Fix CI MPICH issues#258
Merged
Merged
Conversation
caitlinross
force-pushed
the
ci-mpich-issues
branch
2 times, most recently
from
70ef96b to
6c8281a
Compare
make ci tests always run in verbose mode, in case we have situations where tests return success even though they shouldn't. this way we can go back to ci logs and see the output.
caitlinross
force-pushed
the
ci-mpich-issues
branch
from
6c8281a to
fab21ae
Compare
Codecov Report✅ All modified and coverable lines are covered by tests. 📢 Thoughts on this report? Let us know! |
this lets us check if an mpi issue is a codes problem or a CI machine problem
Ubuntu 24.04's apt mpich (4.2.0-5build3, ch4:ucx) is broken on the GitHub
runners in two ways: UCX aborts MPI_Init on the runner's verbs device
("ibv_create_srq() failed"), and even past that mpiexec launches size-1
singletons so optimistic (--sync=3) tests silently run sequential. There's no
fixed mpich in the noble archive, so the affected legs run inside a container
image that ships a source-built mpich (ch4:ofi -- libfabric/TCP, no UCX/verbs)
that forms a real multi-rank world.
- build matrix: a conditional container (empty string => no container, so every
non-24.04-mpich leg runs on the runner exactly as before).
- sanitizers + coverage: unconditional container (both are single-config
24.04 mpich); drop their now-redundant apt install steps. Keeping the
sanitizer lanes on mpich (quiet under ASan/UBSan) now actually exercises the
optimistic paths.
- old-ross-compat is left on apt mpich: it only builds, never runs mpiexec.
The image (codes-ci-ubuntu24-mpich) and its publish workflow live on a separate
branch; it must be published to GHCR and made public before these legs can pull
it. mpi-multirank-sanity going green confirms the launcher forms a real world.
caitlinross
force-pushed
the
ci-mpich-issues
branch
from
fab21ae to
260a273
Compare
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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.
This PR confirms that the image added in #260 fixes the CI issues that popped up today (after a couple of additional PRs to add missed dependencies #261 and #262).