[ET-VK] Fix out-of-range broadcast size calculation for 0-dim tensors

[msluszniak]

Fixes #22331.

calculate_broadcasted_output_size() guards its loop with i >= -out_sizes.size(). That is size_t arithmetic, so when both operands are 0-dimensional the bound evaluates to 0, i promotes to SIZE_MAX, the guard stays true, and the body runs out_sizes.at(0 - 1) on an empty vector, throwing std::out_of_range.

check_binary_op_args() calls it during graph construction, so any binary op with a 0-dimensional output aborts at load rather than failing at execute time.

Non-empty outputs are unaffected and stay bit-identical: the same unsigned wraparound happens to compare correctly there. With size() == 2 the bound is SIZE_MAX - 1, so i = -1, -2 pass and i = -3 fails, exactly as intended. Only the empty case is broken. This holds the bound in a signed local so the loop is simply skipped.

Why it matters

Whisper's log-mel normalisation subtracts one scalar from another, so its whole encode method (mel preprocessor plus encoder) could not be loaded on Vulkan. That is not caused by the dynamic audio dimension (a static 30 s export fails identically) or by multi-method lowering (an encode-only .pte fails identically).

Verification

Snapdragon SM8850 (Adreno 840):

model	before	after
x.max() - (x.max() - 1.0), 0-dim output	abort	exact, max abs diff 0
x - x.max(), 0-dim broadcast to rank 1	ok	exact, max abs diff 0
same arithmetic at rank 1	ok	exact, max abs diff 0
Whisper-tiny encode	abort at load	runs, cosine 0.99998790 vs CPU

test_vulkan_backend_binary_op_zero_dim covers the 0-dim case.

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/22332

• 📄 Preview Python docs built from this PR

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

⚠️ 15 Awaiting Approval

As of commit 7f484a7 with merge base c27baa8 ():

AWAITING APPROVAL - The following workflows need approval before CI can run:

• Build documentation (gh)
• Build Presets (gh)
• Lint (gh)
• periodic (gh)
• pull (gh)
• Test ARM Backend (gh)
• Test CoreML Backend (gh)
• Test Cortex-M Backend (gh)
• Test NXP Backend (gh)
• Test QNN Backend (gh)
• Test QNN Windows MSVC build (gh)
• Test Vulkan Backend (gh)
• Test Vulkan Backend (specialized runners) (gh)
• Test WebGPU Backend (gh)
• Test XNNPACK Backend (gh)

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.