{2023.06} foss/2022b #309
Merged
{2023.06} foss/2022b #309
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
|
Instance
|
|
bot: build repo:eessi-2023.06-software arch:x86_64/generic |
Updates by the bot instance
|
|
bot: build repo:eessi-2023.06-software arch:x86_64/generic |
Updates by the bot instance
|
|
New job on instance
|
|
bot: build repo:eessi-2023.06-software arch:x86_64/generic |
Updates by the bot instance
|
|
New job on instance
|
|
bot: build repo:eessi-2023.06-software arch:x86_64/intel/haswell |
Updates by the bot instance
|
|
New job on instance
|
|
New job on instance
|
|
New job on instance
|
|
New job on instance
|
|
New job on instance
|
|
New job on instance
|
|
New job on instance
|
|
The failure on |
So apparently the increase to 300 here is not enough for this version. Do we increase it a bit more (don't know how much sense it makes to just ignore more and more tests?)? |
|
I don't know either. I guess we can, but it does make one wonder: if numerical results are different, how different are they, and is that still acceptable? We should at the very least check that all failures are numerical failures, I guess. I think the more fundamental question is: what should are tests guarantee?
It's probably good to at least report an issue upstream, see what they say. I'd assume the devs are more adept at judging whether these numerical inconsistencies should be considered problematic or not. One issue with taking the first approach is that issues may also pop up in other packages, such as #306 |
|
More detailed error: So it is all numerical failures. That is a little bit encouraging... |
|
I managed to go into the singularity container by unpacking the tarball at doing a and find the detailed testing results at Now, I'm no expert in these tests, and have no idea what this means. It does not look like small numerical errors to my untrained eye, but I might be completely wrong... |
|
@casparvl To avoid getting stuck on this even longer, I think we should:
We should take a similar approach to unblock other PRs ( We can't reasonable expect that we'll figure out each and every failing test for all software we'll install, especially because we known that some test suite are quite buggy themselves (PyTorch comes to mind, but this also applies to LAPACK, see @bartoldeman's PR which dealt with a non-numerical failure in the LAPACK test suite, see also issue #18017). This procedure should probably be documented as well, and even become part of the contribution policy, with some rules of thumb on when this approach is acceptable (how many failing tests we see, on how many CPU targets, etc.). |
|
bot: build repo:eessi-2023.06-software arch:aarch64/neoverse_v1 |
Updates by the bot instance
|
|
New job on instance
|
|
Created issue for the OpenBLAS test failures: |
boegel
reviewed
Aug 8, 2023
Merged
boegel
added
the
bot:deploy
boegel
approved these changes
Aug 9, 2023
trz42
pushed a commit
to trz42/software-layer
that referenced
this pull request
Apr 7, 2024
remove Lmod cache update
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.
No description provided.