldGV->getThreadLocalMode(),
+OldGV->getType()->getPointerAddressSpace());
GV->setVisibility(OldGV->getVisibility());
GV->setDSOLocal(OldGV->isDSOLocal());
GV->setComdat(OldGV->getComdat());
diff --git a/clang/test/CodeGen/address-space.c
b/clang/test/CodeGen/address-space.c
index c66dfc87c0c0..baefb4b983fc 100644
--- a/clang/test/CodeGen/address-space.c
+++ b/clang/test/CodeGen/address-space.c
@@ -59,3 +59,14 @@ void __attribute__((address_space(1)))*
void_ptr_arithmetic_test(void __attribute__((address_space(1))) *arg) {
return arg + 4;
}
+
+// CHECK-LABEL: define i32* @test5(
+const unsigned *test5() {
+ // Intentionally leave a part of an array uninitialized. This triggers a
+ //
diff erent code path contrary to a fully initialized array.
+ // CHECK: ret i32* getelementptr inbounds ([256 x i32]
+ static const unsigned bars[256] = {
+ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,
+ 11, 12, 13, 14, 15, 16, 17, 18, 19, 20};
+ return &bars[0];
+}
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] 9263931 - [SYCL] Assume SYCL device functions are convergent
Author: Alexey Bader
Date: 2020-09-29T15:23:50+03:00
New Revision: 9263931fcccdc99000c1de668bea330711333729
URL:
https://github.com/llvm/llvm-project/commit/9263931fcccdc99000c1de668bea330711333729
DIFF:
https://github.com/llvm/llvm-project/commit/9263931fcccdc99000c1de668bea330711333729.diff
LOG: [SYCL] Assume SYCL device functions are convergent
SYCL device compiler (similar to other SPMD compilers) assumes that
functions are convergent by default to avoid invalid transformations.
This attribute can be removed if compiler can prove that function does
not have convergent operations.
Reviewed By: Naghasan
Differential Revision: https://reviews.llvm.org/D87282
Added:
clang/test/CodeGenSYCL/convergent.cpp
Modified:
clang/lib/Frontend/CompilerInvocation.cpp
Removed:
diff --git a/clang/lib/Frontend/CompilerInvocation.cpp
b/clang/lib/Frontend/CompilerInvocation.cpp
index 2d008d8a3fbe..42224339250d 100644
--- a/clang/lib/Frontend/CompilerInvocation.cpp
+++ b/clang/lib/Frontend/CompilerInvocation.cpp
@@ -2882,7 +2882,8 @@ static void ParseLangArgs(LangOptions &Opts, ArgList
&Args, InputKind IK,
Opts.Coroutines = Opts.CPlusPlus20 || Args.hasArg(OPT_fcoroutines_ts);
Opts.ConvergentFunctions = Opts.OpenCL || (Opts.CUDA && Opts.CUDAIsDevice) ||
-Args.hasArg(OPT_fconvergent_functions);
+ Opts.SYCLIsDevice ||
+ Args.hasArg(OPT_fconvergent_functions);
Opts.DoubleSquareBracketAttributes =
Args.hasFlag(OPT_fdouble_square_bracket_attributes,
diff --git a/clang/test/CodeGenSYCL/convergent.cpp
b/clang/test/CodeGenSYCL/convergent.cpp
new file mode 100644
index ..784fb8976c27
--- /dev/null
+++ b/clang/test/CodeGenSYCL/convergent.cpp
@@ -0,0 +1,20 @@
+// RUN: %clang_cc1 -fsycl -fsycl-is-device -emit-llvm -disable-llvm-passes \
+// RUN: -triple spir64-unknown-unknown-sycldevice -emit-llvm %s -o - | \
+// RUN: FileCheck %s
+
+// CHECK-DAG: Function Attrs:
+// CHECK-DAG-SAME: convergent
+// CHECK-DAG-NEXT: define void @_Z3foov
+void foo() {
+ int a = 1;
+}
+
+template
+__attribute__((sycl_kernel)) void kernel_single_task(const Func &kernelFunc) {
+ kernelFunc();
+}
+
+int main() {
+ kernel_single_task([] { foo(); });
+ return 0;
+}
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] 2ab513c - [SYCL] Enable `opencl_global_[host,device]` attributes for SYCL
Author: Alexey Bader
Date: 2021-05-18T10:27:35+03:00
New Revision: 2ab513cd3e0648806db7ed1f8170ad4a3d4e7749
URL:
https://github.com/llvm/llvm-project/commit/2ab513cd3e0648806db7ed1f8170ad4a3d4e7749
DIFF:
https://github.com/llvm/llvm-project/commit/2ab513cd3e0648806db7ed1f8170ad4a3d4e7749.diff
LOG: [SYCL] Enable `opencl_global_[host,device]` attributes for SYCL
Differential Revision: https://reviews.llvm.org/D100396
Added:
Modified:
clang/docs/SYCLSupport.rst
clang/include/clang/AST/Type.h
clang/include/clang/Basic/AddressSpaces.h
clang/include/clang/Sema/ParsedAttr.h
clang/lib/AST/ASTContext.cpp
clang/lib/AST/ItaniumMangle.cpp
clang/lib/AST/TypePrinter.cpp
clang/lib/Basic/Targets/AMDGPU.cpp
clang/lib/Basic/Targets/NVPTX.h
clang/lib/Basic/Targets/SPIR.h
clang/lib/Basic/Targets/TCE.h
clang/lib/Basic/Targets/X86.h
clang/test/CodeGenSYCL/address-space-conversions.cpp
clang/test/SemaSYCL/address-space-conversions.cpp
clang/test/SemaTemplate/address_space-dependent.cpp
Removed:
diff --git a/clang/docs/SYCLSupport.rst b/clang/docs/SYCLSupport.rst
index d45ecfbea53f8..6b529e3eb0127 100644
--- a/clang/docs/SYCLSupport.rst
+++ b/clang/docs/SYCLSupport.rst
@@ -34,10 +34,20 @@ the address space qualifier inference as detailed in
The default address space is "generic-memory", which is a virtual address space
that overlaps the global, local, and private address spaces. SYCL mode enables
-explicit conversions to/from the default address space from/to the address
-space-attributed type and implicit conversions from the address
space-attributed
-type to the default address space. All named address spaces are disjoint and
-sub-sets of default address space.
+following conversions:
+
+- explicit conversions to/from the default address space from/to the address
+ space-attributed type
+- implicit conversions from the address space-attributed type to the default
+ address space
+- explicit conversions to/from the global address space from/to the
+ ``__attribute__((opencl_global_device))`` or
+ ``__attribute__((opencl_global_host))`` address space-attributed type
+- implicit conversions from the ``__attribute__((opencl_global_device))`` or
+ ``__attribute__((opencl_global_host))`` address space-attributed type to the
+ global address space
+
+All named address spaces are disjoint and sub-sets of default address space.
The SPIR target allocates SYCL namespace scope variables in the global address
space.
@@ -93,6 +103,10 @@ space attributes for pointers:
- SYCL address_space enumeration
* - ``__attribute__((opencl_global))``
- global_space, constant_space
+ * - ``__attribute__((opencl_global_device))``
+ - global_space
+ * - ``__attribute__((opencl_global_host))``
+ - global_space
* - ``__attribute__((opencl_local))``
- local_space
* - ``__attribute__((opencl_private))``
diff --git a/clang/include/clang/AST/Type.h b/clang/include/clang/AST/Type.h
index 84de51aba5b2d..9f46d53378976 100644
--- a/clang/include/clang/AST/Type.h
+++ b/clang/include/clang/AST/Type.h
@@ -486,13 +486,16 @@ class Qualifiers {
// allocated on device, which are a subset of __global.
(A == LangAS::opencl_global && (B == LangAS::opencl_global_device ||
B == LangAS::opencl_global_host)) ||
+ (A == LangAS::sycl_global && (B == LangAS::sycl_global_device ||
+ B == LangAS::sycl_global_host)) ||
// Consider pointer size address spaces to be equivalent to default.
((isPtrSizeAddressSpace(A) || A == LangAS::Default) &&
(isPtrSizeAddressSpace(B) || B == LangAS::Default)) ||
// Default is a superset of SYCL address spaces.
(A == LangAS::Default &&
(B == LangAS::sycl_private || B == LangAS::sycl_local ||
- B == LangAS::sycl_global));
+ B == LangAS::sycl_global || B == LangAS::sycl_global_device ||
+ B == LangAS::sycl_global_host));
}
/// Returns true if the address space in these qualifiers is equal to or
diff --git a/clang/include/clang/Basic/AddressSpaces.h
b/clang/include/clang/Basic/AddressSpaces.h
index 5fa031250a629..99bb67fd26d19 100644
--- a/clang/include/clang/Basic/AddressSpaces.h
+++ b/clang/include/clang/Basic/AddressSpaces.h
@@ -46,6 +46,8 @@ enum class LangAS : unsigned {
// SYCL specific address spaces.
sycl_global,
+ sycl_global_device,
+ sycl_global_host,
sycl_local,
sycl_private,
diff --git a/clang/include/clang/Sema/ParsedAttr.h
b/clang/include/clang/Sema/ParsedAttr.h
index 347925873b30f..f47f557adeb10 100644
--- a/clang/include/clang/Sema/ParsedAttr.h
+++ b/clang/include/clang/Sema/ParsedAttr.h
@@ -657,6 +657,10 @@ class ParsedAttr final
switch (getKind()) {
case
[clang] 15feaaa - Add myself as a code owner for SYCL support
Author: Alexey Bader Date: 2021-09-20T09:32:25+03:00 New Revision: 15feaaa359c7245bb59ff0a2aa3b806682f44286 URL: https://github.com/llvm/llvm-project/commit/15feaaa359c7245bb59ff0a2aa3b806682f44286 DIFF: https://github.com/llvm/llvm-project/commit/15feaaa359c7245bb59ff0a2aa3b806682f44286.diff LOG: Add myself as a code owner for SYCL support Added: Modified: clang/CODE_OWNERS.TXT Removed: diff --git a/clang/CODE_OWNERS.TXT b/clang/CODE_OWNERS.TXT index faf9575a2898..740aeeb7125b 100644 --- a/clang/CODE_OWNERS.TXT +++ b/clang/CODE_OWNERS.TXT @@ -8,6 +8,10 @@ beautification by scripts. The fields are: name (N), email (E), web-address (W), PGP key ID and fingerprint (P), description (D), and snail-mail address (S). +N: Alexey Bader +E: alexey.ba...@intel.com +D: SYCL support + N: Aaron Ballman E: aa...@aaronballman.com D: Clang attributes ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] c064ba3 - [NFC] Add commas in code comments.
Author: Alexey Bader
Date: 2021-08-13T08:59:47+03:00
New Revision: c064ba34c7d867c1180279fe3dca8817d835cb28
URL:
https://github.com/llvm/llvm-project/commit/c064ba34c7d867c1180279fe3dca8817d835cb28
DIFF:
https://github.com/llvm/llvm-project/commit/c064ba34c7d867c1180279fe3dca8817d835cb28.diff
LOG: [NFC] Add commas in code comments.
Added:
Modified:
clang/include/clang/Sema/ParsedAttr.h
Removed:
diff --git a/clang/include/clang/Sema/ParsedAttr.h
b/clang/include/clang/Sema/ParsedAttr.h
index 61a36ddcc1db..408032cec7e8 100644
--- a/clang/include/clang/Sema/ParsedAttr.h
+++ b/clang/include/clang/Sema/ParsedAttr.h
@@ -627,7 +627,7 @@ class ParsedAttr final
/// a Spelling enumeration, the value UINT_MAX is returned.
unsigned getSemanticSpelling() const;
- /// If this is an OpenCL address space attribute returns its representation
+ /// If this is an OpenCL address space attribute, returns its representation
/// in LangAS, otherwise returns default address space.
LangAS asOpenCLLangAS() const {
switch (getParsedKind()) {
@@ -650,7 +650,7 @@ class ParsedAttr final
}
}
- /// If this is an OpenCL address space attribute returns its SYCL
+ /// If this is an OpenCL address space attribute, returns its SYCL
/// representation in LangAS, otherwise returns default address space.
LangAS asSYCLLangAS() const {
switch (getKind()) {
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] d754b97 - [NFC] Drop idle compiler option from the test.
Author: Alexey Bader
Date: 2021-08-13T13:20:11+03:00
New Revision: d754b970eddb83b1c68346a36b067391d93166b0
URL:
https://github.com/llvm/llvm-project/commit/d754b970eddb83b1c68346a36b067391d93166b0
DIFF:
https://github.com/llvm/llvm-project/commit/d754b970eddb83b1c68346a36b067391d93166b0.diff
LOG: [NFC] Drop idle compiler option from the test.
Differential Revision: https://reviews.llvm.org/D108020
Added:
Modified:
clang/test/AST/ast-print-sycl-unique-stable-name.cpp
Removed:
diff --git a/clang/test/AST/ast-print-sycl-unique-stable-name.cpp
b/clang/test/AST/ast-print-sycl-unique-stable-name.cpp
index 3f49ea9ee733c..b55d01aa1b6b6 100644
--- a/clang/test/AST/ast-print-sycl-unique-stable-name.cpp
+++ b/clang/test/AST/ast-print-sycl-unique-stable-name.cpp
@@ -1,4 +1,4 @@
-// RUN: %clang_cc1 -ast-print -fsycl-is-device %s -o - -triple
spir64-sycldevice | FileCheck %s
+// RUN: %clang_cc1 -ast-print -fsycl-is-device %s -o - | FileCheck %s
template
void WrappedInTemplate(T t) {
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] 49682f1 - [SPIR-V] Add translator tool
Author: Henry Linjamäki
Date: 2021-11-18T03:41:24+03:00
New Revision: 49682f14bf3fb8db5e2721d9896b27bb4c2bd635
URL:
https://github.com/llvm/llvm-project/commit/49682f14bf3fb8db5e2721d9896b27bb4c2bd635
DIFF:
https://github.com/llvm/llvm-project/commit/49682f14bf3fb8db5e2721d9896b27bb4c2bd635.diff
LOG: [SPIR-V] Add translator tool
Add a tool for constructing commands for translating LLVM IR to
SPIR-V.
Used by HIPSPV tool chain (D110618).
Reviewed By: bader
Differential Revision: https://reviews.llvm.org/D112404
Added:
clang/lib/Driver/ToolChains/SPIRV.cpp
clang/lib/Driver/ToolChains/SPIRV.h
Modified:
clang/lib/Driver/CMakeLists.txt
Removed:
diff --git a/clang/lib/Driver/CMakeLists.txt b/clang/lib/Driver/CMakeLists.txt
index 61a1cd04e088..31e357b67361 100644
--- a/clang/lib/Driver/CMakeLists.txt
+++ b/clang/lib/Driver/CMakeLists.txt
@@ -69,6 +69,7 @@ add_clang_library(clangDriver
ToolChains/PS4CPU.cpp
ToolChains/RISCVToolchain.cpp
ToolChains/Solaris.cpp
+ ToolChains/SPIRV.cpp
ToolChains/TCE.cpp
ToolChains/VEToolchain.cpp
ToolChains/WebAssembly.cpp
diff --git a/clang/lib/Driver/ToolChains/SPIRV.cpp
b/clang/lib/Driver/ToolChains/SPIRV.cpp
new file mode 100644
index ..3f942478e2ad
--- /dev/null
+++ b/clang/lib/Driver/ToolChains/SPIRV.cpp
@@ -0,0 +1,48 @@
+//===--- SPIRV.cpp - SPIR-V Tool Implementations *- C++
-*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM
Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===--===//
+#include "SPIRV.h"
+#include "CommonArgs.h"
+#include "clang/Driver/Compilation.h"
+#include "clang/Driver/Driver.h"
+#include "clang/Driver/InputInfo.h"
+#include "clang/Driver/Options.h"
+
+using namespace clang::driver::tools;
+using namespace llvm::opt;
+
+void SPIRV::constructTranslateCommand(Compilation &C, const Tool &T,
+ const JobAction &JA,
+ const InputInfo &Output,
+ const InputInfo &Input,
+ const llvm::opt::ArgStringList &Args) {
+ llvm::opt::ArgStringList CmdArgs(Args);
+ CmdArgs.push_back(Input.getFilename());
+
+ if (Input.getType() == types::TY_PP_Asm)
+CmdArgs.push_back("-to-binary");
+ if (Output.getType() == types::TY_PP_Asm)
+CmdArgs.push_back("-spirv-text");
+
+ CmdArgs.append({"-o", Output.getFilename()});
+
+ const char *Exec =
+ C.getArgs().MakeArgString(T.getToolChain().GetProgramPath("llvm-spirv"));
+ C.addCommand(std::make_unique(JA, T, ResponseFileSupport::None(),
+ Exec, CmdArgs, Input, Output));
+}
+
+void SPIRV::Translator::ConstructJob(Compilation &C, const JobAction &JA,
+ const InputInfo &Output,
+ const InputInfoList &Inputs,
+ const ArgList &Args,
+ const char *LinkingOutput) const {
+ claimNoWarnArgs(Args);
+ if (Inputs.size() != 1)
+llvm_unreachable("Invalid number of input files.");
+ constructTranslateCommand(C, *this, JA, Output, Inputs[0], {});
+}
diff --git a/clang/lib/Driver/ToolChains/SPIRV.h
b/clang/lib/Driver/ToolChains/SPIRV.h
new file mode 100644
index ..35d0446bd8b8
--- /dev/null
+++ b/clang/lib/Driver/ToolChains/SPIRV.h
@@ -0,0 +1,46 @@
+//===--- SPIRV.h - SPIR-V Tool Implementations --*- C++
-*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM
Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===--===//
+
+#ifndef LLVM_CLANG_LIB_DRIVER_TOOLCHAINS_SPIRV_H
+#define LLVM_CLANG_LIB_DRIVER_TOOLCHAINS_SPIRV_H
+
+#include "clang/Driver/Tool.h"
+#include "clang/Driver/ToolChain.h"
+
+namespace clang {
+namespace driver {
+namespace tools {
+namespace SPIRV {
+
+void addTranslatorArgs(const llvm::opt::ArgList &InArgs,
+ llvm::opt::ArgStringList &OutArgs);
+
+void constructTranslateCommand(Compilation &C, const Tool &T,
+ const JobAction &JA, const InputInfo &Output,
+ const InputInfo &Input,
+ const llvm::opt::ArgStringList &Args);
+
+class LLVM_LIBRARY_VISIBILITY Translator : public Tool {
+public:
+ Translator(const ToolChain &TC)
+ : Tool("SPIR-V::Translator", "llvm-spirv", TC) {}
+
+ bool hasIntegratedCPP() const override { return false; }
+ bool hasIntegratedAssembler() const override { return true; }
+
+
[clang] 66bd607 - [Attr][Doc] Fix pragma unroll documentation.
Author: Alexey Bader Date: 2022-10-19T03:50:47-07:00 New Revision: 66bd6074c133402e45075b591c062c22f308ef26 URL: https://github.com/llvm/llvm-project/commit/66bd6074c133402e45075b591c062c22f308ef26 DIFF: https://github.com/llvm/llvm-project/commit/66bd6074c133402e45075b591c062c22f308ef26.diff LOG: [Attr][Doc] Fix pragma unroll documentation. There is a contradiction in the #pragma unroll behavior documentation. It says that specifying `#pragma unroll` without a parameter directs the loop unroller to attempt to partially unroll the loop if the trip count is not known at compile time. At the same time later it states that `#pragma unroll` has identical semantics to `#pragma clang loop unroll(full)`, which doesn't attempt to unroll partially if the trip count is not known at compile time. pragma clang loop unroll(enable): If unroll(enable) is specified the unroller will attempt to fully unroll the loop if the trip count is known at compile time. If the fully unrolled code size is greater than an internal limit the loop will be partially unrolled up to this limit. If the trip count is not known at compile time the loop will be partially unrolled with a heuristically chosen unroll factor. pragma clang loop unroll(full): If unroll(full) is specified the unroller will attempt to fully unroll the loop if the trip count is known at compile time identically to unroll(enable). However, with unroll(full) the loop will not be unrolled if the loop count is not known at compile time. Differential Revision: https://reviews.llvm.org/D136160 Added: Modified: clang/include/clang/Basic/AttrDocs.td Removed: diff --git a/clang/include/clang/Basic/AttrDocs.td b/clang/include/clang/Basic/AttrDocs.td index 33a18ac033252..484052f4db8ad 100644 --- a/clang/include/clang/Basic/AttrDocs.td +++ b/clang/include/clang/Basic/AttrDocs.td @@ -3569,7 +3569,7 @@ Specifying ``#pragma nounroll`` indicates that the loop should not be unrolled: } ``#pragma unroll`` and ``#pragma unroll _value_`` have identical semantics to -``#pragma clang loop unroll(full)`` and +``#pragma clang loop unroll(enable)`` and ``#pragma clang loop unroll_count(_value_)`` respectively. ``#pragma nounroll`` is equivalent to ``#pragma clang loop unroll(disable)``. See `language extensions ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] a2e2f47 - [Clang][Docs] Fix typo in clang-offload-packager documentation
Author: Alexey Bader Date: 2023-09-11T16:24:37-07:00 New Revision: a2e2f471c51f0aea836a4929dfdf50523dcff1eb URL: https://github.com/llvm/llvm-project/commit/a2e2f471c51f0aea836a4929dfdf50523dcff1eb DIFF: https://github.com/llvm/llvm-project/commit/a2e2f471c51f0aea836a4929dfdf50523dcff1eb.diff LOG: [Clang][Docs] Fix typo in clang-offload-packager documentation Fixed formatting of the section violating 80-char line limit. Added: Modified: clang/docs/ClangOffloadPackager.rst Removed: diff --git a/clang/docs/ClangOffloadPackager.rst b/clang/docs/ClangOffloadPackager.rst index 1d6ded952e8adfe..2b985e260e302de 100644 --- a/clang/docs/ClangOffloadPackager.rst +++ b/clang/docs/ClangOffloadPackager.rst @@ -124,7 +124,7 @@ array of the :ref:`string entry` format. +--+--+---+ | Type | Identifier | Description | +==+==+===+ -| uint64_t | key offset | Absolute byte offset of the key in th string table| +| uint64_t | key offset | Absolute byte offset of the key in the string table | +--+--+---+ | uint64_t | value offset | Absolute byte offset of the value in the string table | +--+--+---+ @@ -145,9 +145,9 @@ Usage = This tool can be used with the following arguments. Generally information is -passed as a key-value pair to the ``image=`` argument. The ``file`` and ``triple``, -arguments are considered mandatory to make a valid image. The ``arch`` argument -is suggested. +passed as a key-value pair to the ``image=`` argument. The ``file`` and +``triple``, arguments are considered mandatory to make a valid image. +The ``arch`` argument is suggested. .. code-block:: console ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] bd97704 - [SYCL] Driver option to select SYCL version
Author: Ruyman
Date: 2020-02-27T15:08:42+03:00
New Revision: bd97704eb5a95ecb048ce343c1a4be5d94e5
URL:
https://github.com/llvm/llvm-project/commit/bd97704eb5a95ecb048ce343c1a4be5d94e5
DIFF:
https://github.com/llvm/llvm-project/commit/bd97704eb5a95ecb048ce343c1a4be5d94e5.diff
LOG: [SYCL] Driver option to select SYCL version
Summary:
User can select the version of SYCL the compiler will
use via the flag -sycl-std, similar to -cl-std.
The flag defines the LangOpts.SYCLVersion option to the
version of SYCL. The default value is undefined.
If driver is building SYCL code, flag is set to the default SYCL
version (1.2.1)
The preprocessor uses this variable to define CL_SYCL_LANGUAGE_VERSION macro,
which should be defined according to SYCL 1.2.1 standard.
Only valid value at this point for the flag is 1.2.1.
Co-Authored-By: David Wood
Signed-off-by: Ruyman Reyes
Subscribers: ebevhan, Anastasia, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D72857
Signed-off-by: Alexey Bader
Added:
Modified:
clang/include/clang/Basic/LangOptions.def
clang/include/clang/Driver/Options.td
clang/lib/Driver/ToolChains/Clang.cpp
clang/lib/Frontend/CompilerInvocation.cpp
clang/lib/Frontend/InitPreprocessor.cpp
clang/test/Driver/sycl.c
clang/test/Frontend/sycl-aux-triple.cpp
clang/test/Preprocessor/sycl-macro.cpp
clang/test/SemaSYCL/kernel-attribute.cpp
Removed:
diff --git a/clang/include/clang/Basic/LangOptions.def
b/clang/include/clang/Basic/LangOptions.def
index 3cc7c384ac10..53b87b737568 100644
--- a/clang/include/clang/Basic/LangOptions.def
+++ b/clang/include/clang/Basic/LangOptions.def
@@ -230,7 +230,9 @@ LANGOPT(GPURelocatableDeviceCode, 1, 0, "generate
relocatable device code")
LANGOPT(GPUAllowDeviceInit, 1, 0, "allowing device side global init functions
for HIP")
LANGOPT(GPUMaxThreadsPerBlock, 32, 256, "default max threads per block for
kernel launch bounds for HIP")
+LANGOPT(SYCL , 1, 0, "SYCL")
LANGOPT(SYCLIsDevice , 1, 0, "Generate code for SYCL device")
+LANGOPT(SYCLVersion , 32, 0, "Version of the SYCL standard used")
LANGOPT(HIPUseNewLaunchAPI, 1, 0, "Use new kernel launching API for HIP")
diff --git a/clang/include/clang/Driver/Options.td
b/clang/include/clang/Driver/Options.td
index f1801e3e89e7..36626342e364 100644
--- a/clang/include/clang/Driver/Options.td
+++ b/clang/include/clang/Driver/Options.td
@@ -3420,10 +3420,12 @@ defm underscoring : BooleanFFlag<"underscoring">,
Group;
defm whole_file : BooleanFFlag<"whole-file">, Group;
// C++ SYCL options
-def fsycl : Flag<["-"], "fsycl">, Group,
+def fsycl : Flag<["-"], "fsycl">, Group, Flags<[CC1Option,
CoreOption]>,
HelpText<"Enable SYCL kernels compilation for device">;
-def fno_sycl : Flag<["-"], "fno-sycl">, Group,
+def fno_sycl : Flag<["-"], "fno-sycl">, Group, Flags<[CoreOption]>,
HelpText<"Disable SYCL kernels compilation for device">;
+def sycl_std_EQ : Joined<["-"], "sycl-std=">, Group,
Flags<[CC1Option, NoArgumentUnused, CoreOption]>,
+ HelpText<"SYCL language standard to compile for.">, Values<"2017, 121,
1.2.1, sycl-1.2.1">;
include "CC1Options.td"
diff --git a/clang/lib/Driver/ToolChains/Clang.cpp
b/clang/lib/Driver/ToolChains/Clang.cpp
index d387a1dc2079..87596dd19c5a 100644
--- a/clang/lib/Driver/ToolChains/Clang.cpp
+++ b/clang/lib/Driver/ToolChains/Clang.cpp
@@ -4040,9 +4040,18 @@ void Clang::ConstructJob(Compilation &C, const JobAction
&JA,
CmdArgs.push_back(Args.MakeArgString(NormalizedTriple));
}
- if (Args.hasFlag(options::OPT_fsycl, options::OPT_fno_sycl, false))
+ if (Args.hasFlag(options::OPT_fsycl, options::OPT_fno_sycl, false)) {
+CmdArgs.push_back("-fsycl");
CmdArgs.push_back("-fsycl-is-device");
+if (Arg *A = Args.getLastArg(options::OPT_sycl_std_EQ)) {
+ A->render(Args, CmdArgs);
+} else {
+ // Ensure the default version in SYCL mode is 1.2.1 (aka 2017)
+ CmdArgs.push_back("-sycl-std=2017");
+}
+ }
+
if (IsOpenMPDevice) {
// We have to pass the triple of the host if compiling for an OpenMP
device.
std::string NormalizedTriple =
diff --git a/clang/lib/Frontend/CompilerInvocation.cpp
b/clang/lib/Frontend/CompilerInvocation.cpp
index 9cc41c9d96f8..76f63d065017 100644
--- a/clang/lib/Frontend/CompilerInvocation.cpp
+++ b/clang/lib/Frontend/CompilerInvocation.cpp
@@ -2535,6 +2535,24 @@ static void ParseLangArgs(LangOptions &Opts, ArgList
&Args, InputKind IK,
LangStd = OpenCLLangStd;
}
+ Opts.SYCL = Args.hasArg(options::OPT_fsycl);
+ Opts.SYCLIsDevice = Opts.SYCL && Args.hasArg(options::OPT_fsycl_is_device);
+ if (Opts.SYCL) {
+// -sycl-std applies to any SYCL source, not only those containing kernels,
+// but also those using the SYCL API
+if (const Arg *A = Args.getLastArg(OPT_sycl_std_EQ)) {
+
[clang] 740ed61 - Revert "[SYCL] Driver option to select SYCL version"
Author: Alexey Bader
Date: 2020-02-27T16:23:54+03:00
New Revision: 740ed617f7d4d16e7883636c5eff994f8be7eba4
URL:
https://github.com/llvm/llvm-project/commit/740ed617f7d4d16e7883636c5eff994f8be7eba4
DIFF:
https://github.com/llvm/llvm-project/commit/740ed617f7d4d16e7883636c5eff994f8be7eba4.diff
LOG: Revert "[SYCL] Driver option to select SYCL version"
This reverts commit bd97704eb5a95ecb048ce343c1a4be5d94e5.
