Update Neutron SW to 3.1.3

[MartinPavella]

Summary

Update Neutron SW to the latest version (3.1.3) and update related tests accordingly.

Test plan

Tested by existing NXP backend tests.

cc @robert-kalmar @JakeStevens @digantdesai @rascani

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/20616

• 📄 Preview Python docs built from this PR

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

⏳ No Failures, 10 Pending

As of commit f1f2ddf with merge base 2ec218c ():
💚 Looks good so far! There are no failures yet. 💚

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

[linux-foundation-easycla]

• ❌ - login: @MartinPavella / name: Martin Pavella. The commit (f1f2ddf) is not authorized under a signed CLA. Please click here to be authorized. For further assistance with EasyCLA, please visit our EasyCLA portal and chat with our support bot.

[robert-kalmar]

@MartinPavella temporarily increasing the time limit is OK, but please file a ticket internally to divide our tests into 2 parts - short running and long running.

The PR unit test executes all the time, multiple times when PR is raised. So in next sprint lets address this, to move the long running tests from pull to trunk workflow. So it will run only after PR is merged.

[MartinPavella]

@MartinPavella temporarily increasing the time limit is OK, but please file a ticket internally to divide our tests into 2 parts - short running and long running.

The PR unit test executes all the time, multiple times when PR is raised. So in next sprint lets address this, to move the long running tests from pull to trunk workflow. So it will run only after PR is merged.

@robert-kalmar I'm sorry, I misread the time stamps. It is actually not necessary to increase the time limit. It appears that some test is hanging on the GitHub CI. The internal CI is passing, and all test pass locally too. On a recent standup you were discussing some issue with the mlperf tiny repo being cloned by the parallel workers, which caused some issues. Could that be the problem here? I assume the problem was discovered when I was on holiday.

[robert-kalmar]

@MartinPavella temporarily increasing the time limit is OK, but please file a ticket internally to divide our tests into 2 parts - short running and long running.
The PR unit test executes all the time, multiple times when PR is raised. So in next sprint lets address this, to move the long running tests from pull to trunk workflow. So it will run only after PR is merged.

@robert-kalmar I'm sorry, I misread the time stamps. It is actually not necessary to increase the time limit. It appears that some test is hanging on the GitHub CI. The internal CI is passing, and all test pass locally too. On a recent standup you were discussing some issue with the mlperf tiny repo being cloned by the parallel workers, which caused some issues. Could that be the problem here? I assume the problem was discovered when I was on holiday.

• Hanging.. So it it not caused by introducing the the tests for big convolutions, like test__depthwise__big. Make sense as those tests were xfailed not skipped, so have run even before.
• Definitelly not with the mlperf tiny repo, as it is not accessible outside, so no test clones it. But in general might be a problem with concurent access. We addressed where the tests write their artefacts. Maybe we have artefacts or resources we missed.

[MartinPavella]

The skipped checks seem unrelated and the internal CI is passing. Merging.