Arm/Cortex-M backend: Share deterministic test seed handling

[zingo]

Use TEST_SEED for both Arm and Cortex-M tests. Choose a single session seed before collection so module-level random tensors are reproducible, report that seed once in the pytest header, and reuse it in the auto use seed fixture before each test.

cc @digantdesai @freddan80 @per @oscarandersson8218 @mansnils @Sebastian-Larsson @robell @rascani @psiddh @AdrianLundell

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/20279

• 📄 Preview Python docs built from this PR

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

❌ 3 New Failures, 2 Unclassified Failures

As of commit 6ae3d12 with merge base e88fd04 ():

NEW FAILURES - The following jobs have failed:

• Cadence Build & Test / hifi-build / hifi4 (gh)
  Input required and not supplied: aws-region
• Cadence Build & Test / vision-build / vision (gh)
  Input required and not supplied: aws-region
• pull / unittest / macos / macos-job (gh)
  export/tests/test_target_recipes.py::TestTargetRecipes::test_ic4_model

UNCLASSIFIED FAILURES - DrCI could not classify the following jobs because the workflow did not run on the merge base. The failures may be pre-existing on trunk or introduced by this PR:

• trunk / test-models-macos-coreml (mobilebert) / macos-job (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
  RuntimeError: Command bash /Users/ec2-user/runner/_work/_temp/exec_script failed with exit code 1
• trunk / test-models-macos-cpu (mv3, xnnpack-quantization-delegation) / macos-job (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
  RuntimeError: Command bash /Users/ec2-user/runner/_work/_temp/exec_script failed with exit code 1

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

[Copilot]

Pull request overview

This PR standardizes deterministic random seed handling across the Arm and Cortex-M pytest suites by using a shared TEST_SEED environment variable and selecting a single session seed early (before test collection) so module-level random tensors are reproducible, while re-seeding consistently before each test.

Changes:

• Added session-level seed selection via TEST_SEED (including RANDOM) and stored the chosen seed/label on the pytest config object.
• Updated both suites to reseed per-test using the session seed for repeatable behavior.
• Report the chosen seed once in pytest’s header output (Arm: seed label; Cortex-M: targets + seed label).

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 2 comments.

File	Description
backends/cortex_m/test/conftest.py	Adds session seed setup + autouse reseeding, and includes seed info in the header alongside selected targets.
backends/arm/test/conftest.py	Switches from ARM_TEST_SEED to TEST_SEED, chooses one session seed, reseeds per test, and reports the seed in the pytest header.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.