Improve cuda::memcpy_async codegen - #5996
Merged
bernhardmgruber merged 12 commits intoOct 27, 2025
Merged
Conversation
Contributor
|
Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually. Contributors can view more details about this message here. |
cuda::memcpy_async to improve codegen
bernhardmgruber
force-pushed
the
memcpy_async_hacks
branch
from
September 23, 2025 20:53
5707d3c to
6487733
Compare
bernhardmgruber
commented
Sep 23, 2025
Comment on lines
+62
to
+65
| (if (__elect_from_group(__g)) { | ||
| ::cuda::ptx::cp_async_bulk( | ||
| ::cuda::ptx::space_cluster, ::cuda::ptx::space_global, __dest, __src, __size, __bar_handle); | ||
| ::cuda::ptx::mbarrier_arrive_expect_tx(ptx::sem_release, ptx::scope_cta, ptx::space_shared, __bar_handle, __size); |
Contributor
Author
There was a problem hiding this comment.
We put the transaction count update into the same branch as the bulk copy. Improves codegen.
Contributor
Author
|
/ok to test 446535d |
This comment has been minimized.
This comment has been minimized.
miscco
reviewed
Sep 24, 2025
bernhardmgruber
force-pushed
the
memcpy_async_hacks
branch
from
September 24, 2025 15:10
446535d to
bf64704
Compare
bernhardmgruber
force-pushed
the
memcpy_async_hacks
branch
from
October 3, 2025 14:08
bf64704 to
cce3315
Compare
Contributor
Author
|
/ok to test cce3315 |
This comment has been minimized.
This comment has been minimized.
bernhardmgruber
force-pushed
the
memcpy_async_hacks
branch
3 times, most recently
from
October 23, 2025 22:36
6cf386b to
5f4e3f9
Compare
cuda::memcpy_async to improve codegencuda::memcpy_async codegen
bernhardmgruber
marked this pull request as ready for review
October 23, 2025 23:29
This comment has been minimized.
This comment has been minimized.
miscco
reviewed
Oct 24, 2025
miscco
approved these changes
Oct 27, 2025
…ync.rst Co-authored-by: Michael Schellenberger Costa <miscco@nvidia.com>
…obal.h Co-authored-by: Michael Schellenberger Costa <miscco@nvidia.com>
bernhardmgruber
force-pushed
the
memcpy_async_hacks
branch
from
October 27, 2025 07:50
c24f4e7 to
4e4700f
Compare
bernhardmgruber
enabled auto-merge (squash)
October 27, 2025 07:53
This comment has been minimized.
This comment has been minimized.
Contributor
🥳 CI Workflow Results🟩 Finished in 2h 37m: Pass: 100%/89 | Total: 18h 11m | Max: 1h 05m | Hits: 97%/212780See results here. |
1 task
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
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 PR improves the codegen for
cuda::memcpy_asyncby ensuring a uniform data path for the bulk copy instruction and updating the mbarrier's tx count in the same branch.Fixes: #5995
(SASS of
cuda::memcpy_asyncversion is now identical to hand-writtencuda::ptx)