fix: wake acceptor select via self-pipe on close#46
Merged
Conversation
Closing listen fds from another thread does not reliably wake select on Linux (POSIX leaves it unspecified). #45's stop path therefore still blocked until the accept timeout — reproduced by the new wake tests (~30s). Close the self-pipe write end so the read end becomes readable and select returns promptly. Co-authored-by: Kaius Ruokonen <ruoka@users.noreply.github.com>
Keep the assertion well below the 30s accept timeout while allowing scheduler noise when the full net suite runs concurrently. Co-authored-by: Kaius Ruokonen <ruoka@users.noreply.github.com>
ruoka
marked this pull request as ready for review
July 22, 2026 06:14
Contributor
Author
|
Bugbot is not enabled for your account, so this pull request was not reviewed. Enable Bugbot in the Cursor dashboard to get automatic reviews on future PRs. |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Bug and impact
http::server::stop()/acceptor::close()from #45 closes listen fds to abort a blocked accept wait, but on Linux that does not wakeselect(). Idlestop()+join()still blocks for the full accept timeout (reproduced ~30s with the new wake tests; Docker SIGTERM grace can expire).Root cause
POSIX leaves cross-thread
close()of a selected fd unspecified. On Linux,select()keeps blocking on the (now closed) listen fds until the timeval expires. Closing listen sockets alone is not a portable wake.Fix
net::acceptorselectsetclose(), close the write end so the read end becomes readable (EOF) andselectreturns promptlyValidation
close wakes accept wait promptlyandstop wakes accept wait promptly with no clientnow pass on Linux (previously hung ~30s)< 2s(still ≪ 30s accept timeout) for scheduler noise under full-suite load