Fix deadlock when handling open channel on contended handler - #734
Merged
Conversation
EpicEric
changed the title
Before Eugeny#723, synchronous channel acceptance/rejection would happen regardless of message dispatch. Now, the channel reply uses the same bounded mpsc channel that the handler consumes from. However, under contention (for example, with multiple simultaneous channels opening and writing over the same handler), the awaits in `ChannelOpenHandleInner::accept` and `ChannelOpenHandleInner::reject` may hang while the consumer is the handler itself, thus leading to a self-deadlock. This PR changes channel confirmation to use its own unbounded channel on both server and client implementations, avoiding any awaits (since send is sync), and thus, any deadlocks under load. Because of the change to `ChannelOpenHandleInner::accept` and `ChannelOpenHandleInner::reject` into sync methods, this is backwards-incompatible.
EpicEric
force-pushed
the
fix-channel-open-starvation
branch
from
July 6, 2026 12:16
53070fd to
7298d48
Compare
Contributor
Author
|
I've added a test_contention regression test that sometimes locks up on russh 0.62.1 due to the race condition, but runs correctly with these changes. |
Contributor
Author
|
Despite the accept/reject methods being fully sync, I could change them back to async as to not break backwards compatibility. |
Owner
|
Thanks for the PR! Please change the accept/reject fns back to async - it's barely an issue for client code, and I'd like to avoid a breaking change for a bugfix |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Before #723, synchronous channel acceptance/rejection would happen regardless of message dispatch. Now, the channel reply uses the same bounded mpsc channel that the handler consumes from.
However, under contention (for example, with multiple simultaneous channels opening and writing over the same handler), the awaits in
ChannelOpenHandleInner::acceptandChannelOpenHandleInner::rejectmay hang since the consumer is the handler itself, thus leading to a self-deadlock.This PR changes channel confirmation to use its own unbounded channel on both server and client implementations, avoiding any awaits (since send is sync), and thus, any deadlocks under load.
Because of the change to
ChannelOpenHandleInner::acceptandChannelOpenHandleInner::rejectinto sync methods, this is backwards-incompatible.