mpich
mpich copied to clipboard
ofi: fix a bug in gpu pipelining with potentially out of order chunks
Receiver posts receive for each chunk, and write to GPU buffers in the order of the receive event completion. This however can potentially leads to out of order writing because the completion events may be out of order.
Pull Request Description
Author Checklist
- [x] Provide Description Particularly focus on why, not what. Reference background, issues, test failures, xfail entries, etc.
- [ ] Commits Follow Good Practice
Commits are self-contained and do not do two things at once.
Commit message is of the form:
module: short descriptionCommit message explains what's in the commit. - [ ] Passes All Tests Whitespace checker. Warnings test. Additional tests via comments.
- [ ] Contribution Agreement For non-Argonne authors, check contribution agreement. If necessary, request an explicit comment from your companies PR approval manager.