Ensure all MPI communication to is handled by master thread #3090
Conversation
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
MPI calls have to be done from the *same* thread (`master`) instead of from a `single` thread in MPI_FUNNELED mode according to the specifications.
heplesser
approved these changes
Feb 5, 2024
Contributor
heplesser
left a comment
heplesser
left a comment
There was a problem hiding this comment.
@terhorstd Thank you for this fix. While strictly speaking the resize calls could be done in single, since they do not perform MPI communication, it makes sense to place them in master so we are certain that the buffer data structures always are allocated on the thread performing MPI communication.
mlober
approved these changes
Feb 6, 2024
heplesser
changed the title
terhorstd
added
T: Bug
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Outdated suggestions cannot be applied.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.
According to the MPI specifications, calls have to be done from the same thread (
master) instead of from asinglethread in MPI_FUNNELED mode . This means a number ofomp singlepragmas needed to be changed toomp master. This should not impact performance overly much, but needs checking!omp master, I added explicit ones where not given anyway.