Restore RNG states for composable reentrant activation checkpointing

[mrshenli]

Stack from ghstack (oldest at bottom):

• -> Restore RNG states for composable reentrant activation checkpointing #91265

This allows ops like randperm to behave the same during re-computation.

Differential Revision: D42196758

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/91265

• 📄 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.

❌ 2 Failures

As of commit 39c18fc:

NEW FAILURES - The following jobs have failed:

• linux-focal-rocm5.3-py3.8 / test (default, 1, 2, linux.rocm.gpu)
• linux-focal-rocm5.3-py3.8 / test (default, 2, 2, linux.rocm.gpu)

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

[mrshenli]

rocm failures are irrelevant:

2022-12-21T22:41:05.7702218Z Running test_cpp_extensions_jit ... [2022-12-21 22:41:05.713862]
2022-12-21T22:41:05.7702710Z Executing ['/opt/conda/bin/python', '-bb', 'test_cpp_extensions_jit.py', '-v', '--import-slow-tests', '--import-disabled-tests'] ... [2022-12-21 22:41:05.714987]
2022-12-21T22:42:24.2644440Z ##[error]The runner has received a shutdown signal. This can happen when the runner service is stopped, or a manually started runner is canceled.
2022-12-21T22:42:29.3919178Z ##[error]The operation was canceled.

[facebook-github-bot]

@pytorchbot merge

(Initiating merge automatically since Phabricator Diff has merged)

[pytorchmergebot]

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

Advanced Debugging

Check the merge workflow status
here