fix(nxp): check the conv/linear node's user count in BatchNorm fusion

[durvesh1992]

Summary

The _is_conv / _is_linear closures in the NXP BatchNorm fusion passes check len(node.users) instead of len(node_.users).

node_ is the closure's parameter (the conv/linear node being tested), but node resolves to the enclosing for node in graph_module.graph.nodes loop variable — which at the call site (_is_conv(bn_node.args[0])) is always the BatchNorm node.

The single-user guard exists to prevent fusion when the conv/linear output feeds consumers other than the BatchNorm, because folding the BatchNorm into the conv/linear weights changes that output for all consumers. Due to the typo the guard inspected the BatchNorm's user count rather than the conv/linear's, so a conv/linear with multiple consumers was still fused — corrupting the other consumers' inputs.

Affected files:

• backends/nxp/aten_passes/fuse_batch_norm_with_conv_pass.py
• backends/nxp/aten_passes/fuse_batch_norm_with_linear_pass.py

Test plan

Reproduced on a small torch.fx graph where the conv has two consumers (BatchNorm + a relu), so fusion must be skipped:

• Before: the pass fuses anyway → model output max-abs-diff vs. reference ≈ 76.2 (corrupted).
• After: the pass is correctly a no-op → diff 0.0.
• After, single-consumer conv: still fuses correctly, numerics preserved (diff ≈ 3.8e-6).

The linear pass contains the identical typo and the identical one-character fix.

cc @robert-kalmar @JakeStevens @digantdesai @rascani

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/20601

• 📄 Preview Python docs built from this PR

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

❌ 1 Cancelled Job

As of commit 4537786 with merge base b331ebd ():

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

• pull / unittest-editable / macos / macos-job (gh)
  ##[error]The operation was canceled.

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

[linux-foundation-easycla]

• ❌ - login: @durvesh1992 / name: Durvesh Pilankar. The commit (4537786, d14a781) 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.

[github-actions]

This PR needs a release notes: label

If your change should be included in the release notes (i.e. would users of this library care about this change?), please use a label starting with release notes:. This helps us keep track and include your important work in the next release notes.

To add a label, you can comment to pytorchbot, for example
@pytorchbot label "release notes: none"

For more information, see
https://github.com/pytorch/pytorch/wiki/PyTorch-AutoLabel-Bot#why-categorize-for-release-notes-and-how-does-it-work.

[robert-kalmar]

CC @MartinPavella @roman-janik-nxp , @novak-vaclav

@durvesh1992 , do you mind to add a corresponding unit test as you described in the Test plan?

[durvesh1992]

Thanks for the review @robert-kalmar! I've added unit tests in backends/nxp/tests/test_fuse_batch_norm_single_user.py covering both passes: a conv/linear with a single user has its BatchNorm fused away, and a conv/linear whose output feeds an additional consumer keeps its BatchNorm (fusion skipped) with the numerical result preserved. The multi-user cases fail without the node_.users fix in this PR.

[durvesh1992]

Heads up on the red EasyCLA check: the Meta CLA check passes, and I'm actively sorting out the Linux Foundation EasyCLA corporate authorization on my end (getting allowlisted under the company CCLA). The code and tests are ready for review in the meantime — I'll get the EasyCLA green as soon as the corporate approval lands. This same authorization also covers my other two ExecuTorch PRs (#20603, #20608).