[
https://issues.apache.org/jira/browse/PROTON-2292?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17229306#comment-17229306
]
Andrew Stitcher commented on PROTON-2292:
-----------------------------------------
{quote}8c8a8bc - TSAN sees a possible collision with mutex use (pni_timer_set)
against mutex destruction (pni_timer_free). This is not actually possible. For
connection timers, set and free are serialized calls from the working context.
For the proactor timer, the free call is bound to pn_proactor_free which is
only valid when all worker threads have stopped proactor work and no competing
calls to any proactor APIs can happen
{quote}
I'm pretty sure this is incorrect. TSAN is not seeing a *possible* "collision"
it is seeing an actual unlock when the lock has already been deleted. The way
TSAN works is to instrument the code so if it sees an unlock after free then I
don't see a way that could just be possible it is an *actual* unloock of a
freed lock.
Of course I'm not arguing that your logic is incorrect, just that my
understanding of how TSAN works seems to preclude this conclusion.
> [c] Proactor: TSAN failure in pni_timer_manager
> -----------------------------------------------
>
> Key: PROTON-2292
> URL: https://issues.apache.org/jira/browse/PROTON-2292
> Project: Qpid Proton
> Issue Type: Bug
> Components: proton-c
> Affects Versions: proton-c-0.33.0
> Reporter: Andrew Stitcher
> Assignee: Clifford Jansen
> Priority: Critical
>
> See Test errors starting at:
> https://travis-ci.org/github/apache/qpid-proton/jobs/742285891#L3018
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@qpid.apache.org
For additional commands, e-mail: dev-h...@qpid.apache.org
