Skip XPU manywheel builds for Python 3.15 by atalman · Pull Request #184906 · pytorch/pytorch

atalman · 2026-05-22T16:33:25Z

Summary

Skip XPU manywheel builds for Python 3.15 / 3.15t in the same way #184600 skipped ROCm. Triton XPU 3.15 wheel builds are currently broken (see the exclusion in .github/workflows/build-triton-wheel.yml); without working triton-xpu, manywheel XPU 3.15 builds cannot link triton and fail.

Change

.github/scripts/generate_binary_build_matrix.py — adds 8 lines mirroring the ROCm skip block:

# TODO: Re-enable XPU for python 3.15 once the triton XPU 3.15
# wheel build is fixed (tracked in #184901). Triton XPU is
# currently skipped for 3.15/3.15t in build-triton-wheel.yml.
if arch_version in XPU_ARCHES and (
    python_version == "3.15" or python_version == "3.15t"
):
    continue

Notes

The generated workflow YAML (generated-linux-binary-manywheel-nightly.yml) will need regenerating to drop the manywheel-py3_15-xpu / manywheel-py3_15t-xpu entries — that's 44 lines removed, fully derived from this script change via:

python .github/scripts/generate_ci_workflows.py

Will follow up with the regen commit, or happy for a maintainer to push it on this branch.

References

Umbrella: Python 3.15 support for PyTorch #184352 (Python 3.15 support for PyTorch)
Tracking: XPU: Triton wheel builds for Python 3.15 are currently broken; skipped in build-triton-wheel.yml #184901 (XPU triton 3.15 build failure)
Related: Add 3.15 triton wheel build #184829 (triton 3.15 wheel build that skipped XPU)
Precedent: Linux py 3.15 wheel builds #184600 (ROCm 3.15 skip)

Triton XPU 3.15 wheel builds are currently broken and are skipped in build-triton-wheel.yml. Skip XPU manywheel builds for 3.15 and 3.15t in the same way #184600 skipped ROCm, until tracking issue #184901 is resolved. Refs: #184352, #184901, #184829

pytorch-bot · 2026-05-22T16:33:29Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/184906

📄 Preview Python docs built from this PR
📄 Preview C++ docs built from this PR
❓ Need help or want to give feedback on the CI? Visit the bot commands wiki

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

❌ 91 Pending, 1 Unclassified Failure

As of commit 7de9eea with merge base 9ae2b85 ():

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

pull / linux-jammy-py3.10-clang18 / test (onnx, 1, 1, linux.4xlarge) (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
test/onnx/ops/test_ops.py::NativeOnnxOpsTest::test_rotary_embedding_opcheck

UNSTABLE - The following jobs are marked as unstable, possibly due to flakiness on trunk:

⏳ pull / linux-jammy-aarch64-py3.10 / test-osdc (default, 2, 5, mt-l-arm64g4-16-62) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-aarch64-py3.10 / test-osdc (default, 4, 5, mt-l-arm64g4-16-62) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-aarch64-py3.10 / test-osdc (default, 5, 5, mt-l-arm64g4-16-62) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-clang18 / test-osdc (crossref, 1, 2, mt-l-x86iavx512-8-64) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-clang18 / test-osdc (crossref, 2, 2, mt-l-x86iavx512-8-64) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-clang18 / test-osdc (default, 1, 5, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-clang18 / test-osdc (default, 2, 5, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-clang18 / test-osdc (default, 3, 5, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-clang18 / test-osdc (default, 4, 5, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-clang18 / test-osdc (default, 5, 5, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-clang18 / test-osdc (dynamo_wrapped, 1, 3, mt-l-x86iavx512-8-64) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-clang18 / test-osdc (dynamo_wrapped, 2, 3, mt-l-x86iavx512-8-64) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-clang18-asan / test-osdc (default, 1, 7, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-clang18-asan / test-osdc (default, 2, 7, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-clang18-asan / test-osdc (default, 3, 7, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-clang18-asan / test-osdc (default, 4, 7, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-clang18-asan / test-osdc (default, 5, 7, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-clang18-asan / test-osdc (default, 6, 7, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-clang18-asan / test-osdc (default, 7, 7, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-gcc11 / test-osdc (default, 1, 5, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-gcc11 / test-osdc (default, 2, 5, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-gcc11 / test-osdc (default, 3, 5, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-gcc11 / test-osdc (default, 4, 5, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-gcc11 / test-osdc (default, 5, 5, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.10-gcc11 / test-osdc (distributed, 1, 3, mt-l-x86iamx-8-64) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.14-clang18 / test-osdc (crossref, 1, 2, mt-l-x86iavx512-8-64) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.14-clang18 / test-osdc (crossref, 2, 2, mt-l-x86iavx512-8-64) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.14-clang18 / test-osdc (default, 1, 5, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.14-clang18 / test-osdc (default, 2, 5, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.14-clang18 / test-osdc (default, 3, 5, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.14-clang18 / test-osdc (default, 4, 5, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.14-clang18 / test-osdc (default, 5, 5, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.14-clang18 / test-osdc (dynamo_wrapped, 1, 3, mt-l-x86iavx512-8-64) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.14-clang18 / test-osdc (dynamo_wrapped, 2, 3, mt-l-x86iavx512-8-64) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.14t-clang18 / test-osdc (crossref, 1, 2, mt-l-x86iavx512-8-64) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.14t-clang18 / test-osdc (crossref, 2, 2, mt-l-x86iavx512-8-64) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.14t-clang18 / test-osdc (default, 1, 5, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.14t-clang18 / test-osdc (default, 2, 5, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.14t-clang18 / test-osdc (default, 3, 5, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.14t-clang18 / test-osdc (default, 4, 5, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.14t-clang18 / test-osdc (default, 5, 5, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.14t-clang18 / test-osdc (dynamo_wrapped, 1, 3, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-py3.14t-clang18 / test-osdc (dynamo_wrapped, 2, 3, mt-l-x86iavx512-16-128) (gh) (PyTorch OSDC job marked as unstable)
⏳ pull / linux-jammy-rocm-py3.10 / build-osdc (gh) (PyTorch OSDC job marked as unstable)

