Hi Dan,
On Tue, 2003-09-30 at 18:59, Dan Winship wrote:
> I can reproduce this reliably. If I change any gconf key that the
> summary is listening to, it will throw away and recreate its CalClients,
> resulting in connector wedging like so:
...
> #1 0x41364198 in __pthread_wait_for_restart_signal ()
> from /lib/libpthread.so.0
> #2 0x41365cc9 in __pthread_alt_lock () from /lib/libpthread.so.0
> #3 0x413629f7 in pthread_mutex_lock () from /lib/libpthread.so.0
> #4 0x41301ddd in ORBit_POA_handle_request (poa=0x81139c0,
> recv_buffer=0x8257578, objkey=0x8257590) at poa.c:1336
...
> #24 0x413011bb in ORBit_POA_deactivate_object (poa=0x8113808, pobj=0x8250478,
> do_etherealize=1 '\001', is_cleanup=160 '\x{FFFF}') at poa.c:997
I fixed a deadlock very similar to this in HEAD ORBit2 yesterday from a
similar trace on gnome-components-list; can you update and try again:
2003-09-29 Michael Meeks <[EMAIL PROTECTED]>
* test/test-mem.c (main_func): release refs we own.
implies the fix;
Thanks,
Michael.
--
[EMAIL PROTECTED] <><, Pseudo Engineer, itinerant idiot
_______________________________________________
evolution-hackers maillist - [EMAIL PROTECTED]
http://lists.ximian.com/mailman/listinfo/evolution-hackers
