TL/MLX5: generate schedule for zcopy allgather #1059
+192
−0
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
samnordmann
reviewed
Jan 6, 2025
nsarka
approved these changes
Jan 6, 2025
|
@samnordmann @nsarka Thanks for the comments! Please let me know if you have any more. I added a new commit. |
MamziB
force-pushed
the
mamzi/mcast-allgather-3-part-2
branch
2 times, most recently
from
1c7ff90 to
2156eac
Compare
samnordmann
approved these changes
Feb 6, 2025
MamziB
force-pushed
the
mamzi/mcast-allgather-3-part-2
branch
from
2156eac to
e83dc20
Compare
|
@samnordmann Thanks for the new comments, I have addressed them. |
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Outdated suggestions cannot be applied.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.
The PR focuses on optimizing allgather communication by leveraging a pipelined multicast mechanism with configurable concurrency and prepost bucket sizes for recv buffers. It implements a new function,
ucc_tl_mlx5_mcast_prepare_zero_copy_allgather, which calculates and sets up a pipelined schedule for the zero-copy allgather. The function validates parameters, allocates schedules, and registers the receive buffers as well.