Avoid broken build artifacts on build failure by bybrooks · Pull Request #17276 · astral-sh/uv

bybrooks · 2025-12-31T09:05:28Z

Summary

When uv build fails (e.g., due to missing __init__.py), partial distribution
files were being left in the dist/ directory.

Changes

Use NamedTempFile to write build output to a temporary file first
Only persist the file to the final location on successful build
If build fails, the temporary file is automatically cleaned up
Added Error::Persist variant for handling persistence failures

Test Plan

Added no_partial_files_on_build_failure test that verifies:

build_source_dist fails when __init__.py is missing
build_wheel fails when __init__.py is missing
The dist/ directory remains empty after both failures (no partial files)

EliteTK

Thanks for this!

I left a few comments, I'm somewhat new here so my colleagues might have additional comments.

It would be good to add an integration test in crates/uv/tests/it/build_backend.rs to check the error messages.

There's something to consider, this changes the behaviour for handling pre-existing built files. Currently we start by overwriting them. This means that in the case of a failure, the existing files are lost and replaced with incomplete new ones. Now we are only overwriting the files if we succeed. Personally it's what I would expect here.

But when reporting errors, the user will see something like "Failed to ... dist/project-0.1.0-py2.py3-none-any.whl", implying the specified file was not able to be built, but if they check, they will see there is a file there.

Could this be confusing and should we do something about it? I see a couple of options aside from just leaving it as is:

Delete any target files before attempting to write the new ones
Mention in error reporting that the (or we could make it easier for ourselves and just say "any") existing file has not been overwritten

crates/uv-build-backend/src/lib.rs

crates/uv-build-backend/src/source_dist.rs

crates/uv-build-backend/src/wheel.rs

crates/uv-build-backend/src/lib.rs

bybrooks · 2025-12-31T18:35:04Z

@EliteTK
Thank you for the feedback!

Delete any target files before attempting to write the new ones
Mention in error reporting that the (or we could make it easier for ourselves and just say "any") existing file has not been overwritten

I'll go with Delete any target files before attempting to write the new ones.

This ensures a consistent state where the existence of the target file directly indicates a successful build.

I'll implement this by deleting the target file (if it exists) before creating the temporary file.

EliteTK · 2025-12-31T18:38:55Z

@bybrooks regarding the deleting - feel free, although once we discuss this further, we might not want to go with that option, so also feel free to wait for a verdict there. There is no rush at least on my end.

Main issue is that everyone else is on annual leave :)

codspeed-hq · 2025-12-31T18:47:41Z

CodSpeed Performance Report

Merging #17276 will not alter performance

_{Comparing bybrooks:bybrooks/atomic-build-output (bc49b6e) with main (82a6a66)}

Summary

✅ 5 untouched

konstin · 2026-01-05T12:07:53Z

There's something to consider, this changes the behaviour for handling pre-existing built files. Currently we start by overwriting them. This means that in the case of a failure, the existing files are lost and replaced with incomplete new ones. Now we are only overwriting the files if we succeed. Personally it's what I would expect here.

But when reporting errors, the user will see something like "Failed to ... dist/project-0.1.0-py2.py3-none-any.whl", implying the specified file was not able to be built, but if they check, they will see there is a file there.

Could this be confusing and should we do something about it? I see a couple of options aside from just leaving it as is:
* Delete any target files before attempting to write the new ones

* Mention in error reporting that the (or we could make it easier for ourselves and just say "any") existing file has not been overwritten

Deleting the target file(s) before starting the build seems to be the easiest. No strong opinion, users shouldn't rely on a specific the state of dist/ after a non-zero exist code anyway.

bybrooks · 2026-01-06T12:54:45Z

Deleting the target file(s) before starting the build seems to be the easiest. No strong opinion, users shouldn't rely on a specific the state of dist/ after a non-zero exist code anyway.

Thanks for the reply.
I’ve implemented the logic to delete existing files before the build starts. （ bc49b6e）