It broke tests on mac: http://45.33.8.238/mac/9011/step_7.txt
Added:
Modified:
clang/include/clang/Basic/LangOptions.def
clang/include/clang/Driver/Options.td
clang/lib/Driver/ToolChains/Clang.cpp
clang/lib/Frontend/CompilerInvocation.cpp
clang/lib/Frontend/InitPreprocessor.cpp
clang/test/Driver/sycl.c
clang/test/Frontend/sycl-aux-triple.cpp
clang/test/Preprocessor/sycl-macro.cpp
clang/test/SemaSYCL/kernel-attribute.cpp
Removed:
diff --git a/clang/include/clang/Basic/LangOptions.def
b/clang/include/clang/Basic/LangOptions.def
index 53b87b737568..3cc7c384ac10 100644
--- a/clang/include/clang/Basic/LangOptions.def
+++ b/clang/include/clang/Basic/LangOptions.def
@@ -230,9 +230,7 @@ LANGOPT(GPURelocatableDeviceCode, 1, 0, "generate
relocatable device code")
LANGOPT(GPUAllowDeviceInit, 1, 0, "allowing device side global init functions
for HIP")
LANGOPT(GPUMaxThreadsPerBlock, 32, 256, "default max threads per block for
kernel launch bounds for HIP")
-LANGOPT(SYCL , 1, 0, "SYCL")
LANGOPT(SYCLIsDevice , 1, 0, "Generate code for SYCL device")
-LANGOPT(SYCLVersion , 32, 0, "Version of the SYCL standard used")
LANGOPT(HIPUseNewLaunchAPI, 1, 0, "Use new kernel launching API for HIP")
diff --git a/clang/include/clang/Driver/Options.td
b/clang/include/clang/Driver/Options.td
index 36626342e364..f1801e3e89e7 100644
--- a/clang/include/clang/Driver/Options.td
+++ b/clang/include/clang/Driver/Options.td
@@ -3420,12 +3420,10 @@ defm underscoring : BooleanFFlag<"underscoring">,
Group;
defm whole_file : BooleanFFlag<"whole-file">, Group;
// C++ SYCL options
-def fsycl : Flag<["-"], "fsycl">, Group, Flags<[CC1Option,
CoreOption]>,
+def fsycl : Flag<["-"], "fsycl">, Group,
HelpText<"Enable SYCL kernels compilation for device">;
-def fno_sycl : Flag<["-"], "fno-sycl">, Group, Flags<[CoreOption]>,
+def fno_sycl : Flag<["-"], "fno-sycl">, Group,
HelpText<"Disable SYCL kernels compilation for device">;
-def sycl_std_EQ : Joined<["-"], "sycl-std=">, Group,
Flags<[CC1Option, NoArgumentUnused, CoreOption]>,
- HelpText<"SYCL language standard to compile for.">, Values<"2017, 121,
1.2.1, sycl-1.2.1">;
include "CC1Options.td"
diff --git a/clang/lib/Driver/ToolChains/Clang.cpp
b/clang/lib/Driver/ToolChains/Clang.cpp
index 87596dd19c5a..d387a1dc2079 100644
--- a/clang/lib/Driver/ToolChains/Clang.cpp
+++ b/clang/lib/Driver/ToolChains/Clang.cpp
@@ -4040,18 +4040,9 @@ void Clang::ConstructJob(Compilation &C, const JobAction
&JA,
CmdArgs.push_back(Args.MakeArgString(NormalizedTriple));
}
- if (Args.hasFlag(options::OPT_fsycl, options::OPT_fno_sycl, false)) {
-CmdArgs.push_back("-fsycl");
+ if (Args.hasFlag(options::OPT_fsycl, options::OPT_fno_sycl, false))
CmdArgs.push_back("-fsycl-is-device");
-if (Arg *A = Args.getLastArg(options::OPT_sycl_std_EQ)) {
- A->render(Args, CmdArgs);
-} else {
- // Ensure the default version in SYCL mode is 1.2.1 (aka 2017)
- CmdArgs.push_back("-sycl-std=2017");
-}
- }
-
if (IsOpenMPDevice) {
// We have to pass the triple of the host if compiling for an OpenMP
device.
std::string NormalizedTriple =
diff --git a/clang/lib/Frontend/CompilerInvocation.cpp
b/clang/lib/Frontend/CompilerInvocation.cpp
index 76f63d065017..9cc41c9d96f8 100644
--- a/clang/lib/Frontend/CompilerInvocation.cpp
+++ b/clang/lib/Frontend/CompilerInvocation.cpp
@@ -2535,24 +2535,6 @@ static void ParseLangArgs(LangOptions &Opts, ArgList
&Args, InputKind IK,
LangStd = OpenCLLangStd;
}
- Opts.SYCL = Args.hasArg(options::OPT_fsycl);
- Opts.SYCLIsDevice = Opts.SYCL && Args.hasArg(options::OPT_fsycl_is_device);
- if (Opts.SYCL) {
-// -sycl-std applies to any SYCL source, not only those containing kernels,
-// but also those using the SYCL API
-if (const Arg *A = Args.getLastArg(OPT_sycl_std_EQ)) {
- Opts.SYCLVersion = llvm::StringSwitch(A->getValue())
- .Cases("2017", "1.2.1", "121", "sycl-1.2.1", 2017)
- .Default(0U);
-
- if (Opts.SYCLVersion == 0U) {
-// User has passed an invalid value to the flag, this is an error
-Diags.Report(diag::err_drv_invalid_value)
-<< A->getAsString(Args) << A->getValue();
- }
-}
- }
-
Opts.IncludeDefaultHeader = Args.hasArg(OPT_finclude_default_header);
Opts.DeclareOpenCLBuiltins = Args.hasArg(OPT_fdeclare_ope
[clang] c094e7d - [SYCL] Add sycl_kernel attribute for accelerated code outlining
Author: Mariya Podchishchaeva
Date: 2019-12-03T16:13:22+03:00
New Revision: c094e7dc4b3f9d1c1e590b008bb1cc46e3496abd
URL:
https://github.com/llvm/llvm-project/commit/c094e7dc4b3f9d1c1e590b008bb1cc46e3496abd
DIFF:
https://github.com/llvm/llvm-project/commit/c094e7dc4b3f9d1c1e590b008bb1cc46e3496abd.diff
LOG: [SYCL] Add sycl_kernel attribute for accelerated code outlining
SYCL is single source offload programming model relying on compiler to
separate device code (i.e. offloaded to an accelerator) from the code
executed on the host.
Here is code example of the SYCL program to demonstrate compiler
outlining work:
```
int foo(int x) { return ++x; }
int bar(int x) { throw std::exception("CPU code only!"); }
...
using namespace cl::sycl;
queue Q;
buffer a(range<1>{1024});
Q.submit([&](handler& cgh) {
auto A = a.get_access(cgh);
cgh.parallel_for(range<1>{1024}, [=](id<1> index) {
A[index] = index[0] + foo(42);
});
}
...
```
SYCL device compiler must compile lambda expression passed to
cl::sycl::handler::parallel_for method and function foo called from this
lambda expression for an "accelerator". SYCL device compiler also must
ignore bar function as it's not required for offloaded code execution.
This patch adds the sycl_kernel attribute, which is used to mark code
passed to cl::sycl::handler::parallel_for as "accelerated code".
Attribute must be applied to function templates which parameters include
at least "kernel name" and "kernel function object". These parameters
will be used to establish an ABI between the host application and
offloaded part.
Reviewers: jlebar, keryell, Naghasan, ABataev, Anastasia, bader, aaron.ballman,
rjmccall, rsmith
Reviewed By: keryell, bader
Subscribers: mgorny, OlegM, ArturGainullin, agozillon, aaron.ballman, ebevhan,
Anastasia, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D60455
Signed-off-by: Alexey Bader
Added:
clang/test/SemaSYCL/kernel-attribute-on-non-sycl.cpp
clang/test/SemaSYCL/kernel-attribute.cpp
Modified:
clang/include/clang/Basic/Attr.td
clang/include/clang/Basic/AttrDocs.td
clang/include/clang/Basic/DiagnosticSemaKinds.td
clang/lib/Sema/SemaDeclAttr.cpp
Removed:
diff --git a/clang/include/clang/Basic/Attr.td
b/clang/include/clang/Basic/Attr.td
index 21cf53f0a815..4ea1c9f58beb 100644
--- a/clang/include/clang/Basic/Attr.td
+++ b/clang/include/clang/Basic/Attr.td
@@ -121,6 +121,11 @@ def GlobalVar : SubsetSubjectisInlineSpecified()}], "inline functions">;
+def FunctionTmpl
+: SubsetSubjectgetTemplatedKind() ==
+ FunctionDecl::TK_FunctionTemplate}],
+"function templates">;
+
// FIXME: this hack is needed because DeclNodes.td defines the base Decl node
// type to be a class, not a definition. This makes it impossible to create an
// attribute subject which accepts a Decl. Normally, this is not a problem,
@@ -296,6 +301,7 @@ def MicrosoftExt : LangOpt<"MicrosoftExt">;
def Borland : LangOpt<"Borland">;
def CUDA : LangOpt<"CUDA">;
def HIP : LangOpt<"HIP">;
+def SYCL : LangOpt<"SYCLIsDevice">;
def COnly : LangOpt<"COnly", "!LangOpts.CPlusPlus">;
def CPlusPlus : LangOpt<"CPlusPlus">;
def OpenCL : LangOpt<"OpenCL">;
@@ -1056,6 +1062,13 @@ def CUDAShared : InheritableAttr {
let Documentation = [Undocumented];
}
+def SYCLKernel : InheritableAttr {
+ let Spellings = [Clang<"sycl_kernel">];
+ let Subjects = SubjectList<[FunctionTmpl]>;
+ let LangOpts = [SYCL];
+ let Documentation = [SYCLKernelDocs];
+}
+
def C11NoReturn : InheritableAttr {
let Spellings = [Keyword<"_Noreturn">];
let Subjects = SubjectList<[Function], ErrorDiag>;
diff --git a/clang/include/clang/Basic/AttrDocs.td
b/clang/include/clang/Basic/AttrDocs.td
index f35199f81c3e..d47315a3f518 100644
--- a/clang/include/clang/Basic/AttrDocs.td
+++ b/clang/include/clang/Basic/AttrDocs.td
@@ -253,6 +253,79 @@ any option of a multiversioned function is undefined.
}];
}
+def SYCLKernelDocs : Documentation {
+ let Category = DocCatFunction;
+ let Content = [{
+The ``sycl_kernel`` attribute specifies that a function template will be used
+to outline device code and to generate an OpenCL kernel.
+Here is a code example of the SYCL program, which demonstrates the compiler's
+outlining job:
+.. code-block:: c++
+
+ int foo(int x) { return ++x; }
+
+ using namespace cl::sycl;
+ queue Q;
+ buffer a(range<1>{1024});
+ Q.submit([&](handler& cgh) {
+auto A = a.get_access(cgh);
+cgh.parallel_for(range<1>{1024}, [=](id<1> index) {
+ A[index] = index[0] + foo(42);
+});
+ }
+
+A C++ function object passed to the ``parallel_for`` is called a "SYCL kernel".
+A SYCL kernel defines the entry point to the "device part" of the code. The
+compiler will emit all symbols accessible from a "kernel". In this code
+example, the compiler will emit "foo" function. More
[clang] 11a9bae - [AST] Enable expression of OpenCL language address spaces an attribute
Author: Victor Lomuller
Date: 2019-12-05T12:24:06+03:00
New Revision: 11a9bae8f66986751078501988b4414f24dbe37e
URL:
https://github.com/llvm/llvm-project/commit/11a9bae8f66986751078501988b4414f24dbe37e
DIFF:
https://github.com/llvm/llvm-project/commit/11a9bae8f66986751078501988b4414f24dbe37e.diff
LOG: [AST] Enable expression of OpenCL language address spaces an attribute
Summary:
Enable a way to set OpenCL language address space using attributes
in addition to existing keywords.
Signed-off-by: Victor Lomuller vic...@codeplay.com
Reviewers: aaron.ballman, Anastasia
Subscribers: yaxunl, ebevhan, cfe-commits, Naghasan
Tags: #clang
Differential Revision: https://reviews.llvm.org/D71005
Signed-off-by: Alexey Bader
Added:
clang/test/AST/language_address_space_attribute.cpp
Modified:
clang/include/clang/Basic/Attr.td
clang/lib/Sema/SemaType.cpp
clang/test/SemaOpenCL/address-spaces.cl
Removed:
diff --git a/clang/include/clang/Basic/Attr.td
b/clang/include/clang/Basic/Attr.td
index 4ea1c9f58beb..9ca4be0e07c8 100644
--- a/clang/include/clang/Basic/Attr.td
+++ b/clang/include/clang/Basic/Attr.td
@@ -1120,27 +1120,27 @@ def OpenCLAccess : Attr {
}
def OpenCLPrivateAddressSpace : TypeAttr {
- let Spellings = [Keyword<"__private">, Keyword<"private">];
+ let Spellings = [Keyword<"__private">, Keyword<"private">,
Clang<"opencl_private">];
let Documentation = [OpenCLAddressSpacePrivateDocs];
}
def OpenCLGlobalAddressSpace : TypeAttr {
- let Spellings = [Keyword<"__global">, Keyword<"global">];
+ let Spellings = [Keyword<"__global">, Keyword<"global">,
Clang<"opencl_global">];
let Documentation = [OpenCLAddressSpaceGlobalDocs];
}
def OpenCLLocalAddressSpace : TypeAttr {
- let Spellings = [Keyword<"__local">, Keyword<"local">];
+ let Spellings = [Keyword<"__local">, Keyword<"local">,
Clang<"opencl_local">];
let Documentation = [OpenCLAddressSpaceLocalDocs];
}
def OpenCLConstantAddressSpace : TypeAttr {
- let Spellings = [Keyword<"__constant">, Keyword<"constant">];
+ let Spellings = [Keyword<"__constant">, Keyword<"constant">,
Clang<"opencl_constant">];
let Documentation = [OpenCLAddressSpaceConstantDocs];
}
def OpenCLGenericAddressSpace : TypeAttr {
- let Spellings = [Keyword<"__generic">, Keyword<"generic">];
+ let Spellings = [Keyword<"__generic">, Keyword<"generic">,
Clang<"opencl_generic">];
let Documentation = [OpenCLAddressSpaceGenericDocs];
}
diff --git a/clang/lib/Sema/SemaType.cpp b/clang/lib/Sema/SemaType.cpp
index 52a0581643bc..1375ccbabc50 100644
--- a/clang/lib/Sema/SemaType.cpp
+++ b/clang/lib/Sema/SemaType.cpp
@@ -7407,6 +7407,16 @@ static void HandleLifetimeBoundAttr(TypeProcessingState
&State,
}
}
+static bool isAddressSpaceKind(const ParsedAttr &attr) {
+ auto attrKind = attr.getKind();
+
+ return attrKind == ParsedAttr::AT_AddressSpace ||
+ attrKind == ParsedAttr::AT_OpenCLPrivateAddressSpace ||
+ attrKind == ParsedAttr::AT_OpenCLGlobalAddressSpace ||
+ attrKind == ParsedAttr::AT_OpenCLLocalAddressSpace ||
+ attrKind == ParsedAttr::AT_OpenCLConstantAddressSpace ||
+ attrKind == ParsedAttr::AT_OpenCLGenericAddressSpace;
+}
static void processTypeAttrs(TypeProcessingState &state, QualType &type,
TypeAttrLocation TAL,
@@ -7445,11 +7455,11 @@ static void processTypeAttrs(TypeProcessingState
&state, QualType &type,
if (!IsTypeAttr)
continue;
}
- } else if (TAL != TAL_DeclChunk &&
- attr.getKind() != ParsedAttr::AT_AddressSpace) {
+ } else if (TAL != TAL_DeclChunk && !isAddressSpaceKind(attr)) {
// Otherwise, only consider type processing for a C++11 attribute if
// it's actually been applied to a type.
-// We also allow C++11 address_space attributes to pass through.
+// We also allow C++11 address_space and
+// OpenCL language address space attributes to pass through.
continue;
}
}
diff --git a/clang/test/AST/language_address_space_attribute.cpp
b/clang/test/AST/language_address_space_attribute.cpp
new file mode 100644
index ..7c6bdca06c06
--- /dev/null
+++ b/clang/test/AST/language_address_space_attribute.cpp
@@ -0,0 +1,36 @@
+// RUN: %clang_cc1 %s -ast-dump | FileCheck %s
+
+// Verify that the language address space attribute is
+// understood correctly by clang.
+
+void langas() {
+ // CHECK: VarDecl {{.*}} x_global '__global int *'
+ __attribute__((opencl_global)) int *x_global;
+
+ // CHECK: VarDecl {{.*}} z_global '__global int *'
+ [[clang::opencl_global]] int *z_global;
+
+ // CHECK: VarDecl {{.*}} x_local '__local int *'
+ __attribute__((opencl_local)) int *x_local;
+
+ // CHECK: VarDecl {{.*}} z_local '__local int *'
+ [[clang::opencl_local]] int *z_local;
+
+ // CHECK: VarDecl {{.*}} x_constant '_
[clang] cb30ad7 - [SYCL] Add support for auxiliary triple specification to Frontend
Author: Alexey Bader
Date: 2019-12-11T12:40:43+03:00
New Revision: cb30ad728f0b791c72a6a1399f36ebc60ad5
URL:
https://github.com/llvm/llvm-project/commit/cb30ad728f0b791c72a6a1399f36ebc60ad5
DIFF:
https://github.com/llvm/llvm-project/commit/cb30ad728f0b791c72a6a1399f36ebc60ad5.diff
LOG: [SYCL] Add support for auxiliary triple specification to Frontend
Summary:
Add host predefined macros to compilation for SYCL device, which is
required for pre-processing host specific includes (e.g. system
headers).
Reviewers: ABataev, jdoerfert
Subscribers: ebevhan, Anastasia, cfe-commits, keryell, Naghasan, Fznamznon
Tags: #clang
Differential Revision: https://reviews.llvm.org/D71286
Signed-off-by: Alexey Bader
Added:
clang/test/Frontend/sycl-aux-triple.cpp
Modified:
clang/lib/Frontend/CompilerInstance.cpp
clang/lib/Frontend/InitPreprocessor.cpp
Removed:
diff --git a/clang/lib/Frontend/CompilerInstance.cpp
b/clang/lib/Frontend/CompilerInstance.cpp
index b69dc4f854ff..05ecc3f447cc 100644
--- a/clang/lib/Frontend/CompilerInstance.cpp
+++ b/clang/lib/Frontend/CompilerInstance.cpp
@@ -917,8 +917,9 @@ bool CompilerInstance::ExecuteAction(FrontendAction &Act) {
if (!hasTarget())
return false;
- // Create TargetInfo for the other side of CUDA and OpenMP compilation.
- if ((getLangOpts().CUDA || getLangOpts().OpenMPIsDevice) &&
+ // Create TargetInfo for the other side of CUDA/OpenMP/SYCL compilation.
+ if ((getLangOpts().CUDA || getLangOpts().OpenMPIsDevice ||
+ getLangOpts().SYCLIsDevice) &&
!getFrontendOpts().AuxTriple.empty()) {
auto TO = std::make_shared();
TO->Triple = llvm::Triple::normalize(getFrontendOpts().AuxTriple);
diff --git a/clang/lib/Frontend/InitPreprocessor.cpp
b/clang/lib/Frontend/InitPreprocessor.cpp
index c27c33c530f6..b1c7aed0d44e 100644
--- a/clang/lib/Frontend/InitPreprocessor.cpp
+++ b/clang/lib/Frontend/InitPreprocessor.cpp
@@ -1124,7 +1124,8 @@ void clang::InitializePreprocessor(
if (InitOpts.UsePredefines) {
// FIXME: This will create multiple definitions for most of the predefined
// macros. This is not the right way to handle this.
-if ((LangOpts.CUDA || LangOpts.OpenMPIsDevice) && PP.getAuxTargetInfo())
+if ((LangOpts.CUDA || LangOpts.OpenMPIsDevice || LangOpts.SYCLIsDevice) &&
+PP.getAuxTargetInfo())
InitializePredefinedMacros(*PP.getAuxTargetInfo(), LangOpts, FEOpts,
PP.getPreprocessorOpts(), Builder);
diff --git a/clang/test/Frontend/sycl-aux-triple.cpp
b/clang/test/Frontend/sycl-aux-triple.cpp
new file mode 100644
index ..38b6a24fb3ce
--- /dev/null
+++ b/clang/test/Frontend/sycl-aux-triple.cpp
@@ -0,0 +1,5 @@
+// RUN: %clang_cc1 %s -triple spir -aux-triple x86_64-unknown-linux-gnu -E -dM
| FileCheck %s
+// RUN: %clang_cc1 %s -fsycl-is-device -triple spir -aux-triple
x86_64-unknown-linux-gnu -E -dM | FileCheck --check-prefix=CHECK-SYCL %s
+
+// CHECK-NOT:#define __x86_64__ 1
+// CHECK-SYCL:#define __x86_64__ 1
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] 2d6a5e4 - [OpenMP][Test] Add check for aux-triple predefined macros
Author: Alexey Bader Date: 2019-12-12T14:36:11+03:00 New Revision: 2d6a5e4fe45d0a1f1c94df6b3422ffb0d676fb6d URL: https://github.com/llvm/llvm-project/commit/2d6a5e4fe45d0a1f1c94df6b3422ffb0d676fb6d DIFF: https://github.com/llvm/llvm-project/commit/2d6a5e4fe45d0a1f1c94df6b3422ffb0d676fb6d.diff LOG: [OpenMP][Test] Add check for aux-triple predefined macros Summary: Make sure that auxiliary target specific macros are defined in OpenMP mode. Reviewers: ABataev, jdoerfert Subscribers: guansong, ebevhan, cfe-commits Tags: #clang Differential Revision: https://reviews.llvm.org/D71413 Added: clang/test/OpenMP/aux-triple-macros.cpp Modified: Removed: diff --git a/clang/test/OpenMP/aux-triple-macros.cpp b/clang/test/OpenMP/aux-triple-macros.cpp new file mode 100644 index ..73b74bd92f6d --- /dev/null +++ b/clang/test/OpenMP/aux-triple-macros.cpp @@ -0,0 +1,5 @@ +// RUN: %clang_cc1 %s -triple nvptx64 -aux-triple x86_64-unknown-linux-gnu -E -dM | FileCheck %s +// RUN: %clang_cc1 %s -fopenmp -fopenmp-is-device -triple nvptx64 -aux-triple x86_64-unknown-linux-gnu -E -dM | FileCheck --check-prefix=CHECK-OMP-DEVICE %s + +// CHECK-NOT:#define __x86_64__ 1 +// CHECK-OMP-DEVICE:#define __x86_64__ 1 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] 118b057 - [SYCL] Driver option to select SYCL version
Author: Ruyman
Date: 2020-03-07T18:28:54+03:00
New Revision: 118b057f1268d1789e40ffceb214e73772df04f4
URL:
https://github.com/llvm/llvm-project/commit/118b057f1268d1789e40ffceb214e73772df04f4
DIFF:
https://github.com/llvm/llvm-project/commit/118b057f1268d1789e40ffceb214e73772df04f4.diff
LOG: [SYCL] Driver option to select SYCL version
Summary:
User can select the version of SYCL the compiler will
use via the flag -sycl-std, similar to -cl-std.
The flag defines the LangOpts.SYCLVersion option to the
version of SYCL. The default value is undefined.
If driver is building SYCL code, flag is set to the default SYCL
version (1.2.1)
The preprocessor uses this variable to define CL_SYCL_LANGUAGE_VERSION macro,
which should be defined according to SYCL 1.2.1 standard.
Only valid value at this point for the flag is 1.2.1.
Co-Authored-By: David Wood
Signed-off-by: Ruyman Reyes
Subscribers: ebevhan, Anastasia, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D72857
Added:
Modified:
clang/include/clang/Basic/LangOptions.def
clang/include/clang/Driver/Options.td
clang/lib/Driver/ToolChains/Clang.cpp
clang/lib/Frontend/CompilerInvocation.cpp
clang/lib/Frontend/InitPreprocessor.cpp
clang/test/Driver/sycl.c
clang/test/Frontend/sycl-aux-triple.cpp
clang/test/Preprocessor/sycl-macro.cpp
clang/test/SemaSYCL/kernel-attribute.cpp
Removed:
diff --git a/clang/include/clang/Basic/LangOptions.def
b/clang/include/clang/Basic/LangOptions.def
index 3cc7c384ac10..53b87b737568 100644
--- a/clang/include/clang/Basic/LangOptions.def
+++ b/clang/include/clang/Basic/LangOptions.def
@@ -230,7 +230,9 @@ LANGOPT(GPURelocatableDeviceCode, 1, 0, "generate
relocatable device code")
LANGOPT(GPUAllowDeviceInit, 1, 0, "allowing device side global init functions
for HIP")
LANGOPT(GPUMaxThreadsPerBlock, 32, 256, "default max threads per block for
kernel launch bounds for HIP")
+LANGOPT(SYCL , 1, 0, "SYCL")
LANGOPT(SYCLIsDevice , 1, 0, "Generate code for SYCL device")
+LANGOPT(SYCLVersion , 32, 0, "Version of the SYCL standard used")
LANGOPT(HIPUseNewLaunchAPI, 1, 0, "Use new kernel launching API for HIP")
diff --git a/clang/include/clang/Driver/Options.td
b/clang/include/clang/Driver/Options.td
index 2aaf85434214..0d5cba8d682a 100644
--- a/clang/include/clang/Driver/Options.td
+++ b/clang/include/clang/Driver/Options.td
@@ -3422,10 +3422,12 @@ defm underscoring : BooleanFFlag<"underscoring">,
Group;
defm whole_file : BooleanFFlag<"whole-file">, Group;
// C++ SYCL options
-def fsycl : Flag<["-"], "fsycl">, Group,
+def fsycl : Flag<["-"], "fsycl">, Group, Flags<[CC1Option,
CoreOption]>,
HelpText<"Enable SYCL kernels compilation for device">;
-def fno_sycl : Flag<["-"], "fno-sycl">, Group,
+def fno_sycl : Flag<["-"], "fno-sycl">, Group, Flags<[CoreOption]>,
HelpText<"Disable SYCL kernels compilation for device">;
+def sycl_std_EQ : Joined<["-"], "sycl-std=">, Group,
Flags<[CC1Option, NoArgumentUnused, CoreOption]>,
+ HelpText<"SYCL language standard to compile for.">, Values<"2017, 121,
1.2.1, sycl-1.2.1">;
include "CC1Options.td"
diff --git a/clang/lib/Driver/ToolChains/Clang.cpp
b/clang/lib/Driver/ToolChains/Clang.cpp
index 3ca034e69b3b..99faae396be3 100644
--- a/clang/lib/Driver/ToolChains/Clang.cpp
+++ b/clang/lib/Driver/ToolChains/Clang.cpp
@@ -4048,9 +4048,18 @@ void Clang::ConstructJob(Compilation &C, const JobAction
&JA,
CmdArgs.push_back(Args.MakeArgString(NormalizedTriple));
}
- if (Args.hasFlag(options::OPT_fsycl, options::OPT_fno_sycl, false))
+ if (Args.hasFlag(options::OPT_fsycl, options::OPT_fno_sycl, false)) {
+CmdArgs.push_back("-fsycl");
CmdArgs.push_back("-fsycl-is-device");
+if (Arg *A = Args.getLastArg(options::OPT_sycl_std_EQ)) {
+ A->render(Args, CmdArgs);
+} else {
+ // Ensure the default version in SYCL mode is 1.2.1 (aka 2017)
+ CmdArgs.push_back("-sycl-std=2017");
+}
+ }
+
if (IsOpenMPDevice) {
// We have to pass the triple of the host if compiling for an OpenMP
device.
std::string NormalizedTriple =
diff --git a/clang/lib/Frontend/CompilerInvocation.cpp
b/clang/lib/Frontend/CompilerInvocation.cpp
index 48c65aded817..9f3522a3e654 100644
--- a/clang/lib/Frontend/CompilerInvocation.cpp
+++ b/clang/lib/Frontend/CompilerInvocation.cpp
@@ -2544,6 +2544,24 @@ static void ParseLangArgs(LangOptions &Opts, ArgList
&Args, InputKind IK,
LangStd = OpenCLLangStd;
}
+ Opts.SYCL = Args.hasArg(options::OPT_fsycl);
+ Opts.SYCLIsDevice = Opts.SYCL && Args.hasArg(options::OPT_fsycl_is_device);
+ if (Opts.SYCL) {
+// -sycl-std applies to any SYCL source, not only those containing kernels,
+// but also those using the SYCL API
+if (const Arg *A = Args.getLastArg(OPT_sycl_std_EQ)) {
+ Opts.SYCLVersion = llvm::Strin
[clang] 863d975 - [SYCL][Driver] Add clang driver option to enable SYCL compilation mode
Author: Alexey Bader
Date: 2020-02-06T08:42:31+03:00
New Revision: 863d9752105f390b31b3d08d1980d2888c15b034
URL:
https://github.com/llvm/llvm-project/commit/863d9752105f390b31b3d08d1980d2888c15b034
DIFF:
https://github.com/llvm/llvm-project/commit/863d9752105f390b31b3d08d1980d2888c15b034.diff
LOG: [SYCL][Driver] Add clang driver option to enable SYCL compilation mode
Summary:
As a first step this implementation enables compilation of the offload
code.
