[
https://issues.apache.org/jira/browse/PROTON-764?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14236076#comment-14236076
]
Ted Ross commented on PROTON-764:
---------------------------------
I've lowered the severity of this to minor. I think Dispatch is the only user
of this driver that will see this problem and the issue has been worked around
in Dispatch.
> Driver is multi-thread-unsafe when connections are closed uncleanly
> -------------------------------------------------------------------
>
> Key: PROTON-764
> URL: https://issues.apache.org/jira/browse/PROTON-764
> Project: Qpid Proton
> Issue Type: Bug
> Components: proton-c
> Affects Versions: 0.8
> Reporter: Ted Ross
> Priority: Minor
>
> In a multi-threaded environment (Qpid Dispatch), lost connections can result
> in multiple calls to pn_connector_close for a given connector. This results
> in driver->close_count never returning to zero and the poll loop forever
> running with a zero timeout.
> The reason for this is that pn_driver_wait_3 calls pn_connector_close when it
> sees an error, even if that connector is currently being processed by another
> thread. I think that pn_driver_wait_3 should only flag the connector as
> failed so the user can close it in the proper thread.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
