Repository navigation
check_versions.py: deprecate as soon as upstream ships any riscv64 wheel - #3514
Merged
justeph merged 1 commit intoOct 20, 2026
Merged
Conversation
justeph
force-pushed
the
deprecate-when-upstream-ships-at-least-a-riscv64-wheel
branch
from
October 19, 2026 14:58
fbba14b to
b4ca613
Compare
Previously we only deprecated once upstream published riscv64 wheels for every target Python (3.12, 3.13, 3.14, 3.14t). If even one was missing we kept building the package ourselves for all target versions, including the ones upstream already covers. xmhuffman is an example: upstream ships riscv64 wheels for 3.12, 3.13 and 3.14, but not 3.14t. We don't want to keep building those covered versions ourselves, but we also shouldn't start building 3.14t (or any other upstream-unsupported version) on upstream's behalf just to fill the gap — a free-threaded build succeeding doesn't mean the resulting wheel is actually free-threading-safe, and the same applies to any other Python version upstream chose not to support. That's a correctness guarantee only upstream can make, so we defer to whatever matrix upstream decides to publish instead of completing it for them. has_riscv64_wheel (require every target version) becomes has_any_riscv64_wheel (require at least one), used both to stop declaring new versions for us to build once upstream has any riscv64 coverage, and to trigger deprecation on partial upstream coverage. Signed-off-by: Julien Stephan <jstephan@baylibre.com>
Collaborator
|
I think this is OK to merge - pip_audit actually fails despite the status because PyAV isn't shipping wheels. Looking into why that didn't happen (I've previously submitted a PR). |
justeph
deleted the
deprecate-when-upstream-ships-at-least-a-riscv64-wheel
branch
October 20, 2026 07:14
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
2 participants
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.
Previously we only deprecated once upstream published riscv64 wheels for every target Python (3.12, 3.13, 3.14, 3.14t). If even one was missing we kept building the package ourselves for all target versions, including the ones upstream already covers.
xmhuffman is an example: upstream ships riscv64 wheels for 3.12, 3.13 and 3.14, but not 3.14t. We don't want to keep building those covered versions ourselves, but we also shouldn't start building 3.14t (or any other upstream-unsupported version) on upstream's behalf just to fill the gap — a free-threaded build succeeding doesn't mean the resulting wheel is actually free-threading-safe, and the same applies to any other Python version upstream chose not to support. That's a correctness guarantee only upstream can make, so we defer to whatever matrix upstream decides to publish instead of completing it for them.
has_riscv64_wheel (require every target version) becomes has_any_riscv64_wheel (require at least one), used both to stop declaring new versions for us to build once upstream has any riscv64 coverage, and to trigger deprecation on partial upstream coverage.