Reviewers: ABataev
Subscribers: ebevhan, Anastasia, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D74048
Added:
clang/test/Driver/sycl.c
Modified:
clang/include/clang/Driver/Options.td
clang/lib/Driver/ToolChains/Clang.cpp
Removed:
diff --git a/clang/include/clang/Driver/Options.td
b/clang/include/clang/Driver/Options.td
index 2c925d018da7..2dea0ac2ba5c 100644
--- a/clang/include/clang/Driver/Options.td
+++ b/clang/include/clang/Driver/Options.td
@@ -124,6 +124,9 @@ def pedantic_Group : OptionGroup<"">,
Group,
def opencl_Group : OptionGroup<"">, Group,
DocName<"OpenCL flags">;
+def sycl_Group : OptionGroup<"">, Group,
+ DocName<"SYCL flags">;
+
def m_Group : OptionGroup<"">, Group,
DocName<"Target-dependent compilation options">;
@@ -3407,6 +3410,11 @@ defm stack_arrays : BooleanFFlag<"stack-arrays">,
Group;
defm underscoring : BooleanFFlag<"underscoring">, Group;
defm whole_file : BooleanFFlag<"whole-file">, Group;
+// C++ SYCL options
+def fsycl : Flag<["-"], "fsycl">, Group,
+ HelpText<"Enable SYCL kernels compilation for device">;
+def fno_sycl : Flag<["-"], "fno-sycl">, Group,
+ HelpText<"Disable SYCL kernels compilation for device">;
include "CC1Options.td"
diff --git a/clang/lib/Driver/ToolChains/Clang.cpp
b/clang/lib/Driver/ToolChains/Clang.cpp
index ccdfbe8c604f..0bed933185f6 100644
--- a/clang/lib/Driver/ToolChains/Clang.cpp
+++ b/clang/lib/Driver/ToolChains/Clang.cpp
@@ -4023,6 +4023,9 @@ void Clang::ConstructJob(Compilation &C, const JobAction
&JA,
CmdArgs.push_back(Args.MakeArgString(NormalizedTriple));
}
+ if (Args.hasFlag(options::OPT_fsycl, options::OPT_fno_sycl, false))
+CmdArgs.push_back("-fsycl-is-device");
+
if (IsOpenMPDevice) {
// We have to pass the triple of the host if compiling for an OpenMP
device.
std::string NormalizedTriple =
diff --git a/clang/test/Driver/sycl.c b/clang/test/Driver/sycl.c
new file mode 100644
index ..6c4b291f387b
--- /dev/null
+++ b/clang/test/Driver/sycl.c
@@ -0,0 +1,10 @@
+// RUN: %clang -### -fsycl -c %s 2>&1 | FileCheck %s --check-prefix=ENABLED
+// RUN: %clang -### -fsycl %s 2>&1 | FileCheck %s --check-prefix=ENABLED
+// RUN: %clang -### -fno-sycl -fsycl %s 2>&1 | FileCheck %s
--check-prefix=ENABLED
+// RUN: %clangxx -### -fsycl %s 2>&1 | FileCheck %s --check-prefix=ENABLED
+// RUN: %clangxx -### -fno-sycl %s 2>&1 | FileCheck %s --check-prefix=DISABLED
+// RUN: %clangxx -### -fsycl -fno-sycl %s 2>&1 | FileCheck %s
--check-prefix=DISABLED
+// RUN: %clangxx -### %s 2>&1 | FileCheck %s --check-prefix=DISABLED
+
+// ENABLED: "-cc1"{{.*}} "-fsycl-is-device"
+// DISABLED-NOT: "-fsycl-is-device"
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] dd18729 - [Attr][Doc][NFC] Fix code snippet formatting for attribute documentation
Author: Alexey Bader
Date: 2020-01-21T18:30:56+03:00
New Revision: dd18729b2a7a23b76b8d74fbf4f4bb4efbe8aa97
URL:
https://github.com/llvm/llvm-project/commit/dd18729b2a7a23b76b8d74fbf4f4bb4efbe8aa97
DIFF:
https://github.com/llvm/llvm-project/commit/dd18729b2a7a23b76b8d74fbf4f4bb4efbe8aa97.diff
LOG: [Attr][Doc][NFC] Fix code snippet formatting for attribute documentation
Reviewers: aaron.ballman, Fznamznon
Subscribers: ebevhan, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D73104
Added:
Modified:
clang/include/clang/Basic/AttrDocs.td
Removed:
diff --git a/clang/include/clang/Basic/AttrDocs.td
b/clang/include/clang/Basic/AttrDocs.td
index 456edd1daafc..8fd59a0c9b38 100644
--- a/clang/include/clang/Basic/AttrDocs.td
+++ b/clang/include/clang/Basic/AttrDocs.td
@@ -260,6 +260,7 @@ The ``sycl_kernel`` attribute specifies that a function
template will be used
to outline device code and to generate an OpenCL kernel.
Here is a code example of the SYCL program, which demonstrates the compiler's
outlining job:
+
.. code-block:: c++
int foo(int x) { return ++x; }
@@ -282,27 +283,29 @@ compilation of functions for the device part can be found
in the SYCL 1.2.1
specification Section 6.4.
To show to the compiler entry point to the "device part" of the code, the SYCL
runtime can use the ``sycl_kernel`` attribute in the following way:
+
.. code-block:: c++
-namespace cl {
-namespace sycl {
-class handler {
- template
- __attribute__((sycl_kernel)) void sycl_kernel_function(KernelType
KernelFuncObj) {
-// ...
-KernelFuncObj();
- }
- template
- void parallel_for(range NumWorkItems, KernelType KernelFunc) {
-#ifdef __SYCL_DEVICE_ONLY__
-sycl_kernel_function(KernelFunc);
-#else
-// Host implementation
-#endif
- }
-};
-} // namespace sycl
-} // namespace cl
+ namespace cl {
+ namespace sycl {
+ class handler {
+template
+__attribute__((sycl_kernel)) void sycl_kernel_function(KernelType
KernelFuncObj) {
+ // ...
+ KernelFuncObj();
+}
+
+template
+void parallel_for(range NumWorkItems, KernelType KernelFunc) {
+ #ifdef __SYCL_DEVICE_ONLY__
+ sycl_kernel_function(KernelFunc);
+ #else
+ // Host implementation
+ #endif
+}
+ };
+ } // namespace sycl
+ } // namespace cl
The compiler will also generate an OpenCL kernel using the function marked with
the ``sycl_kernel`` attribute.
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
r318290 - [OpenCL] Fix code generation of function-scope constant samplers.
Author: bader
Date: Wed Nov 15 03:38:17 2017
New Revision: 318290
URL: http://llvm.org/viewvc/llvm-project?rev=318290&view=rev
Log:
[OpenCL] Fix code generation of function-scope constant samplers.
Summary:
Constant samplers are handled as static variables and clang's code generation
library, which leads to llvm::unreachable. We bypass emitting sampler variable
as static since it's translated to a function call later.
Reviewers: yaxunl, Anastasia
Reviewed By: yaxunl, Anastasia
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D34342
Modified:
cfe/trunk/lib/CodeGen/CGDecl.cpp
cfe/trunk/test/CodeGenOpenCL/sampler.cl
cfe/trunk/test/SemaOpenCL/sampler_t.cl
Modified: cfe/trunk/lib/CodeGen/CGDecl.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/CodeGen/CGDecl.cpp?rev=318290&r1=318289&r2=318290&view=diff
==
--- cfe/trunk/lib/CodeGen/CGDecl.cpp (original)
+++ cfe/trunk/lib/CodeGen/CGDecl.cpp Wed Nov 15 03:38:17 2017
@@ -162,6 +162,10 @@ void CodeGenFunction::EmitVarDecl(const
// needs to be emitted like a static variable, e.g. a function-scope
// variable in constant address space in OpenCL.
if (D.getStorageDuration() != SD_Automatic) {
+// Static sampler variables translated to function calls.
+if (D.getType()->isSamplerT())
+ return;
+
llvm::GlobalValue::LinkageTypes Linkage =
CGM.getLLVMLinkageVarDefinition(&D, /*isConstant=*/false);
Modified: cfe/trunk/test/CodeGenOpenCL/sampler.cl
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/CodeGenOpenCL/sampler.cl?rev=318290&r1=318289&r2=318290&view=diff
==
--- cfe/trunk/test/CodeGenOpenCL/sampler.cl (original)
+++ cfe/trunk/test/CodeGenOpenCL/sampler.cl Wed Nov 15 03:38:17 2017
@@ -20,6 +20,8 @@
constant sampler_t glb_smp = CLK_ADDRESS_CLAMP_TO_EDGE |
CLK_NORMALIZED_COORDS_TRUE | CLK_FILTER_LINEAR;
// CHECK-NOT: glb_smp
+int get_sampler_initializer(void);
+
void fnc4smp(sampler_t s) {}
// CHECK: define spir_func void @fnc4smp(%opencl.sampler_t addrspace(2)* %
@@ -58,4 +60,20 @@ kernel void foo(sampler_t smp_par) {
fnc4smp(5);
// CHECK: [[SAMP:%[0-9]+]] = call %opencl.sampler_t addrspace(2)*
@__translate_sampler_initializer(i32 5)
// CHECK: call spir_func void @fnc4smp(%opencl.sampler_t addrspace(2)*
[[SAMP]])
+
+ const sampler_t const_smp = CLK_ADDRESS_CLAMP_TO_EDGE |
CLK_NORMALIZED_COORDS_TRUE | CLK_FILTER_LINEAR;
+ fnc4smp(const_smp);
+ // CHECK: [[CONST_SAMP:%[0-9]+]] = call %opencl.sampler_t addrspace(2)*
@__translate_sampler_initializer(i32 35)
+ // CHECK: store %opencl.sampler_t addrspace(2)* [[CONST_SAMP]],
%opencl.sampler_t addrspace(2)** [[CONST_SMP_PTR:%[a-zA-Z0-9]+]]
+ fnc4smp(const_smp);
+ // CHECK: [[SAMP:%[0-9]+]] = load %opencl.sampler_t addrspace(2)*,
%opencl.sampler_t addrspace(2)** [[CONST_SMP_PTR]]
+ // CHECK: call spir_func void @fnc4smp(%opencl.sampler_t addrspace(2)*
[[SAMP]])
+
+ constant sampler_t constant_smp = CLK_ADDRESS_CLAMP_TO_EDGE |
CLK_NORMALIZED_COORDS_TRUE | CLK_FILTER_LINEAR;
+ fnc4smp(constant_smp);
+ // CHECK: [[SAMP:%[0-9]+]] = call %opencl.sampler_t addrspace(2)*
@__translate_sampler_initializer(i32 35)
+ // CHECK: call spir_func void @fnc4smp(%opencl.sampler_t addrspace(2)*
[[SAMP]])
+
+ // TODO: enable sampler initialization with non-constant integer.
+ //const sampler_t const_smp_func_init = get_sampler_initializer();
}
Modified: cfe/trunk/test/SemaOpenCL/sampler_t.cl
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/SemaOpenCL/sampler_t.cl?rev=318290&r1=318289&r2=318290&view=diff
==
--- cfe/trunk/test/SemaOpenCL/sampler_t.cl (original)
+++ cfe/trunk/test/SemaOpenCL/sampler_t.cl Wed Nov 15 03:38:17 2017
@@ -46,36 +46,11 @@ const constant sampler_t glb_smp11 = CLK
void kernel ker(sampler_t argsmp) {
local sampler_t smp; // expected-error{{sampler type cannot be used with the
__local and __global address space qualifiers}}
- const sampler_t const_smp = CLK_ADDRESS_CLAMP_TO_EDGE |
CLK_NORMALIZED_COORDS_TRUE | CLK_FILTER_LINEAR;
- const sampler_t const_smp2;
- const sampler_t const_smp3 = const_smp;
- const sampler_t const_smp4 = f();
const sampler_t const_smp5 = 1.0f; // expected-error{{initializing 'const
sampler_t' with an expression of incompatible type 'float'}}
const sampler_t const_smp6 = 0x1LL; // expected-error{{sampler_t
initialization requires 32-bit integer, not 'long long'}}
- foo(glb_smp);
- foo(glb_smp2);
- foo(glb_smp3);
- foo(glb_smp4);
- foo(glb_smp5);
- foo(glb_smp6);
- foo(glb_smp7);
- foo(glb_smp8);
- foo(glb_smp9);
- foo(smp);
- foo(sampler_str.smp);
- foo(const_smp);
- foo(const_smp2);
- foo(const_smp3);
- foo(const_smp4);
- foo(const_smp5);
- foo(const_smp6);
- foo(argsmp);
- foo(5);
Re: [PATCH] D23914: [OpenCL] Make is_valid_event overloadable.
bader accepted this revision. bader added a comment. This revision is now accepted and ready to land. LGTM https://reviews.llvm.org/D23914 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: [PATCH] D23915: [OpenCL] Remove access qualifiers on images in arg info metadata.
bader accepted this revision. bader added a comment. This revision is now accepted and ready to land. LGTM https://reviews.llvm.org/D23915 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[PATCH] D23992: [OpenCL] Augment pipe built-ins with pipe packet size and alignment.
bader created this revision.
bader added reviewers: Anastasia, vpykhtin.
bader added a subscriber: cfe-commits.
Keep pipe packet size and alignment in LLVM IR layer to enable
translation from LLVM IR to SPIR-V format.
https://reviews.llvm.org/D23992
Files:
lib/CodeGen/CGBuiltin.cpp
lib/CodeGen/CGOpenCLRuntime.cpp
lib/CodeGen/CGOpenCLRuntime.h
test/CodeGenOpenCL/pipe_builtin.cl
test/CodeGenOpenCL/pipe_types.cl
Index: test/CodeGenOpenCL/pipe_types.cl
===
--- test/CodeGenOpenCL/pipe_types.cl
+++ test/CodeGenOpenCL/pipe_types.cl
@@ -41,7 +41,7 @@
read_only pipe struct Person SPipe) {
// CHECK: define void @test_reserved_read_pipe
read_pipe (SPipe, SDst);
- // CHECK: call i32 @__read_pipe_2(%opencl.pipe_t* %{{.*}}, i8* %{{.*}})
+ // CHECK: call i32 @__read_pipe_2(%opencl.pipe_t* %{{.*}}, i8* %{{.*}}, i32 16, i32 8)
read_pipe (SPipe, SDst);
- // CHECK: call i32 @__read_pipe_2(%opencl.pipe_t* %{{.*}}, i8* %{{.*}})
+ // CHECK: call i32 @__read_pipe_2(%opencl.pipe_t* %{{.*}}, i8* %{{.*}}, i32 16, i32 8)
}
Index: test/CodeGenOpenCL/pipe_builtin.cl
===
--- test/CodeGenOpenCL/pipe_builtin.cl
+++ test/CodeGenOpenCL/pipe_builtin.cl
@@ -4,58 +4,58 @@
// CHECK: %opencl.reserve_id_t = type opaque
void test1(read_only pipe int p, global int *ptr) {
- // CHECK: call i32 @__read_pipe_2(%opencl.pipe_t* %{{.*}}, i8* %{{.*}})
+ // CHECK: call i32 @__read_pipe_2(%opencl.pipe_t* %{{.*}}, i8* %{{.*}}, i32 4, i32 4)
read_pipe(p, ptr);
- // CHECK: call %opencl.reserve_id_t* @__reserve_read_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}})
+ // CHECK: call %opencl.reserve_id_t* @__reserve_read_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}}, i32 4, i32 4)
reserve_id_t rid = reserve_read_pipe(p, 2);
- // CHECK: call i32 @__read_pipe_4(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 {{.*}}, i8* %{{.*}})
+ // CHECK: call i32 @__read_pipe_4(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 {{.*}}, i8* %{{.*}}, i32 4, i32 4)
read_pipe(p, rid, 2, ptr);
- // CHECK: call void @__commit_read_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}})
+ // CHECK: call void @__commit_read_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 4, i32 4)
commit_read_pipe(p, rid);
}
void test2(write_only pipe int p, global int *ptr) {
- // CHECK: call i32 @__write_pipe_2(%opencl.pipe_t* %{{.*}}, i8* %{{.*}})
+ // CHECK: call i32 @__write_pipe_2(%opencl.pipe_t* %{{.*}}, i8* %{{.*}}, i32 4, i32 4)
write_pipe(p, ptr);
- // CHECK: call %opencl.reserve_id_t* @__reserve_write_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}})
+ // CHECK: call %opencl.reserve_id_t* @__reserve_write_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}}, i32 4, i32 4)
reserve_id_t rid = reserve_write_pipe(p, 2);
- // CHECK: call i32 @__write_pipe_4(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 {{.*}}, i8* %{{.*}})
+ // CHECK: call i32 @__write_pipe_4(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 {{.*}}, i8* %{{.*}}, i32 4, i32 4)
write_pipe(p, rid, 2, ptr);
- // CHECK: call void @__commit_write_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}})
+ // CHECK: call void @__commit_write_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 4, i32 4)
commit_write_pipe(p, rid);
}
void test3(read_only pipe int p, global int *ptr) {
- // CHECK: call %opencl.reserve_id_t* @__work_group_reserve_read_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}})
+ // CHECK: call %opencl.reserve_id_t* @__work_group_reserve_read_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}}, i32 4, i32 4)
reserve_id_t rid = work_group_reserve_read_pipe(p, 2);
- // CHECK: call void @__work_group_commit_read_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}})
+ // CHECK: call void @__work_group_commit_read_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 4, i32 4)
work_group_commit_read_pipe(p, rid);
}
void test4(write_only pipe int p, global int *ptr) {
- // CHECK: call %opencl.reserve_id_t* @__work_group_reserve_write_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}})
+ // CHECK: call %opencl.reserve_id_t* @__work_group_reserve_write_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}}, i32 4, i32 4)
reserve_id_t rid = work_group_reserve_write_pipe(p, 2);
- // CHECK: call void @__work_group_commit_write_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}})
+ // CHECK: call void @__work_group_commit_write_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 4, i32 4)
work_group_commit_write_pipe(p, rid);
}
void test5(read_only pipe int p, global int *ptr) {
- // CHECK: call %opencl.reserve_id_t* @__sub_group_reserve_read_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}})
+ // CHECK: call %opencl.reserve_id_t* @__sub_group_reserve_read_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}}, i32 4, i32 4)
reserve_id_t rid =
Re: [PATCH] D23914: [OpenCL] Make is_valid_event, create_user_event overloadable.
This revision was automatically updated to reflect the committed changes. Closed by commit rL280097: [OpenCL] Make is_valid_event, create_user_event overloadable. (authored by bader). Changed prior to commit: https://reviews.llvm.org/D23914?vs=69650&id=69682#toc Repository: rL LLVM https://reviews.llvm.org/D23914 Files: cfe/trunk/lib/Headers/opencl-c.h Index: cfe/trunk/lib/Headers/opencl-c.h === --- cfe/trunk/lib/Headers/opencl-c.h +++ cfe/trunk/lib/Headers/opencl-c.h @@ -16729,11 +16729,11 @@ void __ovld release_event(clk_event_t); -clk_event_t create_user_event(void); +clk_event_t __ovld create_user_event(void); void __ovld set_user_event_status(clk_event_t e, int state); -bool is_valid_event (clk_event_t event); +bool __ovld is_valid_event (clk_event_t event); void __ovld capture_event_profiling_info(clk_event_t, clk_profiling_info, __global void* value); Index: cfe/trunk/lib/Headers/opencl-c.h === --- cfe/trunk/lib/Headers/opencl-c.h +++ cfe/trunk/lib/Headers/opencl-c.h @@ -16729,11 +16729,11 @@ void __ovld release_event(clk_event_t); -clk_event_t create_user_event(void); +clk_event_t __ovld create_user_event(void); void __ovld set_user_event_status(clk_event_t e, int state); -bool is_valid_event (clk_event_t event); +bool __ovld is_valid_event (clk_event_t event); void __ovld capture_event_profiling_info(clk_event_t, clk_profiling_info, __global void* value); ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
r280097 - [OpenCL] Make is_valid_event, create_user_event overloadable.
Author: bader Date: Tue Aug 30 09:42:54 2016 New Revision: 280097 URL: http://llvm.org/viewvc/llvm-project?rev=280097&view=rev Log: [OpenCL] Make is_valid_event, create_user_event overloadable. Summary: Make is_valid_event and create_user_event overloadable like other built-ins. Patch by Evgeniy Tyurin. Reviewers: bader, yaxunl Subscribers: Anastasia, cfe-commits Differential Revision: https://reviews.llvm.org/D23914 Modified: cfe/trunk/lib/Headers/opencl-c.h Modified: cfe/trunk/lib/Headers/opencl-c.h URL: http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Headers/opencl-c.h?rev=280097&r1=280096&r2=280097&view=diff == --- cfe/trunk/lib/Headers/opencl-c.h (original) +++ cfe/trunk/lib/Headers/opencl-c.h Tue Aug 30 09:42:54 2016 @@ -16729,11 +16729,11 @@ void __ovld retain_event(clk_event_t); void __ovld release_event(clk_event_t); -clk_event_t create_user_event(void); +clk_event_t __ovld create_user_event(void); void __ovld set_user_event_status(clk_event_t e, int state); -bool is_valid_event (clk_event_t event); +bool __ovld is_valid_event (clk_event_t event); void __ovld capture_event_profiling_info(clk_event_t, clk_profiling_info, __global void* value); ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[PATCH] D24136: [OpenCL] Fix pipe built-in functions return type.
bader created this revision.
bader added a reviewer: Anastasia.
bader added subscribers: yaxunl, cfe-commits.
By default return type of call expressions calling built-in
functions is set to bool.
Fixes https://llvm.org/bugs/show_bug.cgi?id=30219.
https://reviews.llvm.org/D24136
Files:
lib/Sema/SemaChecking.cpp
test/CodeGenOpenCL/pipe_builtin.cl
Index: test/CodeGenOpenCL/pipe_builtin.cl
===
--- test/CodeGenOpenCL/pipe_builtin.cl
+++ test/CodeGenOpenCL/pipe_builtin.cl
@@ -59,3 +59,13 @@
// CHECK: call i32 @__get_pipe_max_packets(%opencl.pipe_t* %{{.*}})
*ptr = get_pipe_max_packets(p);
}
+
+void test8(read_only pipe int r, write_only pipe int w, global int *ptr) {
+ // verify that return type is correctly casted to i1 value
+ // CHECK: %[[R:[0-9]+]] = call i32 @__read_pipe_2
+ // CHECK: icmp ne i32 %[[R]], 0
+ if (read_pipe(r, ptr)) *ptr = -1;
+ // CHECK: %[[W:[0-9]+]] = call i32 @__write_pipe_2
+ // CHECK: icmp ne i32 %[[W]], 0
+ if (write_pipe(w, ptr)) *ptr = -1;
+}
Index: lib/Sema/SemaChecking.cpp
===
--- lib/Sema/SemaChecking.cpp
+++ lib/Sema/SemaChecking.cpp
@@ -1020,6 +1020,7 @@
// check for the argument.
if (SemaBuiltinRWPipe(*this, TheCall))
return ExprError();
+TheCall->setType(Context.IntTy);
break;
case Builtin::BIreserve_read_pipe:
case Builtin::BIreserve_write_pipe:
@@ -1047,6 +1048,7 @@
case Builtin::BIget_pipe_max_packets:
if (SemaBuiltinPipePackets(*this, TheCall))
return ExprError();
+TheCall->setType(Context.UnsignedIntTy);
break;
case Builtin::BIto_global:
case Builtin::BIto_local:
Index: test/CodeGenOpenCL/pipe_builtin.cl
===
--- test/CodeGenOpenCL/pipe_builtin.cl
+++ test/CodeGenOpenCL/pipe_builtin.cl
@@ -59,3 +59,13 @@
// CHECK: call i32 @__get_pipe_max_packets(%opencl.pipe_t* %{{.*}})
*ptr = get_pipe_max_packets(p);
}
+
+void test8(read_only pipe int r, write_only pipe int w, global int *ptr) {
+ // verify that return type is correctly casted to i1 value
+ // CHECK: %[[R:[0-9]+]] = call i32 @__read_pipe_2
+ // CHECK: icmp ne i32 %[[R]], 0
+ if (read_pipe(r, ptr)) *ptr = -1;
+ // CHECK: %[[W:[0-9]+]] = call i32 @__write_pipe_2
+ // CHECK: icmp ne i32 %[[W]], 0
+ if (write_pipe(w, ptr)) *ptr = -1;
+}
Index: lib/Sema/SemaChecking.cpp
===
--- lib/Sema/SemaChecking.cpp
+++ lib/Sema/SemaChecking.cpp
@@ -1020,6 +1020,7 @@
// check for the argument.
if (SemaBuiltinRWPipe(*this, TheCall))
return ExprError();
+TheCall->setType(Context.IntTy);
break;
case Builtin::BIreserve_read_pipe:
case Builtin::BIreserve_write_pipe:
@@ -1047,6 +1048,7 @@
case Builtin::BIget_pipe_max_packets:
if (SemaBuiltinPipePackets(*this, TheCall))
return ExprError();
+TheCall->setType(Context.UnsignedIntTy);
break;
case Builtin::BIto_global:
case Builtin::BIto_local:
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: [PATCH] D24054: Do not validate pch when -fno-validate-pch is set
bader added a comment. Could you a regression test, please? https://reviews.llvm.org/D24054 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: [PATCH] D23992: [OpenCL] Augment pipe built-ins with pipe packet size and alignment.
bader updated this revision to Diff 70366.
bader added a comment.
Applied code review comment from Anastasia.
Use getTypeSizeInChars instead of getTypeSize to get type size in bytes. Assume
that char size is always one byte.
https://reviews.llvm.org/D23992
Files:
lib/CodeGen/CGBuiltin.cpp
lib/CodeGen/CGOpenCLRuntime.cpp
lib/CodeGen/CGOpenCLRuntime.h
test/CodeGenOpenCL/pipe_builtin.cl
test/CodeGenOpenCL/pipe_types.cl
Index: test/CodeGenOpenCL/pipe_types.cl
===
--- test/CodeGenOpenCL/pipe_types.cl
+++ test/CodeGenOpenCL/pipe_types.cl
@@ -41,7 +41,7 @@
read_only pipe struct Person SPipe) {
// CHECK: define void @test_reserved_read_pipe
read_pipe (SPipe, SDst);
- // CHECK: call i32 @__read_pipe_2(%opencl.pipe_t* %{{.*}}, i8* %{{.*}})
+ // CHECK: call i32 @__read_pipe_2(%opencl.pipe_t* %{{.*}}, i8* %{{.*}}, i32 16, i32 8)
read_pipe (SPipe, SDst);
- // CHECK: call i32 @__read_pipe_2(%opencl.pipe_t* %{{.*}}, i8* %{{.*}})
+ // CHECK: call i32 @__read_pipe_2(%opencl.pipe_t* %{{.*}}, i8* %{{.*}}, i32 16, i32 8)
}
Index: test/CodeGenOpenCL/pipe_builtin.cl
===
--- test/CodeGenOpenCL/pipe_builtin.cl
+++ test/CodeGenOpenCL/pipe_builtin.cl
@@ -4,58 +4,58 @@
// CHECK: %opencl.reserve_id_t = type opaque
void test1(read_only pipe int p, global int *ptr) {
- // CHECK: call i32 @__read_pipe_2(%opencl.pipe_t* %{{.*}}, i8* %{{.*}})
+ // CHECK: call i32 @__read_pipe_2(%opencl.pipe_t* %{{.*}}, i8* %{{.*}}, i32 4, i32 4)
read_pipe(p, ptr);
- // CHECK: call %opencl.reserve_id_t* @__reserve_read_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}})
+ // CHECK: call %opencl.reserve_id_t* @__reserve_read_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}}, i32 4, i32 4)
reserve_id_t rid = reserve_read_pipe(p, 2);
- // CHECK: call i32 @__read_pipe_4(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 {{.*}}, i8* %{{.*}})
+ // CHECK: call i32 @__read_pipe_4(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 {{.*}}, i8* %{{.*}}, i32 4, i32 4)
read_pipe(p, rid, 2, ptr);
- // CHECK: call void @__commit_read_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}})
+ // CHECK: call void @__commit_read_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 4, i32 4)
commit_read_pipe(p, rid);
}
void test2(write_only pipe int p, global int *ptr) {
- // CHECK: call i32 @__write_pipe_2(%opencl.pipe_t* %{{.*}}, i8* %{{.*}})
+ // CHECK: call i32 @__write_pipe_2(%opencl.pipe_t* %{{.*}}, i8* %{{.*}}, i32 4, i32 4)
write_pipe(p, ptr);
- // CHECK: call %opencl.reserve_id_t* @__reserve_write_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}})
+ // CHECK: call %opencl.reserve_id_t* @__reserve_write_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}}, i32 4, i32 4)
reserve_id_t rid = reserve_write_pipe(p, 2);
- // CHECK: call i32 @__write_pipe_4(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 {{.*}}, i8* %{{.*}})
+ // CHECK: call i32 @__write_pipe_4(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 {{.*}}, i8* %{{.*}}, i32 4, i32 4)
write_pipe(p, rid, 2, ptr);
- // CHECK: call void @__commit_write_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}})
+ // CHECK: call void @__commit_write_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 4, i32 4)
commit_write_pipe(p, rid);
}
void test3(read_only pipe int p, global int *ptr) {
- // CHECK: call %opencl.reserve_id_t* @__work_group_reserve_read_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}})
+ // CHECK: call %opencl.reserve_id_t* @__work_group_reserve_read_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}}, i32 4, i32 4)
reserve_id_t rid = work_group_reserve_read_pipe(p, 2);
- // CHECK: call void @__work_group_commit_read_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}})
+ // CHECK: call void @__work_group_commit_read_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 4, i32 4)
work_group_commit_read_pipe(p, rid);
}
void test4(write_only pipe int p, global int *ptr) {
- // CHECK: call %opencl.reserve_id_t* @__work_group_reserve_write_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}})
+ // CHECK: call %opencl.reserve_id_t* @__work_group_reserve_write_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}}, i32 4, i32 4)
reserve_id_t rid = work_group_reserve_write_pipe(p, 2);
- // CHECK: call void @__work_group_commit_write_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}})
+ // CHECK: call void @__work_group_commit_write_pipe(%opencl.pipe_t* %{{.*}}, %opencl.reserve_id_t* %{{.*}}, i32 4, i32 4)
work_group_commit_write_pipe(p, rid);
}
void test5(read_only pipe int p, global int *ptr) {
- // CHECK: call %opencl.reserve_id_t* @__sub_group_reserve_read_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}})
+ // CHECK: call %opencl.reserve_id_t* @__sub_group_reserve_read_pipe(%opencl.pipe_t* %{{.*}}, i32 {{.*}}, i32 4, i32 4)
reserve_i
Re: [PATCH] D24136: [OpenCL] Fix pipe built-in functions return type.
bader updated this revision to Diff 70368.
bader added a comment.
