🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/111377

• 📄 Preview Python docs built from this PR
• 📄 Preview C++ docs built from this PR
• ❓ Need help or want to give feedback on the CI? Visit the bot commands wiki or our office hours

Note: Links to docs will display an error until the docs builds have been completed.

✅ You can merge normally! (3 Unrelated Failures)

As of commit a0731c0 with merge base 7fe51e3 ():

This comment was automatically generated by Dr. CI and updates every 15 minutes.

@pytorchbot merge

Merge started

Your 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 failed

Reason: Command git -C /home/runner/work/pytorch/pytorch rebase origin/main returned non-zero exit code 1

Rebasing (1/1)
Auto-merging torch/utils/__init__.py
CONFLICT (content): Merge conflict in torch/utils/__init__.py
error: could not apply fee9e3f5f92... Add `torch.utils.deterministic.fill_uninitialized_memory` flag (#111377)
hint: Resolve all conflicts manually, mark them as resolved with
hint: "git add/rm <conflicted_files>", then run "git rebase --continue".
hint: You can instead skip this commit: run "git rebase --skip".
hint: To abort and get back to the state before "git rebase", run "git rebase --abort".
Could not apply fee9e3f5f92... Add `torch.utils.deterministic.fill_uninitialized_memory` flag (#111377)

@pytorchbot merge

Merge started

Your 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

This is causing test/profiler/test_profiler.py: test_tensor_properties to timeout and fail internally. Any chance you can help take a look? Thank you.

https://github.com/pytorch/pytorch/blob/main/test/profiler/test_profiler.py#L2308

@pytorchbot revert -m="Diff reverted internally" -c="ghfirst"

This Pull Request has been reverted by a revert inside Meta. To re-land this change, please open another pull request, assign the same reviewers, fix the CI failures that caused the revert and make sure that the failing CI runs on the PR by applying the proper ciflow label (e.g., ciflow/trunk).)

@pytorchbot successfully started a revert job. Check the current status here.
Questions? Feedback? Please reach out to the PyTorch DevX Team

@kurtamohler your PR has been successfully reverted.

Thanks for letting me know, I will take a look this week

@aaronenyeshi, when I run python test/profiler/test_profiler.py -k test_tensor_properties using the changes from this PR, I don't get any timeout. In fact, it seems that none of the tests in test/profiler/test_profiler.py even turn deterministic mode on.

I tried adding torch.use_deterministic_algorithms(True) to the top of the test/profiler/test_profiler.py file after the imports, but that didn't cause any timeouts on my machine either. However, it did cause a few of the other tests in that file (but not the test_tensor_properties test itself) to fail due to a different pattern of memory allocations in deterministic vs nondeterministic mode. Those failing tests explicitly check memory allocations against an expected pattern.

Are you sure that this PR is really the reason for the internal failure? Maybe it's just a flaky failure. Or maybe the internal testing infrastructure is doing something to turn deterministic mode on before running some of the profiler tests. Could you verify that deterministic mode is actually turned on when the internal CI runs the test_tensor_properties test? If the test is not being run with deterministic mode turned on, I'm not sure how this PR could be the cause of the failure

Sorry, I think its unrelated, and the automatic bisect flagged this PR. It looks like that test is still having timeout issues. See D50763398 . Please feel free to re-land this diff. Sorry for the inconvenience!

Also, it seems that its failing only in dev-tsan mode, rather than opt or dev modes. https://www.internalfb.com/intern/test/281475086418213/

@pytorchbot merge

Merge started

Your 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

Huh, for some reason CI did not rerun after I rebased, so the merge command merged this immediately. I guess it's probably ok, since this PR passed CI before

@pytorchbot drci

(please ignore this comment, I'm testing Dr.CI)