Make arbor-build-catalogue part of pyarb. #2416
Open
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
boeschf
reviewed
Oct 16, 2024
Comment on lines
+39
to
+57
| @property | ||
| def bin(self): | ||
| return self._bin | ||
|
|
||
| @bin.setter | ||
| def bin(self, val): | ||
| self._pakdir = None | ||
| self._bindir = None | ||
| self._bin = val | ||
|
|
||
| @property | ||
| def lib(self): | ||
| return self._lib | ||
|
|
||
| @lib.setter | ||
| def lib(self, val): | ||
| self._pakdir = None | ||
| self._bindir = None | ||
| self._lib = val |
|
|
||
| def make_gpu_support(gpu, cxx): | ||
| if gpu == "cuda": | ||
| gpu_support = f""" | ||
| include(FindCUDAToolkit) | ||
| add_compile_definitions(ARB_CUDA) | ||
| add_compile_definitions(ARB_HAVE_GPU) | ||
| find_package(CUDAToolkit) | ||
| enable_language(CUDA) | ||
| set(CMAKE_CUDA_STANDARD 14) |
| print("CUDA-Clang support is currently considered experimental only.") | ||
| gpu_support = f""" | ||
| elif gpu == "cuda-clang": | ||
| gpu_support = f""" | ||
| add_compile_definitions(ARB_CUDA) | ||
| add_compile_definitions(ARB_HAVE_GPU) | ||
| find_package(CUDAToolkit) | ||
| enable_language(CUDA) | ||
| set(CMAKE_CUDA_STANDARD 14) |
| add_compile_definitions(ARB_CUDA) | ||
| add_compile_definitions(ARB_HAVE_GPU) | ||
| find_package(CUDAToolkit) | ||
| enable_language(CUDA) | ||
| set(CMAKE_CUDA_STANDARD 14) | ||
| add_compile_options(-xcuda --cuda-gpu-arch=sm_60 --cuda-gpu-arch=sm_70 --cuda-gpu-arch=sm_80 --cuda-path=${CUDA_TOOLKIT_ROOT_DIR})) | ||
| add_compile_options(-xcuda --cuda-gpu-arch=sm_60 --cuda-gpu-arch=sm_70 --cuda-gpu-arch=sm_80 --cuda-path=${{CUDA_TOOLKIT_ROOT_DIR}})) |
There was a problem hiding this comment.
probably worth to add sm_90
boeschf
reviewed
Oct 16, 2024
There was a problem hiding this comment.
I thought I commented on this file already, but seems my comment has evaporated... Anyhow: I am not fond of the long inline CMake strings in here - ideally we should have a single source of truth so a) we don't diverge and b) we can test. Not sure how feasible this is, though
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Integrate a-b-c into Arbor's Python bindings.
Currently, this is not officially supported, thus undocumented, due to the rather crude interface.
I do plan on changing this, though, in the midterm future, and make both a-b-c and modcc callable
as Python functions. For now these modules just serve as entry points for
pyproject/pip.No longer use (and install)
BuildModules.cmakeina-b-c.Skip tests requiring a dynamic catalogue if it cannot be built. Most likely these issue occur due to the
dependency of
Tests > a-b-c > CMake InstallCloses #2395