Added get_pipe_num_packets and get_pipe_max_packets test cases.
https://reviews.llvm.org/D24136
Files:
lib/Sema/SemaChecking.cpp
test/CodeGenOpenCL/pipe_builtin.cl
Index: test/CodeGenOpenCL/pipe_builtin.cl
===
--- test/CodeGenOpenCL/pipe_builtin.cl
+++ test/CodeGenOpenCL/pipe_builtin.cl
@@ -59,3 +59,19 @@
// CHECK: call i32 @__get_pipe_max_packets(%opencl.pipe_t* %{{.*}})
*ptr = get_pipe_max_packets(p);
}
+
+void test8(read_only pipe int r, write_only pipe int w, global int *ptr) {
+ // verify that return type is correctly casted to i1 value
+ // CHECK: %[[R:[0-9]+]] = call i32 @__read_pipe_2
+ // CHECK: icmp ne i32 %[[R]], 0
+ if (read_pipe(r, ptr)) *ptr = -1;
+ // CHECK: %[[W:[0-9]+]] = call i32 @__write_pipe_2
+ // CHECK: icmp ne i32 %[[W]], 0
+ if (write_pipe(w, ptr)) *ptr = -1;
+ // CHECK: %[[N:[0-9]+]] = call i32 @__get_pipe_num_packets
+ // CHECK: icmp ne i32 %[[N]], 0
+ if (get_pipe_num_packets(r)) *ptr = -1;
+ // CHECK: %[[M:[0-9]+]] = call i32 @__get_pipe_max_packets
+ // CHECK: icmp ne i32 %[[M]], 0
+ if (get_pipe_max_packets(w)) *ptr = -1;
+}
Index: lib/Sema/SemaChecking.cpp
===
--- lib/Sema/SemaChecking.cpp
+++ lib/Sema/SemaChecking.cpp
@@ -1020,6 +1020,7 @@
// check for the argument.
if (SemaBuiltinRWPipe(*this, TheCall))
return ExprError();
+TheCall->setType(Context.IntTy);
break;
case Builtin::BIreserve_read_pipe:
case Builtin::BIreserve_write_pipe:
@@ -1047,6 +1048,7 @@
case Builtin::BIget_pipe_max_packets:
if (SemaBuiltinPipePackets(*this, TheCall))
return ExprError();
+TheCall->setType(Context.UnsignedIntTy);
break;
case Builtin::BIto_global:
case Builtin::BIto_local:
Index: test/CodeGenOpenCL/pipe_builtin.cl
===
--- test/CodeGenOpenCL/pipe_builtin.cl
+++ test/CodeGenOpenCL/pipe_builtin.cl
@@ -59,3 +59,19 @@
// CHECK: call i32 @__get_pipe_max_packets(%opencl.pipe_t* %{{.*}})
*ptr = get_pipe_max_packets(p);
}
+
+void test8(read_only pipe int r, write_only pipe int w, global int *ptr) {
+ // verify that return type is correctly casted to i1 value
+ // CHECK: %[[R:[0-9]+]] = call i32 @__read_pipe_2
+ // CHECK: icmp ne i32 %[[R]], 0
+ if (read_pipe(r, ptr)) *ptr = -1;
+ // CHECK: %[[W:[0-9]+]] = call i32 @__write_pipe_2
+ // CHECK: icmp ne i32 %[[W]], 0
+ if (write_pipe(w, ptr)) *ptr = -1;
+ // CHECK: %[[N:[0-9]+]] = call i32 @__get_pipe_num_packets
+ // CHECK: icmp ne i32 %[[N]], 0
+ if (get_pipe_num_packets(r)) *ptr = -1;
+ // CHECK: %[[M:[0-9]+]] = call i32 @__get_pipe_max_packets
+ // CHECK: icmp ne i32 %[[M]], 0
+ if (get_pipe_max_packets(w)) *ptr = -1;
+}
Index: lib/Sema/SemaChecking.cpp
===
--- lib/Sema/SemaChecking.cpp
+++ lib/Sema/SemaChecking.cpp
@@ -1020,6 +1020,7 @@
// check for the argument.
if (SemaBuiltinRWPipe(*this, TheCall))
return ExprError();
+TheCall->setType(Context.IntTy);
break;
case Builtin::BIreserve_read_pipe:
case Builtin::BIreserve_write_pipe:
@@ -1047,6 +1048,7 @@
case Builtin::BIget_pipe_max_packets:
if (SemaBuiltinPipePackets(*this, TheCall))
return ExprError();
+TheCall->setType(Context.UnsignedIntTy);
break;
case Builtin::BIto_global:
case Builtin::BIto_local:
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
r280699 - [OpenCL] Remove access qualifiers on images in arg info metadata.
Author: bader
Date: Tue Sep 6 05:10:28 2016
New Revision: 280699
URL: http://llvm.org/viewvc/llvm-project?rev=280699&view=rev
Log:
[OpenCL] Remove access qualifiers on images in arg info metadata.
Summary:
Remove access qualifiers on images in arg info metadata:
* kernel_arg_type
* kernel_arg_base_type
Image access qualifiers are inseparable from type in clang implementation,
but OpenCL spec provides a special query to get access qualifier
via clGetKernelArgInfo with CL_KERNEL_ARG_ACCESS_QUALIFIER.
Besides that OpenCL conformance test_api get_kernel_arg_info expects
image types without access qualifier.
Patch by Evgeniy Tyurin.
Reviewers: bader, yaxunl, Anastasia
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D23915
Modified:
cfe/trunk/lib/CodeGen/CodeGenFunction.cpp
cfe/trunk/test/CodeGenOpenCL/kernel-arg-info.cl
Modified: cfe/trunk/lib/CodeGen/CodeGenFunction.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/CodeGen/CodeGenFunction.cpp?rev=280699&r1=280698&r2=280699&view=diff
==
--- cfe/trunk/lib/CodeGen/CodeGenFunction.cpp (original)
+++ cfe/trunk/lib/CodeGen/CodeGenFunction.cpp Tue Sep 6 05:10:28 2016
@@ -428,6 +428,26 @@ void CodeGenFunction::EmitFunctionInstru
EmitNounwindRuntimeCall(F, args);
}
+static void removeImageAccessQualifier(std::string& TyName) {
+ std::string ReadOnlyQual("__read_only");
+ std::string::size_type ReadOnlyPos = TyName.find(ReadOnlyQual);
+ if (ReadOnlyPos != std::string::npos)
+// "+ 1" for the space after access qualifier.
+TyName.erase(ReadOnlyPos, ReadOnlyQual.size() + 1);
+ else {
+std::string WriteOnlyQual("__write_only");
+std::string::size_type WriteOnlyPos = TyName.find(WriteOnlyQual);
+if (WriteOnlyPos != std::string::npos)
+ TyName.erase(WriteOnlyPos, WriteOnlyQual.size() + 1);
+else {
+ std::string ReadWriteQual("__read_write");
+ std::string::size_type ReadWritePos = TyName.find(ReadWriteQual);
+ if (ReadWritePos != std::string::npos)
+TyName.erase(ReadWritePos, ReadWriteQual.size() + 1);
+}
+ }
+}
+
// OpenCL v1.2 s5.6.4.6 allows the compiler to store kernel argument
// information in the program executable. The argument information stored
// includes the argument name, its type, the address and access qualifiers
used.
@@ -524,8 +544,6 @@ static void GenOpenCLArgMetadata(const F
if (ty.isCanonical() && pos != std::string::npos)
typeName.erase(pos+1, 8);
- argTypeNames.push_back(llvm::MDString::get(Context, typeName));
-
std::string baseTypeName;
if (isPipe)
baseTypeName = ty.getCanonicalType()->getAs()
@@ -535,6 +553,17 @@ static void GenOpenCLArgMetadata(const F
baseTypeName =
ty.getUnqualifiedType().getCanonicalType().getAsString(Policy);
+ // Remove access qualifiers on images
+ // (as they are inseparable from type in clang implementation,
+ // but OpenCL spec provides a special query to get access qualifier
+ // via clGetKernelArgInfo with CL_KERNEL_ARG_ACCESS_QUALIFIER):
+ if (ty->isImageType()) {
+removeImageAccessQualifier(typeName);
+removeImageAccessQualifier(baseTypeName);
+ }
+
+ argTypeNames.push_back(llvm::MDString::get(Context, typeName));
+
// Turn "unsigned type" to "utype"
pos = baseTypeName.find("unsigned");
if (pos != std::string::npos)
Modified: cfe/trunk/test/CodeGenOpenCL/kernel-arg-info.cl
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/CodeGenOpenCL/kernel-arg-info.cl?rev=280699&r1=280698&r2=280699&view=diff
==
--- cfe/trunk/test/CodeGenOpenCL/kernel-arg-info.cl (original)
+++ cfe/trunk/test/CodeGenOpenCL/kernel-arg-info.cl Tue Sep 6 05:10:28 2016
@@ -1,5 +1,5 @@
-// RUN: %clang_cc1 %s -emit-llvm -o - -triple spir-unknown-unknown | FileCheck
%s
-// RUN: %clang_cc1 %s -emit-llvm -o - -triple spir-unknown-unknown
-cl-kernel-arg-info | FileCheck %s -check-prefix ARGINFO
+// RUN: %clang_cc1 %s -cl-std=CL2.0 -emit-llvm -o - -triple
spir-unknown-unknown | FileCheck %s
+// RUN: %clang_cc1 %s -cl-std=CL2.0 -emit-llvm -o - -triple
spir-unknown-unknown -cl-kernel-arg-info | FileCheck %s -check-prefix ARGINFO
kernel void foo(__global int * restrict X, const int Y,
volatile int anotherArg, __constant float * restrict Z) {
@@ -14,7 +14,7 @@ kernel void foo(__global int * restrict
// CHECK-NOT: !kernel_arg_name
// ARGINFO: !kernel_arg_name ![[MD15:[0-9]+]]
-kernel void foo2(read_only image1d_t img1, image2d_t img2, write_only
image2d_array_t img3) {
+kernel void foo2(read_only image1d_t img1, image2d_t img2, write_only
image2d_array_t img3, read_write image1d_t img4) {
}
// CHECK: define spir_kernel void @foo2{{[^!]+}}
// CHECK: !kernel_arg_addr_space ![[MD21:[0-9]+]]
@@ -65,11 +65,11 @@ kern
Re: [PATCH] D23915: [OpenCL] Remove access qualifiers on images in arg info metadata.
This revision was automatically updated to reflect the committed changes.
Closed by commit rL280699: [OpenCL] Remove access qualifiers on images in arg
info metadata. (authored by bader).
Changed prior to commit:
https://reviews.llvm.org/D23915?vs=69649&id=70372#toc
Repository:
rL LLVM
https://reviews.llvm.org/D23915
Files:
cfe/trunk/lib/CodeGen/CodeGenFunction.cpp
cfe/trunk/test/CodeGenOpenCL/kernel-arg-info.cl
Index: cfe/trunk/test/CodeGenOpenCL/kernel-arg-info.cl
===
--- cfe/trunk/test/CodeGenOpenCL/kernel-arg-info.cl
+++ cfe/trunk/test/CodeGenOpenCL/kernel-arg-info.cl
@@ -1,5 +1,5 @@
-// RUN: %clang_cc1 %s -emit-llvm -o - -triple spir-unknown-unknown | FileCheck %s
-// RUN: %clang_cc1 %s -emit-llvm -o - -triple spir-unknown-unknown -cl-kernel-arg-info | FileCheck %s -check-prefix ARGINFO
+// RUN: %clang_cc1 %s -cl-std=CL2.0 -emit-llvm -o - -triple spir-unknown-unknown | FileCheck %s
+// RUN: %clang_cc1 %s -cl-std=CL2.0 -emit-llvm -o - -triple spir-unknown-unknown -cl-kernel-arg-info | FileCheck %s -check-prefix ARGINFO
kernel void foo(__global int * restrict X, const int Y,
volatile int anotherArg, __constant float * restrict Z) {
@@ -14,7 +14,7 @@
// CHECK-NOT: !kernel_arg_name
// ARGINFO: !kernel_arg_name ![[MD15:[0-9]+]]
-kernel void foo2(read_only image1d_t img1, image2d_t img2, write_only image2d_array_t img3) {
+kernel void foo2(read_only image1d_t img1, image2d_t img2, write_only image2d_array_t img3, read_write image1d_t img4) {
}
// CHECK: define spir_kernel void @foo2{{[^!]+}}
// CHECK: !kernel_arg_addr_space ![[MD21:[0-9]+]]
@@ -65,11 +65,11 @@
// CHECK: ![[MD13]] = !{!"int*", !"int", !"int", !"float*"}
// CHECK: ![[MD14]] = !{!"restrict", !"const", !"volatile", !"restrict const"}
// ARGINFO: ![[MD15]] = !{!"X", !"Y", !"anotherArg", !"Z"}
-// CHECK: ![[MD21]] = !{i32 1, i32 1, i32 1}
-// CHECK: ![[MD22]] = !{!"read_only", !"read_only", !"write_only"}
-// CHECK: ![[MD23]] = !{!"__read_only image1d_t", !"__read_only image2d_t", !"__write_only image2d_array_t"}
-// CHECK: ![[MD24]] = !{!"", !"", !""}
-// ARGINFO: ![[MD25]] = !{!"img1", !"img2", !"img3"}
+// CHECK: ![[MD21]] = !{i32 1, i32 1, i32 1, i32 1}
+// CHECK: ![[MD22]] = !{!"read_only", !"read_only", !"write_only", !"read_write"}
+// CHECK: ![[MD23]] = !{!"image1d_t", !"image2d_t", !"image2d_array_t", !"image1d_t"}
+// CHECK: ![[MD24]] = !{!"", !"", !"", !""}
+// ARGINFO: ![[MD25]] = !{!"img1", !"img2", !"img3", !"img4"}
// CHECK: ![[MD31]] = !{i32 1}
// CHECK: ![[MD32]] = !{!"none"}
// CHECK: ![[MD33]] = !{!"half*"}
@@ -82,7 +82,7 @@
// CHECK: ![[MD45]] = !{!"", !""}
// ARGINFO: ![[MD46]] = !{!"X", !"Y"}
// CHECK: ![[MD51]] = !{!"read_only", !"write_only"}
-// CHECK: ![[MD52]] = !{!"myImage", !"__write_only image1d_t"}
-// CHECK: ![[MD53]] = !{!"__read_only image1d_t", !"__write_only image1d_t"}
+// CHECK: ![[MD52]] = !{!"myImage", !"image1d_t"}
+// CHECK: ![[MD53]] = !{!"image1d_t", !"image1d_t"}
// ARGINFO: ![[MD54]] = !{!"img1", !"img2"}
Index: cfe/trunk/lib/CodeGen/CodeGenFunction.cpp
===
--- cfe/trunk/lib/CodeGen/CodeGenFunction.cpp
+++ cfe/trunk/lib/CodeGen/CodeGenFunction.cpp
@@ -428,6 +428,26 @@
EmitNounwindRuntimeCall(F, args);
}
+static void removeImageAccessQualifier(std::string& TyName) {
+ std::string ReadOnlyQual("__read_only");
+ std::string::size_type ReadOnlyPos = TyName.find(ReadOnlyQual);
+ if (ReadOnlyPos != std::string::npos)
+// "+ 1" for the space after access qualifier.
+TyName.erase(ReadOnlyPos, ReadOnlyQual.size() + 1);
+ else {
+std::string WriteOnlyQual("__write_only");
+std::string::size_type WriteOnlyPos = TyName.find(WriteOnlyQual);
+if (WriteOnlyPos != std::string::npos)
+ TyName.erase(WriteOnlyPos, WriteOnlyQual.size() + 1);
+else {
+ std::string ReadWriteQual("__read_write");
+ std::string::size_type ReadWritePos = TyName.find(ReadWriteQual);
+ if (ReadWritePos != std::string::npos)
+TyName.erase(ReadWritePos, ReadWriteQual.size() + 1);
+}
+ }
+}
+
// OpenCL v1.2 s5.6.4.6 allows the compiler to store kernel argument
// information in the program executable. The argument information stored
// includes the argument name, its type, the address and access qualifiers used.
@@ -524,8 +544,6 @@
if (ty.isCanonical() && pos != std::string::npos)
typeName.erase(pos+1, 8);
- argTypeNames.push_back(llvm::MDString::get(Context, typeName));
-
std::string baseTypeName;
if (isPipe)
baseTypeName = ty.getCanonicalType()->getAs()
@@ -535,6 +553,17 @@
baseTypeName =
ty.getUnqualifiedType().getCanonicalType().getAsString(Policy);
+ // Remove access qualifiers on images
+ // (as they are inseparable from type in clang implementation,
+ // but OpenCL spec provides a special query to get access qua
r280800 - [OpenCL] Fix pipe built-in functions return type.
Author: bader
Date: Wed Sep 7 05:32:03 2016
New Revision: 280800
URL: http://llvm.org/viewvc/llvm-project?rev=280800&view=rev
Log:
[OpenCL] Fix pipe built-in functions return type.
By default return type of call expressions calling built-in
functions is set to bool.
Fixes https://llvm.org/bugs/show_bug.cgi?id=30219.
Reviewers: Anastasia
Subscribers: dmitry, cfe-commits, yaxunl
Differential Revision: https://reviews.llvm.org/D24136
Modified:
cfe/trunk/lib/Sema/SemaChecking.cpp
cfe/trunk/test/CodeGenOpenCL/pipe_builtin.cl
Modified: cfe/trunk/lib/Sema/SemaChecking.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Sema/SemaChecking.cpp?rev=280800&r1=280799&r2=280800&view=diff
==
--- cfe/trunk/lib/Sema/SemaChecking.cpp (original)
+++ cfe/trunk/lib/Sema/SemaChecking.cpp Wed Sep 7 05:32:03 2016
@@ -1020,6 +1020,7 @@ Sema::CheckBuiltinFunctionCall(FunctionD
// check for the argument.
if (SemaBuiltinRWPipe(*this, TheCall))
return ExprError();
+TheCall->setType(Context.IntTy);
break;
case Builtin::BIreserve_read_pipe:
case Builtin::BIreserve_write_pipe:
@@ -1047,6 +1048,7 @@ Sema::CheckBuiltinFunctionCall(FunctionD
case Builtin::BIget_pipe_max_packets:
if (SemaBuiltinPipePackets(*this, TheCall))
return ExprError();
+TheCall->setType(Context.UnsignedIntTy);
break;
case Builtin::BIto_global:
case Builtin::BIto_local:
Modified: cfe/trunk/test/CodeGenOpenCL/pipe_builtin.cl
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/CodeGenOpenCL/pipe_builtin.cl?rev=280800&r1=280799&r2=280800&view=diff
==
--- cfe/trunk/test/CodeGenOpenCL/pipe_builtin.cl (original)
+++ cfe/trunk/test/CodeGenOpenCL/pipe_builtin.cl Wed Sep 7 05:32:03 2016
@@ -59,3 +59,19 @@ void test7(write_only pipe int p, global
// CHECK: call i32 @__get_pipe_max_packets(%opencl.pipe_t* %{{.*}})
*ptr = get_pipe_max_packets(p);
}
+
+void test8(read_only pipe int r, write_only pipe int w, global int *ptr) {
+ // verify that return type is correctly casted to i1 value
+ // CHECK: %[[R:[0-9]+]] = call i32 @__read_pipe_2
+ // CHECK: icmp ne i32 %[[R]], 0
+ if (read_pipe(r, ptr)) *ptr = -1;
+ // CHECK: %[[W:[0-9]+]] = call i32 @__write_pipe_2
+ // CHECK: icmp ne i32 %[[W]], 0
+ if (write_pipe(w, ptr)) *ptr = -1;
+ // CHECK: %[[N:[0-9]+]] = call i32 @__get_pipe_num_packets
+ // CHECK: icmp ne i32 %[[N]], 0
+ if (get_pipe_num_packets(r)) *ptr = -1;
+ // CHECK: %[[M:[0-9]+]] = call i32 @__get_pipe_max_packets
+ // CHECK: icmp ne i32 %[[M]], 0
+ if (get_pipe_max_packets(w)) *ptr = -1;
+}
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: [PATCH] D24136: [OpenCL] Fix pipe built-in functions return type.
This revision was automatically updated to reflect the committed changes.
Closed by commit rL280800: [OpenCL] Fix pipe built-in functions return type.
(authored by bader).
Changed prior to commit:
https://reviews.llvm.org/D24136?vs=70368&id=70516#toc
Repository:
rL LLVM
https://reviews.llvm.org/D24136
Files:
cfe/trunk/lib/Sema/SemaChecking.cpp
cfe/trunk/test/CodeGenOpenCL/pipe_builtin.cl
Index: cfe/trunk/test/CodeGenOpenCL/pipe_builtin.cl
===
--- cfe/trunk/test/CodeGenOpenCL/pipe_builtin.cl
+++ cfe/trunk/test/CodeGenOpenCL/pipe_builtin.cl
@@ -59,3 +59,19 @@
// CHECK: call i32 @__get_pipe_max_packets(%opencl.pipe_t* %{{.*}})
*ptr = get_pipe_max_packets(p);
}
+
+void test8(read_only pipe int r, write_only pipe int w, global int *ptr) {
+ // verify that return type is correctly casted to i1 value
+ // CHECK: %[[R:[0-9]+]] = call i32 @__read_pipe_2
+ // CHECK: icmp ne i32 %[[R]], 0
+ if (read_pipe(r, ptr)) *ptr = -1;
+ // CHECK: %[[W:[0-9]+]] = call i32 @__write_pipe_2
+ // CHECK: icmp ne i32 %[[W]], 0
+ if (write_pipe(w, ptr)) *ptr = -1;
+ // CHECK: %[[N:[0-9]+]] = call i32 @__get_pipe_num_packets
+ // CHECK: icmp ne i32 %[[N]], 0
+ if (get_pipe_num_packets(r)) *ptr = -1;
+ // CHECK: %[[M:[0-9]+]] = call i32 @__get_pipe_max_packets
+ // CHECK: icmp ne i32 %[[M]], 0
+ if (get_pipe_max_packets(w)) *ptr = -1;
+}
Index: cfe/trunk/lib/Sema/SemaChecking.cpp
===
--- cfe/trunk/lib/Sema/SemaChecking.cpp
+++ cfe/trunk/lib/Sema/SemaChecking.cpp
@@ -1020,6 +1020,7 @@
// check for the argument.
if (SemaBuiltinRWPipe(*this, TheCall))
return ExprError();
+TheCall->setType(Context.IntTy);
break;
case Builtin::BIreserve_read_pipe:
case Builtin::BIreserve_write_pipe:
@@ -1047,6 +1048,7 @@
case Builtin::BIget_pipe_max_packets:
if (SemaBuiltinPipePackets(*this, TheCall))
return ExprError();
+TheCall->setType(Context.UnsignedIntTy);
break;
case Builtin::BIto_global:
case Builtin::BIto_local:
Index: cfe/trunk/test/CodeGenOpenCL/pipe_builtin.cl
===
--- cfe/trunk/test/CodeGenOpenCL/pipe_builtin.cl
+++ cfe/trunk/test/CodeGenOpenCL/pipe_builtin.cl
@@ -59,3 +59,19 @@
// CHECK: call i32 @__get_pipe_max_packets(%opencl.pipe_t* %{{.*}})
*ptr = get_pipe_max_packets(p);
}
+
+void test8(read_only pipe int r, write_only pipe int w, global int *ptr) {
+ // verify that return type is correctly casted to i1 value
+ // CHECK: %[[R:[0-9]+]] = call i32 @__read_pipe_2
+ // CHECK: icmp ne i32 %[[R]], 0
+ if (read_pipe(r, ptr)) *ptr = -1;
+ // CHECK: %[[W:[0-9]+]] = call i32 @__write_pipe_2
+ // CHECK: icmp ne i32 %[[W]], 0
+ if (write_pipe(w, ptr)) *ptr = -1;
+ // CHECK: %[[N:[0-9]+]] = call i32 @__get_pipe_num_packets
+ // CHECK: icmp ne i32 %[[N]], 0
+ if (get_pipe_num_packets(r)) *ptr = -1;
+ // CHECK: %[[M:[0-9]+]] = call i32 @__get_pipe_max_packets
+ // CHECK: icmp ne i32 %[[M]], 0
+ if (get_pipe_max_packets(w)) *ptr = -1;
+}
Index: cfe/trunk/lib/Sema/SemaChecking.cpp
===
--- cfe/trunk/lib/Sema/SemaChecking.cpp
+++ cfe/trunk/lib/Sema/SemaChecking.cpp
@@ -1020,6 +1020,7 @@
// check for the argument.
if (SemaBuiltinRWPipe(*this, TheCall))
return ExprError();
+TheCall->setType(Context.IntTy);
break;
case Builtin::BIreserve_read_pipe:
case Builtin::BIreserve_write_pipe:
@@ -1047,6 +1048,7 @@
case Builtin::BIget_pipe_max_packets:
if (SemaBuiltinPipePackets(*this, TheCall))
return ExprError();
+TheCall->setType(Context.UnsignedIntTy);
break;
case Builtin::BIto_global:
case Builtin::BIto_local:
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: [PATCH] D23992: [OpenCL] Augment pipe built-ins with pipe packet size and alignment.
bader added a comment. Valery, do you have any additional comments/questions regarding this patch? https://reviews.llvm.org/D23992 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: [PATCH] D24626: [OpenCL] Diagnose assignment to dereference of half type pointer
bader accepted this revision. bader added a comment. This revision is now accepted and ready to land. LGTM. Thanks. https://reviews.llvm.org/D24626 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: [PATCH] D24666: [OpenCL] Allow half type kernel argument when cl_khr_fp16 is enabled
bader added inline comments.
Comment at: test/SemaOpenCL/half.cl:29
@@ +28,3 @@
+kernel void half_disabled_kernel(global half *p,
+ half h); // expected-error{{'half' cannot be
used as the type of a kernel parameter}} // expected-error{{declaring function
parameter of type 'half' is not allowed; did you forget * ?}}
+
It's not related to your change, but it looks like there are two checks that
report the same error.
Could you remove the duplication, please?
https://reviews.llvm.org/D24666
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: [PATCH] D24666: [OpenCL] Allow half type kernel argument when cl_khr_fp16 is enabled
bader added inline comments. Comment at: lib/Sema/SemaDecl.cpp:7599-7602 @@ -7595,3 +7598,6 @@ // of event_t type. -S.Diag(Param->getLocation(), diag::err_bad_kernel_param_type) << PT; +// Do not diagnose half type since it is diagnosed as invalid argument +// type for any function eleswhere. +if (!PT->isHalfType()) + S.Diag(Param->getLocation(), diag::err_bad_kernel_param_type) << PT; D.setInvalidType(); It looks strange to me. First we check if parameter type is half - we set InvalidKernelParam status, later we check again and do not report an error. I think it would be simpler just return ValidKernelParam for half data type from getOpenCLKernelParameterType, I think the whole patch should two deleted lines from getOpenCLKernelParameterType function + test case. https://reviews.llvm.org/D24666 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
r306653 - [OpenCL] Allow function declaration with empty argument list.
Author: bader
Date: Thu Jun 29 01:44:10 2017
New Revision: 306653
URL: http://llvm.org/viewvc/llvm-project?rev=306653&view=rev
Log:
[OpenCL] Allow function declaration with empty argument list.
Summary:
does it make sense to enable K&R function declaration style for OpenCL?
clang throws following error message for the declaration w/o arguments:
```
int my_func();
error: function with no prototype cannot use the spir_function calling
convention
```
Current way to fix this issue is to specify that parameter list is empty by
using 'void':
```
int my_func(void);
```
Let me know what do you think about this patch.
Reviewers: Anastasia, yaxunl
Reviewed By: Anastasia
Subscribers: cfe-commits, echuraev
Differential Revision: https://reviews.llvm.org/D33681
Added:
cfe/trunk/test/SemaOpenCL/function-no-args.cl
Modified:
cfe/trunk/lib/Sema/SemaType.cpp
cfe/trunk/test/SemaOpenCL/invalid-pipes-cl2.0.cl
Modified: cfe/trunk/lib/Sema/SemaType.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Sema/SemaType.cpp?rev=306653&r1=306652&r2=306653&view=diff
==
--- cfe/trunk/lib/Sema/SemaType.cpp (original)
+++ cfe/trunk/lib/Sema/SemaType.cpp Thu Jun 29 01:44:10 2017
@@ -4355,7 +4355,7 @@ static TypeSourceInfo *GetFullTypeForDec
FunctionType::ExtInfo EI(getCCForDeclaratorChunk(S, D, FTI, chunkIndex));
- if (!FTI.NumParams && !FTI.isVariadic && !LangOpts.CPlusPlus) {
+ if (!FTI.NumParams && !FTI.isVariadic && !LangOpts.CPlusPlus &&
!LangOpts.OpenCL) {
// Simple void foo(), where the incoming T is the result type.
T = Context.getFunctionNoProtoType(T, EI);
} else {
Added: cfe/trunk/test/SemaOpenCL/function-no-args.cl
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/SemaOpenCL/function-no-args.cl?rev=306653&view=auto
==
--- cfe/trunk/test/SemaOpenCL/function-no-args.cl (added)
+++ cfe/trunk/test/SemaOpenCL/function-no-args.cl Thu Jun 29 01:44:10 2017
@@ -0,0 +1,9 @@
+// RUN: %clang_cc1 -verify -pedantic -fsyntax-only -cl-std=CL2.0 %s
+// expected-no-diagnostics
+
+global int gi;
+int my_func();
+int my_func() {
+ gi = 2;
+ return gi;
+}
Modified: cfe/trunk/test/SemaOpenCL/invalid-pipes-cl2.0.cl
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/SemaOpenCL/invalid-pipes-cl2.0.cl?rev=306653&r1=306652&r2=306653&view=diff
==
--- cfe/trunk/test/SemaOpenCL/invalid-pipes-cl2.0.cl (original)
+++ cfe/trunk/test/SemaOpenCL/invalid-pipes-cl2.0.cl Thu Jun 29 01:44:10 2017
@@ -3,7 +3,7 @@
global pipe int gp;// expected-error {{type '__global read_only
pipe int' can only be used as a function parameter in OpenCL}}
global reserve_id_t rid; // expected-error {{the '__global
reserve_id_t' type cannot be used to declare a program scope variable}}
-extern pipe write_only int get_pipe(); // expected-error {{type '__global
write_only pipe int ()' can only be used as a function parameter in OpenCL}}
+extern pipe write_only int get_pipe(); // expected-error {{type '__global
write_only pipe int (void)' can only be used as a function parameter in OpenCL}}
kernel void test_invalid_reserved_id(reserve_id_t ID) { // expected-error
{{'reserve_id_t' cannot be used as the type of a kernel parameter}}
}
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
r323522 - [OpenCL] Add "cles_khr_int64" extension.
Author: bader
Date: Fri Jan 26 03:48:46 2018
New Revision: 323522
URL: http://llvm.org/viewvc/llvm-project?rev=323522&view=rev
Log:
[OpenCL] Add "cles_khr_int64" extension.