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

atalman · 2026-05-22T19:45:04Z

@pytorchmergebot merge -f "lint and builds look good"

pytorchmergebot · 2026-05-22T19:47:09Z

Merge started

Your change will be merged immediately since you used the force (-f) flag, bypassing any CI checks (ETA: 1-5 minutes). Please use -f as last resort and instead consider -i/--ignore-current to continue the merge ignoring current failures. This will allow currently pending tests to finish and report signal before the merge.

Learn more about merging in the wiki.

Questions? Feedback? Please reach out to the PyTorch DevX Team

Advanced Debugging

Check the merge workflow status
here

## Summary Skip XPU manywheel builds for Python 3.15 / 3.15t in the same way pytorch#184600 skipped ROCm. Triton XPU 3.15 wheel builds are currently broken (see the exclusion in `.github/workflows/build-triton-wheel.yml`); without working triton-xpu, manywheel XPU 3.15 builds cannot link triton and fail. ## Change `.github/scripts/generate_binary_build_matrix.py` — adds 8 lines mirroring the ROCm skip block: ```py # TODO: Re-enable XPU for python 3.15 once the triton XPU 3.15 # wheel build is fixed (tracked in pytorch#184901). Triton XPU is # currently skipped for 3.15/3.15t in build-triton-wheel.yml. if arch_version in XPU_ARCHES and ( python_version == "3.15" or python_version == "3.15t" ): continue ``` ## Notes The generated workflow YAML (`generated-linux-binary-manywheel-nightly.yml`) will need regenerating to drop the `manywheel-py3_15-xpu` / `manywheel-py3_15t-xpu` entries — that's 44 lines removed, fully derived from this script change via: ```bash python .github/scripts/generate_ci_workflows.py ``` Will follow up with the regen commit, or happy for a maintainer to push it on this branch. ## References - Umbrella: pytorch#184352 (Python 3.15 support for PyTorch) - Tracking: pytorch#184901 (XPU triton 3.15 build failure) - Related: pytorch#184829 (triton 3.15 wheel build that skipped XPU) - Precedent: pytorch#184600 (ROCm 3.15 skip) Pull Request resolved: pytorch#184906 Approved by: https://github.com/albanD, https://github.com/malfet

The XPU triton 3.15 wheel build issue (tracked in pytorch#184901) has been fixed. Remove the exclusion added in pytorch#184829 and the manywheel skip added in pytorch#184906 so that XPU Python 3.15/3.15t builds run again. Changes: - .github/workflows/build-triton-wheel.yml: remove exclude block for xpu+3.15/3.15t - .github/scripts/generate_binary_build_matrix.py: remove XPU 3.15 skip - Regenerated generated-linux-binary-manywheel-nightly.yml Test Plan: CI will validate via build-triton-wheel.yml and manywheel nightly. Authored by Claude.

## Summary Re-enable XPU triton wheel builds and manywheel builds for Python 3.15/3.15t on Linux. fixes #184901 This reverts the XPU-specific exclusions added in #184829 and #184906. ## Changes - `.github/workflows/build-triton-wheel.yml`: remove `exclude` block for xpu + 3.15/3.15t - `.github/scripts/generate_binary_build_matrix.py`: remove XPU 3.15 skip block - Regenerated `generated-linux-binary-manywheel-nightly.yml` to re-add `manywheel-py3_15-xpu` / `manywheel-py3_15t-xpu` entries ## Test Plan CI will validate via `build-triton-wheel.yml` and manywheel nightly workflows. Pull Request resolved: #185094 Approved by: https://github.com/Skylion007, https://github.com/EikanWang

atalman requested a review from a team as a code owner May 22, 2026 16:33

pytorch-bot Bot added the topic: not user facing topic category label May 22, 2026

albanD approved these changes May 22, 2026

View reviewed changes

Regenerate manywheel workflow YAML for XPU 3.15 skip

7de9eea

malfet approved these changes May 22, 2026

View reviewed changes

pytorchmergebot added the merging label May 22, 2026

pytorchmergebot added the Merged label May 22, 2026

pytorchmergebot closed this in c314c2c May 22, 2026

pytorchmergebot removed the merging label May 22, 2026

chuanqi129 mentioned this pull request May 25, 2026

Re-enable triton XPU Python 3.15 Linux build #185094

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Skip XPU manywheel builds for Python 3.15#184906

Skip XPU manywheel builds for Python 3.15#184906
atalman wants to merge 2 commits into
mainfrom
atalman/skip-xpu-py315-manywheel

atalman commented May 22, 2026

Uh oh!

pytorch-bot Bot commented May 22, 2026 •

edited

Loading

Uh oh!

atalman commented May 22, 2026

Uh oh!

pytorchmergebot commented May 22, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Conversation

atalman commented May 22, 2026

Summary

Change

Notes

References

Uh oh!

pytorch-bot Bot commented May 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/184906

❌ 91 Pending, 1 Unclassified Failure

Uh oh!

atalman commented May 22, 2026

Uh oh!

pytorchmergebot commented May 22, 2026

Merge started

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

pytorch-bot Bot commented May 22, 2026 •

edited

Loading