[Inductor] Add 0 initialization to Triton masked loads #127311
Closed
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
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/127311
Note: Links to docs will display an error until the docs builds have been completed. ❗ 1 Active SEVsThere are 1 currently active SEVs. If your PR is affected, please view them below: ✅ You can merge normally! (2 Unrelated Failures)As of commit 11e6287 with merge base cbb79a2 ( UNSTABLE - The following jobs failed but were 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. |
soulitzer
added
the
triaged
jansel
approved these changes
May 29, 2024
|
@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 |
Merge failedReason: 1 jobs have failed, first few of them are: trunk / macos-13-py3-arm64 / test (default, 1, 3, macos-m1-stable) Details for Dev Infra teamRaised by workflow job |
|
@pytorchbot merge -i |
Merge startedYour change will be merged while ignoring the following 2 checks: pull / linux-focal-cuda12.4-py3.10-gcc9-sm86 / test (default, 4, 5, linux.g5.4xlarge.nvidia.gpu, unstable), inductor / linux-jammy-cpu-py3.8-gcc11-inductor / test (inductor_torchbench_cpu_smoketest_perf, 1, 1, linux.24xl.spr-metal, unstable) Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
pytorch-bot bot
pushed a commit
that referenced
this pull request
May 30, 2024
For a masked `tl.load` operation, the Triton language specifies that values masked out (i.e. where the mask evaluates to false) are undefined in the output of the load. Triton provides an optional `other` parameter which, when included, provides an explicit value to use for masked out values from the load. If the output from a masked load without the `other` parameter is used in a conditional, unexpected behavior can occur. Despite the language specification, all Triton backends currently in use by PyTorch Inductor (NVIDIA, AMD, and Intel) 0-initialize masked loads if `other` is not present (we recently changed the Intel backend behavior to match NVIDIA and AMD because that's what our users expect, even if we are not following the Triton spec to the tee). This PR attempts to "future-proof" Inductor for new backends (or perhaps changes in the current backends? - we did not see any performance change from 0-initializing in the Intel XPU backend but one could imagine compiler optimizations to remove paths that depend on undefined) to add an explicit `other` in instances where later conditionals depend on the `tl.load` output. I also removed an exception to `other` behavior for boolean loads, which was put in place for a Triton bug that should be fixed. I added `other` to the getting started documentation as a clue that masked load behavior requires explicit initialization if, even though I don't expect `undef` values to cause the example code to fail if the underlying output is not 0-initialized. Finally, I added other to the `make_load` function in `select_algorithm.py`, though I wasn't able to determine if that function was actually being called. Fixes #126535 Pull Request resolved: #127311 Approved by: https://github.com/jansel
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 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.
For a masked
tl.loadoperation, the Triton language specifies that values masked out (i.e. where the mask evaluates to false) are undefined in the output of the load. Triton provides an optionalotherparameter which, when included, provides an explicit value to use for masked out values from the load. If the output from a masked load without theotherparameter is used in a conditional, unexpected behavior can occur.Despite the language specification, all Triton backends currently in use by PyTorch Inductor (NVIDIA, AMD, and Intel) 0-initialize masked loads if
otheris not present (we recently changed the Intel backend behavior to match NVIDIA and AMD because that's what our users expect, even if we are not following the Triton spec to the tee). This PR attempts to "future-proof" Inductor for new backends (or perhaps changes in the current backends? - we did not see any performance change from 0-initializing in the Intel XPU backend but one could imagine compiler optimizations to remove paths that depend on undefined) to add an explicitotherin instances where later conditionals depend on thetl.loadoutput. I also removed an exception tootherbehavior for boolean loads, which was put in place for a Triton bug that should be fixed. I addedotherto the getting started documentation as a clue that masked load behavior requires explicit initialization if, even though I don't expectundefvalues to cause the example code to fail if the underlying output is not 0-initialized. Finally, I added other to themake_loadfunction inselect_algorithm.py, though I wasn't able to determine if that function was actually being called.Fixes #126535
cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @peterbell10 @ipiszy @yf225 @chenyang78 @kadeng @muchulee8 @ColinPeppler @amjames @desertfire @chauhang