Summary:
For OpenCL 1.1 embedded profile 64 bit integers i.e. long,
ulong including the appropriate vector data types and operations
on 64-bit integers are optional. The "cles_khr_int64" extension
string will be reported if the embedded profile implementation
supports 64-bit integers.
Reviewers: Anastasia, bader
Reviewed By: Anastasia, bader
Subscribers: bader, yaxunl, Anastasia, cfe-commits
Differential Revision: https://reviews.llvm.org/D42532
Modified:
cfe/trunk/include/clang/Basic/OpenCLExtensions.def
cfe/trunk/test/SemaOpenCL/extension-version.cl
Modified: cfe/trunk/include/clang/Basic/OpenCLExtensions.def
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/include/clang/Basic/OpenCLExtensions.def?rev=323522&r1=323521&r2=323522&view=diff
==
--- cfe/trunk/include/clang/Basic/OpenCLExtensions.def (original)
+++ cfe/trunk/include/clang/Basic/OpenCLExtensions.def Fri Jan 26 03:48:46 2018
@@ -53,6 +53,9 @@ OPENCLEXT_INTERNAL(cl_khr_icd, 100, ~0U)
OPENCLEXT_INTERNAL(cl_khr_gl_event, 110, ~0U)
OPENCLEXT_INTERNAL(cl_khr_d3d10_sharing, 110, ~0U)
+// EMBEDDED_PROFILE
+OPENCLEXT_INTERNAL(cles_khr_int64, 110, ~0U)
+
// OpenCL 1.2.
OPENCLEXT_INTERNAL(cl_khr_context_abort, 120, ~0U)
OPENCLEXT_INTERNAL(cl_khr_d3d11_sharing, 120, ~0U)
Modified: cfe/trunk/test/SemaOpenCL/extension-version.cl
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/SemaOpenCL/extension-version.cl?rev=323522&r1=323521&r2=323522&view=diff
==
--- cfe/trunk/test/SemaOpenCL/extension-version.cl (original)
+++ cfe/trunk/test/SemaOpenCL/extension-version.cl Fri Jan 26 03:48:46 2018
@@ -131,6 +131,15 @@
#endif
#pragma OPENCL EXTENSION cl_khr_d3d10_sharing: enable
+#if (__OPENCL_C_VERSION__ >= 110)
+#ifndef cles_khr_int64
+#error "Missing cles_khr_int64 define"
+#endif
+#else
+// expected-warning@+2{{unsupported OpenCL extension 'cles_khr_int64' -
ignoring}}
+#endif
+#pragma OPENCL EXTENSION cles_khr_int64: enable
+
#if (__OPENCL_C_VERSION__ >= 120)
#ifndef cl_khr_context_abort
#error "Missing cl_context_abort define"
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[PATCH] D26746: [OpenCL] Split PipeType into ReadPipe/WritePipe
bader accepted this revision.
bader added a comment.
This revision is now accepted and ready to land.
LGTM. Thanks!
A few minor comments regarding outdated comments and style.
Comment at: include/clang/AST/ASTContext.h:1124
/// \brief Return pipe type for the specified type.
+ QualType getReadPipeType(QualType T) const;
Please, update the comment to specify that this function return pipe type with
'__read_only' access qualifier.
Comment at: lib/AST/ASTContext.cpp:3341
/// Return pipe type for the specified type.
+QualType ASTContext::getReadPipeType(QualType T) const {
Please, remove this comment. It's a copy of the comment from the header file.
Comment at: lib/Serialization/ASTReader.cpp:5806-5807
+return Context.getReadPipeType(ElementType);
+ }
+ case TYPE_WRITE_PIPE: {
+if (Record.size() != 1) {
Please, separate case with an empty line.
Repository:
rL LLVM
https://reviews.llvm.org/D26746
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
r288126 - [OpenCL] Prohibit using reserve_id_t in program scope.
Author: bader
Date: Tue Nov 29 04:21:40 2016
New Revision: 288126
URL: http://llvm.org/viewvc/llvm-project?rev=288126&view=rev
Log:
[OpenCL] Prohibit using reserve_id_t in program scope.
Patch by Egor Churaev (echuraev).
Reviewers: Anastasia
Subscribers: cfe-commits, yaxunl, bader
Differential Revision: https://reviews.llvm.org/D27099
Added:
cfe/trunk/test/SemaOpenCL/invalid-clk-events-cl2.0.cl
Modified:
cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td
cfe/trunk/lib/Sema/SemaDecl.cpp
cfe/trunk/test/SemaOpenCL/event_t.cl
cfe/trunk/test/SemaOpenCL/invalid-pipes-cl2.0.cl
Modified: cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td?rev=288126&r1=288125&r2=288126&view=diff
==
--- cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td (original)
+++ cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td Tue Nov 29 04:21:40
2016
@@ -8075,8 +8075,6 @@ def note_within_field_of_type : Note<
"within field of type %0 declared here">;
def note_illegal_field_declared_here : Note<
"field of illegal %select{type|pointer type}0 %1 declared here">;
-def err_event_t_global_var : Error<
- "the event_t type cannot be used to declare a program scope variable">;
def err_opencl_type_struct_or_union_field : Error<
"the %0 type cannot be used to declare a structure or union field">;
def err_event_t_addr_space_qual : Error<
@@ -8589,6 +8587,8 @@ def note_related_result_type_inferred :
def note_related_result_type_explicit : Note<
"%select{overridden|current}0 method is explicitly declared 'instancetype'"
"%select{| and is expected to return an instance of its class type}0">;
+def err_invalid_type_for_program_scope_var : Error<
+ "the %0 type cannot be used to declare a program scope variable">;
}
Modified: cfe/trunk/lib/Sema/SemaDecl.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Sema/SemaDecl.cpp?rev=288126&r1=288125&r2=288126&view=diff
==
--- cfe/trunk/lib/Sema/SemaDecl.cpp (original)
+++ cfe/trunk/lib/Sema/SemaDecl.cpp Tue Nov 29 04:21:40 2016
@@ -5909,32 +5909,31 @@ NamedDecl *Sema::ActOnVariableDeclarator
return nullptr;
}
- // OpenCL v2.0 s6.9.b - Image type can only be used as a function argument.
- // OpenCL v2.0 s6.13.16.1 - Pipe type can only be used as a function
- // argument.
- if (getLangOpts().OpenCL && (R->isImageType() || R->isPipeType())) {
-Diag(D.getIdentifierLoc(),
- diag::err_opencl_type_can_only_be_used_as_function_parameter)
-<< R;
-D.setInvalidType();
-return nullptr;
- }
-
- DeclSpec::SCS SCSpec = D.getDeclSpec().getStorageClassSpec();
- StorageClass SC = StorageClassSpecToVarDeclStorageClass(D.getDeclSpec());
-
- // dllimport globals without explicit storage class are treated as extern. We
- // have to change the storage class this early to get the right DeclContext.
- if (SC == SC_None && !DC->isRecord() &&
- hasParsedAttr(S, D, AttributeList::AT_DLLImport) &&
- !hasParsedAttr(S, D, AttributeList::AT_DLLExport))
-SC = SC_Extern;
+ if (getLangOpts().OpenCL) {
+// OpenCL v2.0 s6.9.b - Image type can only be used as a function argument.
+// OpenCL v2.0 s6.13.16.1 - Pipe type can only be used as a function
+// argument.
+if (R->isImageType() || R->isPipeType()) {
+ Diag(D.getIdentifierLoc(),
+ diag::err_opencl_type_can_only_be_used_as_function_parameter)
+ << R;
+ D.setInvalidType();
+ return nullptr;
+}
- DeclContext *OriginalDC = DC;
- bool IsLocalExternDecl = SC == SC_Extern &&
- adjustContextForLocalExternDecl(DC);
+// OpenCL v1.2 s6.9.r:
+// The event type cannot be used to declare a program scope variable.
+// OpenCL v2.0 s6.9.q:
+// The clk_event_t and reserve_id_t types cannot be declared in program
scope.
+if (NULL == S->getParent()) {
+ if (R->isReserveIDT() || R->isClkEventT() || R->isEventT()) {
+Diag(D.getIdentifierLoc(),
+ diag::err_invalid_type_for_program_scope_var) << R;
+D.setInvalidType();
+return nullptr;
+ }
+}
- if (getLangOpts().OpenCL) {
// OpenCL v1.0 s6.8.a.3: Pointers to functions are not allowed.
QualType NR = R;
while (NR->isPointerType()) {
@@ -5954,8 +5953,40 @@ NamedDecl *Sema::ActOnVariableDeclarator
D.setInvalidType();
}
}
+
+// OpenCL v1.2 s6.9.b p4:
+// The sampler type cannot be used with the __local and __global address
+// space qualifiers.
+if (R->isSamplerT() && (R.getAddressSpace() == LangAS::opencl_local ||
+ R.getAddressSpace() == LangAS::opencl_global)) {
+ Diag(D.getIdentifierLoc(), diag::err_wrong_sampler_addressspace);
+}
+
+// OpenCL v1.2 s6.9.r:
+// The event
r288890 - [OpenCL] Fix SPIR version generation.
Author: bader
Date: Wed Dec 7 02:38:24 2016
New Revision: 288890
URL: http://llvm.org/viewvc/llvm-project?rev=288890&view=rev
Log:
[OpenCL] Fix SPIR version generation.
Patch by Egor Churaev (echuraev).
Reviewers: Anastasia
Subscribers: bader, yaxunl, cfe-commits
Differential Revision: https://reviews.llvm.org/D27300
Modified:
cfe/trunk/lib/CodeGen/TargetInfo.cpp
cfe/trunk/test/CodeGenOpenCL/spir_version.cl
Modified: cfe/trunk/lib/CodeGen/TargetInfo.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/CodeGen/TargetInfo.cpp?rev=288890&r1=29&r2=288890&view=diff
==
--- cfe/trunk/lib/CodeGen/TargetInfo.cpp (original)
+++ cfe/trunk/lib/CodeGen/TargetInfo.cpp Wed Dec 7 02:38:24 2016
@@ -7801,8 +7801,10 @@ void SPIRTargetCodeGenInfo::emitTargetMD
// SPIR v2.0 s2.12 - The SPIR version used by the module is stored in the
// opencl.spir.version named metadata.
llvm::Metadata *SPIRVerElts[] = {
- llvm::ConstantAsMetadata::get(llvm::ConstantInt::get(Int32Ty, 2)),
- llvm::ConstantAsMetadata::get(llvm::ConstantInt::get(Int32Ty, 0))};
+ llvm::ConstantAsMetadata::get(
+ llvm::ConstantInt::get(Int32Ty, CGM.getLangOpts().OpenCLVersion /
100)),
+ llvm::ConstantAsMetadata::get(llvm::ConstantInt::get(
+ Int32Ty, (CGM.getLangOpts().OpenCLVersion / 100 > 1) ? 0 : 2))};
llvm::NamedMDNode *SPIRVerMD =
M.getOrInsertNamedMetadata("opencl.spir.version");
SPIRVerMD->addOperand(llvm::MDNode::get(Ctx, SPIRVerElts));
Modified: cfe/trunk/test/CodeGenOpenCL/spir_version.cl
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/CodeGenOpenCL/spir_version.cl?rev=288890&r1=29&r2=288890&view=diff
==
--- cfe/trunk/test/CodeGenOpenCL/spir_version.cl (original)
+++ cfe/trunk/test/CodeGenOpenCL/spir_version.cl Wed Dec 7 02:38:24 2016
@@ -13,19 +13,22 @@ kernel void foo() {}
// CHECK-SPIR-CL10: !opencl.spir.version = !{[[SPIR:![0-9]+]]}
// CHECK-SPIR-CL10: !opencl.ocl.version = !{[[OCL:![0-9]+]]}
-// CHECK-SPIR-CL10: [[SPIR]] = !{i32 2, i32 0}
+// CHECK-SPIR-CL10: [[SPIR]] = !{i32 1, i32 2}
// CHECK-SPIR-CL10: [[OCL]] = !{i32 1, i32 0}
-// CHECK-SPIR-CL12: !opencl.spir.version = !{[[SPIR:![0-9]+]]}
-// CHECK-SPIR-CL12: !opencl.ocl.version = !{[[OCL:![0-9]+]]}
-// CHECK-SPIR-CL12: [[SPIR]] = !{i32 2, i32 0}
-// CHECK-SPIR-CL12: [[OCL]] = !{i32 1, i32 2}
-// CHECK-SPIR-CL20: !opencl.spir.version = !{[[SPIR:![0-9]+]]}
-// CHECK-SPIR-CL20: !opencl.ocl.version = !{[[SPIR:![0-9]+]]}
-// CHECK-SPIR-CL20: [[SPIR]] = !{i32 2, i32 0}
+// CHECK-SPIR-CL12: !opencl.spir.version = !{[[VER:![0-9]+]]}
+// CHECK-SPIR-CL12: !opencl.ocl.version = !{[[VER]]}
+// CHECK-SPIR-CL12: [[VER]] = !{i32 1, i32 2}
+// CHECK-SPIR-CL20: !opencl.spir.version = !{[[VER:![0-9]+]]}
+// CHECK-SPIR-CL20: !opencl.ocl.version = !{[[VER]]}
+// CHECK-SPIR-CL20: [[VER]] = !{i32 2, i32 0}
+
+// CHECK-AMDGCN-CL10-NOT: !opencl.spir.version
// CHECK-AMDGCN-CL10: !opencl.ocl.version = !{[[OCL:![0-9]+]]}
// CHECK-AMDGCN-CL10: [[OCL]] = !{i32 1, i32 0}
+// CHECK-AMDGCN-CL12-NOT: !opencl.spir.version
// CHECK-AMDGCN-CL12: !opencl.ocl.version = !{[[OCL:![0-9]+]]}
// CHECK-AMDGCN-CL12: [[OCL]] = !{i32 1, i32 2}
+// CHECK-AMDGCN-CL20-NOT: !opencl.spir.version
// CHECK-AMDGCN-CL20: !opencl.ocl.version = !{[[OCL:![0-9]+]]}
-// CHECK-AMDGCN-CL20: [[OCL]] = !{i32 2, i32 0}
\ No newline at end of file
+// CHECK-AMDGCN-CL20: [[OCL]] = !{i32 2, i32 0}
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
r288891 - [OpenCL] Added a LIT test for ensuring address space mangling is done the same both in OpenCL1.2 and OpenCL2.0.
Author: bader
Date: Wed Dec 7 02:43:49 2016
New Revision: 288891
URL: http://llvm.org/viewvc/llvm-project?rev=288891&view=rev
Log:
[OpenCL] Added a LIT test for ensuring address space mangling is done the same
both in OpenCL1.2 and OpenCL2.0.
Patch by Egor Churaev (echuraev).
Reviewers: Anastasia
Subscribers: yaxunl, cfe-commits, bader
Differential Revision: https://reviews.llvm.org/D27403
Modified:
cfe/trunk/test/CodeGenOpenCL/address-spaces-mangling.cl
Modified: cfe/trunk/test/CodeGenOpenCL/address-spaces-mangling.cl
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/CodeGenOpenCL/address-spaces-mangling.cl?rev=288891&r1=288890&r2=288891&view=diff
==
--- cfe/trunk/test/CodeGenOpenCL/address-spaces-mangling.cl (original)
+++ cfe/trunk/test/CodeGenOpenCL/address-spaces-mangling.cl Wed Dec 7 02:43:49
2016
@@ -1,6 +1,10 @@
// RUN: %clang_cc1 %s -ffake-address-space-map
-faddress-space-map-mangling=yes -triple %itanium_abi_triple -emit-llvm -o - |
FileCheck -check-prefix=ASMANG %s
// RUN: %clang_cc1 %s -ffake-address-space-map -faddress-space-map-mangling=no
-triple %itanium_abi_triple -emit-llvm -o - | FileCheck -check-prefix=NOASMANG
%s
+// We check that the address spaces are mangled the same in both version of
OpenCL
+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -cl-std=CL2.0 -emit-llvm -o
- | FileCheck -check-prefix=OCL-20 %s
+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -cl-std=CL1.2 -emit-llvm -o
- | FileCheck -check-prefix=OCL-12 %s
+
// We can't name this f as private is equivalent to default
// no specifier given address space so we get multiple definition
// warnings, but we do want it for comparison purposes.
@@ -8,23 +12,33 @@ __attribute__((overloadable))
void ff(int *arg) { }
// ASMANG: @_Z2ffPi
// NOASMANG: @_Z2ffPi
+// OCL-20-DAG: @_Z2ffPU3AS4i
+// OCL-12-DAG: @_Z2ffPi
__attribute__((overloadable))
void f(private int *arg) { }
// ASMANG: @_Z1fPi
// NOASMANG: @_Z1fPi
+// OCL-20-DAG: @_Z1fPi
+// OCL-12-DAG: @_Z1fPi
__attribute__((overloadable))
void f(global int *arg) { }
// ASMANG: @_Z1fPU3AS1i
// NOASMANG: @_Z1fPU8CLglobali
+// OCL-20-DAG: @_Z1fPU3AS1i
+// OCL-12-DAG: @_Z1fPU3AS1i
__attribute__((overloadable))
void f(local int *arg) { }
// ASMANG: @_Z1fPU3AS2i
// NOASMANG: @_Z1fPU7CLlocali
+// OCL-20-DAG: @_Z1fPU3AS2i
+// OCL-12-DAG: @_Z1fPU3AS2i
__attribute__((overloadable))
void f(constant int *arg) { }
// ASMANG: @_Z1fPU3AS3i
// NOASMANG: @_Z1fPU10CLconstanti
+// OCL-20-DAG: @_Z1fPU3AS3i
+// OCL-12-DAG: @_Z1fPU3AS3i
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[PATCH] D25305: [OpenCL] Setting constant address space for array initializers
bader accepted this revision. bader added a comment. LGTM https://reviews.llvm.org/D25305 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[PATCH] D25305: [OpenCL] Setting constant address space for array initializers
bader added inline comments.
Comment at: lib/CodeGen/CGDecl.cpp:1272
+if (getLangOpts().OpenCL) {
+ UA = llvm::GlobalValue::UnnamedAddr::None;
+ AS = CGM.getContext().getTargetAddressSpace(LangAS::opencl_constant);
AlexeySotkin wrote:
> Anastasia wrote:
> > Why this change?
> Without this change, global variables with unnamed address space are
> translated to SPIR-V as variables with "Function" storage class, which is
> wrong.
> This should fix this issue:
> https://github.com/KhronosGroup/SPIRV-LLVM/issues/50
There is inconsistency with how Clang maps initializers to OpenCL memory model.
Consider example from the test case:
```
__private int arr[] = {1, 2, 3};
```
This code allocates a global constant for initializer {1, 2, 3} and later
copies values from this global constant to the private array using memcpy
intrinsic. The global constant must be allocated in constant address space, but
old code do assign any address space, which is considered to be a private
memory region.
This patch puts global constant objects to constant address space.
https://reviews.llvm.org/D25305
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[PATCH] D25305: [OpenCL] Setting constant address space for array initializers
bader added inline comments.
Comment at: lib/CodeGen/CGDecl.cpp:1272
+if (getLangOpts().OpenCL) {
+ UA = llvm::GlobalValue::UnnamedAddr::None;
+ AS = CGM.getContext().getTargetAddressSpace(LangAS::opencl_constant);
Anastasia wrote:
> bader wrote:
> > AlexeySotkin wrote:
> > > Anastasia wrote:
> > > > Why this change?
> > > Without this change, global variables with unnamed address space are
> > > translated to SPIR-V as variables with "Function" storage class, which is
> > > wrong.
> > > This should fix this issue:
> > > https://github.com/KhronosGroup/SPIRV-LLVM/issues/50
> > There is inconsistency with how Clang maps initializers to OpenCL memory
> > model.
> > Consider example from the test case:
> > ```
> > __private int arr[] = {1, 2, 3};
> > ```
> > This code allocates a global constant for initializer {1, 2, 3} and later
> > copies values from this global constant to the private array using memcpy
> > intrinsic. The global constant must be allocated in constant address space,
> > but old code do assign any address space, which is considered to be a
> > private memory region.
> > This patch puts global constant objects to constant address space.
> Yes, this is perfectly fine. I was specifically asking about setting
> llvm::GlobalValue::UnnamedAddr::None attribute. I can't see why this has to
> be done or is it because we now assign concrete address space and private was
> treated as no address space at all?
This attribute has nothing to do with address spaces.
See http://llvm.org/docs/LangRef.html#global-variables for local_unnamed_addr
and unnamed_addr attributes description.
My understanding is that llvm::GlobalValue::UnnamedAddr::Global should be fine
here.
https://reviews.llvm.org/D25305
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
r285557 - [OpenCL] Setting constant address space for array initializers
Author: bader
Date: Mon Oct 31 05:26:31 2016
New Revision: 285557
URL: http://llvm.org/viewvc/llvm-project?rev=285557&view=rev
Log:
[OpenCL] Setting constant address space for array initializers
Summary: Setting constant address space for global constants used for
memcpy-initialization of arrays.
Patch by Alexey Sotkin.
Reviewers: bader, yaxunl, Anastasia
Subscribers: cfe-commits, AlexeySotkin
Differential Revision: https://reviews.llvm.org/D25305
Added:
cfe/trunk/test/CodeGenOpenCL/private-array-initialization.cl
Modified:
cfe/trunk/lib/CodeGen/CGDecl.cpp
cfe/trunk/test/CodeGenOpenCL/partial_initializer.cl
Modified: cfe/trunk/lib/CodeGen/CGDecl.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/CodeGen/CGDecl.cpp?rev=285557&r1=285556&r2=285557&view=diff
==
--- cfe/trunk/lib/CodeGen/CGDecl.cpp (original)
+++ cfe/trunk/lib/CodeGen/CGDecl.cpp Mon Oct 31 05:26:31 2016
@@ -1225,10 +1225,16 @@ void CodeGenFunction::EmitAutoVarInit(co
// Otherwise, create a temporary global with the initializer then
// memcpy from the global to the alloca.
std::string Name = getStaticDeclName(CGM, D);
+unsigned AS = 0;
+if (getLangOpts().OpenCL) {
+ AS = CGM.getContext().getTargetAddressSpace(LangAS::opencl_constant);
+ BP = llvm::PointerType::getInt8PtrTy(getLLVMContext(), AS);
+}
llvm::GlobalVariable *GV =
new llvm::GlobalVariable(CGM.getModule(), constant->getType(), true,
llvm::GlobalValue::PrivateLinkage,
- constant, Name);
+ constant, Name, nullptr,
+ llvm::GlobalValue::NotThreadLocal, AS);
GV->setAlignment(Loc.getAlignment().getQuantity());
GV->setUnnamedAddr(llvm::GlobalValue::UnnamedAddr::Global);
Modified: cfe/trunk/test/CodeGenOpenCL/partial_initializer.cl
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/CodeGenOpenCL/partial_initializer.cl?rev=285557&r1=285556&r2=285557&view=diff
==
--- cfe/trunk/test/CodeGenOpenCL/partial_initializer.cl (original)
+++ cfe/trunk/test/CodeGenOpenCL/partial_initializer.cl Mon Oct 31 05:26:31 2016
@@ -24,7 +24,7 @@ int4 GV1 = (int4)((int2)(1,2),3,4);
// CHECK: @GV2 = addrspace(1) global <4 x i32> ,
align 16
int4 GV2 = (int4)(1);
-// CHECK: @f.S = private unnamed_addr constant %struct.StrucTy { i32 1, i32 2,
i32 0 }, align 4
+// CHECK: @f.S = private unnamed_addr addrspace(2) constant %struct.StrucTy {
i32 1, i32 2, i32 0 }, align 4
// CHECK-LABEL: define spir_func void @f()
void f(void) {
@@ -46,7 +46,7 @@ void f(void) {
float A[6][6] = {1.0f, 2.0f};
// CHECK: %[[v5:.*]] = bitcast %struct.StrucTy* %S to i8*
- // CHECK: call void @llvm.memcpy.p0i8.p0i8.i32(i8* %[[v5]], i8* bitcast
(%struct.StrucTy* @f.S to i8*), i32 12, i32 4, i1 false)
+ // CHECK: call void @llvm.memcpy.p0i8.p2i8.i32(i8* %[[v5]], i8 addrspace(2)*
bitcast (%struct.StrucTy addrspace(2)* @f.S to i8 addrspace(2)*), i32 12, i32
4, i1 false)
StrucTy S = {1, 2};
// CHECK: store <2 x i32> , <2 x i32>* %[[compoundliteral1]],
align 8
Added: cfe/trunk/test/CodeGenOpenCL/private-array-initialization.cl
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/CodeGenOpenCL/private-array-initialization.cl?rev=285557&view=auto
==
--- cfe/trunk/test/CodeGenOpenCL/private-array-initialization.cl (added)
+++ cfe/trunk/test/CodeGenOpenCL/private-array-initialization.cl Mon Oct 31
05:26:31 2016
@@ -0,0 +1,9 @@
+// RUN: %clang_cc1 %s -triple spir-unknown-unknown -O0 -emit-llvm -o - |
FileCheck %s
+
+// CHECK: @test.arr = private unnamed_addr addrspace(2) constant [3 x i32]
[i32 1, i32 2, i32 3], align 4
+
+void test() {
+ __private int arr[] = {1, 2, 3};
+// CHECK: %[[arr_i8_ptr:[0-9]+]] = bitcast [3 x i32]* %arr to i8*
+// CHECK: call void @llvm.memcpy.p0i8.p2i8.i32(i8* %[[arr_i8_ptr]], i8
addrspace(2)* bitcast ([3 x i32] addrspace(2)* @test.arr to i8 addrspace(2)*),
i32 12, i32 4, i1 false)
+}
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[PATCH] D25305: [OpenCL] Setting constant address space for array initializers
bader added a comment. Committed at 285557 with updated tests. https://reviews.llvm.org/D25305 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
r285700 - [OpenCL] Override supported OpenCL extensions with -cl-ext option
Author: bader
Date: Tue Nov 1 10:50:52 2016
New Revision: 285700
URL: http://llvm.org/viewvc/llvm-project?rev=285700&view=rev
Log:
[OpenCL] Override supported OpenCL extensions with -cl-ext option
Summary:
This patch adds a command line option '-cl-ext' to control a set of
supported OpenCL extensions. Option accepts a comma-separated list
of extensions prefixed with '+' or '-'.
It can be used together with a target triple to override support for some
extensions:
// spir target supports all extensions, but we want to disable fp64
clang -cc1 -triple spir-unknown-unknown -cl-ext=-cl_khr_fp64
Special 'all' extension allows to enable or disable all possible
extensions:
// only fp64 will be supported
clang -cc1 -triple spir-unknown-unknown -cl-ext=-all,+cl_khr_fp64
Patch by asavonic (Andrew Savonichev).
Reviewers: joey, yaxunl
Subscribers: yaxunl, bader, Anastasia, cfe-commits
Differential Revision: https://reviews.llvm.org/D23712
Modified:
cfe/trunk/include/clang/Basic/OpenCLOptions.h
cfe/trunk/include/clang/Basic/TargetInfo.h
cfe/trunk/include/clang/Basic/TargetOptions.h
cfe/trunk/include/clang/Driver/CC1Options.td
cfe/trunk/lib/Basic/Targets.cpp
cfe/trunk/lib/Frontend/CompilerInvocation.cpp
cfe/trunk/test/SemaOpenCL/extensions.cl
Modified: cfe/trunk/include/clang/Basic/OpenCLOptions.h
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/include/clang/Basic/OpenCLOptions.h?rev=285700&r1=285699&r2=285700&view=diff
==
--- cfe/trunk/include/clang/Basic/OpenCLOptions.h (original)
+++ cfe/trunk/include/clang/Basic/OpenCLOptions.h Tue Nov 1 10:50:52 2016
@@ -15,6 +15,8 @@
#ifndef LLVM_CLANG_BASIC_OPENCLOPTIONS_H
#define LLVM_CLANG_BASIC_OPENCLOPTIONS_H
+#include "llvm/ADT/StringRef.h"
+
namespace clang {
/// \brief OpenCL supported extensions and optional core features
@@ -28,9 +30,39 @@ public:
#include "clang/Basic/OpenCLExtensions.def"
}
- // Enable all options.
- void setAll() {
-#define OPENCLEXT(nm) nm = 1;
+ // Enable or disable all options.
+ void setAll(bool Enable = true) {
+#define OPENCLEXT(nm) nm = Enable;
+#include "clang/Basic/OpenCLExtensions.def"
+ }
+
+ /// \brief Enable or disable support for OpenCL extensions
+ /// \param Ext name of the extension optionally prefixed with
+ ///'+' or '-'
+ /// \param Enable used when \p Ext is not prefixed by '+' or '-'
+ void set(llvm::StringRef Ext, bool Enable = true) {
+assert(!Ext.empty() && "Extension is empty.");
+
+switch (Ext[0]) {
+case '+':
+ Enable = true;
+ Ext = Ext.drop_front();
+ break;
+case '-':
+ Enable = false;
+ Ext = Ext.drop_front();
+ break;
+}
+
+if (Ext.equals("all")) {
+ setAll(Enable);
+ return;
+}
+
+#define OPENCLEXT(nm) \
+if (Ext.equals(#nm)) { \
+ nm = Enable; \
+}
#include "clang/Basic/OpenCLExtensions.def"
}
Modified: cfe/trunk/include/clang/Basic/TargetInfo.h
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/include/clang/Basic/TargetInfo.h?rev=285700&r1=285699&r2=285700&view=diff
==
--- cfe/trunk/include/clang/Basic/TargetInfo.h (original)
+++ cfe/trunk/include/clang/Basic/TargetInfo.h Tue Nov 1 10:50:52 2016
@@ -995,6 +995,13 @@ public:
/// \brief Set supported OpenCL extensions and optional core features.
virtual void setSupportedOpenCLOpts() {}
+ /// \brief Set supported OpenCL extensions as written on command line
+ virtual void setOpenCLExtensionOpts() {
+for (const auto &Ext : getTargetOpts().OpenCLExtensionsAsWritten) {
+ getTargetOpts().SupportedOpenCLOptions.set(Ext);
+}
+ }
+
/// \brief Get supported OpenCL extensions and optional core features.