This MR contains the following updates: | Package | Update | Change | |---|---|---| | [astral-sh/uv](https://github.com/astral-sh/uv) | patch | `0.9.22` → `0.9.24` | MR created with the help of [el-capitano/tools/renovate-bot](https://gitlab.com/el-capitano/tools/renovate-bot). **Proposed changes to behavior should be submitted there as MRs.** --- ### Release Notes <details> <summary>astral-sh/uv (astral-sh/uv)</summary> ### [`v0.9.24`](https://github.com/astral-sh/uv/blob/HEAD/CHANGELOG.md#0924) [Compare Source](astral-sh/uv@0.9.23...0.9.24) Released on 2026-01-09. ##### Bug fixes - Fix handling of `UV_NO_SYNC=1 uv run ...` ([#17391](astral-sh/uv#17391)) - Rebuild dynamic distribution when version changes with `--no-cache` ([#17387](astral-sh/uv#17387)) ##### Documentation - Add Rust language classifier ([#17389](astral-sh/uv#17389)) ### [`v0.9.23`](https://github.com/astral-sh/uv/blob/HEAD/CHANGELOG.md#0923) [Compare Source](astral-sh/uv@0.9.22...0.9.23) Released on 2026-01-09. ##### Enhancements - Only write portable paths in `RECORD` files ([#17339](astral-sh/uv#17339)) - Support relative paths in `UV_PYTHON_BIN_DIR` and `UV_TOOL_BIN_DIR` ([#17367](astral-sh/uv#17367)) ##### Preview features - Enable uploads to S3 via pre-signed URLs ([#17349](astral-sh/uv#17349)) ##### Configuration - Allow setting proxy variables via global / user configuration ([#16918](astral-sh/uv#16918)) - Manually parse and reconcile Boolean environment variables ([#17321](astral-sh/uv#17321)) ##### Bug fixes - Avoid broken build artifacts on build failure ([#17276](astral-sh/uv#17276)) - Fix missing dependencies on synthetic root in SBOM export ([#17363](astral-sh/uv#17363)) - Recognize `armv8l` as an alias for `armv7l` in platform tag parsing ([#17384](astral-sh/uv#17384)) - Fix redaction of a URL in a middleware trace log ([#17346](astral-sh/uv#17346)) ##### Documentation - Add `index.md` suggestion to `llms.txt` ([#17362](astral-sh/uv#17362)) - Clarify that `uv run` uses inexact syncing by default ([#17366](astral-sh/uv#17366)) </details> --- ### Configuration 📅 **Schedule**: Branch creation - At any time (no schedule defined), Automerge - At any time (no schedule defined). 🚦 **Automerge**: Disabled by config. Please merge this manually once you are satisfied. ♻ **Rebasing**: Whenever MR becomes conflicted, or you tick the rebase/retry checkbox. 🔕 **Ignore**: Close this MR and you won't be reminded about this update again. --- - [ ] If you want to rebase/retry this MR, check this box --- This MR has been generated by [Renovate Bot](https://github.com/renovatebot/renovate).

…ith `0o666 & !umask` permissions on `unix` platforms again (#18020) ## Summary Fix #18004. #17276 introduced a regression since `NamedTempFile::new_in` defaults to `0o600` permissions. Fortunately `uv_fs` already has a wrapper for a `new_in` with "normal" permissions this so this PR uses that. ## Test Plan Reliable testing would require us to support setting a umask for a test context process which we currently don't have support for and implementing it is not trivial so for now I just tested this manually. To test, you can just use: ```bash #!/usr/bin/env bash set -e tempdir=$(mktemp -d) trap 'rm -rf "$tempdir"' EXIT uv init --lib "$tempdir" cd "$tempdir" uv build umask ls -al dist/ ```

Hayashi Reo added 5 commits December 31, 2025 08:28

Atomic build output

ee70181

Add test for atomic build output

2f752c3

Replace std::fs::File with fs_err::File in uv-build-backend

797f73f

Add tempfile to dependencies

0180b54

Resolve Clippy execution errors

3444279

EliteTK reviewed Dec 31, 2025

View reviewed changes

EliteTK self-assigned this Dec 31, 2025

Hayashi Reo added 2 commits December 31, 2025 23:35

Extract inner io::Error from tempfile::PersistError

c3f88c3

Remove unnecessary try_clone() calls in build backend writers

2bc47a5

zanieb requested a review from konstin January 3, 2026 14:02

Add test for overwriting existing build artifacts on successful build

3b70a11

konstin approved these changes Jan 5, 2026

View reviewed changes

konstin added the bug Something isn't working label Jan 5, 2026

Delete existing target files before build to ensure consistent state

bc49b6e

konstin changed the title ~~Clean up partial sdist/wheel on build failure~~ Avoid broken build artifacts on build failure Jan 6, 2026

konstin merged commit cc1ca8b into astral-sh:main Jan 6, 2026
102 checks passed

BrewTestBot mentioned this pull request Jan 9, 2026

uv 0.9.23 Homebrew/homebrew-core#262077

Merged

This was referenced Feb 15, 2026

Dist files have mod 600 since v0.9.23 #18004

Closed

Make the uv_build build backend produce wheel and sdist files with 0o666 & !umask permissions on unix platforms again #18020

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Comments

Avoid broken build artifacts on build failure#17276

Avoid broken build artifacts on build failure#17276
konstin merged 9 commits intoastral-sh:mainfrom
bybrooks:bybrooks/atomic-build-output

bybrooks commented Dec 31, 2025

Uh oh!

EliteTK left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

bybrooks commented Dec 31, 2025 •

edited

Loading

Uh oh!

EliteTK commented Dec 31, 2025

Uh oh!

codspeed-hq bot commented Dec 31, 2025 •

edited

Loading

Uh oh!

konstin commented Jan 5, 2026

Uh oh!

bybrooks commented Jan 6, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Comments

Conversation

bybrooks commented Dec 31, 2025

Summary

Changes

Test Plan

Uh oh!

EliteTK left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

bybrooks commented Dec 31, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

EliteTK commented Dec 31, 2025

Uh oh!

codspeed-hq bot commented Dec 31, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

CodSpeed Performance Report

Merging #17276 will not alter performance

Summary

Uh oh!

konstin commented Jan 5, 2026

Uh oh!

bybrooks commented Jan 6, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

bybrooks commented Dec 31, 2025 •

edited

Loading

codspeed-hq bot commented Dec 31, 2025 •

edited

Loading