fix: Flush vfox download buffer #6969
Merged
Conversation
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
I suspect (but can't confirm) the issue in jdx#6527 is due to a race condition when writing the downloaded archive to disk. If the write isn't fully flushed to disk, it's possible for readers to see an incomplete/invalid archive. As a precaution, we should explicitly flush before returning. The issue is probably more likely on CI runners with slower file systems. As for why we seem to only see this with `yarn`, I'm guessing its smaller size (1.2M) makes it more likely to sit in OS buffers longer. This is all just a theory! Again I haven't been able to reliably reproduce the issue. Adding a flush seems like it can't hurt, though.
Contributor
There was a problem hiding this comment.
Pull Request Overview
This PR adds an explicit flush operation to the vfox download function to prevent potential race conditions when writing downloaded archives to disk. The change addresses issue #6527 where incomplete file writes might cause archive corruption, particularly on CI systems with slower file systems or for smaller files like yarn (1.2MB) that may remain in OS buffers longer.
- Adds
sync_all()call after writing downloaded bytes to disk to ensure complete flush before returning
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
Owner
|
bugbot run |
![cursor[bot]](https://avatars.githubusercontent.com/in/1210556?s=60&v=4){kind=small}
jdx
pushed a commit
that referenced
this pull request
Nov 15, 2025
### 🚀 Features - **(http)** Add 'format' to http backend by @thejcannon in [#6957](#6957) ### 🐛 Bug Fixes - **(bootstrap)** wrong directory on first run by @vmeurisse in [#6971](#6971) - **(tasks)** fix nested colons with `mise task edit` by @jdx in [#6978](#6978) - Use compatible env flags by @thejcannon in [#6964](#6964) - Flush vfox download buffer by @blampe in [#6969](#6969) ### 📚 Documentation - `arch()` template is `x64` by @thejcannon in [#6967](#6967) - update section headers in getting-started.md by @JunichiroKohari in [#6980](#6980) ### New Contributors - @JunichiroKohari made their first contribution in [#6980](#6980) - @blampe made their first contribution in [#6969](#6969) - @thejcannon made their first contribution in [#6964](#6964) ## 📦 Aqua Registry Updates #### New Packages (1) - [`cirruslabs/cirrus-cli`](https://github.com/cirruslabs/cirrus-cli) #### Updated Packages (1) - [`axodotdev/cargo-dist`](https://github.com/axodotdev/cargo-dist)
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.
I suspect (but can't confirm) the issue in #6527 is due to a race condition when writing the downloaded archive to disk.
If the write isn't fully flushed to disk, it's possible for readers to see an incomplete/invalid archive. As a precaution, we should explicitly flush before returning.
The issue is probably more likely on CI runners with slower file systems.
As for why we seem to only see this with
yarn, I'm guessing its smaller size (1.2M) makes it more likely to sit in OS buffers longer.This is all just a theory! Again I haven't been able to reliably reproduce the issue. Adding a flush seems like it can't hurt, though.
Note
Flushes the downloaded file to disk with
file.sync_all()after write to prevent incomplete reads.Written by Cursor Bugbot for commit ec82db4. This will update automatically on new commits. Configure here.