OpenCLOptions &getSupportedOpenCLOpts() {
return getTargetOpts().SupportedOpenCLOptions;
Modified: cfe/trunk/include/clang/Basic/TargetOptions.h
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/include/clang/Basic/TargetOptions.h?rev=285700&r1=285699&r2=285700&view=diff
==
--- cfe/trunk/include/clang/Basic/TargetOptions.h (original)
+++ cfe/trunk/include/clang/Basic/TargetOptions.h Tue Nov 1 10:50:52 2016
@@ -58,6 +58,10 @@ public:
/// Supported OpenCL extensions and optional core features.
OpenCLOptions SupportedOpenCLOptions;
+
+ /// \brief The list of OpenCL extensions to enable or disable, as written on
+ /// the command line.
+ std::vector OpenCLExtensionsAsWritten;
};
} // end namespace clang
Modified: cfe/trunk/include/clang/Driver/CC1Options.td
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/include/clang/Driver/CC1Options.td?rev=285700&r1=285699&r2=285700&view=diff
==
--- cfe/trunk/include/clang/Driver/CC1Options.td (o
[PATCH] D23712: [OpenCL] Override supported OpenCL extensions with -cl-ext option
This revision was automatically updated to reflect the committed changes.
Closed by commit rL285700: [OpenCL] Override supported OpenCL extensions with
-cl-ext option (authored by bader).
Changed prior to commit:
https://reviews.llvm.org/D23712?vs=75568&id=76571#toc
Repository:
rL LLVM
https://reviews.llvm.org/D23712
Files:
cfe/trunk/include/clang/Basic/OpenCLOptions.h
cfe/trunk/include/clang/Basic/TargetInfo.h
cfe/trunk/include/clang/Basic/TargetOptions.h
cfe/trunk/include/clang/Driver/CC1Options.td
cfe/trunk/lib/Basic/Targets.cpp
cfe/trunk/lib/Frontend/CompilerInvocation.cpp
cfe/trunk/test/SemaOpenCL/extensions.cl
Index: cfe/trunk/include/clang/Driver/CC1Options.td
===
--- cfe/trunk/include/clang/Driver/CC1Options.td
+++ cfe/trunk/include/clang/Driver/CC1Options.td
@@ -688,6 +688,13 @@
HelpText<"include a detailed record of preprocessing actions">;
//===--===//
+// OpenCL Options
+//===--===//
+
+def cl_ext_EQ : CommaJoined<["-"], "cl-ext=">,
+ HelpText<"OpenCL only. Enable or disable OpenCL extensions. The argument is a comma-separated sequence of one or more extension names, each prefixed by '+' or '-'.">;
+
+//===--===//
// CUDA Options
//===--===//
Index: cfe/trunk/include/clang/Basic/TargetInfo.h
===
--- cfe/trunk/include/clang/Basic/TargetInfo.h
+++ cfe/trunk/include/clang/Basic/TargetInfo.h
@@ -995,6 +995,13 @@
/// \brief Set supported OpenCL extensions and optional core features.
virtual void setSupportedOpenCLOpts() {}
+ /// \brief Set supported OpenCL extensions as written on command line
+ virtual void setOpenCLExtensionOpts() {
+for (const auto &Ext : getTargetOpts().OpenCLExtensionsAsWritten) {
+ getTargetOpts().SupportedOpenCLOptions.set(Ext);
+}
+ }
+
/// \brief Get supported OpenCL extensions and optional core features.
OpenCLOptions &getSupportedOpenCLOpts() {
return getTargetOpts().SupportedOpenCLOptions;
Index: cfe/trunk/include/clang/Basic/TargetOptions.h
===
--- cfe/trunk/include/clang/Basic/TargetOptions.h
+++ cfe/trunk/include/clang/Basic/TargetOptions.h
@@ -58,6 +58,10 @@
/// Supported OpenCL extensions and optional core features.
OpenCLOptions SupportedOpenCLOptions;
+
+ /// \brief The list of OpenCL extensions to enable or disable, as written on
+ /// the command line.
+ std::vector OpenCLExtensionsAsWritten;
};
} // end namespace clang
Index: cfe/trunk/include/clang/Basic/OpenCLOptions.h
===
--- cfe/trunk/include/clang/Basic/OpenCLOptions.h
+++ cfe/trunk/include/clang/Basic/OpenCLOptions.h
@@ -15,6 +15,8 @@
#ifndef LLVM_CLANG_BASIC_OPENCLOPTIONS_H
#define LLVM_CLANG_BASIC_OPENCLOPTIONS_H
+#include "llvm/ADT/StringRef.h"
+
namespace clang {
/// \brief OpenCL supported extensions and optional core features
@@ -28,9 +30,39 @@
#include "clang/Basic/OpenCLExtensions.def"
}
- // Enable all options.
- void setAll() {
-#define OPENCLEXT(nm) nm = 1;
+ // Enable or disable all options.
+ void setAll(bool Enable = true) {
+#define OPENCLEXT(nm) nm = Enable;
+#include "clang/Basic/OpenCLExtensions.def"
+ }
+
+ /// \brief Enable or disable support for OpenCL extensions
+ /// \param Ext name of the extension optionally prefixed with
+ ///'+' or '-'
+ /// \param Enable used when \p Ext is not prefixed by '+' or '-'
+ void set(llvm::StringRef Ext, bool Enable = true) {
+assert(!Ext.empty() && "Extension is empty.");
+
+switch (Ext[0]) {
+case '+':
+ Enable = true;
+ Ext = Ext.drop_front();
+ break;
+case '-':
+ Enable = false;
+ Ext = Ext.drop_front();
+ break;
+}
+
+if (Ext.equals("all")) {
+ setAll(Enable);
+ return;
+}
+
+#define OPENCLEXT(nm) \
+if (Ext.equals(#nm)) { \
+ nm = Enable; \
+}
#include "clang/Basic/OpenCLExtensions.def"
}
Index: cfe/trunk/test/SemaOpenCL/extensions.cl
===
--- cfe/trunk/test/SemaOpenCL/extensions.cl
+++ cfe/trunk/test/SemaOpenCL/extensions.cl
@@ -2,7 +2,26 @@
// RUN: %clang_cc1 %s -triple spir-unknown-unknown -verify -pedantic -fsyntax-only -cl-std=CL1.1
// Test with a target not supporting fp64.
-// RUN: %clang_cc1 %s -triple r600-unknown-unknown -target-cpu r600 -verify -pedantic -fsyntax-only -DNOFP64
+// RUN: %clang_cc1 %s -triple r600-unknown-unknown -target-cpu r600 -verify -pedantic -fsyntax-only -DNOFP64 -DNOFP16
+
+// Test with some extensions ena
[PATCH] D26157: [OpenCL] always use SPIR address spaces for kernel_arg_addr_space MD
bader added a comment. @pekka.jaaskelainen, I have related question: what is the problem with retaining OpenCL address space information in LLVM IR? My understanding is that target CodeGen can ignore this information for the machines with 'flat' address space model. On the other hand I would expect this information be useful for the Optimizer to resolve pointer aliasing. Does it make any sense to keep OpenCL address space information in LLVM IR generated for the machines with 'flat' address space (e.g. CPU)? Repository: rL LLVM https://reviews.llvm.org/D26157 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[PATCH] D12855: [OpenCL] Add new types for OpenCL 2.0.
bader created this revision. bader added a reviewer: pekka.jaaskelainen. bader added a subscriber: cfe-commits. Patch by Pedro Ferreira. This patch was reviewed here: http://lists.llvm.org/pipermail/cfe-commits/Week-of-Mon-20150629/thread.html#131992, but wasn't submitted into the tree. Rebased to rev. 247586. http://reviews.llvm.org/D12855 Files: include/clang/AST/ASTContext.h include/clang/AST/BuiltinTypes.def include/clang/AST/Type.h include/clang/Basic/OpenCLExtensions.def include/clang/Serialization/ASTBitCodes.h lib/AST/ASTContext.cpp lib/AST/ItaniumMangle.cpp lib/AST/MicrosoftMangle.cpp lib/AST/NSAPI.cpp lib/AST/Type.cpp lib/AST/TypeLoc.cpp lib/CodeGen/CGDebugInfo.cpp lib/CodeGen/CGDebugInfo.h lib/CodeGen/CGOpenCLRuntime.cpp lib/CodeGen/CodeGenTypes.cpp lib/CodeGen/ItaniumCXXABI.cpp lib/Index/USRGeneration.cpp lib/Sema/Sema.cpp lib/Sema/SemaType.cpp lib/Serialization/ASTCommon.cpp lib/Serialization/ASTReader.cpp test/PCH/ocl_types.h tools/libclang/CIndex.cpp Index: tools/libclang/CIndex.cpp === --- tools/libclang/CIndex.cpp +++ tools/libclang/CIndex.cpp @@ -1460,9 +1460,19 @@ case BuiltinType::OCLImage1dBuffer: case BuiltinType::OCLImage2d: case BuiltinType::OCLImage2dArray: + case BuiltinType::OCLImage2dDepth: + case BuiltinType::OCLImage2dArrayDepth: + case BuiltinType::OCLImage2dMSAA: + case BuiltinType::OCLImage2dArrayMSAA: + case BuiltinType::OCLImage2dMSAADepth: + case BuiltinType::OCLImage2dArrayMSAADepth: case BuiltinType::OCLImage3d: case BuiltinType::OCLSampler: case BuiltinType::OCLEvent: + case BuiltinType::OCLClkEvent: + case BuiltinType::OCLQueue: + case BuiltinType::OCLNDRange: + case BuiltinType::OCLReserveID: #define BUILTIN_TYPE(Id, SingletonId) #define SIGNED_TYPE(Id, SingletonId) case BuiltinType::Id: #define UNSIGNED_TYPE(Id, SingletonId) case BuiltinType::Id: Index: test/PCH/ocl_types.h === --- test/PCH/ocl_types.h +++ test/PCH/ocl_types.h @@ -23,3 +23,37 @@ // event_t typedef event_t evt_t; + +#if __OPENCL_VERSION__ >= 200 + +// clk_event_t +typedef clk_event_t clkevt_t; + +// queue_t +typedef queue_t q_t; + +// ndrange_t +typedef ndrange_t range_t; + +// reserve_id_t +typedef reserve_id_t reserveid_t; + +// image2d_depth_t +typedef image2d_depth_t img2ddep_t; + +// image2d_array_depth_t +typedef image2d_array_depth_t img2darr_dep_t; + +// image2d_msaa_t +typedef image2d_msaa_t img2dmsaa_t; + +// image2d_array_msaa_t +typedef image2d_array_msaa_t img2darrmsaa_t; + +// image2d_msaa_depth_t +typedef image2d_msaa_depth_t img2dmsaadep_t; + +// image2d_array_msaa_depth_t +typedef image2d_array_msaa_depth_t img2darrmsaadep_t; + +#endif Index: lib/Serialization/ASTReader.cpp === --- lib/Serialization/ASTReader.cpp +++ lib/Serialization/ASTReader.cpp @@ -5824,9 +5824,39 @@ case PREDEF_TYPE_IMAGE1D_BUFF_ID: T = Context.OCLImage1dBufferTy; break; case PREDEF_TYPE_IMAGE2D_ID:T = Context.OCLImage2dTy; break; case PREDEF_TYPE_IMAGE2D_ARR_ID: T = Context.OCLImage2dArrayTy; break; +case PREDEF_TYPE_IMAGE2D_DEP_ID: + T = Context.OCLImage2dDepthTy; + break; +case PREDEF_TYPE_IMAGE2D_ARR_DEP_ID: + T = Context.OCLImage2dArrayDepthTy; + break; +case PREDEF_TYPE_IMAGE2D_MSAA_ID: + T = Context.OCLImage2dMSAATy; + break; +case PREDEF_TYPE_IMAGE2D_ARR_MSAA_ID: + T = Context.OCLImage2dArrayMSAATy; + break; +case PREDEF_TYPE_IMAGE2D_MSAA_DEP_ID: + T = Context.OCLImage2dMSAADepthTy; + break; +case PREDEF_TYPE_IMAGE2D_ARR_MSAA_DEPTH_ID: + T = Context.OCLImage2dArrayMSAADepthTy; + break; case PREDEF_TYPE_IMAGE3D_ID:T = Context.OCLImage3dTy; break; case PREDEF_TYPE_SAMPLER_ID:T = Context.OCLSamplerTy; break; case PREDEF_TYPE_EVENT_ID: T = Context.OCLEventTy; break; +case PREDEF_TYPE_CLK_EVENT_ID: + T = Context.OCLClkEventTy; + break; +case PREDEF_TYPE_QUEUE_ID: + T = Context.OCLQueueTy; + break; +case PREDEF_TYPE_NDRANGE_ID: + T = Context.OCLNDRangeTy; + break; +case PREDEF_TYPE_RESERVE_ID_ID: + T = Context.OCLReserveIDTy; + break; case PREDEF_TYPE_AUTO_DEDUCT: T = Context.getAutoDeductType(); break; case PREDEF_TYPE_AUTO_RREF_DEDUCT: Index: lib/Serialization/ASTCommon.cpp === --- lib/Serialization/ASTCommon.cpp +++ lib/Serialization/ASTCommon.cpp @@ -67,9 +67,39 @@ case BuiltinType::OCLImage1dBuffer: ID = PREDEF_TYPE_IMAGE1D_BUFF_ID; break; case BuiltinType::OCLImage2d: ID = PREDEF_TYPE_IMAGE2D_ID; break; case BuiltinType::OCLImage2dArray: ID = PREDEF_TYPE_IMAGE2D_ARR_ID; break; + case BuiltinType::OCLImage2dDepth
Re: [PATCH] D12855: [OpenCL] Add new types for OpenCL 2.0.
bader added a comment. In http://reviews.llvm.org/D12855#246004, @pekka.jaaskelainen wrote: > The patch seems straightforward enough. BTW does someone know if anyone has > worked on the 'pipe' qualifier? I'm currently supporting Intel's implementation of 'pipe' qualifier and I'd like to contribute this implementation to the community if there is any interest. Comment at: lib/Serialization/ASTCommon.cpp:70 @@ -69,1 +69,3 @@ case BuiltinType::OCLImage2dArray: ID = PREDEF_TYPE_IMAGE2D_ARR_ID; break; + case BuiltinType::OCLImage2dDepth: +ID = PREDEF_TYPE_IMAGE2D_DEP_ID; pekka.jaaskelainen wrote: > A nitpick, but why deviate from the style where each case has been stuffed to > a single line? My guess it's effect of clang-format. Original patch sent by Pedro matched the style, but violated 80-character limit, so he was asked to run clang-format over the patch. Do you want me to run clang-format on surrounding code to unify the style? http://reviews.llvm.org/D12855 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
r247676 - [OpenCL] Add new types for OpenCL 2.0.
Author: bader Date: Tue Sep 15 06:18:52 2015 New Revision: 247676 URL: http://llvm.org/viewvc/llvm-project?rev=247676&view=rev Log: [OpenCL] Add new types for OpenCL 2.0. Patch by Pedro Ferreira. Reviewers: pekka.jaaskelainen Differential Revision: http://reviews.llvm.org/D12855 Modified: cfe/trunk/include/clang/AST/ASTContext.h cfe/trunk/include/clang/AST/BuiltinTypes.def cfe/trunk/include/clang/AST/Type.h cfe/trunk/include/clang/Basic/OpenCLExtensions.def cfe/trunk/include/clang/Serialization/ASTBitCodes.h cfe/trunk/lib/AST/ASTContext.cpp cfe/trunk/lib/AST/ItaniumMangle.cpp cfe/trunk/lib/AST/MicrosoftMangle.cpp cfe/trunk/lib/AST/NSAPI.cpp cfe/trunk/lib/AST/Type.cpp cfe/trunk/lib/AST/TypeLoc.cpp cfe/trunk/lib/CodeGen/CGDebugInfo.cpp cfe/trunk/lib/CodeGen/CGDebugInfo.h cfe/trunk/lib/CodeGen/CGOpenCLRuntime.cpp cfe/trunk/lib/CodeGen/CodeGenTypes.cpp cfe/trunk/lib/CodeGen/ItaniumCXXABI.cpp cfe/trunk/lib/Index/USRGeneration.cpp cfe/trunk/lib/Sema/Sema.cpp cfe/trunk/lib/Sema/SemaType.cpp cfe/trunk/lib/Serialization/ASTCommon.cpp cfe/trunk/lib/Serialization/ASTReader.cpp cfe/trunk/test/PCH/ocl_types.h cfe/trunk/tools/libclang/CIndex.cpp Modified: cfe/trunk/include/clang/AST/ASTContext.h URL: http://llvm.org/viewvc/llvm-project/cfe/trunk/include/clang/AST/ASTContext.h?rev=247676&r1=247675&r2=247676&view=diff == --- cfe/trunk/include/clang/AST/ASTContext.h (original) +++ cfe/trunk/include/clang/AST/ASTContext.h Tue Sep 15 06:18:52 2015 @@ -839,9 +839,12 @@ public: CanQualType ObjCBuiltinIdTy, ObjCBuiltinClassTy, ObjCBuiltinSelTy; CanQualType ObjCBuiltinBoolTy; CanQualType OCLImage1dTy, OCLImage1dArrayTy, OCLImage1dBufferTy; - CanQualType OCLImage2dTy, OCLImage2dArrayTy; + CanQualType OCLImage2dTy, OCLImage2dArrayTy, OCLImage2dDepthTy; + CanQualType OCLImage2dArrayDepthTy, OCLImage2dMSAATy, OCLImage2dArrayMSAATy; + CanQualType OCLImage2dMSAADepthTy, OCLImage2dArrayMSAADepthTy; CanQualType OCLImage3dTy; - CanQualType OCLSamplerTy, OCLEventTy; + CanQualType OCLSamplerTy, OCLEventTy, OCLClkEventTy; + CanQualType OCLQueueTy, OCLNDRangeTy, OCLReserveIDTy; CanQualType OMPArraySectionTy; // Types for deductions in C++0x [stmt.ranged]'s desugaring. Built on demand. Modified: cfe/trunk/include/clang/AST/BuiltinTypes.def URL: http://llvm.org/viewvc/llvm-project/cfe/trunk/include/clang/AST/BuiltinTypes.def?rev=247676&r1=247675&r2=247676&view=diff == --- cfe/trunk/include/clang/AST/BuiltinTypes.def (original) +++ cfe/trunk/include/clang/AST/BuiltinTypes.def Tue Sep 15 06:18:52 2015 @@ -160,6 +160,12 @@ BUILTIN_TYPE(OCLImage1dArray, OCLImage1d BUILTIN_TYPE(OCLImage1dBuffer, OCLImage1dBufferTy) BUILTIN_TYPE(OCLImage2d, OCLImage2dTy) BUILTIN_TYPE(OCLImage2dArray, OCLImage2dArrayTy) +BUILTIN_TYPE(OCLImage2dDepth, OCLImage2dDepthTy) +BUILTIN_TYPE(OCLImage2dArrayDepth, OCLImage2dArrayDepthTy) +BUILTIN_TYPE(OCLImage2dMSAA, OCLImage2dMSAATy) +BUILTIN_TYPE(OCLImage2dArrayMSAA, OCLImage2dArrayMSAATy) +BUILTIN_TYPE(OCLImage2dMSAADepth, OCLImage2dMSAADepthTy) +BUILTIN_TYPE(OCLImage2dArrayMSAADepth, OCLImage2dArrayMSAADepthTy) BUILTIN_TYPE(OCLImage3d, OCLImage3dTy) // OpenCL sampler_t. @@ -168,6 +174,18 @@ BUILTIN_TYPE(OCLSampler, OCLSamplerTy) // OpenCL event_t. BUILTIN_TYPE(OCLEvent, OCLEventTy) +// OpenCL clk_event_t. +BUILTIN_TYPE(OCLClkEvent, OCLClkEventTy) + +// OpenCL queue_t. +BUILTIN_TYPE(OCLQueue, OCLQueueTy) + +// OpenCL ndrange_t. +BUILTIN_TYPE(OCLNDRange, OCLNDRangeTy) + +// OpenCL reserve_id_t. +BUILTIN_TYPE(OCLReserveID, OCLReserveIDTy) + // This represents the type of an expression whose type is // totally unknown, e.g. 'T::foo'. It is permitted for this to // appear in situations where the structure of the type is Modified: cfe/trunk/include/clang/AST/Type.h URL: http://llvm.org/viewvc/llvm-project/cfe/trunk/include/clang/AST/Type.h?rev=247676&r1=247675&r2=247676&view=diff == --- cfe/trunk/include/clang/AST/Type.h (original) +++ cfe/trunk/include/clang/AST/Type.h Tue Sep 15 06:18:52 2015 @@ -1683,17 +1683,27 @@ public: bool isNullPtrType() const; // C++0x nullptr_t bool isAtomicType() const;// C11 _Atomic() - bool isImage1dT() const; // OpenCL image1d_t - bool isImage1dArrayT() const; // OpenCL image1d_array_t - bool isImage1dBufferT() const;// OpenCL image1d_buffer_t - bool isImage2dT() const; // OpenCL image2d_t - bool isImage2dArrayT() const; // OpenCL image2d_array_t - bool isImage3dT() const; // OpenCL image3d_t + bool isImage1dT() const; // OpenCL image1d_t + bool isImage1dArr
Re: [PATCH] D12855: [OpenCL] Add new types for OpenCL 2.0.
This revision was automatically updated to reflect the committed changes. Closed by commit rL247676: [OpenCL] Add new types for OpenCL 2.0. (authored by bader). Changed prior to commit: http://reviews.llvm.org/D12855?vs=34709&id=34796#toc Repository: rL LLVM http://reviews.llvm.org/D12855 Files: cfe/trunk/include/clang/AST/ASTContext.h cfe/trunk/include/clang/AST/BuiltinTypes.def cfe/trunk/include/clang/AST/Type.h cfe/trunk/include/clang/Basic/OpenCLExtensions.def cfe/trunk/include/clang/Serialization/ASTBitCodes.h cfe/trunk/lib/AST/ASTContext.cpp cfe/trunk/lib/AST/ItaniumMangle.cpp cfe/trunk/lib/AST/MicrosoftMangle.cpp cfe/trunk/lib/AST/NSAPI.cpp cfe/trunk/lib/AST/Type.cpp cfe/trunk/lib/AST/TypeLoc.cpp cfe/trunk/lib/CodeGen/CGDebugInfo.cpp cfe/trunk/lib/CodeGen/CGDebugInfo.h cfe/trunk/lib/CodeGen/CGOpenCLRuntime.cpp cfe/trunk/lib/CodeGen/CodeGenTypes.cpp cfe/trunk/lib/CodeGen/ItaniumCXXABI.cpp cfe/trunk/lib/Index/USRGeneration.cpp cfe/trunk/lib/Sema/Sema.cpp cfe/trunk/lib/Sema/SemaType.cpp cfe/trunk/lib/Serialization/ASTCommon.cpp cfe/trunk/lib/Serialization/ASTReader.cpp cfe/trunk/test/PCH/ocl_types.h cfe/trunk/tools/libclang/CIndex.cpp Index: cfe/trunk/tools/libclang/CIndex.cpp === --- cfe/trunk/tools/libclang/CIndex.cpp +++ cfe/trunk/tools/libclang/CIndex.cpp @@ -1460,9 +1460,19 @@ case BuiltinType::OCLImage1dBuffer: case BuiltinType::OCLImage2d: case BuiltinType::OCLImage2dArray: + case BuiltinType::OCLImage2dDepth: + case BuiltinType::OCLImage2dArrayDepth: + case BuiltinType::OCLImage2dMSAA: + case BuiltinType::OCLImage2dArrayMSAA: + case BuiltinType::OCLImage2dMSAADepth: + case BuiltinType::OCLImage2dArrayMSAADepth: case BuiltinType::OCLImage3d: case BuiltinType::OCLSampler: case BuiltinType::OCLEvent: + case BuiltinType::OCLClkEvent: + case BuiltinType::OCLQueue: + case BuiltinType::OCLNDRange: + case BuiltinType::OCLReserveID: #define BUILTIN_TYPE(Id, SingletonId) #define SIGNED_TYPE(Id, SingletonId) case BuiltinType::Id: #define UNSIGNED_TYPE(Id, SingletonId) case BuiltinType::Id: Index: cfe/trunk/include/clang/AST/BuiltinTypes.def === --- cfe/trunk/include/clang/AST/BuiltinTypes.def +++ cfe/trunk/include/clang/AST/BuiltinTypes.def @@ -160,14 +160,32 @@ BUILTIN_TYPE(OCLImage1dBuffer, OCLImage1dBufferTy) BUILTIN_TYPE(OCLImage2d, OCLImage2dTy) BUILTIN_TYPE(OCLImage2dArray, OCLImage2dArrayTy) +BUILTIN_TYPE(OCLImage2dDepth, OCLImage2dDepthTy) +BUILTIN_TYPE(OCLImage2dArrayDepth, OCLImage2dArrayDepthTy) +BUILTIN_TYPE(OCLImage2dMSAA, OCLImage2dMSAATy) +BUILTIN_TYPE(OCLImage2dArrayMSAA, OCLImage2dArrayMSAATy) +BUILTIN_TYPE(OCLImage2dMSAADepth, OCLImage2dMSAADepthTy) +BUILTIN_TYPE(OCLImage2dArrayMSAADepth, OCLImage2dArrayMSAADepthTy) BUILTIN_TYPE(OCLImage3d, OCLImage3dTy) // OpenCL sampler_t. BUILTIN_TYPE(OCLSampler, OCLSamplerTy) // OpenCL event_t. BUILTIN_TYPE(OCLEvent, OCLEventTy) +// OpenCL clk_event_t. +BUILTIN_TYPE(OCLClkEvent, OCLClkEventTy) + +// OpenCL queue_t. +BUILTIN_TYPE(OCLQueue, OCLQueueTy) + +// OpenCL ndrange_t. +BUILTIN_TYPE(OCLNDRange, OCLNDRangeTy) + +// OpenCL reserve_id_t. +BUILTIN_TYPE(OCLReserveID, OCLReserveIDTy) + // This represents the type of an expression whose type is // totally unknown, e.g. 'T::foo'. It is permitted for this to // appear in situations where the structure of the type is Index: cfe/trunk/include/clang/AST/Type.h === --- cfe/trunk/include/clang/AST/Type.h +++ cfe/trunk/include/clang/AST/Type.h @@ -1683,17 +1683,27 @@ bool isNullPtrType() const; // C++0x nullptr_t bool isAtomicType() const;// C11 _Atomic() - bool isImage1dT() const; // OpenCL image1d_t - bool isImage1dArrayT() const; // OpenCL image1d_array_t - bool isImage1dBufferT() const;// OpenCL image1d_buffer_t - bool isImage2dT() const; // OpenCL image2d_t - bool isImage2dArrayT() const; // OpenCL image2d_array_t - bool isImage3dT() const; // OpenCL image3d_t + bool isImage1dT() const; // OpenCL image1d_t + bool isImage1dArrayT() const; // OpenCL image1d_array_t + bool isImage1dBufferT() const; // OpenCL image1d_buffer_t + bool isImage2dT() const; // OpenCL image2d_t + bool isImage2dArrayT() const; // OpenCL image2d_array_t + bool isImage2dDepthT() const; // OpenCL image_2d_depth_t + bool isImage2dArrayDepthT() const; // OpenCL image_2d_array_depth_t + bool isImage2dMSAAT() const; // OpenCL image_2d_msaa_t + bool isImage2dArrayMSAAT() const; // OpenCL image_2d_array_msaa_t + bool isImage2dMSAATDepth() const; // OpenCL image_2d_msaa_depth_t + bool isImage2dA
r247678 - Run clang-format to unify the switch statement style as suggest here: http://reviews.llvm.org/D12855#246073.
Author: bader
Date: Tue Sep 15 07:18:29 2015
New Revision: 247678
URL: http://llvm.org/viewvc/llvm-project?rev=247678&view=rev
Log:
Run clang-format to unify the switch statement style as suggest here:
http://reviews.llvm.org/D12855#246073.
NFC.
