[
https://issues.apache.org/jira/browse/DISPATCH-2238?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17404462#comment-17404462
]
ASF GitHub Bot commented on DISPATCH-2238:
------------------------------------------
franz1981 commented on pull request #1349:
URL: https://github.com/apache/qpid-dispatch/pull/1349#issuecomment-905515315
> It appears to reduce to average wait time for the action_lock, likely due
to the reduction of false cond_wait() wakeups as you explained - thanks!
Yep, it should help a bit if the core thread is busy processing actions, to
further speed up workers to offer new ones.
In my old tests it was helping with performance too :D
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
> Avoid holding the core action lock when waking the core thread
> --------------------------------------------------------------
>
> Key: DISPATCH-2238
> URL: https://issues.apache.org/jira/browse/DISPATCH-2238
> Project: Qpid Dispatch
> Issue Type: Improvement
> Components: Router Node
> Affects Versions: 1.17.0
> Reporter: Ken Giusti
> Assignee: Ken Giusti
> Priority: Minor
> Fix For: 1.18.0
>
>
> Currently when an action is queued for the core to process, the queuing code
> calls pthread_cond_signal() while holding the associated mutex. This can
> cause the core thread to be woken then immediately put sleep while taking the
> mutex.
>
> It is legal to call pthread_cond_signal() without holding the lock. Doing
> this will avoid contention on the associated mutex.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
