Skip to content

[clang][OMPIRBuilder] Fix reduction codegen for SPIR-V #162529

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Oct 9, 2025
Merged

[clang][OMPIRBuilder] Fix reduction codegen for SPIR-V #162529

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
67ae789
[OMPIRBuilder] Fix reduction codegen for SPIR-V
sarnex Oct 8, 2025
a8b2dd2
use existing test
sarnex Oct 8, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 6 additions & 1 deletion clang/lib/CodeGen/CGOpenMPRuntimeGPU.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -869,6 +869,8 @@ CGOpenMPRuntimeGPU::CGOpenMPRuntimeGPU(CodeGenModule &CGM)
CGM.getLangOpts().OpenMPOffloadMandatory,
/*HasRequiresReverseOffload*/ false, /*HasRequiresUnifiedAddress*/ false,
hasRequiresUnifiedSharedMemory(), /*HasRequiresDynamicAllocators*/ false);
Config.setDefaultTargetAS(
CGM.getContext().getTargetInfo().getTargetAddressSpace(LangAS::Default));
OMPBuilder.setConfig(Config);

if (!CGM.getLangOpts().OpenMPIsTargetDevice)
Expand Down Expand Up @@ -1243,7 +1245,10 @@ void CGOpenMPRuntimeGPU::emitParallelCall(
llvm::Value *ID = llvm::ConstantPointerNull::get(CGM.Int8PtrTy);
if (WFn)
ID = Bld.CreateBitOrPointerCast(WFn, CGM.Int8PtrTy);
llvm::Value *FnPtr = Bld.CreateBitOrPointerCast(OutlinedFn, CGM.Int8PtrTy);
llvm::Type *FnPtrTy = llvm::PointerType::get(
CGF.getLLVMContext(), CGM.getDataLayout().getProgramAddressSpace());

llvm::Value *FnPtr = Bld.CreateBitOrPointerCast(OutlinedFn, FnPtrTy);

// Create a private scope that will globalize the arguments
// passed from the outside of the target region.
Expand Down
22 changes: 22 additions & 0 deletions clang/test/OpenMP/spirv_reduction.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,22 @@
// RUN: %clang_cc1 -verify -fopenmp -x c++ -triple x86_64-unknown-linux -fopenmp-targets=spirv64-intel -emit-llvm-bc %s -o %t-host.bc
// RUN: %clang_cc1 -verify -fopenmp -x c++ -triple spirv64-intel -fopenmp-targets=spirv64-intel -emit-llvm %s -fopenmp-is-target-device -fopenmp-host-ir-file-path %t-host.bc -o - | FileCheck %s
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is there really not another reduction test we can just slap spirv onto?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Seems reasonable, I'm assuming the other targets have the same behavior as before since we don't split on function AS.

Right, for other targets default target AS and program AS are 0 so none of these addrspace changes have any effect

Is there really not another reduction test we can just slap spirv onto?

Good point, I'm sure there is, let me take a look

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Just pushed a commit that modifies an existing test instead, thanks


// expected-no-diagnostics

// CHECK: call spir_func addrspace(9) void @__kmpc_parallel_51(ptr addrspace(4) addrspacecast (ptr addrspace(1) @{{.*}} to ptr addrspace(4)),
// CHECK-SAME: i32 %{{.*}}, i32 {{.*}}, i32 {{.*}}, i32 {{.*}}, ptr addrspace(9) @{{.*}}, ptr addrspace(4) {{.*}}, ptr addrspace(4) %{{.*}}, i64 {{.*}})

// CHECK: call addrspace(9) i32 @__kmpc_nvptx_teams_reduce_nowait_v2(ptr addrspace(4) addrspacecast (ptr addrspace(1) @{{.*}} to ptr addrspace(4)),
// CHECK-SAME: ptr addrspace(4) %{{.*}}, i32 1024, i64 4, ptr addrspace(4) %{{.*}}, ptr addrspace(9) @{{.*}}, ptr addrspace(9) @{{.*}}, ptr addrspace(9) @{{.*}}, ptr addrspace(9) @{{.*}}, ptr addrspace(9) @{{.*}}, ptr addrspace(9) @{{.*}})

int main() {
int matrix_sum = 0;
#pragma omp target teams distribute parallel for \
reduction(+:matrix_sum) \
map(tofrom:matrix_sum)
for (int i = 0; i < 100; i++) {

}

return 0;
}
3 changes: 2 additions & 1 deletion llvm/include/llvm/Frontend/OpenMP/OMPKinds.def
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -42,6 +42,7 @@ __OMP_TYPE(Double)

