CMake: overwork std::mdspan support for nvcc
#2445
Merged
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
SimeonEhrig
added
Type:Enhancement
Backend:CUDA
Type:CMake
SimeonEhrig
force-pushed
the
mdspanCUDAExperimentalLamba
branch
2 times, most recently
from
e03c7b8 to
c46c15c
Compare
…AMBDA is disabled - Instead automatically disabling mdspan if alpaka_CUDA_EXPT_EXTENDED_LAMBDA is off throw an error. That's more expected behavior than an enabled feature is off. Also the old behavior does not worked, because mdspan was still compiled and throws a compiler error. - Fix bug, that the Clang as CUDA compiler was also affected by the CMake options alpaka_CUDA_EXPT_EXTENDED_LAMBDA, which sets a nvcc flag.
SimeonEhrig
force-pushed
the
mdspanCUDAExperimentalLamba
branch
from
c46c15c to
fbe8081
Compare
SimeonEhrig
changed the title
SimeonEhrig
changed the title
std::mdspan support for nvcc
psychocoderHPC
approved these changes
Jan 9, 2025
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Outdated suggestions cannot be applied.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.
TLTR;
MdSpan throws a readable compiler error if the nvcc experimental extension is disabled. Actual the CMake should avoid it, because it should automatically disable mdspan but it does not work. Also the behavior feels wrong. An explicit enabled feature is automatically disabled. I think it is better, if the user knows, that is not working and ether disable mdspan or find another solution.
Also the PR enables using nvcc with host compiler clang and mdspan. This was disabled with initial support of mdspan. There is no documentation about what was the actual compiler problem. For me, it was less work to remove the restriction in CMake than writing a job generator filter rule. So tried it and the test pass. So I will keep it, because it is the better solution.
Compile time error message: https://gitlab.com/hzdr/crp/alpaka/-/jobs/8788083447
separated from PR: #2315