[
https://issues.apache.org/jira/browse/QPID-6416?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Chuck Rolke closed QPID-6416.
-----------------------------
Resolution: Fixed
In the example code the derived class (that contained the qpidmessaging
objects) never had its destructor called. Adding a virtual destructor to the
base class
{noformat}
virtual ~MsgPublisher() {}
{noformat}
gets MsgPublisher to call destructors on the derived objects. This in turn gets
qpidmessaging to close it's internal threads in an orderly manner.
* The reason that it happens intermittently is that there's a race. If the
worker threads close before the main thread then everything is OK. If the main
thread closes first then this crash happens.
* The reason it does not happen in protocol version 0-10 is because threads
are owned by a static singleton that is destroyed by the main thread only.
An improvement in how threads are handled will be discussed in a separate issue.
> Mutex.h:116: void qpid::sys::Mutex::lock(): Assertion `0' failed. - during
> Messaging shutdown
> ---------------------------------------------------------------------------------------------
>
> Key: QPID-6416
> URL: https://issues.apache.org/jira/browse/QPID-6416
> Project: Qpid
> Issue Type: Bug
> Components: C++ Client
> Affects Versions: 0.31
> Environment: Raspberry Pi Pidora Linux 3.12.23-1 armv61 32-bit
> Qpid / Proton trunk circa 2015-02-10
> Messaging client test program
> Reporter: Chuck Rolke
> Assignee: Chuck Rolke
> Fix For: 0.31
>
> Attachments: 6416-fails.cpp, 6416-works.cpp, QpidPublisher.cpp,
> qpid-6416-bt.txt, testQpidFactory.cpp
>
>
> The C++ code uses Qpid Messaging and fails with the summary message about
> half the time.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@qpid.apache.org
For additional commands, e-mail: dev-h...@qpid.apache.org
