Fix partial K tile masking in int_scaled_matmul

[bm1016bm-svg]

Summary

• mask the final K tile unless K is divisible by the selected BLOCK_K
• add a sentinel-padded CUDA regression for K=34, BLOCK_K=32
• pin the Triton config in the test so autotuning cannot hide the bug

Fixes #4633

Test plan

• uvx --from ruff==0.11.6 ruff check test/kernel/test_autotuner.py
• uvx --from ruff==0.11.6 ruff format test/kernel/test_autotuner.py
• git diff --check
• CUDA regression not run locally because Torch/Triton are not installed in this Windows environment; CI provides the required runtime

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/4635

• 📄 Preview Python docs built from this PR

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

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

[meta-cla]

Hi @bm1016bm-svg!

Thank you for your pull request and welcome to our community.

Action Required

In order to merge any pull request (code, docs, etc.), we require contributors to sign our Contributor License Agreement, and we don't seem to have one on file for you.

Process

In order for us to review and merge your suggested changes, please sign at https://code.facebook.com/cla. If you are contributing on behalf of someone else (eg your employer), the individual CLA may not be sufficient and your employer may need to sign the corporate CLA.

Once the CLA is signed, our tooling will perform checks and validations. Afterwards, the pull request will be tagged with CLA signed. The tagging process may take up to 1 hour after signing. Please give it that time before contacting us about it.

If you have received this in error or have any questions, please contact us at cla@meta.com. Thanks!

[meta-cla]

Thank you for signing our Contributor License Agreement. We can now accept your code for this (and any) Meta Open Source project. Thanks!