Modified:
cfe/trunk/lib/AST/ItaniumMangle.cpp
cfe/trunk/lib/AST/MicrosoftMangle.cpp
cfe/trunk/lib/AST/Type.cpp
cfe/trunk/lib/Serialization/ASTCommon.cpp
cfe/trunk/lib/Serialization/ASTReader.cpp
Modified: cfe/trunk/lib/AST/ItaniumMangle.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/AST/ItaniumMangle.cpp?rev=247678&r1=247677&r2=247678&view=diff
==
--- cfe/trunk/lib/AST/ItaniumMangle.cpp (original)
+++ cfe/trunk/lib/AST/ItaniumMangle.cpp Tue Sep 15 07:18:29 2015
@@ -1985,34 +1985,79 @@ void CXXNameMangler::mangleType(const Bu
// ::= Dn # std::nullptr_t (i.e., decltype(nullptr))
// ::= u # vendor extended type
switch (T->getKind()) {
- case BuiltinType::Void: Out << 'v'; break;
- case BuiltinType::Bool: Out << 'b'; break;
- case BuiltinType::Char_U: case BuiltinType::Char_S: Out << 'c'; break;
- case BuiltinType::UChar: Out << 'h'; break;
- case BuiltinType::UShort: Out << 't'; break;
- case BuiltinType::UInt: Out << 'j'; break;
- case BuiltinType::ULong: Out << 'm'; break;
- case BuiltinType::ULongLong: Out << 'y'; break;
- case BuiltinType::UInt128: Out << 'o'; break;
- case BuiltinType::SChar: Out << 'a'; break;
+ case BuiltinType::Void:
+Out << 'v';
+break;
+ case BuiltinType::Bool:
+Out << 'b';
+break;
+ case BuiltinType::Char_U:
+ case BuiltinType::Char_S:
+Out << 'c';
+break;
+ case BuiltinType::UChar:
+Out << 'h';
+break;
+ case BuiltinType::UShort:
+Out << 't';
+break;
+ case BuiltinType::UInt:
+Out << 'j';
+break;
+ case BuiltinType::ULong:
+Out << 'm';
+break;
+ case BuiltinType::ULongLong:
+Out << 'y';
+break;
+ case BuiltinType::UInt128:
+Out << 'o';
+break;
+ case BuiltinType::SChar:
+Out << 'a';
+break;
case BuiltinType::WChar_S:
- case BuiltinType::WChar_U: Out << 'w'; break;
- case BuiltinType::Char16: Out << "Ds"; break;
- case BuiltinType::Char32: Out << "Di"; break;
- case BuiltinType::Short: Out << 's'; break;
- case BuiltinType::Int: Out << 'i'; break;
- case BuiltinType::Long: Out << 'l'; break;
- case BuiltinType::LongLong: Out << 'x'; break;
- case BuiltinType::Int128: Out << 'n'; break;
- case BuiltinType::Half: Out << "Dh"; break;
- case BuiltinType::Float: Out << 'f'; break;
- case BuiltinType::Double: Out << 'd'; break;
+ case BuiltinType::WChar_U:
+Out << 'w';
+break;
+ case BuiltinType::Char16:
+Out << "Ds";
+break;
+ case BuiltinType::Char32:
+Out << "Di";
+break;
+ case BuiltinType::Short:
+Out << 's';
+break;
+ case BuiltinType::Int:
+Out << 'i';
+break;
+ case BuiltinType::Long:
+Out << 'l';
+break;
+ case BuiltinType::LongLong:
+Out << 'x';
+break;
+ case BuiltinType::Int128:
+Out << 'n';
+break;
+ case BuiltinType::Half:
+Out << "Dh";
+break;
+ case BuiltinType::Float:
+Out << 'f';
+break;
+ case BuiltinType::Double:
+Out << 'd';
+break;
case BuiltinType::LongDouble:
Out << (getASTContext().getTargetInfo().useFloat128ManglingForLongDouble()
? 'g'
: 'e');
break;
- case BuiltinType::NullPtr: Out << "Dn"; break;
+ case BuiltinType::NullPtr:
+Out << "Dn";
+break;
#define BUILTIN_TYPE(Id, SingletonId)
#define PLACEHOLDER_TYPE(Id, SingletonId) \
@@ -2020,14 +2065,30 @@ void CXXNameMangler::mangleType(const Bu
#include "clang/AST/BuiltinTypes.def"
case BuiltinType::Dependent:
llvm_unreachable("mangling a placeholder type");
- case BuiltinType::ObjCId: Out << "11objc_object"; break;
- case BuiltinType::ObjCClass: Out << "10objc_class"; break;
- case BuiltinType::ObjCSel: Out << "13objc_selector"; break;
- case BuiltinType::OCLImage1d: Out << "11ocl_image1d"; break;
- case BuiltinType::OCLImage1dArray: Out << "16ocl_image1darray"; break;
- case BuiltinType::OCLImage1dBuffer: Out << "17ocl_image1dbuffer"; break;
- case BuiltinType::OCLImage2d: Out << "11ocl_image2d"; break;
- case BuiltinType::OCLImage2dArray: Out << "16ocl_image2darray"; break;
+ case BuiltinType::ObjCId:
+Out << "11objc_object";
+break;
+ case BuiltinType::ObjCClass:
+Out << "10objc_class";
+break;
+ case BuiltinType::ObjCSel:
+Out << "13objc_selector";
+break;
+ case BuiltinType::OCLImage1d:
+Out << "11ocl_image1d";
+break;
+ case BuiltinType::OCLImage1dArray:
+Out << "16ocl_image1darray";
+break;
+ case BuiltinType::OCLImage1dBuffer:
+Out << "17ocl_image1dbuffer";
+break;
+ case BuiltinType::OCLImage2d:
+Out << "11ocl_image2d";
+break;
+ case BuiltinType::OCLImage2
Re: [PATCH] D13105: [OpenCL] Enable program scope variables for OpenCL2.0
bader added a subscriber: bader.
Comment at: include/clang/Basic/DiagnosticSemaKinds.td:7472-7473
@@ -7469,4 +7471,4 @@
"sampler type cannot be used with the __local and __global address space
qualifiers">;
def err_opencl_global_invalid_addr_space : Error<
- "global variables must have a constant address space qualifier">;
+ "program scope variables has wrong address space qualifier">;
def err_opencl_no_main : Error<"%select{function|kernel}0 cannot be called
'main'">;
I think we can do better diagnostics. Original message hinted on way to fix
that error - add 'constant' qualifier.
In our implementation we have 4 kinds of diagnostics (probably we need only 2
of them):
def err_opencl_global_invalid_addr_space : Error<
"program scope variables must reside in constant address space">;
def err_opencl20_global_invalid_addr_space : Error<
"program scope variables must reside in global or constant address space">;
def err_program_scope_variable_non_constant : Error<
"program scope variables are required to be declared in constant address
space">;
def err_program_scope_variable_non_constant_or_global : Error<
"program scope variables are required to be declared either in constant or
global address space">;
They used for OpenCL 1.2 and OpenCL 2.0 sources correspondingly.
Comment at: test/SemaOpenCL/storageclass.cl:9-11
@@ +8,5 @@
+global int G2 = 0;
+#ifndef CL20
+// expected-error@-2{{program scope variables has wrong address space
qualifier}}
+#endif
+
Don't you think it's better to put OpenCL 2.0 test into separate file?
http://reviews.llvm.org/D13105
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[PATCH] D12378: [OpenCL] Improve diagnostics detecting implicit vector conversion.
bader created this revision.
bader added a subscriber: cfe-commits.
http://reviews.llvm.org/D12378
Files:
include/clang/Basic/DiagnosticSemaKinds.td
lib/Sema/SemaExpr.cpp
test/SemaOpenCL/cond.cl
Index: test/SemaOpenCL/cond.cl
===
--- test/SemaOpenCL/cond.cl
+++ test/SemaOpenCL/cond.cl
@@ -84,7 +84,7 @@
float2 ntest04(int2 C, int2 X, float2 Y)
{
- return C ? X : Y; // expected-error {{cannot convert between vector values
of different size ('int2' (vector of 2 'int' values) and 'float2' (vector of 2
'float' values))}}
+ return C ? X : Y; // expected-error {{implicit conversions between vector
types ('int2' (vector of 2 'int' values) and 'float2' (vector of 2 'float'
values)) are not permitted}}
}
float2 ntest05(int2 C, int2 X, float Y)
@@ -115,7 +115,7 @@
char3 ntest10(char C, char3 X, char2 Y)
{
- return C ? X : Y; // expected-error {{cannot convert between vector values
of different size ('char3' (vector of 3 'char' values) and 'char2' (vector of 2
'char' values))}}
+ return C ? X : Y; // expected-error {{implicit conversions between vector
types ('char3' (vector of 3 'char' values) and 'char2' (vector of 2 'char'
values)) are not permitted}}
}
char3 ntest11(char2 C, char3 X, char Y)
Index: lib/Sema/SemaExpr.cpp
===
--- lib/Sema/SemaExpr.cpp
+++ lib/Sema/SemaExpr.cpp
@@ -7534,6 +7534,19 @@
return QualType();
}
+ // OpenCL V1.1 6.2.6.p1:
+ // If the operands are of more than one vector type, then an error shall
+ // occur. Implicit conversions between vector types are not permitted, per
+ // section 6.2.1.
+ if (getLangOpts().OpenCL &&
+ RHSVecType && isa(RHSVecType) &&
+ LHSVecType && isa(LHSVecType)) {
+assert(!Context.hasSameType(LHSType, RHSType));
+Diag(Loc, diag::err_opencl_implicit_vector_conversion) << LHSType
+ << RHSType;
+return QualType();
+ }
+
// Otherwise, use the generic diagnostic.
Diag(Loc, diag::err_typecheck_vector_not_convertable)
<< LHSType << RHSType
Index: include/clang/Basic/DiagnosticSemaKinds.td
===
--- include/clang/Basic/DiagnosticSemaKinds.td
+++ include/clang/Basic/DiagnosticSemaKinds.td
@@ -7445,6 +7445,8 @@
"return value cannot be qualified with address space">;
def err_opencl_constant_no_init : Error<
"variable in constant address space must be initialized">;
+def err_opencl_implicit_vector_conversion : Error<
+ "implicit conversions between vector types (%0 and %1) are not permitted">;
} // end of sema category
let CategoryName = "OpenMP Issue" in {
Index: test/SemaOpenCL/cond.cl
===
--- test/SemaOpenCL/cond.cl
+++ test/SemaOpenCL/cond.cl
@@ -84,7 +84,7 @@
float2 ntest04(int2 C, int2 X, float2 Y)
{
- return C ? X : Y; // expected-error {{cannot convert between vector values of different size ('int2' (vector of 2 'int' values) and 'float2' (vector of 2 'float' values))}}
+ return C ? X : Y; // expected-error {{implicit conversions between vector types ('int2' (vector of 2 'int' values) and 'float2' (vector of 2 'float' values)) are not permitted}}
}
float2 ntest05(int2 C, int2 X, float Y)
@@ -115,7 +115,7 @@
char3 ntest10(char C, char3 X, char2 Y)
{
- return C ? X : Y; // expected-error {{cannot convert between vector values of different size ('char3' (vector of 3 'char' values) and 'char2' (vector of 2 'char' values))}}
+ return C ? X : Y; // expected-error {{implicit conversions between vector types ('char3' (vector of 3 'char' values) and 'char2' (vector of 2 'char' values)) are not permitted}}
}
char3 ntest11(char2 C, char3 X, char Y)
Index: lib/Sema/SemaExpr.cpp
===
--- lib/Sema/SemaExpr.cpp
+++ lib/Sema/SemaExpr.cpp
@@ -7534,6 +7534,19 @@
return QualType();
}
+ // OpenCL V1.1 6.2.6.p1:
+ // If the operands are of more than one vector type, then an error shall
+ // occur. Implicit conversions between vector types are not permitted, per
+ // section 6.2.1.
+ if (getLangOpts().OpenCL &&
+ RHSVecType && isa(RHSVecType) &&
+ LHSVecType && isa(LHSVecType)) {
+assert(!Context.hasSameType(LHSType, RHSType));
+Diag(Loc, diag::err_opencl_implicit_vector_conversion) << LHSType
+ << RHSType;
+return QualType();
+ }
+
// Otherwise, use the generic diagnostic.
Diag(Loc, diag::err_typecheck_vector_not_convertable)
<< LHSType << RHSType
Index: include/clang/Basic/DiagnosticSemaKinds.td
===
--- include/clang/Basic/DiagnosticSemaKinds.td
+++ include/clang/Basic/DiagnosticSemaKinds.td
@@ -7445,6 +7445,8 @@
"return value cannot
r246393 - [OpenCL] Improve diagnostics detecting implicit vector conversion.
Author: bader
Date: Sun Aug 30 13:06:39 2015
New Revision: 246393
URL: http://llvm.org/viewvc/llvm-project?rev=246393&view=rev
Log:
[OpenCL] Improve diagnostics detecting implicit vector conversion.
Reviewers: pekka.jaaskelainen
Differential Revision: http://reviews.llvm.org/D12470
Modified:
cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td
cfe/trunk/lib/Sema/SemaExpr.cpp
cfe/trunk/test/SemaOpenCL/cond.cl
Modified: cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td?rev=246393&r1=246392&r2=246393&view=diff
==
--- cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td (original)
+++ cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td Sun Aug 30 13:06:39
2015
@@ -7445,6 +7445,8 @@ def err_opencl_return_value_with_address
"return value cannot be qualified with address space">;
def err_opencl_constant_no_init : Error<
"variable in constant address space must be initialized">;
+def err_opencl_implicit_vector_conversion : Error<
+ "implicit conversions between vector types (%0 and %1) are not permitted">;
} // end of sema category
let CategoryName = "OpenMP Issue" in {
Modified: cfe/trunk/lib/Sema/SemaExpr.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Sema/SemaExpr.cpp?rev=246393&r1=246392&r2=246393&view=diff
==
--- cfe/trunk/lib/Sema/SemaExpr.cpp (original)
+++ cfe/trunk/lib/Sema/SemaExpr.cpp Sun Aug 30 13:06:39 2015
@@ -7534,6 +7534,18 @@ QualType Sema::CheckVectorOperands(ExprR
return QualType();
}
+ // OpenCL V1.1 6.2.6.p1:
+ // If the operands are of more than one vector type, then an error shall
+ // occur. Implicit conversions between vector types are not permitted, per
+ // section 6.2.1.
+ if (getLangOpts().OpenCL &&
+ RHSVecType && isa(RHSVecType) &&
+ LHSVecType && isa(LHSVecType)) {
+Diag(Loc, diag::err_opencl_implicit_vector_conversion) << LHSType
+ << RHSType;
+return QualType();
+ }
+
// Otherwise, use the generic diagnostic.
Diag(Loc, diag::err_typecheck_vector_not_convertable)
<< LHSType << RHSType
Modified: cfe/trunk/test/SemaOpenCL/cond.cl
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/SemaOpenCL/cond.cl?rev=246393&r1=246392&r2=246393&view=diff
==
--- cfe/trunk/test/SemaOpenCL/cond.cl (original)
+++ cfe/trunk/test/SemaOpenCL/cond.cl Sun Aug 30 13:06:39 2015
@@ -84,7 +84,7 @@ uchar2 ntest03(int2 C, uchar X, uchar Y)
float2 ntest04(int2 C, int2 X, float2 Y)
{
- return C ? X : Y; // expected-error {{cannot convert between vector values
of different size ('int2' (vector of 2 'int' values) and 'float2' (vector of 2
'float' values))}}
+ return C ? X : Y; // expected-error {{implicit conversions between vector
types ('int2' (vector of 2 'int' values) and 'float2' (vector of 2 'float'
values)) are not permitted}}
}
float2 ntest05(int2 C, int2 X, float Y)
@@ -115,7 +115,7 @@ int2 ntest09(int2 C, global int *X, glob
char3 ntest10(char C, char3 X, char2 Y)
{
- return C ? X : Y; // expected-error {{cannot convert between vector values
of different size ('char3' (vector of 3 'char' values) and 'char2' (vector of 2
'char' values))}}
+ return C ? X : Y; // expected-error {{implicit conversions between vector
types ('char3' (vector of 3 'char' values) and 'char2' (vector of 2 'char'
values)) are not permitted}}
}
char3 ntest11(char2 C, char3 X, char Y)
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: [PATCH] D12378: [OpenCL] Improve diagnostics detecting implicit vector conversion.
bader closed this revision. bader added a comment. Thanks! Committed at rev. 246393. http://reviews.llvm.org/D12378 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[PATCH] D14441: [OpenCL] Pipe types support.
bader created this revision.
bader added reviewers: pekka.jaaskelainen, gbenyei.
bader added a subscriber: cfe-commits.
Initial support for OpenCL 2.0 feature: pipe types.
http://reviews.llvm.org/D14441
Files:
include/clang/AST/ASTContext.h
include/clang/AST/DataRecursiveASTVisitor.h
include/clang/AST/RecursiveASTVisitor.h
include/clang/AST/Type.h
include/clang/AST/TypeLoc.h
include/clang/AST/TypeNodes.def
include/clang/Basic/DiagnosticSemaKinds.td
include/clang/Basic/Specifiers.h
include/clang/Basic/TokenKinds.def
include/clang/Sema/DeclSpec.h
include/clang/Sema/Sema.h
include/clang/Serialization/ASTBitCodes.h
lib/AST/ASTContext.cpp
lib/AST/ASTImporter.cpp
lib/AST/ItaniumMangle.cpp
lib/AST/MicrosoftMangle.cpp
lib/AST/Type.cpp
lib/AST/TypePrinter.cpp
lib/CodeGen/CGDebugInfo.cpp
lib/CodeGen/CGDebugInfo.h
lib/CodeGen/CodeGenFunction.cpp
lib/CodeGen/CodeGenTypes.cpp
lib/CodeGen/ItaniumCXXABI.cpp
lib/Parse/ParseDecl.cpp
lib/Sema/DeclSpec.cpp
lib/Sema/SemaDeclCXX.cpp
lib/Sema/SemaExpr.cpp
lib/Sema/SemaLookup.cpp
lib/Sema/SemaTemplate.cpp
lib/Sema/SemaTemplateDeduction.cpp
lib/Sema/SemaTemplateVariadic.cpp
lib/Sema/SemaType.cpp
lib/Sema/TreeTransform.h
lib/Serialization/ASTReader.cpp
lib/Serialization/ASTWriter.cpp
test/CodeGenOpenCL/pipe_types.cl
test/PCH/ocl_types.cl
test/PCH/ocl_types.h
tools/libclang/CIndex.cpp
Index: lib/Serialization/ASTWriter.cpp
===
--- lib/Serialization/ASTWriter.cpp
+++ lib/Serialization/ASTWriter.cpp
@@ -447,6 +447,12 @@
Code = TYPE_ATOMIC;
}
+void
+ASTTypeWriter::VisitPipeType(const PipeType *T) {
+ Writer.AddTypeRef(T->getElementType(), Record);
+ Code = TYPE_PIPE;
+}
+
namespace {
class TypeLocWriter : public TypeLocVisitor {
@@ -672,6 +678,9 @@
Writer.AddSourceLocation(TL.getKWLoc(), Record);
Writer.AddSourceLocation(TL.getLParenLoc(), Record);
Writer.AddSourceLocation(TL.getRParenLoc(), Record);
+}
+void TypeLocWriter::VisitPipeTypeLoc(PipeTypeLoc TL) {
+ Writer.AddSourceLocation(TL.getKWLoc(), Record);
}
void ASTWriter::WriteTypeAbbrevs() {
Index: lib/Serialization/ASTReader.cpp
===
--- lib/Serialization/ASTReader.cpp
+++ lib/Serialization/ASTReader.cpp
@@ -5631,6 +5631,17 @@
QualType ValueType = readType(*Loc.F, Record, Idx);
return Context.getAtomicType(ValueType);
}
+
+ case TYPE_PIPE: {
+if (Record.size() != 1) {
+ Error("Incorrect encoding of pipe type");
+ return QualType();
+}
+
+// Reading the pipe element type.
+QualType ElementType = readType(*Loc.F, Record, Idx);
+return Context.getPipeType(ElementType);
+ }
}
llvm_unreachable("Invalid TypeCode!");
}
@@ -5901,6 +5912,9 @@
TL.setKWLoc(ReadSourceLocation(Record, Idx));
TL.setLParenLoc(ReadSourceLocation(Record, Idx));
TL.setRParenLoc(ReadSourceLocation(Record, Idx));
+}
+void TypeLocReader::VisitPipeTypeLoc(PipeTypeLoc TL) {
+ TL.setKWLoc(ReadSourceLocation(Record, Idx));
}
TypeSourceInfo *ASTReader::GetTypeSourceInfo(ModuleFile &F,
Index: lib/Parse/ParseDecl.cpp
===
--- lib/Parse/ParseDecl.cpp
+++ lib/Parse/ParseDecl.cpp
@@ -3321,6 +3321,15 @@
case tok::kw___bool:
isInvalid = DS.SetTypeAltiVecBool(true, Loc, PrevSpec, DiagID, Policy);
break;
+case tok::kw_pipe:
+ if (!getLangOpts().OpenCL || (getLangOpts().OpenCLVersion < 200)) {
+// OpenCL 2.0 defined this keyword. OpenCL 1.2 and earlier should
+// support the "pipe" word as identifier.
+Tok.getIdentifierInfo()->revertTokenIDToIdentifier();
+goto DoneWithDeclSpec;
+ }
+ isInvalid = DS.SetTypePipe(true, Loc, PrevSpec, DiagID, Policy);
+ break;
case tok::kw___unknown_anytype:
isInvalid = DS.SetTypeSpecType(TST_unknown_anytype, Loc,
PrevSpec, DiagID, Policy);
@@ -4396,6 +4405,9 @@
switch (Tok.getKind()) {
default: return false;
+ case tok::kw_pipe:
+return getLangOpts().OpenCL && (getLangOpts().OpenCLVersion >= 200);
+
case tok::identifier: // foo::bar
// Unfortunate hack to support "Class.factoryMethod" notation.
if (getLangOpts().ObjC1 && NextToken().is(tok::period))
@@ -4841,6 +4853,9 @@
if (Kind == tok::star || Kind == tok::caret)
return true;
+ if ((Kind == tok::kw_pipe) && Lang.OpenCL && (Lang.OpenCLVersion >= 200))
+return true;
+
if (!Lang.CPlusPlus)
return false;
@@ -4935,6 +4950,16 @@
}
tok::TokenKind Kind = Tok.getKind();
+
+ if ( D.getDeclSpec().isTypeSpecPipe() ) {
+DeclSpec &DS = D.getMutableDeclSpec();
+
+ D.AddTypeInfo(DeclaratorChunk::getPipe(DS.getTypeQualifiers(),
+ DS.getPipeLoc()),
+ DS.getAttributes(),
+ SourceLocation());
Re: [PATCH] D23086: [OpenCL] Generate concrete struct type for ndrange_t
bader added inline comments. Comment at: test/CodeGenOpenCL/cl20-device-side-enqueue.cl:12 @@ -11,3 +11,3 @@ unsigned flags = 0; - // CHECK: %ndrange = alloca %opencl.ndrange_t* + // CHECK: %ndrange = alloca %ndrange_t ndrange_t ndrange; Could you also add a regression test to validate that ndrange_t is defined in LLVM as struct type, please? Repository: rL LLVM https://reviews.llvm.org/D23086 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[PATCH] D23120: [OpenCL] Added underscores to the names of 'to_addr' OpenCL built-ins.
bader created this revision.
bader added reviewers: Anastasia, yaxunl.
bader added a subscriber: cfe-commits.
In order to re-define OpenCL built-in functions
'to_{private,local,global}' in OpenCL run-time library LLVM names must
be different from the clang built-in function names.
https://reviews.llvm.org/D23120
Files:
lib/CodeGen/CGBuiltin.cpp
test/CodeGenOpenCL/to_addr_builtin.cl
Index: test/CodeGenOpenCL/to_addr_builtin.cl
===
--- test/CodeGenOpenCL/to_addr_builtin.cl
+++ test/CodeGenOpenCL/to_addr_builtin.cl
@@ -14,74 +14,74 @@
generic int *gen;
//CHECK: %[[ARG:.*]] = addrspacecast i32 addrspace(1)* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @to_global(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @__to_global(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(1)* %[[RET]] to i32 addrspace(1)*
glob = to_global(glob);
//CHECK: %[[ARG:.*]] = addrspacecast i32 addrspace(3)* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @to_global(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @__to_global(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(1)* %[[RET]] to i32 addrspace(1)*
glob = to_global(loc);
//CHECK: %[[ARG:.*]] = addrspacecast i32* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @to_global(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @__to_global(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(1)* %[[RET]] to i32 addrspace(1)*
glob = to_global(priv);
//CHECK: %[[ARG:.*]] = bitcast i32 addrspace(4)* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @to_global(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @__to_global(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(1)* %[[RET]] to i32 addrspace(1)*
glob = to_global(gen);
//CHECK: %[[ARG:.*]] = addrspacecast i32 addrspace(1)* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @to_local(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @__to_local(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(3)* %[[RET]] to i32 addrspace(3)*
loc = to_local(glob);
//CHECK: %[[ARG:.*]] = addrspacecast i32 addrspace(3)* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @to_local(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @__to_local(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(3)* %[[RET]] to i32 addrspace(3)*
loc = to_local(loc);
//CHECK: %[[ARG:.*]] = addrspacecast i32* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @to_local(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @__to_local(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(3)* %[[RET]] to i32 addrspace(3)*
loc = to_local(priv);
//CHECK: %[[ARG:.*]] = bitcast i32 addrspace(4)* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @to_local(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @__to_local(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(3)* %[[RET]] to i32 addrspace(3)*
loc = to_local(gen);
//CHECK: %[[ARG:.*]] = addrspacecast i32 addrspace(1)* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8* @to_private(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8* @__to_private(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8* %[[RET]] to i32*
priv = to_private(glob);
//CHECK: %[[ARG:.*]] = addrspacecast i32 addrspace(3)* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8* @to_private(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8* @__to_private(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8* %[[RET]] to i32*
priv = to_private(loc);
//CHECK: %[[ARG:.*]] = addrspacecast i32* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8* @to_private(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8* @__to_private(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8* %[[RET]] to i32*
priv = to_private(priv);
//CHECK: %[[ARG:.*]] = bitcast i32 addrspace(4)* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8* @to_private(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8* @__to_private(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8* %[[RET]] to i32*
priv = to_private(gen);
//CHECK: %[[ARG:.*]] = addrspacecast %[[A]]* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @to_global(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @__to_global(i8 addrspace(4)* %[[
r277743 - [OpenCL] Added underscores to the names of 'to_addr' OpenCL built-ins.
Author: bader
Date: Thu Aug 4 13:06:27 2016
New Revision: 277743
URL: http://llvm.org/viewvc/llvm-project?rev=277743&view=rev
Log:
[OpenCL] Added underscores to the names of 'to_addr' OpenCL built-ins.
Summary:
In order to re-define OpenCL built-in functions
'to_{private,local,global}' in OpenCL run-time library LLVM names must
be different from the clang built-in function names.
Reviewers: yaxunl, Anastasia
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D23120
Modified:
cfe/trunk/lib/CodeGen/CGBuiltin.cpp
cfe/trunk/test/CodeGenOpenCL/to_addr_builtin.cl
Modified: cfe/trunk/lib/CodeGen/CGBuiltin.cpp
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/CodeGen/CGBuiltin.cpp?rev=277743&r1=277742&r2=277743&view=diff
==
--- cfe/trunk/lib/CodeGen/CGBuiltin.cpp (original)
+++ cfe/trunk/lib/CodeGen/CGBuiltin.cpp Thu Aug 4 13:06:27 2016
@@ -2209,8 +2209,9 @@ RValue CodeGenFunction::EmitBuiltinExpr(
NewArg = Builder.CreateAddrSpaceCast(Arg0, NewArgT);
else
NewArg = Builder.CreateBitOrPointerCast(Arg0, NewArgT);
-auto NewCall = Builder.CreateCall(CGM.CreateRuntimeFunction(FTy,
- E->getDirectCallee()->getName()), {NewArg});
+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(;
}
Modified: cfe/trunk/test/CodeGenOpenCL/to_addr_builtin.cl
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/test/CodeGenOpenCL/to_addr_builtin.cl?rev=277743&r1=277742&r2=277743&view=diff
==
--- cfe/trunk/test/CodeGenOpenCL/to_addr_builtin.cl (original)
+++ cfe/trunk/test/CodeGenOpenCL/to_addr_builtin.cl Thu Aug 4 13:06:27 2016
@@ -14,74 +14,74 @@ void test(void) {
generic int *gen;
//CHECK: %[[ARG:.*]] = addrspacecast i32 addrspace(1)* %{{.*}} to i8
addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @to_global(i8 addrspace(4)*
%[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @__to_global(i8 addrspace(4)*
%[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(1)* %[[RET]] to i32 addrspace(1)*
glob = to_global(glob);
//CHECK: %[[ARG:.*]] = addrspacecast i32 addrspace(3)* %{{.*}} to i8
addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @to_global(i8 addrspace(4)*
%[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @__to_global(i8 addrspace(4)*
%[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(1)* %[[RET]] to i32 addrspace(1)*
glob = to_global(loc);
//CHECK: %[[ARG:.*]] = addrspacecast i32* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @to_global(i8 addrspace(4)*
%[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @__to_global(i8 addrspace(4)*
%[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(1)* %[[RET]] to i32 addrspace(1)*
glob = to_global(priv);
//CHECK: %[[ARG:.*]] = bitcast i32 addrspace(4)* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @to_global(i8 addrspace(4)*
%[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @__to_global(i8 addrspace(4)*
%[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(1)* %[[RET]] to i32 addrspace(1)*
glob = to_global(gen);
//CHECK: %[[ARG:.*]] = addrspacecast i32 addrspace(1)* %{{.*}} to i8
addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @to_local(i8 addrspace(4)*
%[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @__to_local(i8 addrspace(4)*
%[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(3)* %[[RET]] to i32 addrspace(3)*
loc = to_local(glob);
//CHECK: %[[ARG:.*]] = addrspacecast i32 addrspace(3)* %{{.*}} to i8
addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @to_local(i8 addrspace(4)*
%[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @__to_local(i8 addrspace(4)*
%[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(3)* %[[RET]] to i32 addrspace(3)*
loc = to_local(loc);
//CHECK: %[[ARG:.*]] = addrspacecast i32* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @to_local(i8 addrspace(4)*
%[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @__to_local(i8 addrspace(4)*
%[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(3)* %[[RET]] to i32 addrspace(3)*
loc = to_local(priv);
//CHECK: %[[ARG:.*]] = bitcast i32 addrspace(4)* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @to_local(i8 addrspace(4)*
%[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @__to_local(i8 addrspace(4)*
%[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(3)* %[[RET]] to i32 addrspace(3)*
loc = to_local(gen);
//CHECK: %[[ARG:.*]] = addrspacecast i32 addrspace(1)* %{{.*}} to
Re: [PATCH] D23120: [OpenCL] Added underscores to the names of 'to_addr' OpenCL built-ins.
This revision was automatically updated to reflect the committed changes.
Closed by commit rL277743: [OpenCL] Added underscores to the names of 'to_addr'
OpenCL built-ins. (authored by bader).
