Fix statebuffer race#669
Open
josephdviviano wants to merge 3 commits intoFarama-Foundation:masterfrom
Open
Conversation
Member
|
Thanks for the PR @josephdviviano, I was in the process of completely rewriting the vectorizer to fix this problem in #652 however didn't observe significant throughput increase so that sadly stalled. |
Author
|
No problem! Aside from style any changes required? |
The sync_env_ids variable was extracted from info but never used, masking a missing sanity check. Add assertions that sync env IDs are always [0, 1, ..., N-1] after reset and each step, validating the baseline the async comparison relies on. Co-Authored-By: Claude Opus 4.6 <[email protected]>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The test failing in #668 is actually due to a flakey race condition that has failed in previous runs of CI - on master — on Windows, in CI run 20594583659 from December 30.
I think, in unordered mode, write_idx_.fetch_add(1) assigns a circular buffer position, but the actual data write happens after the position is claimed. Another thread can finish its write and increment count_ past the batch boundary, causing collect() to read a position that the first thread hasn't written to yet.
I was able to reproduce on my machine with the following test on the current codebase, but it should pass currently.