Fix Decs Processor #423
Merged
Fix Decs Processor #423
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
marron-at-work
approved these changes
Dec 13, 2025
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Outdated suggestions cannot be applied.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 decided to rework the decs processor a fair bit after I realized CI hanging was ultimately based on manually locking the decs processor mutex. There were also a few races I discovered, mainly the decs thread was starting before the whole decs processor object was initialized and was resuming too early in a collection.
So, we now store state as an atomic, don't manually lock anywhere, and ensure the decs thread is paused for the duration of a whole collection (including computing memstats).
I believe this design is a fair bit simpler - it passes CI on my fork and I ran into no problems when testing with thread sanitizer.