Arm backend: Test fixes for TOSA on Arm64

[zingo]

cc @digantdesai @freddan80 @per @oscarandersson8218 @mansnils @Sebastian-Larsson @robell @rascani

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/20237

• 📄 Preview Python docs built from this PR

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

❌ 2 New Failures, 1 Cancelled Job, 2 Unclassified Failures

As of commit df80895 with merge base 8dbbd9d ():

NEW FAILURES - The following jobs have failed:

• pull / test-arm-backend-no-driver (test_pytest_ops_tosa) / linux-job (gh)
  RuntimeError: Command docker exec -t ec265912bbf2dfa0337c879147b6b7c1c4c7ce043c8f9c052ec138a0a320e6ec /exec failed with exit code 1
• pull / test-qnn-models-linux (mv2) / linux-job (gh)
  RuntimeError: Command docker exec -t 50eb2936e78989f6df1fa683dc7e446accd4876c73c4ee4a4dc062d09ee896c7 /exec failed with exit code 92

UNCLASSIFIED FAILURES - DrCI could not classify the following jobs because the workflow did not run on the merge base. The failures may be pre-existing on trunk or introduced by this PR:

• Cadence Build & Test / hifi-build / hifi4 (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
  Input required and not supplied: aws-region
• Cadence Build & Test / vision-build / vision (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
  Input required and not supplied: aws-region

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

• pull / unittest-nxp-neutron / linux-job (gh)

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

[Copilot]

Pull request overview

Adjusts Arm backend model tests to account for known numeric mismatches on aarch64 when running TOSA FP/BF16 paths, reducing spurious CI failures on Arm64 hosts.

Changes:

• Add an aarch64/arm64 host detection helper (common.is_aarch64_host()).
• Mark the ResNet18 TOSA BF16 test as xfail on aarch64 due to a known mismatch (MLETORCH-2048).
• Relax/xfail specific aarch64 cases for MobileNetV3 FP16 tolerance and Qwen3-VL BF16 vision patch embedding.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 1 comment.

File	Description
backends/arm/test/common.py	Adds is_aarch64_host() helper for host-conditional test behavior.
backends/arm/test/models/test_resnet18.py	Xfails ResNet18 TOSA BF16 on aarch64 for known mismatch.
backends/arm/test/models/test_mobilenet_v3_arm.py	Adjusts FP16 atol on aarch64 (and adds a clarifying comment).
backends/arm/test/models/Qwen3_VL/test_qwen3_vl_layers.py	Adds per-case xfail for BF16 vision_patch_embed on aarch64.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.