Changed prior to commit:
https://reviews.llvm.org/D23120?vs=66653&id=66832#toc
Repository:
rL LLVM
https://reviews.llvm.org/D23120
Files:
cfe/trunk/lib/CodeGen/CGBuiltin.cpp
cfe/trunk/test/CodeGenOpenCL/to_addr_builtin.cl
Index: cfe/trunk/test/CodeGenOpenCL/to_addr_builtin.cl
===
--- cfe/trunk/test/CodeGenOpenCL/to_addr_builtin.cl
+++ cfe/trunk/test/CodeGenOpenCL/to_addr_builtin.cl
@@ -14,74 +14,74 @@
generic int *gen;
//CHECK: %[[ARG:.*]] = addrspacecast i32 addrspace(1)* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @to_global(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @__to_global(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(1)* %[[RET]] to i32 addrspace(1)*
glob = to_global(glob);
//CHECK: %[[ARG:.*]] = addrspacecast i32 addrspace(3)* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @to_global(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @__to_global(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(1)* %[[RET]] to i32 addrspace(1)*
glob = to_global(loc);
//CHECK: %[[ARG:.*]] = addrspacecast i32* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @to_global(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @__to_global(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(1)* %[[RET]] to i32 addrspace(1)*
glob = to_global(priv);
//CHECK: %[[ARG:.*]] = bitcast i32 addrspace(4)* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @to_global(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @__to_global(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(1)* %[[RET]] to i32 addrspace(1)*
glob = to_global(gen);
//CHECK: %[[ARG:.*]] = addrspacecast i32 addrspace(1)* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @to_local(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @__to_local(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(3)* %[[RET]] to i32 addrspace(3)*
loc = to_local(glob);
//CHECK: %[[ARG:.*]] = addrspacecast i32 addrspace(3)* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @to_local(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @__to_local(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(3)* %[[RET]] to i32 addrspace(3)*
loc = to_local(loc);
//CHECK: %[[ARG:.*]] = addrspacecast i32* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @to_local(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @__to_local(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(3)* %[[RET]] to i32 addrspace(3)*
loc = to_local(priv);
//CHECK: %[[ARG:.*]] = bitcast i32 addrspace(4)* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @to_local(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8 addrspace(3)* @__to_local(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8 addrspace(3)* %[[RET]] to i32 addrspace(3)*
loc = to_local(gen);
//CHECK: %[[ARG:.*]] = addrspacecast i32 addrspace(1)* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8* @to_private(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8* @__to_private(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8* %[[RET]] to i32*
priv = to_private(glob);
//CHECK: %[[ARG:.*]] = addrspacecast i32 addrspace(3)* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8* @to_private(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8* @__to_private(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8* %[[RET]] to i32*
priv = to_private(loc);
//CHECK: %[[ARG:.*]] = addrspacecast i32* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8* @to_private(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8* @__to_private(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8* %[[RET]] to i32*
priv = to_private(priv);
//CHECK: %[[ARG:.*]] = bitcast i32 addrspace(4)* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8* @to_private(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[RET:.*]] = call i8* @__to_private(i8 addrspace(4)* %[[ARG]])
//CHECK: %{{.*}} = bitcast i8* %[[RET]] to i32*
priv = to_private(gen);
//CHECK: %[[ARG:.*]] = addrspacecast %[[A]]* %{{.*}} to i8 addrspace(4)*
- //CHECK: %[[RET:.*]] = call i8 addrspace(1)* @to_global(i8 addrspace(4)* %[[ARG]])
+ //CHECK: %[[
Re: [PATCH] D16682: 19957 - OpenCL incorrectly accepts implicit address space conversion with ternary operator
bader added a subscriber: bader.
Comment at: test/SemaOpenCL/ternary-implicit-casts.cl:10-11
@@ +9,4 @@
+kernel void implicit_cast_generic(global int* gint, local int* lint, int cond)
{
+ // will compile, ptr is generic and can accept global and local
+ int* ptr = cond ? gint : lint; // expected-warning {{pointer type
mismatch ('__global int *' and '__local int *')}}
+}
Note that test doesn't specify OpenCL version. By default it's 1.0, so ptr
points to private AS and it's illegal to cast pointers from different address
spaces in OpenCL v1.x.
http://reviews.llvm.org/D16682
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: [PATCH] D20090: [OPENCL] Fix wrongly vla error for OpenCL array.
bader accepted this revision. bader added a reviewer: bader. bader added a comment. This revision is now accepted and ready to land. LGTM http://reviews.llvm.org/D20090 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: [PATCH] D20133: [OpenCL] Fix __builtin_astype for vec3 types.
bader accepted this revision. bader added a comment. This revision is now accepted and ready to land. LGTM. Comment at: test/CodeGenOpenCL/as_type.cl:6 @@ +5,3 @@ + +//CHECK: define spir_func <3 x i8> @f1(<4 x i8> %[[x:.*]]) +//CHECK: %[[astype:.*]] = shufflevector <4 x i8> %[[x]], <4 x i8> undef, <3 x i32> Could you also one test case that requires bitcast from vector type to 4-element vector type, please? E.g. short2 <-> char3 or char16 <-> int3. http://reviews.llvm.org/D20133 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: [PATCH] D19484: [OpenCL] Add supported OpenCL extensions to target info.
bader added inline comments. Comment at: include/clang/Basic/OpenCLExtensions.def:64 @@ +63,3 @@ +OPENCLEXT_INTERNAL(cl_khr_initialize_memory, 200, ~0U) +OPENCLEXT_INTERNAL(cl_khr_spir, 200, ~0U) +OPENCLEXT_INTERNAL(cl_khr_srgb_image_writes, 200, ~0U) Minimum required version should be 120. cl_khr_spir was available in OpenCL 1.2 as well as cl_khr_depth_images and cl_khr_gl_depth_images. https://www.khronos.org/registry/cl/specs/opencl-1.2-extensions.pdf http://reviews.llvm.org/D19484 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: [PATCH] D19484: [OpenCL] Add supported OpenCL extensions to target info.
bader accepted this revision. bader added a comment. LGTM. Thanks! http://reviews.llvm.org/D19484 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: [PATCH] D17821: [OpenCL] Complete image types support
bader added inline comments. Comment at: include/clang/AST/ASTContext.h:903 @@ +902,3 @@ +#define IMAGE_TYPE(ImgType, Id, SingletonId, Access, Suffix) \ + CanQualType SingletonId; +#include "clang/AST/OpenCLImageTypes.def" mgrang wrote: > remove extra spacing in front of the \ Sorry for delay... I used clang-format tool to format that code. I expect it to be in agreement with LLVM coding style guide. Did I miss something or it's clang-format bug? http://reviews.llvm.org/D17821 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: [PATCH] D17821: [OpenCL] Complete image types support
bader added a comment. In http://reviews.llvm.org/D17821#393387, @Anastasia wrote: > Regarding, extending this approach for OpenCL pipe types too. I was thinking > we could change current implementation to have ReadPipeType and > WritePipeType. They can both be derived from PipeType that we already have > now (we can make it an abstract class to avoid its instantiation?). > > Similarly to images, since read and write pipes will be mapped to different > Clang types, we won't need any extra semantical checking but just need to add > extra code in CodeGen of pipe type and builtins to accept two separate types > for read only and write only cases. > > Would this make sense? Sure. Do you want me to add it here or it's okay if we fix pipes in a separate patch? Comment at: include/clang/AST/ASTContext.h:903 @@ +902,3 @@ +#define IMAGE_TYPE(ImgType, Id, SingletonId, Access, Suffix) \ + CanQualType SingletonId; +#include "clang/AST/OpenCLImageTypes.def" Anastasia wrote: > bader wrote: > > mgrang wrote: > > > remove extra spacing in front of the \ > > Sorry for delay... > > I used clang-format tool to format that code. I expect it to be in > > agreement with LLVM coding style guide. > > Did I miss something or it's clang-format bug? > I would have thought clang-format should be fine. > > But it does look a bit weird here. There are other places in this patch that > have the same formatting. > > I can't find anything relevant in coding style description: > http://llvm.org/docs/CodingStandards.html#source-code-width > > I'll remove extra spacing. http://reviews.llvm.org/D17821 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: [PATCH] D17821: [OpenCL] Complete image types support
bader added a comment. In http://reviews.llvm.org/D17821#394620, @Anastasia wrote: > Yes, I think it's better to go in a separate commit, not to complicate this > one too much. Also since there are not many comment here, I think we should > try to commit it ASAP otherwise rebasing would be an issue. I can commit it tomorrow if there are no other comments. http://reviews.llvm.org/D17821 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
r265783 - [OpenCL] Complete image types support.
Author: bader
Date: Fri Apr 8 08:40:33 2016
New Revision: 265783
URL: http://llvm.org/viewvc/llvm-project?rev=265783&view=rev
Log:
[OpenCL] Complete image types support.
I. Current implementation of images is not conformant to spec in the following
points:
1. It makes no distinction with respect to access qualifiers and therefore
allows to use images with different access type interchangeably. The following
code would compile just fine:
void write_image(write_only image2d_t img);
kernel void foo(read_only image2d_t img) { write_image(img); } //
Accepted code
which is disallowed according to s6.13.14.
2. It discards access qualifier on generated code, which leads to generated
code for the above example:
call void @write_image(%opencl.image2d_t* %img);
In OpenCL2.0 however we can have different calls into write_image with
read_only and wite_only images.
Also generally following compiler steps have no easy way to take different
path depending on the image access: linking to the right implementation of
image types, performing IR opts and backend codegen differently.
3. Image types are language keywords and can't be redeclared s6.1.9, which
can happen currently as they are just typedef names.
4. Default access qualifier read_only is to be added if not provided
explicitly.
II. This patch corrects the above points as follows:
1. All images are encapsulated into a separate .def file that is inserted in
different points where image handling is required. This avoid a lot of code
repetition as all images are handled the same way in the code with no
distinction of their exact type.
2. The Cartesian product of image types and image access qualifiers is added
to the builtin types. This simplifies a lot handling of access type mismatch as
no operations are allowed by default on distinct Builtin types. Also spec
intended access qualifier as special type qualifier that are combined with an
image type to form a distinct type (see statement above - images can't be
created w/o access qualifiers).
3. Improves testing of images in Clang.
Author: Anastasia Stulova
Reviewers: bader, mgrang.
Subscribers: pxli168, pekka.jaaskelainen, yaxunl.
Differential Revision: http://reviews.llvm.org/D17821
Added:
cfe/trunk/include/clang/AST/OpenCLImageTypes.def
cfe/trunk/test/CodeGenOpenCL/images.cl
cfe/trunk/test/SemaOpenCL/images.cl
Modified:
cfe/trunk/include/clang/AST/ASTContext.h
cfe/trunk/include/clang/AST/BuiltinTypes.def
cfe/trunk/include/clang/AST/Type.h
cfe/trunk/include/clang/Basic/Specifiers.h
cfe/trunk/include/clang/Basic/TokenKinds.def
cfe/trunk/include/clang/Sema/DeclSpec.h
cfe/trunk/include/clang/Serialization/ASTBitCodes.h
cfe/trunk/lib/AST/ASTContext.cpp
cfe/trunk/lib/AST/ASTImporter.cpp
cfe/trunk/lib/AST/ExprConstant.cpp
cfe/trunk/lib/AST/ItaniumMangle.cpp
cfe/trunk/lib/AST/MicrosoftMangle.cpp
cfe/trunk/lib/AST/NSAPI.cpp
cfe/trunk/lib/AST/Type.cpp
cfe/trunk/lib/AST/TypeLoc.cpp
cfe/trunk/lib/Analysis/PrintfFormatString.cpp
cfe/trunk/lib/CodeGen/CGDebugInfo.cpp
cfe/trunk/lib/CodeGen/CGDebugInfo.h
cfe/trunk/lib/CodeGen/CGOpenCLRuntime.cpp
cfe/trunk/lib/CodeGen/CodeGenTypes.cpp
cfe/trunk/lib/CodeGen/ItaniumCXXABI.cpp
cfe/trunk/lib/Index/USRGeneration.cpp
cfe/trunk/lib/Parse/ParseDecl.cpp
cfe/trunk/lib/Parse/ParseExpr.cpp
cfe/trunk/lib/Parse/ParseTentative.cpp
cfe/trunk/lib/Sema/DeclSpec.cpp
cfe/trunk/lib/Sema/Sema.cpp
cfe/trunk/lib/Sema/SemaExpr.cpp
cfe/trunk/lib/Sema/SemaTemplateVariadic.cpp
cfe/trunk/lib/Sema/SemaType.cpp
cfe/trunk/lib/Serialization/ASTCommon.cpp
cfe/trunk/lib/Serialization/ASTReader.cpp
cfe/trunk/test/CodeGenOpenCL/opencl_types.cl
cfe/trunk/test/SemaOpenCL/invalid-access-qualifier.cl
cfe/trunk/test/SemaOpenCL/invalid-image.cl
cfe/trunk/test/SemaOpenCL/invalid-kernel-parameters.cl
cfe/trunk/tools/libclang/CIndex.cpp
Modified: cfe/trunk/include/clang/AST/ASTContext.h
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/include/clang/AST/ASTContext.h?rev=265783&r1=265782&r2=265783&view=diff
==
--- cfe/trunk/include/clang/AST/ASTContext.h (original)
+++ cfe/trunk/include/clang/AST/ASTContext.h Fri Apr 8 08:40:33 2016
@@ -902,11 +902,9 @@ public:
CanQualType PseudoObjectTy, ARCUnbridgedCastTy;
CanQualType ObjCBuiltinIdTy, ObjCBuiltinClassTy, ObjCBuiltinSelTy;
CanQualType ObjCBuiltinBoolTy;
- CanQualType OCLImage1dTy, OCLImage1dArrayTy, OCLImage1dBufferTy;
- CanQualType OCLImage2dTy, OCLImage2dArrayTy, OCLImage2dDepthTy;
- CanQualType OCLImage2dArrayDepthTy, OCLImage2dMSAATy, OCLImage2dArrayMSAATy;
- CanQualType OCLImage2dMSAADepthTy, OCLImage2dArrayMSAADepthTy;
- CanQualType OCLImage3dTy;
+#define IMAGE_TYPE(ImgType, Id, SingletonId, Access, Suffix) \
+ CanQualType SingletonId;
+#inclu
Re: [PATCH] D17821: [OpenCL] Complete image types support
bader added a comment. In http://reviews.llvm.org/D17821#394708, @bader wrote: > In http://reviews.llvm.org/D17821#394620, @Anastasia wrote: > > > Yes, I think it's better to go in a separate commit, not to complicate this > > one too much. Also since there are not many comment here, I think we should > > try to commit it ASAP otherwise rebasing would be an issue. > > > I can commit it tomorrow if there are no other comments. Committed @265783 Repository: rL LLVM http://reviews.llvm.org/D17821 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: [PATCH] D17821: [OpenCL] Complete image types support
This revision was automatically updated to reflect the committed changes.
Closed by commit rL265783: [OpenCL] Complete image types support. (authored by
bader).
Changed prior to commit:
http://reviews.llvm.org/D17821?vs=51303&id=53021#toc
Repository:
rL LLVM
http://reviews.llvm.org/D17821
Files:
cfe/trunk/include/clang/AST/ASTContext.h
cfe/trunk/include/clang/AST/BuiltinTypes.def
cfe/trunk/include/clang/AST/OpenCLImageTypes.def
cfe/trunk/include/clang/AST/Type.h
cfe/trunk/include/clang/Basic/Specifiers.h
cfe/trunk/include/clang/Basic/TokenKinds.def
cfe/trunk/include/clang/Sema/DeclSpec.h
cfe/trunk/include/clang/Serialization/ASTBitCodes.h
cfe/trunk/lib/AST/ASTContext.cpp
cfe/trunk/lib/AST/ASTImporter.cpp
cfe/trunk/lib/AST/ExprConstant.cpp
cfe/trunk/lib/AST/ItaniumMangle.cpp
cfe/trunk/lib/AST/MicrosoftMangle.cpp
cfe/trunk/lib/AST/NSAPI.cpp
cfe/trunk/lib/AST/Type.cpp
cfe/trunk/lib/AST/TypeLoc.cpp
cfe/trunk/lib/Analysis/PrintfFormatString.cpp
cfe/trunk/lib/CodeGen/CGDebugInfo.cpp
cfe/trunk/lib/CodeGen/CGDebugInfo.h
cfe/trunk/lib/CodeGen/CGOpenCLRuntime.cpp
cfe/trunk/lib/CodeGen/CodeGenTypes.cpp
cfe/trunk/lib/CodeGen/ItaniumCXXABI.cpp
cfe/trunk/lib/Index/USRGeneration.cpp
cfe/trunk/lib/Parse/ParseDecl.cpp
cfe/trunk/lib/Parse/ParseExpr.cpp
cfe/trunk/lib/Parse/ParseTentative.cpp
cfe/trunk/lib/Sema/DeclSpec.cpp
cfe/trunk/lib/Sema/Sema.cpp
cfe/trunk/lib/Sema/SemaExpr.cpp
cfe/trunk/lib/Sema/SemaTemplateVariadic.cpp
cfe/trunk/lib/Sema/SemaType.cpp
cfe/trunk/lib/Serialization/ASTCommon.cpp
cfe/trunk/lib/Serialization/ASTReader.cpp
cfe/trunk/test/CodeGenOpenCL/images.cl
cfe/trunk/test/CodeGenOpenCL/opencl_types.cl
cfe/trunk/test/SemaOpenCL/images.cl
cfe/trunk/test/SemaOpenCL/invalid-access-qualifier.cl
cfe/trunk/test/SemaOpenCL/invalid-image.cl
cfe/trunk/test/SemaOpenCL/invalid-kernel-parameters.cl
cfe/trunk/tools/libclang/CIndex.cpp
Index: cfe/trunk/lib/CodeGen/ItaniumCXXABI.cpp
===
--- cfe/trunk/lib/CodeGen/ItaniumCXXABI.cpp
+++ cfe/trunk/lib/CodeGen/ItaniumCXXABI.cpp
@@ -2536,18 +2536,9 @@
case BuiltinType::UInt128:
return true;
-case BuiltinType::OCLImage1d:
-case BuiltinType::OCLImage1dArray:
-case BuiltinType::OCLImage1dBuffer:
-case BuiltinType::OCLImage2d:
-case BuiltinType::OCLImage2dArray:
-case BuiltinType::OCLImage2dDepth:
-case BuiltinType::OCLImage2dArrayDepth:
-case BuiltinType::OCLImage2dMSAA:
-case BuiltinType::OCLImage2dArrayMSAA:
-case BuiltinType::OCLImage2dMSAADepth:
-case BuiltinType::OCLImage2dArrayMSAADepth:
-case BuiltinType::OCLImage3d:
+#define IMAGE_TYPE(ImgType, Id, SingletonId, Access, Suffix) \
+case BuiltinType::Id:
+#include "clang/AST/OpenCLImageTypes.def"
case BuiltinType::OCLSampler:
case BuiltinType::OCLEvent:
case BuiltinType::OCLClkEvent:
Index: cfe/trunk/lib/CodeGen/CodeGenTypes.cpp
===
--- cfe/trunk/lib/CodeGen/CodeGenTypes.cpp
+++ cfe/trunk/lib/CodeGen/CodeGenTypes.cpp
@@ -464,18 +464,9 @@
ResultType = llvm::IntegerType::get(getLLVMContext(), 128);
break;
-case BuiltinType::OCLImage1d:
-case BuiltinType::OCLImage1dArray:
-case BuiltinType::OCLImage1dBuffer:
-case BuiltinType::OCLImage2d:
-case BuiltinType::OCLImage2dArray:
-case BuiltinType::OCLImage2dDepth:
-case BuiltinType::OCLImage2dArrayDepth:
-case BuiltinType::OCLImage2dMSAA:
-case BuiltinType::OCLImage2dArrayMSAA:
-case BuiltinType::OCLImage2dMSAADepth:
-case BuiltinType::OCLImage2dArrayMSAADepth:
-case BuiltinType::OCLImage3d:
+#define IMAGE_TYPE(ImgType, Id, SingletonId, Access, Suffix) \
+case BuiltinType::Id:
+#include "clang/AST/OpenCLImageTypes.def"
case BuiltinType::OCLSampler:
case BuiltinType::OCLEvent:
case BuiltinType::OCLClkEvent:
Index: cfe/trunk/lib/CodeGen/CGDebugInfo.cpp
===
--- cfe/trunk/lib/CodeGen/CGDebugInfo.cpp
+++ cfe/trunk/lib/CodeGen/CGDebugInfo.cpp
@@ -463,39 +463,11 @@
return SelTy;
}
- case BuiltinType::OCLImage1d:
-return getOrCreateStructPtrType("opencl_image1d_t", OCLImage1dDITy);
- case BuiltinType::OCLImage1dArray:
-return getOrCreateStructPtrType("opencl_image1d_array_t",
-OCLImage1dArrayDITy);
- case BuiltinType::OCLImage1dBuffer:
-return getOrCreateStructPtrType("opencl_image1d_buffer_t",
-OCLImage1dBufferDITy);
- case BuiltinType::OCLImage2d:
-return getOrCreateStructPtrType("opencl_image2d_t", OCLImage2dDITy);
- case BuiltinType::OCLImage2dArray:
-return getOrCreateStructPtrType("opencl_image2d_array_t",
-OCLImage2dArrayDITy);
- case BuiltinType::OCLImage2dDepth:
-return getOrCreateStructPtrType("ope
r266180 - [OpenCL] Move OpenCLImageTypes.def from clangAST to clangBasic library.
Author: bader Date: Wed Apr 13 03:33:41 2016 New Revision: 266180 URL: http://llvm.org/viewvc/llvm-project?rev=266180&view=rev Log: [OpenCL] Move OpenCLImageTypes.def from clangAST to clangBasic library. Putting OpenCLImageTypes.def to clangAST library violates layering requirement: "It's not OK for a Basic/ header to include an AST/ header". This fixes the modules build. Differential revision: http://reviews.llvm.org/D18954 Reviewers: Richard Smith, Vassil Vassilev. Added: cfe/trunk/include/clang/Basic/OpenCLImageTypes.def Removed: cfe/trunk/include/clang/AST/OpenCLImageTypes.def Modified: cfe/trunk/include/clang/AST/ASTContext.h cfe/trunk/include/clang/AST/Type.h cfe/trunk/include/clang/Basic/Specifiers.h cfe/trunk/include/clang/Basic/TokenKinds.def cfe/trunk/include/clang/Sema/DeclSpec.h cfe/trunk/include/clang/Serialization/ASTBitCodes.h cfe/trunk/lib/AST/ASTContext.cpp cfe/trunk/lib/AST/ASTImporter.cpp cfe/trunk/lib/AST/ExprConstant.cpp cfe/trunk/lib/AST/ItaniumMangle.cpp cfe/trunk/lib/AST/MicrosoftMangle.cpp cfe/trunk/lib/AST/NSAPI.cpp cfe/trunk/lib/AST/Type.cpp cfe/trunk/lib/AST/TypeLoc.cpp cfe/trunk/lib/Analysis/PrintfFormatString.cpp cfe/trunk/lib/CodeGen/CGDebugInfo.cpp cfe/trunk/lib/CodeGen/CGDebugInfo.h cfe/trunk/lib/CodeGen/CGOpenCLRuntime.cpp cfe/trunk/lib/CodeGen/CodeGenTypes.cpp cfe/trunk/lib/CodeGen/ItaniumCXXABI.cpp cfe/trunk/lib/Index/USRGeneration.cpp cfe/trunk/lib/Parse/ParseDecl.cpp cfe/trunk/lib/Parse/ParseExpr.cpp cfe/trunk/lib/Parse/ParseTentative.cpp cfe/trunk/lib/Sema/DeclSpec.cpp cfe/trunk/lib/Sema/SemaExpr.cpp cfe/trunk/lib/Sema/SemaTemplateVariadic.cpp cfe/trunk/lib/Sema/SemaType.cpp cfe/trunk/lib/Serialization/ASTCommon.cpp cfe/trunk/lib/Serialization/ASTReader.cpp cfe/trunk/tools/libclang/CIndex.cpp Modified: cfe/trunk/include/clang/AST/ASTContext.h URL: http://llvm.org/viewvc/llvm-project/cfe/trunk/include/clang/AST/ASTContext.h?rev=266180&r1=266179&r2=266180&view=diff == --- cfe/trunk/include/clang/AST/ASTContext.h (original) +++ cfe/trunk/include/clang/AST/ASTContext.h Wed Apr 13 03:33:41 2016 @@ -904,7 +904,7 @@ public: CanQualType ObjCBuiltinBoolTy; #define IMAGE_TYPE(ImgType, Id, SingletonId, Access, Suffix) \ CanQualType SingletonId; -#include "clang/AST/OpenCLImageTypes.def" +#include "clang/Basic/OpenCLImageTypes.def" CanQualType OCLSamplerTy, OCLEventTy, OCLClkEventTy; CanQualType OCLQueueTy, OCLNDRangeTy, OCLReserveIDTy; CanQualType OMPArraySectionTy; Removed: cfe/trunk/include/clang/AST/OpenCLImageTypes.def URL: http://llvm.org/viewvc/llvm-project/cfe/trunk/include/clang/AST/OpenCLImageTypes.def?rev=266179&view=auto == --- cfe/trunk/include/clang/AST/OpenCLImageTypes.def (original) +++ cfe/trunk/include/clang/AST/OpenCLImageTypes.def (removed) @@ -1,82 +0,0 @@ -//===-- OpenCLImageTypes.def - Metadata about BuiltinTypes --*- C++ -*-===// -// -// The LLVM Compiler Infrastructure -// -// This file is distributed under the University of Illinois Open Source -// License. See LICENSE.TXT for details. -// -//===--===// -// This file extends builtin types database with OpenCL image singleton types. -// Custom code should define one of those two macros: -//GENERIC_IMAGE_TYPE(Type, Id) - a generic image with its Id without an -// access type -//IMAGE_TYPE(Type, Id, SingletonId, AccessType, CGSuffix) - an image type -// with given ID, singleton ID access type and a codegen suffix - -#ifdef GENERIC_IMAGE_TYPE - -#define IMAGE_READ_TYPE(Type, Id) GENERIC_IMAGE_TYPE(Type, Id) -#define IMAGE_WRITE_TYPE(Type, Id) -#define IMAGE_READ_WRITE_TYPE(Type, Id) - -#else - -#ifndef IMAGE_READ_TYPE -#define IMAGE_READ_TYPE(Type, Id) \ - IMAGE_TYPE(Type, Id##RO, Id##ROTy, read_only, ro) -#endif -#ifndef IMAGE_WRITE_TYPE -#define IMAGE_WRITE_TYPE(Type, Id) \ - IMAGE_TYPE(Type, Id##WO, Id##WOTy, write_only, wo) -#endif -#ifndef IMAGE_READ_WRITE_TYPE -#define IMAGE_READ_WRITE_TYPE(Type, Id) \ - IMAGE_TYPE(Type, Id##RW, Id##RWTy, read_write, rw) -#endif - -#endif - -IMAGE_READ_TYPE(image1d, OCLImage1d) -IMAGE_READ_TYPE(image1d_array, OCLImage1dArray) -IMAGE_READ_TYPE(image1d_buffer, OCLImage1dBuffer) -IMAGE_READ_TYPE(image2d, OCLImage2d) -IMAGE_READ_TYPE(image2d_array, OCLImage2dArray) -IMAGE_READ_TYPE(image2d_depth, OCLImage2dDepth) -IMAGE_READ_TYPE(image2d_array_depth, OCLImage2dArrayDepth) -IMAGE_READ_TYPE(image2d_msaa, OCLImage2dMSAA) -IMAGE_READ_TYPE(image2d_array_msaa, OCLImage2dArrayMSAA) -IMAGE_READ_TYPE(image2d_msaa_depth, OCLImage2dMSAADepth) -IMAGE_READ_TYPE(image2d_array_msaa_depth, OCLImage2dArrayMSAADepth) -IMAGE_READ_TYPE(image3d, O
r266187 - [modules] Add OpenCLImageTypes.def to module map to fix the modules build.
Author: bader
Date: Wed Apr 13 04:54:47 2016
New Revision: 266187
URL: http://llvm.org/viewvc/llvm-project?rev=266187&view=rev
Log:
[modules] Add OpenCLImageTypes.def to module map to fix the modules build.
Modified:
cfe/trunk/include/clang/module.modulemap
Modified: cfe/trunk/include/clang/module.modulemap
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/include/clang/module.modulemap?rev=266187&r1=266186&r2=266187&view=diff
==
--- cfe/trunk/include/clang/module.modulemap (original)
+++ cfe/trunk/include/clang/module.modulemap Wed Apr 13 04:54:47 2016
@@ -41,6 +41,7 @@ module Clang_Basic {
textual header "Basic/DiagnosticOptions.def"
textual header "Basic/LangOptions.def"
textual header "Basic/OpenCLExtensions.def"
+ textual header "Basic/OpenCLImageTypes.def"
textual header "Basic/OpenMPKinds.def"
textual header "Basic/OperatorKinds.def"
textual header "Basic/Sanitizers.def"
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: [PATCH] D19071: [OpenCL] Add predefined macros.
bader added a comment. LGTM http://reviews.llvm.org/D19071 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: [PATCH] D18369: [OpenCL] Upstreaming khronos OpenCL header file.
bader added inline comments. Comment at: lib/Headers/opencl.h:14892-14898 @@ +14891,9 @@ + +/** +* Sampler-less Image Access +*/ + +float4 __const_func __attribute__((overloadable)) read_imagef(read_only image1d_t image, int coord); +int4 __const_func __attribute__((overloadable)) read_imagei(read_only image1d_t image, int coord); +uint4 __const_func __attribute__((overloadable)) read_imageui(read_only image1d_t image, int coord); + OpenCL 2.0 allows sampler-less reading from an image with 'read_write' access qualifiers and write_image to images with 'read_write' access qualifiers. Since 265783 clang differentiate image types with different access qualifiers, so now we have explicitly declare new built-in functions. Could you add these declaration, please? http://reviews.llvm.org/D18369 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: [PATCH] D18369: [OpenCL] Upstreaming khronos OpenCL header file.
bader added inline comments. Comment at: lib/Headers/opencl.h:15372-15373 @@ +15371,4 @@ + */ +int __const_func __attribute__((overloadable)) get_image_width(image1d_t image); +int __const_func __attribute__((overloadable)) get_image_width(image1d_buffer_t image); +int __const_func __attribute__((overloadable)) get_image_width(image2d_t image); Please, add image access qualifier to Built-in Image Query Functions declarations. It's required by OpenCL 2.0 specification. OpenCL 1.x spec. defines built-in functions w/o access qualifiers (i.e. 'read_only' is applied). http://reviews.llvm.org/D18369 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: [PATCH] D18369: [OpenCL] Upstreaming khronos OpenCL header file.
bader added a comment. To clarify my last comment: I don't think we should define Image Query built-in functions only for 'read_only' access qualifier in OpenCL 1.x. It's just bug in OpenCL 1.x specifications. I think the intention was to provide these built-in functions for both 'read_only' and 'write_only' images. OpenCL 2.0 also provides 'read_write' flavors. http://reviews.llvm.org/D18369 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
