<Close this diff>

[psiddh]

No description provided.

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/20209

• 📄 Preview Python docs built from this PR

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

❌ 1 New Failure, 29 Pending, 1 Unrelated Failure

As of commit f725c7c with merge base 4519036 ():

NEW FAILURE - The following job has failed:

• pull / unittest-nxp-neutron / linux-job (gh)
  RuntimeError: Command docker exec -t 05104b145dfdd68b1f9942428a5a35da77cacca77c73ae5d41f4c64469239eaa /exec failed with exit code 1

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

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

• pull / android / build-android (gh) (trunk failure)
  Process completed with exit code 1.

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

[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.

[Copilot]

Pull request overview

This PR extends the Qualcomm QNN backend’s cdist decomposition support to also handle the internal aten._cdist_forward op, and adds corresponding test coverage to validate lowering for both FP32 and QDQ flows.

Changes:

• Add a CDistForward test model that directly calls torch.ops.aten._cdist_forward.default.
• Add new delegate tests to exercise lowering for _cdist_forward in both non-quantized and QDQ paths.
• Update QNN pass configuration/documentation to enable and document DecomposeCDist for aten._cdist_forward.

Reviewed changes

Copilot reviewed 6 out of 6 changed files in this pull request and generated no comments.

Show a summary per file

File	Description
extension/android/executorch_android/src/main/java/org/pytorch/executorch/Module.kt	Kotlin formatting update for multi-line exception creation call.
backends/qualcomm/tests/test_qnn_delegate.py	Adds new _cdist_forward backend tests for both FP32 and QDQ flows.
backends/qualcomm/tests/models.py	Adds CDistForward module that calls aten._cdist_forward.
backends/qualcomm/builders/README.md	Documents _cdist_forward as handled by DecomposeCDist.
backends/qualcomm/_passes/qnn_pass_manager.py	Enables DecomposeCDist by default and adds its capture-program dependency.
backends/qualcomm/_passes/decompose_cdist.py	Extends decomposition target matching to include aten._cdist_forward.

Comments suppressed due to low confidence (1)

backends/qualcomm/_passes/decompose_cdist.py:55

• torch.ops.aten._cdist_forward.default passes p as a float (e.g. 2.0), but the decomposition asserts node.args[2] == 2. With the newly supported _cdist_forward target (and even for aten.cdist depending on tracing), this assertion can fail even when p is effectively 2.

Accept both 2 and 2.0 (or generally numeric values equal to 2) to avoid spurious assertion failures while still enforcing p=2 only.

            if node.target in self.cdist_targets:
                if len(node.args) > 2:
                    assert (
                        node.args[2] == 2
                    ), "Currently only p=2 is supported for CDist Decomposition"

---

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

[psiddh]

@qti-horodnic I accidentally pushed one of my commits on your PR, Sorry about that. Can you force push one more time so that author, title and summary etc is set properly. Sorry for the trouble

[qti-horodnic]

@qti-horodnic I accidentally pushed one of my commits on your PR, Sorry about that. Can you force push one more time so that author, title and summary etc is set properly. Sorry for the trouble

Looks like you already fixed this, lmk if you need anything from me