Fix refcounting in non blocking send path. #227
Closed
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Outdated suggestions cannot be applied.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.
We were facing connection fd leak when connection goes in
destroyed state, which causing clients hung, as fd was open and not
rearmed for epoll, client will get tcp zero window and may or
mat not do connection reset to recover based on kernel version.
If connection xprt is in destroyed state, then we may not
be able process POLLOUT event for send task and refs taken will
not go away.
We don't need to register for POLLOUT event with ref taken for send
queue, as we take ref for send task.
We don't need ref for each response we queue, we can just single ref
for send queue and release it when we start processing.
We need to close xp_fd_send on connection destroy.