🔗 Helpful links

• 🧪  See artifacts and rendered test results at hud.pytorch.org/pr/76006
• 📄  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 our office hours
• ↩️  [fb-only] Re-run with SSH instructions

❌ 1 New Failures

As of commit ad73bd7 (more details on the Dr. CI page):

2022-05-07T02:35:03.7439498Z FAILED (errors=1, skipped=110, expected failures=3)
2022-05-07T02:35:03.7439685Z 
2022-05-07T02:35:03.7439803Z Generating XML reports...
2022-05-07T02:35:03.7440229Z Generated XML report: test-reports\python-unittest\test_linalg\TEST-TestLinalgCPU-20220507022901.xml
2022-05-07T02:35:03.7440790Z Generated XML report: test-reports\python-unittest\test_linalg\TEST-TestLinalgCUDA-20220507022901.xml
2022-05-07T02:35:04.5352310Z Traceback (most recent call last):
2022-05-07T02:35:04.5352793Z   File "run_test.py", line 1070, in <module>
2022-05-07T02:35:04.5353050Z     main()
2022-05-07T02:35:04.5353326Z   File "run_test.py", line 1048, in main
2022-05-07T02:35:04.5353630Z     raise RuntimeError(err_message)
2022-05-07T02:35:04.5353894Z RuntimeError: test_linalg failed!
2022-05-07T02:35:04.9393507Z 
2022-05-07T02:35:04.9394477Z (base) C:\actions-runner\_work\pytorch\pytorch\test>popd
2022-05-07T02:35:04.9401267Z 
2022-05-07T02:35:04.9401858Z (base) C:\actions-runner\_work\pytorch\pytorch>if ERRORLEVEL 1 exit /b 1 
2022-05-07T02:35:04.9433590Z + cleanup
2022-05-07T02:35:04.9433895Z + retcode=1
2022-05-07T02:35:04.9434493Z + set +x
2022-05-07T02:35:04.9473072Z ##[error]Process completed with exit code 1.
2022-05-07T02:35:04.9895836Z ##[group]Run pytorch/pytorch/.github/actions/get-workflow-job-id@master
2022-05-07T02:35:04.9896192Z with:

@davidberard98 has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

cc @jjsjann123 did you have a plan on how to flip the switch? I was trying to test it here, but looks like this strategy doesn't work (I think it's due to static initialization order where CUDAHooks needs to get registered before at::globakContext().hasCUDA() will work in cuda_graph_fuser.h)

Do you have another idea on how to do this? I think we could probably try to wrap the registration in some functions and call the functions to enforce registration order... but not sure if there's a nicer way to do it

@davidberard98 has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

@pytorchbot force merge this

(Initiating merge automatically since Phabricator Diff has merged, using force because this PR might not pass merge_rules.json but landed internally)

Merge failed due to Matched rule NVFuser, but PR has not been reviewed yet
Raised by https://github.com/pytorch/pytorch/actions/runs/2301021974

@pytorchbot force merge this

(Initiating merge automatically since Phabricator Diff has merged, using force because this PR might not pass merge_rules.json but landed internally)

Merge failed due to Matched rule NVFuser, but PR has not been reviewed yet
Raised by https://github.com/pytorch/pytorch/actions/runs/2301167574

@pytorchbot force merge this

(Initiating merge automatically since Phabricator Diff has merged, using force because this PR might not pass merge_rules.json but landed internally)

Merge failed due to Matched rule NVFuser, but PR has not been reviewed yet
Raised by https://github.com/pytorch/pytorch/actions/runs/2301240336

@pytorchbot force merge this

(Initiating merge automatically since Phabricator Diff has merged, using force because this PR might not pass merge_rules.json but landed internally)

Merge failed due to Matched rule NVFuser, but PR has not been reviewed yet
Raised by https://github.com/pytorch/pytorch/actions/runs/2301316270

@pytorchbot force merge this

(Initiating merge automatically since Phabricator Diff has merged, using force because this PR might not pass merge_rules.json but landed internally)

Merge failed due to Matched rule NVFuser, but PR has not been reviewed yet
Raised by https://github.com/pytorch/pytorch/actions/runs/2301390729

@pytorchbot force merge this

(Initiating merge automatically since Phabricator Diff has merged, using force because this PR might not pass merge_rules.json but landed internally)

Merge failed due to Matched rule NVFuser, but PR has not been reviewed yet
Raised by https://github.com/pytorch/pytorch/actions/runs/2301689752

@pytorchbot force merge this

(Initiating merge automatically since Phabricator Diff has merged, using force because this PR might not pass merge_rules.json but landed internally)

Merge failed due to Matched rule NVFuser, but PR has not been reviewed yet
Raised by https://github.com/pytorch/pytorch/actions/runs/2301733362

@pytorchbot force merge this

(Initiating merge automatically since Phabricator Diff has merged, using force because this PR might not pass merge_rules.json but landed internally)

Merge failed due to Matched rule NVFuser, but PR has not been reviewed yet
Raised by https://github.com/pytorch/pytorch/actions/runs/2301762179

@pytorchbot force merge this

(Initiating merge automatically since Phabricator Diff has merged, using force because this PR might not pass merge_rules.json but landed internally)

Merge failed due to Matched rule NVFuser, but PR has not been reviewed yet
Raised by https://github.com/pytorch/pytorch/actions/runs/2301795555

@pytorchbot force merge this

(Initiating merge automatically since Phabricator Diff has merged, using force because this PR might not pass merge_rules.json but landed internally)

Merge failed due to Matched rule NVFuser, but PR has not been reviewed yet
Raised by https://github.com/pytorch/pytorch/actions/runs/2301855761

Apologies for the spam here. For Meta engineers the fix is D36284533

retrying in #77213