| FazBrowse GitHub Viewer | Trending | | Home |
| Tools: [Download Repo ZIP] [Original HTTPS Page] |
Sorry, something went wrong.
Signed-off-by: Francisco Javier Arceo <farceo@redhat.com>
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
Sorry, something went wrong.
|
⚠️ Please install the Codecov Report✅ All modified and coverable lines are covered by tests. @@ Coverage Diff @@
## master #6835 +/- ##
==========================================
+ Coverage 47.08% 47.10% +0.01%
==========================================
Files 419 419
Lines 51877 51879 +2
Branches 7525 7525
==========================================
+ Hits 24428 24436 +8
+ Misses 25700 25693 -7
- Partials 1749 1750 +1
Continue to review full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
Sorry, something went wrong.
Signed-off-by: Francisco Javier Arceo <farceo@redhat.com>
| Back | FazBrowse Home | New Git URL |
What this PR does / why we need it:
The FIPS import-order test exceeded its 60-second subprocess deadline in the Python 3.11/macOS unit-test job for #6188. Give the fresh Feast import 180 seconds to complete under CI load while preserving the FIPS and import-order assertions.
Schedule a faulthandler stack dump after 120 seconds and include captured stdout/stderr in timeout failures. This provides evidence of where a future slow or stalled import is spending time. Handle missing output and decode captured bytes with replacement for invalid characters.
Linux CI also fails before tests because --torch-backend cpu selects CPU wheels whose hashes are absent from the existing locks. Add the published CPU wheel hashes for the pinned torch and torchvision versions to all three CI lockfiles, retaining the existing versions and PyPI hashes. These hashes were generated with uv pip compile --no-deps --torch-backend cpu --universal --generate-hashes for the existing pins. The broader lock-generation changes are tracked in #6825.
Which issue(s) this PR fixes:
Mitigates the timeout observed in this CI job. The original log does not identify the stalled import, so the longer deadline is a mitigation rather than proof of the underlying cause.
Checks
Testing Strategy
Ruff lint and formatting checks passed for the changed file. Scoped mypy (--follow-imports=silent tests/unit/test_offline_server.py) also passed.