[CI] Fix Linux aarch64 nightly wheel builds - #1757
Merged
vmoens merged 1 commit intoAug 5, 2026
Merged
Conversation
peterdsharpe
marked this pull request as ready for review
August 4, 2026 22:52
10 tasks
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Today's nightly build for the new Linux aarch64 target (added by my previous PR #1756) failed because of a pretty silly shell command-scoping bug: only the initial
sedcommand ran through test-infra's Conda wrapper. Everything after the first&&, includingpip wheel, ran outside the prepared build environment.Log:
https://github.com/pytorch/tensordict/actions/runs/30919780779
This fixes that by moving the complete nightly wheel build into a small script, so the whole sequence runs inside the Conda environment and
pyproject.tomlis restored even if the build fails. It also builds with--no-deps, since test-infra has already installed nightly PyTorch, and installs the splitlibpython-staticpackage that CMake needs for Linux Python 3.15/3.15t builds.I reproduced the full Linux aarch64 matrix natively (Python 3.10–3.15, including 3.14t and 3.15t). All eight wheels built successfully, passed test-infra's manylinux repair step, and passed
twine check.