Skip to content

make LLVM easyblock aware of amdgcn_capabilities build option #3824

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 3 commits into from
Jul 25, 2025
Merged

make LLVM easyblock aware of amdgcn_capabilities build option #3824

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
ebab7a8
llvm: Use amdgcn_capabilities instead of amd_gfx_list
Thyre Jul 3, 2025
20a9b2d
Merge branch 'develop' into llvm-use-amdgcn-capabilities
Thyre Jul 7, 2025
aca170d
LLVM: Adopt use of template value list for AMDGPU
Thyre Jul 7, 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
17 changes: 15 additions & 2 deletions easybuild/easyblocks/l/llvm.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -204,7 +204,7 @@ class EB_LLVM(CMakeMake):
def extra_options():
extra_vars = CMakeMake.extra_options()
extra_vars.update({
'amd_gfx_list': [None, "List of AMDGPU targets to build for.", CUSTOM],
'amd_gfx_list': [None, "DEPRECATED, list of AMDGPU targets to build for.", CUSTOM],
'assertions': [False, "Enable assertions. Helps to catch bugs in Clang.", CUSTOM],
'bootstrap': [True, "Build LLVM-Clang using itself", CUSTOM],
'build_bolt': [False, "Build the LLVM bolt binary optimizer", CUSTOM],
Expand Down Expand Up @@ -383,9 +383,22 @@ def llvm_src_dir(self):
return self.start_dir

def _configure_build_targets(self):
# list of CUDA compute capabilities to use can be specifed in two ways (where (2) overrules (1)):
# (1) in the easyconfig file, via the custom cuda_compute_capabilities;
# (2) in the EasyBuild configuration, via --cuda-compute-capabilities configuration option;
# Similar rules apply for AMDGCN capabilities
cuda_cc_list = self.cfg.get_cuda_cc_template_value("cuda_cc_space_sep", required=False).split()
cuda_toolchain = hasattr(self.toolchain, 'COMPILER_CUDA_FAMILY')
amd_gfx_list = self.cfg['amd_gfx_list'] or []
amd_gfx_list = self.cfg.get_amdgcn_cc_template_value("amdgcn_cc_space_sep", required=False).split()
if self.cfg['amd_gfx_list'] is not None:
self.log.deprecated("Use of easyconfig parameter 'amd_gfx_list', "
"replace by build option 'amdgcn_capabilities'", '6.0')
if not amd_gfx_list:
amd_gfx_list = self.cfg['amd_gfx_list']
else:
# Do not overwrite value set via the new build option
print_warning("Both 'amd_gfx_list' and build option 'amdgcn_capabilities' "
"are set, please use only the build option 'amdgcn_capabilities'")

# List of (lower-case) dependencies
self.deps = [dep['name'].lower() for dep in self.cfg.dependencies()]
Expand Down