Force smcuda BTL to use CUDA IPC path for all GPU buffers where possible #698
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
Member
|
👍 |
bosilca
added a commit
that referenced
this pull request
Jul 9, 2015
Force smcuda BTL to use CUDA IPC path for all GPU buffers where possible
jsquyres
added a commit
to jsquyres/ompi
that referenced
this pull request
Nov 10, 2015
…plat_ucx BUILD: add support for UCX detection
cinek810
added a commit
to cinek810/ompi
that referenced
this pull request
Dec 14, 2020
Signed-of-by: Marcin Stolarek<cinek@schedmd.com>
cinek810
added a commit
to cinek810/ompi
that referenced
this pull request
Dec 14, 2020
Signed-of-by: Marcin Stolarek<stolarek.marcin@gmail.com>
cinek810
added a commit
to cinek810/ompi
that referenced
this pull request
Dec 14, 2020
Signed-off-by: Marcin Stolarek<stolarek.marcin@gmail.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.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 change leverages some features in the CUDA-aware path to force the smcuda to use CUDA IPC for all size messages rather than just larger ones. This is how most users are expecting it to work.
If you get a change, @bosilca can look at this. Plan to put in 2.x as well.