Revert "Revert "Arm backend: Enable pytest VKML op unit testing in ciflow/trunk""

[zingo]

Reverts #15841

cc @freddan80 @per @oscarandersson8218 @digantdesai

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/15843

• 📄 Preview Python docs built from this PR

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

❌ 8 New Failures, 1 Cancelled Job, 1 Pending, 2 Unrelated Failures

As of commit 76b56b2 with merge base 76d43bc ():

NEW FAILURES - The following jobs have failed:

• Build Presets / apple (llm) / build (gh)
  The process '/opt/homebrew/bin/git' failed with exit code 1
• pull / unittest / windows / windows-job (gh)
  backends/xnnpack/test/recipes/test_xnnpack_recipes.py::TestXnnpackRecipes::test_all_models_with_recipes
• Test CUDA Builds / test-model-cuda-e2e (openai, whisper-large-v3-turbo, non-quantized) / linux-job (gh)
  RuntimeError: Command docker exec -t 7bde5b717c4d862f48ce85a0e2ab160ce83fad35f756137d684213cfb835653b /exec failed with exit code 1
• Test CUDA Builds / test-model-cuda-e2e (openai, whisper-large-v3-turbo, quantized-int4-tile-packed) / linux-job (gh)
  RuntimeError: Command docker exec -t 0db4ac0b2993c2bfd55d60188d23d080de4f171e0b8dfd72399bf5530309472e /exec failed with exit code 1
• Test CUDA Builds / test-model-cuda-e2e (openai, whisper-large-v3-turbo, quantized-int4-weight-only) / linux-job (gh)
  RuntimeError: Command docker exec -t a396fe850657df44961d6c5cbc32b5743b70b9afcdc409aa6bb846b34306b043 /exec failed with exit code 1
• Test CUDA Builds / test-model-cuda-e2e (openai, whisper-small, non-quantized) / linux-job (gh)
  RuntimeError: Command docker exec -t cc984f61c10e877fd3165a972e4b32d3533697455d0acf8aeb3410aabf46a453 /exec failed with exit code 1
• Test CUDA Builds / test-model-cuda-e2e (openai, whisper-small, quantized-int4-tile-packed) / linux-job (gh)
  RuntimeError: Command docker exec -t 7ee0fc4c0662740aff47ccae5a449f3f7b3d1f2f464c63f10001b0def54f6352 /exec failed with exit code 1
• Test CUDA Builds / test-model-cuda-e2e (openai, whisper-small, quantized-int4-weight-only) / linux-job (gh)
  RuntimeError: Command docker exec -t f597354293d20fe8c699eeff15e27b41523ae9e00cd8f34a833db91aa8e6e2fe /exec failed with exit code 1

CANCELLED JOB - The following job was cancelled. Please retry:

• trunk / test-arm-backend (test_pytest_ops_ethosu_fvp) / linux-job (gh)
  ##[error]The operation was canceled.

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

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

• Test Metal Backend / test-model-metal-e2e (openai, whisper-large-v3-turbo, non-quantized) / macos-job (gh) (trunk failure)
  RuntimeError: Command bash /Users/ec2-user/runner/_work/_temp/exec_script failed with exit code 1
• Test Metal Backend / test-model-metal-e2e (openai, whisper-small, non-quantized) / macos-job (gh) (trunk failure)
  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.

[zingo]

@mansnils this put one of the test over the 2h time budget, I seems to take 1h40min with the added flag. We probably have to get the patch splitting this job up in first.

[zingo]

Is replaced by #15847