Repository navigation
fixed lost stream send request when a flow control update races with queueing it - #96
Merged
Merged
Conversation
…queueing it The send request was counted only after it had been queued. If the sender thread executed it in between (finding the stream blocked by flow control) and the receiver thread processed the MAX_STREAM_DATA right then, streamNotBlocked saw a count of -1 and queued nothing: the stream never sent again and the writer blocked forever on the full send buffer. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Owner
|
Thanks! This also fixes #91 ;-). |
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Hi Peter,
we encountered a stalled write (KWIK 0.12) in our tests and Claude found a possible race:
The send request was counted only after it had been queued. If the sender thread executed it in between (finding the stream blocked by flow control) and the receiver thread processed the MAX_STREAM_DATA right then, streamNotBlocked saw a count of -1 and queued nothing: the stream never sent again and the writer blocked forever on the full send buffer.