Skip to content

[SYCL][NVPTX] Optimize ID queries when they fit in int #18999

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
Jun 17, 2025
+35 −0
Merged

[SYCL][NVPTX] Optimize ID queries when they fit in int #18999

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
27a7a4f
[SYCL][NVPTX] Optimize ID queries when they fit in int
frasercrmck Jun 16, 2025
1779200
fix formatting
frasercrmck Jun 16, 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
5 changes: 5 additions & 0 deletions clang/lib/Driver/ToolChains/Cuda.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -968,6 +968,11 @@ void CudaToolChain::addClangTargetOptions(
"--nvptx-prec-sqrtf32=0"});

CC1Args.append({"-mllvm", "-enable-memcpyopt-without-libcalls"});

if (DriverArgs.hasFlag(options::OPT_fsycl_id_queries_fit_in_int,
options::OPT_fno_sycl_id_queries_fit_in_int, false))
CC1Args.append(
{"-mllvm", "-nvvm-reflect-add=__CUDA_ID_QUERIES_FIT_IN_INT=1"});
} else {
CC1Args.append({"-fcuda-is-device", "-mllvm",
"-enable-memcpyopt-without-libcalls",
Expand Down
16 changes: 16 additions & 0 deletions clang/test/Driver/sycl-nvptx-id-queries-fit-in-int.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,16 @@
// REQUIRES: nvptx-registered-target

// RUN: %clang -### -nocudalib \
// RUN: -fsycl -fsycl-targets=nvptx64-nvidia-cuda %s 2>&1 \
// RUN: | FileCheck --check-prefix=CHECK-DEFAULT %s

// RUN: %clang -### -nocudalib \
// RUN: -fsycl -fsycl-targets=nvptx64-nvidia-cuda -fno-sycl-id-queries-fit-in-int %s 2>&1 \
// RUN: | FileCheck --check-prefix=CHECK-DEFAULT %s

// RUN: %clang -### -nocudalib \
// RUN: -fsycl -fsycl-targets=nvptx64-nvidia-cuda -fsycl-id-queries-fit-in-int %s 2>&1 \
// RUN: | FileCheck --check-prefix=CHECK-INT %s

// CHECK-INT: "-mllvm" "-nvvm-reflect-add=__CUDA_ID_QUERIES_FIT_IN_INT=1"
// CHECK-DEFAULT-NOT: "-nvvm-reflect-add=__CUDA_ID_QUERIES_FIT_IN_INT=1"
14 changes: 14 additions & 0 deletions libclc/libspirv/lib/ptx-nvidiacl/workitem/get_global_id.cl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,17 +8,31 @@

#include <libspirv/spirv.h>

extern int __nvvm_reflect_ocl(constant char *);

_CLC_DEF _CLC_OVERLOAD size_t __spirv_GlobalInvocationId_x() {
if (__nvvm_reflect_ocl("__CUDA_ID_QUERIES_FIT_IN_INT")) {
return (uint)__spirv_WorkgroupId_x() * (uint)__spirv_WorkgroupSize_x() +
(uint)__spirv_LocalInvocationId_x() + (uint)__spirv_GlobalOffset_x();
}
return __spirv_WorkgroupId_x() * __spirv_WorkgroupSize_x() +
__spirv_LocalInvocationId_x() + __spirv_GlobalOffset_x();
}

_CLC_DEF _CLC_OVERLOAD size_t __spirv_GlobalInvocationId_y() {
if (__nvvm_reflect_ocl("__CUDA_ID_QUERIES_FIT_IN_INT")) {
return (uint)__spirv_WorkgroupId_y() * (uint)__spirv_WorkgroupSize_y() +
(uint)__spirv_LocalInvocationId_y() + (uint)__spirv_GlobalOffset_y();
}
return __spirv_WorkgroupId_y() * __spirv_WorkgroupSize_y() +
__spirv_LocalInvocationId_y() + __spirv_GlobalOffset_y();
}

_CLC_DEF _CLC_OVERLOAD size_t __spirv_GlobalInvocationId_z() {
if (__nvvm_reflect_ocl("__CUDA_ID_QUERIES_FIT_IN_INT")) {
return (uint)__spirv_WorkgroupId_z() * (uint)__spirv_WorkgroupSize_z() +
(uint)__spirv_LocalInvocationId_z() + (uint)__spirv_GlobalOffset_z();
}
return __spirv_WorkgroupId_z() * __spirv_WorkgroupSize_z() +
__spirv_LocalInvocationId_z() + __spirv_GlobalOffset_z();
}