Bump the PyTorch pin to 20250811

[swolchok]

No description provided.

[swolchok]

Stack from ghstack (oldest at bottom):

• -> Bump the PyTorch pin to 20250811 #13334

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/13334

• 📄 Preview Python docs built from this PR

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

❌ 4 New Failures, 4 Cancelled Jobs

As of commit 8c7ef6e with merge base 8e73c09 ():

NEW FAILURES - The following jobs have failed:

• pull / test-llava-runner-linux / linux-job (gh)
  RuntimeError: Command docker exec -t 07b3c8d12398d7f8f87db7af7e705b48db775996b1a9e0a3a80a3db989681663 /exec failed with exit code 139
• trunk / test-models-arm-zephyr (add) / linux-job (gh)
  /opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/include/torch/headeronly/util/complex.h:543:19: error: redefinition of ‘struct c10::complex<c10::Half>’
• trunk / test-models-arm-zephyr (mv2) / linux-job (gh)
  /opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/include/torch/headeronly/util/complex.h:543:19: error: redefinition of ‘struct c10::complex<c10::Half>’
• trunk / test-models-arm-zephyr (softmax) / linux-job (gh)
  /opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/include/torch/headeronly/util/complex.h:543:19: error: redefinition of ‘struct c10::complex<c10::Half>’

CANCELLED JOBS - The following jobs were cancelled. Please retry:

• trunk / test-arm-backend (test_pytest_models_ethosu_fvp) / linux-job (gh)
  ##[error]The operation was canceled.
• trunk / test-demo-backend-delegation (buck2) / linux-job (gh)
  ##[error]The operation was canceled.
• trunk / test-demo-backend-delegation (cmake) / linux-job (gh)
  ##[error]The operation was canceled.
• trunk / test-models-linux-aarch64 (ic4, portable, linux.arm64.2xlarge) / linux-job (gh)
  ##[error]The operation was canceled.

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

[swolchok]

failing zephyr jobs seem to be using the old version of bit_cast.h somehow. inclined to ignore those and hope they clear on land, but I'll see if I can find anyone else who has seen this behavior first.

unittest-arm-backend-with-no-fvp (test_pytest_ops) / linux-job (pull_request) failure looks like a flake, rerunning.

[swolchok]

happy to ignore test-llava-runner-linux failure; it's known to segfault sometimes

[swolchok]

it looks like zephyr is somehow pinned to an old executorch version. going to talk to @BujSet soon

[swolchok]

@BujSet says we can have zephyr tests point at the current ExecuTorch source and he'll try to get around to that soon.