NXP backend: Add infrastructure for context dependant partitioning

[MartinPavella]

Summary

This PR adds the option to specify delegation conditions which depend on the partition a particular node ends up in. This infrastructure is applied to the view_copy node.

Test plan

Unit tests provided.

cc @robert-kalmar @roman-janik-nxp @StrycekSimon @jirioc

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/14373

• 📄 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, 2 Cancelled Jobs, 2 Unrelated Failures

As of commit 8ad83cf with merge base b3f3111 ():

NEW FAILURES - The following jobs have failed:

• pull / test-qnn-wheel-packages-linux (3.10) / linux-job (gh)
  RuntimeError: Command docker exec -t a99ac0105788f2103af37c2afadd1ff4dae955ed1764c07ad90ec2bad5f11b9c /exec failed with exit code 1
• pull / test-qnn-wheel-packages-linux (3.11) / linux-job (gh)
  RuntimeError: Command docker exec -t 8b195491c10e8ba219682fd8b3ee81dd1ddade78cf51cf61036087749300007a /exec failed with exit code 1
• pull / test-qnn-wheel-packages-linux (3.12) / linux-job (gh)
  RuntimeError: Command docker exec -t 70b32842f1a28b5269ceb978ad989c990dfe897ed6e9f115e3b77645c3bc0c6a /exec failed with exit code 1
• pull / test-samsung-models-linux / linux-job (gh)
  RuntimeError: Command docker exec -t eb113d605097f59f99e3ffe38d3232ad6bf0ca8f7856849d2facd90043150ca2 /exec failed with exit code 1

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

• pull / test-binary-size-linux-gcc / linux-job (gh)
  ##[error]The operation was canceled.
• pull / test-setup-linux-gcc / linux-job (gh)

FLAKY - The following job failed but was likely due to flakiness present on trunk:

• pull / unittest-editable / macos / macos-job (gh) (similar failure)
  exir/tests/test_quant_fusion_pass.py::TestQuantFusionPass::test_embedding_torchao

BROKEN TRUNK - The following job failed but was present on the merge base:

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

• pull / unittest / macos / macos-job (gh) (trunk failure)
  exir/tests/test_quant_fusion_pass.py::TestQuantFusionPass::test_embedding_torchao

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

[MartinPavella]

@pytorchbot label "module: nxp" "release notes: nxp"

[jirioc]

Now that the unit tests restore the original state of supports_partitioning_result, the PR looks OK.

[MartinPavella]

The failing test pull / test-samsung-models-linux / linux-job (pull_request) fails on the command mkdir -p /tmp/exynos_ai_lite_core, which seems unrelated to this PR.
All other failing tests seem unrelated to this PR, and are also failing on main.

[robert-kalmar]

For any practical model the contion is enough, but in theory not right.
Thinking loudly:

• The view_copy (==> Reshape in Neutron IR) cannot be alone in the partition because on Neutron it is a NOOP. Regardless of the count (2 view_copies is also not allowed)
• Another operator is clone, as we convert it to an "identity" tensor, what will be NOOP in neutron too.
  So I would say, the correct condition is there must be at least 1 compute operator, that is the partition must not contain only view_copy and clone ops (any number).

[MartinPavella]

Good catch.
I will create a ticket for this, as it is not trivial to determine which nodes will end up being "noops". And it would require another PR to be merged first.