Bug #100577 for POE: POE::Kernel->stop causes crash

Thu Nov 27 05:10:06 2014 GVL [...] cpan.org - Ticket created

Subject:

POE::Kernel->stop causes crash

Latest changes in 1.366 cause nested sessions running under POE::Wheel::Run to die in certain circumstances. Sadly the code causing this is proprietary and I haven't been able to reproduce it in a code sample yet. This is the change that causes the crash: https://github.com/rcaputo/poe/commit/c6f8e4f9dabbef51a75dbb043c01c92807fc604c lib/POE/Kernel.pm line 1330. It works fine with _data_ses_free and POE < 1.366.

Fri Nov 28 03:03:50 2014 RCAPUTO [...] cpan.org - Correspondence added

On Thu Nov 27 05:10:06 2014, GVL wrote: Show quoted text

> Latest changes in 1.366 cause nested sessions running under > POE::Wheel::Run to die in certain circumstances.

Show quoted text

> This is the change that causes the crash: > > https://github.com/rcaputo/poe/commit/c6f8e4f9dabbef51a75dbb043c01c92807fc604c

The change invokes _stop handlers during POE::Kernel->stop(). Previously _stop handlers would not be invoked. Is it possible that a side effect in one of your _stop handlers is causing the crash?

Fri Nov 28 03:03:51 2014 The RT System itself - Status changed from 'new' to 'open'

Mon Dec 01 08:59:48 2014 GVL [...] cpan.org - Correspondence added

I do not have a stop handler in the child package (that dies) and the stop handler in the parent just calls $log (Log::Log4perl) once to notify that the component has stopped. On Fri Nov 28 03:03:50 2014, RCAPUTO wrote: Show quoted text

> The change invokes _stop handlers during POE::Kernel->stop(). > Previously _stop handlers would not be invoked. > > Is it possible that a side effect in one of your _stop handlers is > causing the crash?

Tue Jun 02 14:15:18 2015 RCAPUTO [...] cpan.org - Correspondence added

On Mon Dec 01 08:59:48 2014, GVL wrote: Show quoted text

> I do not have a stop handler in the child package (that dies) and the > stop handler in the parent just calls $log (Log::Log4perl) once to > notify that the component has stopped.

I'm afraid I can't fix this without more information. I don't seem to have a copy of the crash message(s), and I haven't been able to reproduce the problem so I could track it down.

Tue Dec 13 04:38:43 2016 GVL [...] cpan.org - Correspondence added

There were _stop handlers that got called with the cahnegs in 1.366 and basically shut down the whole process.

Tue Dec 13 04:38:44 2016 GVL [...] cpan.org - Status changed from 'open' to 'rejected'

Tue Dec 13 04:38:44 2016 GVL [...] cpan.org - Fixed in 1.366 added