after using yelp evolution crash on closing

Thanks for your bug. How does it "won't restart"? What happens if you try starting it again? Does it hang? Does it crash? Do you get any command line message about it?

Wow Sebastien, you're fast!!!

Once you view the Evolution help content and close Evolution, Evolution
will never restart by clicking on any icon launcher until you reboot.
Before rebooting, the small circle pops up, indicating that Evolution is
starting, but then dies without displaying the Evolution window or any
errors (AMD64 3200 Venice, Asus A8N-E Nforce4, Ubuntu Dapper AMD64). The
little circular activity indicator just disappears after it runs for a
while.

Once Evolution won't restart and you try to reboot the PC, the PC hangs
for a few minutes before it will actually reboot. There is nothing
displayed about what is going on, or any problems. After a few minutes
the PC will eventually reboot.

After you reboot the PC, you can then use Evolution again until you look
at the help contents. Once viewing help contents again, Evolution will
no longer function until you reboot.

This happens every time on two PCs. One PC (second one listed below) is
a fresh Dapper FLIGHT6 install with nothing else done, this PC might be
FLIGHT5 updated by adding a few packages.

2.6.15-18-AMD64-generic kernel on this one (listed above)
2.6.15-19-AMD64-generic kernel on the other PC (AMD3200 Winchester,MSI
Neo 2 Plat Nforce3, Ubuntu Dapper AMD64)

Jim

On Sun, 2006-04-02 at 14:44 +0000, Sebastien Bacher wrote:
> Thanks for your bug. How does it "won't restart"? What happens if you try starting it again? Does it hang? Does it crash? Do you get any command line message about it?
>
> ** Changed in: evolution (Ubuntu)
> Assignee: (unassigned) => Ubuntu Desktop Bugs
> Status: Unconfirmed => Needs Info

does 'evolution --force-shutdown' help?

Dennis,

I started Evolution, clicked on the help contents then exited. Evolution
can not be restarted as stated before.

I then entered your command "evolution --force-shutdown" and it echoed
back "CalDAV Eplugin Starting Up...." and it's hung at that point. Hope
it helps.

Jim

On Sun, 02 Apr 2006 12:40:09 -0400, Dennis Kaarsemaker
<email address hidden> wrote:

> does 'evolution --force-shutdown' help?
>

--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/

What does happen if you try restarting it from a command line? Does crash or hang? Could you run it with gdb to get a backtrace?
- gdb evolution
(gdb) run
... hang
(gdb) thread apply all bt

Does it use some CPU or just return to the command line or just hang without using CPU?

I can confirm this.
A normal evolution run shows the following warnings:
  CalDAV Eplugin starting up ...
  (evolution-2.6:22253): evolution-mail-WARNING **: ignored this junk plugin: not enabled or we have already loaded one
  (evolution-2.6:22253): e-utils-WARNING **: Plugin 'Bogofilter junk plugin' failed to load hook 'org.gnome.evolution.mail.junk:1.0'
  (evolution-2.6:22253): camel-WARNING **: camel_exception_get_id called with NULL parameter.
Now when the help contents is opened, it shows this
  /usr/share/yelp/info.xml:0: error : unterminated entity reference development tools.
  Unmatched element: br
In that case, when evolution is closed, it warns
  *** glibc detected *** corrupted double-linked list: 0x00000000005daa00 ***

When I attach to evolution with gdb, I see the following:
[reading lots of symbols]
0x00002ba731f5a94b in pthread_setcanceltype ()
 from /lib/libc.so.6
(gdb) thread apply all bt

Thread 6 (Thread 1082132832 (LWP 22306)):
#0 0x00002ba732b5ca9a in pthread_cond_wait@@GLIBC_2.3.2 ()
   from /lib/libpthread.so.0
#1 0x00002ba72e230f22 in e_msgport_wait ()
   from /usr/lib/libedataserver-1.2.so.7
#2 0x00002ba72e23165b in e_thread_busy ()
   from /usr/lib/libedataserver-1.2.so.7
#3 0x00002ba732b5a0fa in start_thread () from /lib/libpthread.so.0
#4 0x00002ba731f4ece2 in clone () from /lib/libc.so.6
#5 0x0000000000000000 in ?? ()

