[inductor] use smaller RBLOCK for expensive reduction kernels #126477
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
[ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/126477
Note: Links to docs will display an error until the docs builds have been completed. ✅ You can merge normally! (4 Unrelated Failures)As of commit 32e418d with merge base 53f73cd ( BROKEN TRUNK - The following jobs failed but were present on the merge base:👉 Rebase onto the `viable/strict` branch to avoid these failures
UNSTABLE - The following job failed but was likely due to flakiness present on trunk and has been marked as unstable:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
…els" cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx peterbell10 ipiszy yf225 chenyang78 kadeng muchulee8 ColinPeppler amjames desertfire chauhang [ghstack-poisoned]
…els" Triton sometimes uses less registers for more expensive kernel which results in worse perf ( #126463 ). This may make inductor end up with a sub-optimal config. Use a smaller max RBLOCK if the reduction potentially need many registers. Will run perf test.. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx peterbell10 ipiszy yf225 chenyang78 kadeng muchulee8 ColinPeppler amjames desertfire chauhang [ghstack-poisoned]
eellison
reviewed
May 17, 2024
| @@ -1480,12 +1485,32 @@ def _reduction_configs( | |||
| assert len(size_hints) == 2 | |||
| rnumel = size_hints[-1] | |||
|
|
|||
| MAX_RBLOCK = 2048 | |||
| if ( | |||
There was a problem hiding this comment.
were there any register spills on old heuristics here ?
There was a problem hiding this comment.
no. I checked CompiledKernel.n_spills, it's 0.
jansel
approved these changes
May 18, 2024
…els" Triton sometimes uses less registers for more expensive kernel which results in worse perf ( #126463 ). This may make inductor end up with a sub-optimal config. Use a smaller max RBLOCK if the reduction potentially need many registers. Will run perf test.. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx peterbell10 ipiszy yf225 chenyang78 kadeng muchulee8 ColinPeppler amjames desertfire chauhang [ghstack-poisoned]
|
Perf test on this PR is roughly neutral. I'll land this one and debug the perf/compilation time regress for the PR tuning partitioner. |
|
@pytorchbot merge |
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
desai0007
pushed a commit
to desai0007/test-repo-pytorch
that referenced
this pull request
Feb 26, 2025
ghstack-source-id: 1456e6a Pull Request resolved: pytorch/pytorch#126477
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.
Stack from ghstack (oldest at bottom):
Triton sometimes uses less registers for more expensive kernel which results in worse perf ( #126463 ). This may make inductor end up with a sub-optimal config. Use a smaller max RBLOCK if the reduction potentially need many registers.
Will run perf test..
cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @peterbell10 @ipiszy @yf225 @chenyang78 @kadeng @muchulee8 @ColinPeppler @amjames @desertfire @chauhang