NXP Backend: Update Neutron Software to version SDK_25.09

[robert-kalmar]

Summary

Update Neutron Software to version SDK 25.09.
Test updated, as there is a known issue with SDK 25.09, tracked internally as AIR-13336.

Test plan

Unit test updated.

CC @JakeStevens

cc @digantdesai @JakeStevens

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/14591

• 📄 Preview Python docs built from this PR

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

❌ 5 New Failures

As of commit 43ce023 with merge base d43cde5 ():

NEW FAILURES - The following jobs have failed:

• pull / test-vulkan-operators-linux / linux-job (gh)
  RuntimeError: Command docker exec -t 5a809c6f93d03bf5dd641ba1142bf22002dbcf60d1f3e3e897d62bb377bdc16f /exec failed with exit code 127
• Test CUDA Builds / check-all-cuda-builds (gh)
  Process completed with exit code 1.
• Test CUDA Builds / test-executorch-cuda-build-12.6 / linux-job (gh)
  RuntimeError: Command docker exec -t 6094c7c77f4a0a5cb16de21df8ad79d7b7870945e1f569a0e0d737daf6f361e4 /exec failed with exit code 1
• Test CUDA Builds / test-executorch-cuda-build-12.8 / linux-job (gh)
  RuntimeError: Command docker exec -t c5c08dcac3b96eaf70621991d8b80695e5d5b138e769d16980b505eab7581a96 /exec failed with exit code 1
• Test CUDA Builds / test-executorch-cuda-build-13.0 / linux-job (gh)
  RuntimeError: Command docker exec -t 81ff13d59817d7d23cfc4c95513069205d670e6b1f1334923278ed69f99ee26e /exec failed with exit code 1

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

[JakeStevens]

Are the prebuilts updated?

What is the plan for keeping these in sync, since OSS mainline is not going to be inline with nxp sdk releases?

[robert-kalmar]

Are the prebuilts updated?

Yes, the Neutron Driver and Firmware to SDK_25.09 is available on the MCUXpresso distribution portal, where the executorch with the examples for RT700 are provided. - https://github.com/nxp-mcuxpresso/mcuxsdk-middleware-executorch/tree/release/mcux/third-party/neutron/rt700 .

What is the plan for keeping these in sync, since OSS mainline is not going to be inline with nxp sdk releases?

• OSS mainline (main) will use the latest Neutron SW package released by NXP, in this case on aforementioned repo and eIQ PyPI.
• The NXP release will use the latest ExecuTorch release (since 1.0) and new Neutron SW package. In some cases the Neutron Backend might have some additional features supported in the new Neutron SW.
  There will be pushed right away to OSS main branch.
• The OSS ExecuTorch release will refer to the latest Neutron SW available at feature freeze (release branch cut off). As the milestones (feature freeze, release) of the MCUXPresso SDK and PyTorch/ExecuTorch happens close to each other, the OSS ExecuTorch might refer the previous Neutron SW release.

The difference btw. MCUXPresso SDK ExecuTorch release and OSS main will be minor, as the Neutron Software will be provided immediately to the OSS main.