Skip to content

[rocm] Raise the unittest-rocm-gfx950 timeout to 150 minutes - #23646

Merged
Gasoonjia merged 1 commit into
mainfrom
gasoonjia/rocm-unittest-timeout
Oct 9, 2026
Merged

Gasoonjia merged 1 commit into
mainfrom
gasoonjia/rocm-unittest-timeout

Conversation

@Gasoonjia

@Gasoonjia Gasoonjia commented Oct 9, 2026 •

Copy link
Copy Markdown
Contributor

Since #23603 the QuantizedGemmFamily kernels compile and run on gfx950 instead of failing fast at the Triton compile. As a result unittest-rocm-gfx950 went from ~28 to ~80 minutes (https://github.com/pytorch/executorch/actions/runs/37914177023/job/113766563170: 10:02–11:22, pytest 67.6 min). At the current 90-minute timeout one run was already cancelled (https://github.com/pytorch/executorch/actions/runs/37848283613/job/113554882943).

Where the time goes (gfx950, with #23642):

minutes
test_int4_quantized_gemm.py 46.4 (of which Int4AutotunePickTest 36.6)
test_int8_quantized_gemm.py 4.4
test_int5_quantized_gemm.py 3.4
everything else < 3 each

Int4AutotunePickTest's time is the CPU-saturated AOTI compile, which is what that test exercises.

Change

  • Raise the job's timeout from 90 to 150 minutes. The other gfx950 jobs in this workflow use 180.
  • No test is dropped or split out. Splitting would be the alternative if the extra runner time per PR is acceptable.

Test plan

CI: unittest-rocm-gfx950 completes instead of being cancelled at 90 minutes.

@pytorch-bot

pytorch-bot Bot commented Oct 9, 2026 •

Copy link
Copy Markdown

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/23646

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

⏳ 2 Pending, 1 Unrelated Failure

As of commit 0928594 with merge base 9875560 (image):

BROKEN TRUNK - The following job failed but were present on the merge base:

👉 Rebase onto the `viable/strict` branch to avoid these failures

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

@meta-cla meta-cla Bot added the CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. label Oct 9, 2026
Since #23603 the QuantizedGemmFamily kernels compile and run on gfx950
instead of failing fast at the Triton compile, and unittest-rocm-gfx950 went
from ~28 to ~80 minutes (job 113766563170: 10:02-11:22, pytest 67.6 min).
About 37 of those minutes are Int4AutotunePickTest's CPU-saturated AOTI
compile, which is the point of that test. At a 90-minute timeout one run was
already cancelled.

Raise the job's timeout to 150 minutes rather than drop or split suites. The
other gfx950 jobs in this workflow use 180.
@Gasoonjia
Gasoonjia force-pushed the gasoonjia/rocm-unittest-timeout branch from 501df8a to 0928594 Compare October 9, 2026 12:30
@github-actions

github-actions Bot commented Oct 9, 2026

Copy link
Copy Markdown

This PR needs a release notes: label

If your change should be included in the release notes (i.e. would users of this library care about this change?), please use a label starting with release notes:. This helps us keep track and include your important work in the next release notes.

To add a label, you can comment to pytorchbot, for example
@pytorchbot label "release notes: none"

For more information, see
https://github.com/pytorch/pytorch/wiki/PyTorch-AutoLabel-Bot#why-categorize-for-release-notes-and-how-does-it-work.

@Gasoonjia
Gasoonjia marked this pull request as ready for review October 9, 2026 19:01
@Gasoonjia
Gasoonjia merged commit a43bac0 into main Oct 9, 2026
@Gasoonjia
Gasoonjia deleted the gasoonjia/rocm-unittest-timeout branch October 9, 2026 19:50

This branch was successfully deployed

1 active deployment
cadence — 0928594e Deployed Oct 9, 2026 by Gasoonjia
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

ciflow/rocm CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. module: rocm

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants