Skip to content

Add hipstdpar support to BabelStream #195

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

Closed
gsitaram wants to merge 5 commits into from
Closed

Add hipstdpar support to BabelStream #195

Show file tree
Hide file tree
Changes from 3 commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
9d4cc72
Add support for hipstdpar
gsitaram Apr 29, 2024
9f67c5f
Remove --hipstdpar-path as it is not needed with ROCm 6.1 onwards
gsitaram Apr 30, 2024
6bd658c
Merge pull request #1 from gsitaram/hipstdpar
afanfa Apr 30, 2024
7e5c410
Change definition names to be more specific to AMD GPUs
gsitaram Jun 13, 2024
cf2e1b3
Merge pull request #2 from gsitaram/hipstdpar
gsitaram Jun 20, 2024
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
24 changes: 24 additions & 0 deletions src/std-data/model.cmake
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -35,6 +35,21 @@ register_flag_optional(USE_ONEDPL
This requires the DPC++ compiler (other SYCL compilers are untested), required SYCL flags are added automatically."
"OFF")

register_flag_optional(CLANG_OFFLOAD
"Enable offloading support (via the non-standard `-stdpar`) for
Clang/LLVM. The values are AMDGPU processors (https://www.llvm.org/docs/AMDGPUUsage.html#processors)
which will be passed in via `--offload-arch=` argument.

Example values are:
gfx906 - Compile for Vega20 GPUs
gfx908 - Compile for CDNA1 GPUs
gfx90a - Compile for CDNA2 GPUs
gfx942 - Compile for CDNA3 GPUs
gfx1030 - Compile for RDNA2 NV21 GPUs
gfx1100 - Compile for RDNA3 NV31 GPUs"
"")


macro(setup)
set(CMAKE_CXX_STANDARD 17)
if (NVHPC_OFFLOAD)
Expand All @@ -50,4 +65,13 @@ macro(setup)
register_definitions(USE_ONEDPL)
register_link_library(oneDPL)
endif ()
if (CLANG_OFFLOAD)
set(CLANG_FLAGS --hipstdpar --offload-arch=${CLANG_OFFLOAD})
if (NOT CLANG_OFFLOAD MATCHES "^gfx9")
list(APPEND CLANG_FLAGS --hipstdpar-interpose-alloc)
endif ()

register_append_cxx_flags(ANY ${CLANG_FLAGS})
register_append_link_flags(--hipstdpar)
endif ()
endmacro()
25 changes: 25 additions & 0 deletions src/std-indices/model.cmake
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -35,6 +35,22 @@ register_flag_optional(USE_ONEDPL
This requires the DPC++ compiler (other SYCL compilers are untested), required SYCL flags are added automatically."
"OFF")

register_flag_optional(CLANG_OFFLOAD
"Enable offloading support (via the non-standard `-stdpar`) for
Clang/LLVM. The values are AMDGPU processors (https://www.llvm.org/docs/AMDGPUUsage.html#processors)
which will be passed in via `--offload-arch=` argument.

Example values are:
gfx906 - Compile for Vega20 GPUs
gfx908 - Compile for CDNA1 GPUs
gfx90a - Compile for CDNA2 GPUs
gfx942 - Compile for CDNA3 GPUs
gfx1030 - Compile for RDNA2 NV21 GPUs
gfx1100 - Compile for RDNA3 NV31 GPUs"
"")



macro(setup)
set(CMAKE_CXX_STANDARD 17)
if (NVHPC_OFFLOAD)
Expand All @@ -50,4 +66,13 @@ macro(setup)
register_definitions(USE_ONEDPL)
register_link_library(oneDPL)
endif ()
if (CLANG_OFFLOAD)

@tomdeakin tomdeakin May 13, 2024

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

I'm not totally sold on the name of this definition as CLANG is not specific to AMD, but the CLANG_FLAGS set below are definitely specific. Can we rename this please to something appropriate?

@gsitaram gsitaram May 13, 2024 •
edited
Loading

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

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

Would replacing CLANG_OFFLOAD with AMDGPU_TARGET_OFFLOAD and CLANG_FLAGS with AMDGPU_TARGET_OFFLOAD_FLAGS be acceptable?

@gsitaram gsitaram May 28, 2024

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

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

Hi @tomdeakin, please confirm if the above suggestion would work or recommend something better. I can make the changes accordingly.

@tomdeakin tomdeakin Sep 5, 2025

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

LGTM

set(CLANG_FLAGS --hipstdpar --offload-arch=${CLANG_OFFLOAD})
if (NOT CLANG_OFFLOAD MATCHES "^gfx9")
list(APPEND CLANG_FLAGS --hipstdpar-interpose-alloc)
endif ()

register_append_cxx_flags(ANY ${CLANG_FLAGS})
register_append_link_flags(--hipstdpar)
endif ()
endmacro()