Thread 5 (Thread 1090525536 (LWP 22309)):
#0 0x00002ba732b5ca9a in pthread_cond_wait@@GLIBC_2.3.2 ()
   from /lib/libpthread.so.0
#1 0x00002ba72e230f22 in e_msgport_wait ()
   from /usr/lib/libedataserver-1.2.so.7
#2 0x00002ba72e23165b in e_thread_busy ()
   from /usr/lib/libedataserver-1.2.so.7
#3 0x00002ba732b5a0fa in start_thread () from /lib/libpthread.so.0
#4 0x00002ba731f4ece2 in clone () from /lib/libc.so.6
#5 0x0000000000000000 in ?? ()
Thread 4 (Thread 1098918240 (LWP 22311)):
#0 0x00002ba732b5ca9a in pthread_cond_wait@@GLIBC_2.3.2 ()
   from /lib/libpthread.so.0
#1 0x00002ba72e230f22 in e_msgport_wait ()
   from /usr/lib/libedataserver-1.2.so.7
#2 0x00002ba72e23165b in e_thread_busy ()
   from /usr/lib/libedataserver-1.2.so.7
#3 0x00002ba732b5a0fa in start_thread () from /lib/libpthread.so.0
#4 0x00002ba731f4ece2 in clone () from /lib/libc.so.6
#5 0x0000000000000000 in ?? ()

Thread 3 (Thread 1099184480 (LWP 22312)):
#0 0x00002ba731f45e22 in poll () from /lib/libc.so.6
#1 0x00002ba730c6ed00 in g_main_context_check ()
   from /usr/lib/libglib-2.0.so.0
#2 0x00002ba730c6f18a in g_main_loop_run () from /usr/lib/libglib-2.0.so.0
#3 0x00002ba72f51bb90 in link_thread_io_context ()
   from /usr/lib/libORBit-2.so.0
#4 0x00002ba730c87b6b in g_static_private_free ()
   from /usr/lib/libglib-2.0.so.0
#5 0x00002ba732b5a0fa in start_thread () from /lib/libpthread.so.0
#6 0x00002ba731f4ece2 in clone () from /lib/libc.so.6
#7 0x0000000000000000 in ?? ()

Thread 2 (Thread 1107577184 (LWP 22314)):
#0 0x00002ba732b5ca9a in pthread_cond_wait@@GLIBC_2.3.2 ()
   from /lib/libpthread.so.0
#1 0x00002ba72e230f22 in e_msgport_wait ()
   from /usr/lib/libedataserver-1.2.so.7
#2 0x00002ba72e23165b in e_thread_busy ()
   from /usr/lib/libedataserver-1.2.so.7
#3 0x00002ba732b5a0fa ...

On dapper you can install evolution-dbg and evolution-data-server-dbg to get debug symbols for it, it would probably make the backtrace better.

" *** glibc detected *** corrupted double-linked list: 0x00000000005daa00 ***" is likely to be the issue. Getting a log with valgrind could be useful. Install the valgrind package and run "valgrind --trace-children=yes evolution", then get the issue ... it makes the program really slow but is useful to catch errors than can give a such message by example

Ok, find attached a gdb session with -dbg packages installed and also the valgrind output with the command-line you proposed. Both are runs consisting of: start program, open help, close help, close evolution. If you need more data, please ask. When time permits, I'll do some more debugging.

Thank you for the effort on that, I've forwarded it upstream: https://launchpad.net/distros/ubuntu/+source/evolution/+bug/37723

upstream http://bugzilla.gnome.org/show_bug.cgi?id=347997

the bug should be fixed in hardy, closing but feel free to reopen if you still get the issue

Confirmed on Hardy: neither glibc nor yelp warning are displayed anymore,and evolution properly terminates on quit. Great!