OMP_TYPE(SizeTy, M.getDataLayout().getIntPtrType(Ctx))
OMP_TYPE(Int63, Type::getIntNTy(Ctx, 63))
OMP_TYPE(FuncPtrTy, PointerType::get(Ctx, M.getDataLayout().getProgramAddressSpace()))

__OMP_PTR_TYPE(VoidPtr)
__OMP_PTR_TYPE(VoidPtrPtr)
Expand Down Expand Up @@ -471,7 +472,7 @@ __OMP_RTL(__kmpc_target_init, false, Int32, KernelEnvironmentPtr, KernelLaunchEn
__OMP_RTL(__kmpc_target_deinit, false, Void,)
__OMP_RTL(__kmpc_kernel_prepare_parallel, false, Void, VoidPtr)
__OMP_RTL(__kmpc_parallel_51, false, Void, IdentPtr, Int32, Int32, Int32, Int32,
VoidPtr, VoidPtr, VoidPtrPtr, SizeTy)
FuncPtrTy, VoidPtr, VoidPtrPtr, SizeTy)
__OMP_RTL(__kmpc_for_static_loop_4, false, Void, IdentPtr, VoidPtr, VoidPtr, Int32, Int32, Int32, Int8)
__OMP_RTL(__kmpc_for_static_loop_4u, false, Void, IdentPtr, VoidPtr, VoidPtr, Int32, Int32, Int32, Int8)
__OMP_RTL(__kmpc_for_static_loop_8, false, Void, IdentPtr, VoidPtr, VoidPtr, Int64, Int64, Int64, Int8)
Expand Down
11 changes: 7 additions & 4 deletions llvm/lib/Frontend/OpenMP/OMPIRBuilder.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3623,7 +3623,9 @@ OpenMPIRBuilder::InsertPointOrErrorTy OpenMPIRBuilder::createReductionsGPU(
// 1. Build a list of reduction variables.
// void *RedList[<n>] = {<ReductionVars>[0], ..., <ReductionVars>[<n>-1]};
auto Size = ReductionInfos.size();
Type *PtrTy = PointerType::getUnqual(Ctx);
Type *PtrTy = PointerType::get(Ctx, Config.getDefaultTargetAS());
Type *FuncPtrTy =
Builder.getPtrTy(M.getDataLayout().getProgramAddressSpace());
Type *RedArrayTy = ArrayType::get(PtrTy, Size);
CodeGenIP = Builder.saveIP();
Builder.restoreIP(AllocaIP);
Expand Down Expand Up @@ -3667,9 +3669,9 @@ OpenMPIRBuilder::InsertPointOrErrorTy OpenMPIRBuilder::createReductionsGPU(
Builder.getInt64(MaxDataSize * ReductionInfos.size());
if (!IsTeamsReduction) {
Value *SarFuncCast =
Builder.CreatePointerBitCastOrAddrSpaceCast(SarFunc, PtrTy);
Builder.CreatePointerBitCastOrAddrSpaceCast(SarFunc, FuncPtrTy);
Value *WcFuncCast =
Builder.CreatePointerBitCastOrAddrSpaceCast(WcFunc, PtrTy);
Builder.CreatePointerBitCastOrAddrSpaceCast(WcFunc, FuncPtrTy);
Value *Args[] = {SrcLocInfo, ReductionDataSize, RL, SarFuncCast,
WcFuncCast};
Function *Pv2Ptr = getOrCreateRuntimeFunctionPtr(
Expand Down Expand Up @@ -10072,13 +10074,14 @@ void OpenMPIRBuilder::initializeTypes(Module &M) {
LLVMContext &Ctx = M.getContext();
StructType *T;
unsigned DefaultTargetAS = Config.getDefaultTargetAS();
unsigned ProgramAS = M.getDataLayout().getProgramAddressSpace();
#define OMP_TYPE(VarName, InitValue) VarName = InitValue;
#define OMP_ARRAY_TYPE(VarName, ElemTy, ArraySize) \
VarName##Ty = ArrayType::get(ElemTy, ArraySize); \
VarName##PtrTy = PointerType::get(Ctx, DefaultTargetAS);
#define OMP_FUNCTION_TYPE(VarName, IsVarArg, ReturnType, ...) \
VarName = FunctionType::get(ReturnType, {__VA_ARGS__}, IsVarArg); \
VarName##Ptr = PointerType::get(Ctx, DefaultTargetAS);
VarName##Ptr = PointerType::get(Ctx, ProgramAS);
#define OMP_STRUCT_TYPE(VarName, StructName, Packed, ...) \
T = StructType::getTypeByName(Ctx, StructName); \
if (!T) \
Expand Down