(gutsy) pidgin exits when changing buddy icon

[ChrisC]

Binary package hint: pidgin

I update gutsy every day, so the version is whatever is current (I'll comment if its fixed!)
this seems independent image type
doesnt happen when off line
might be related to the msn protocol

[Justin M. Wray (wray-justin)]

Thank you for taking the time to report this bug and helping to make Ubuntu better. Please try to obtain a backtrace following the instructions at [WWW] http://wiki.ubuntu.com/DebuggingProgramCrash and upload the backtrace (as an attachment) to the bug report. This will greatly help us in tracking down your problem.

[ChrisC (chris-chris-camacho-deactivatedaccount)]

assuming people who report bugs can follow that complexity of instructions must *really* cut down the number
of bugs you have to deal with...

GNU gdb 6.6-debian
Copyright (C) 2006 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB. Type "show warranty" for details.
This GDB was configured as "i486-linux-gnu"...
Using host libthread_db library "/lib/tls/i686/cmov/libthread_db.so.1".
(gdb) handle SIG33 pass nostop noprint
Signal Stop Print Pass to program Description
SIG33 No No Yes Real-time event 33
(gdb) set pagination 0
(gdb) run
Starting program: /usr/bin/pidgin
[Thread debugging using libthread_db enabled]
[New Thread -1224443696 (LWP 6309)]
[New Thread -1247335536 (LWP 6312)]
[New Thread -1271874672 (LWP 6331)]
[New Thread -1280267376 (LWP 6332)]
[New Thread -1288660080 (LWP 6333)]
[Thread -1280267376 (LWP 6332) exited]
[Thread -1288660080 (LWP 6333) exited]
[Thread -1271874672 (LWP 6331) exited]
[New Thread -1271874672 (LWP 6334)]
[Thread -1271874672 (LWP 6334) exited]
[New Thread -1271874672 (LWP 6335)]
[New Thread -1288660080 (LWP 6336)]
[New Thread -1280267376 (LWP 6337)]
[New Thread -1297052784 (LWP 6338)]
[New Thread -1305445488 (LWP 6339)]
[New Thread -1313838192 (LWP 6340)]
[New Thread -1322230896 (LWP 6341)]
[New Thread -1330623600 (LWP 6342)]
[New Thread -1339016304 (LWP 6343)]
[New Thread -1347409008 (LWP 6344)]
[Thread -1280267376 (LWP 6337) exited]
[Thread -1271874672 (LWP 6335) exited]
[Thread -1313838192 (LWP 6340) exited]
[Thread -1305445488 (LWP 6339) exited]
[Thread -1297052784 (LWP 6338) exited]
[Thread -1288660080 (LWP 6336) exited]
[Thread -1347409008 (LWP 6344) exited]
[Thread -1339016304 (LWP 6343) exited]
[Thread -1330623600 (LWP 6342) exited]
[Thread -1322230896 (LWP 6341) exited]
[New Thread -1322230896 (LWP 6346)]
[Thread -1322230896 (LWP 6346) exited]
[New Thread -1322230896 (LWP 6348)]
[New Thread -1330623600 (LWP 6349)]
[New Thread -1339016304 (LWP 6350)]
[New Thread -1347409008 (LWP 6351)]
[New Thread -1271874672 (LWP 6352)]
[New Thread -1280267376 (LWP 6353)]
[Thread -1330623600 (LWP 6349) exited]
[Thread -1322230896 (LWP 6348) exited]
[Thread -1280267376 (LWP 6353) exited]
[Thread -1271874672 (LWP 6352) exited]
[Thread -1347409008 (LWP 6351) exited]
[Thread -1339016304 (LWP 6350) exited]
[New Thread -1339016304 (LWP 6361)]

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread -1224443696 (LWP 6309)]
0x08095b32 in update_buddy_status_timeout (buddy=0xb76eb8ac) at ../../pidgin/gtkconv.c:7054
7054 ../../pidgin/gtkconv.c: No such file or directory.
 in ../../pidgin/gtkconv.c
(gdb) backtrace full
#0 0x08095b32 in update_buddy_status_timeout (buddy=0xb76eb8ac) at ../../pidgin/gtkconv.c:7054
 conv = <value optimized out>
#1 0x082d8a18 in ?? ()
No symbol table info available.
#2 0xb76eb8ac in __pthread_mutex_unlock_usercnt () from /lib/tls/i686/cmov/libpthread.so.0
No symbol table info available.
Backtrace stopped: previous frame inner to this frame (corrup...

[Saivann Carignan (oxmosys)]

Thanks to everybody who worked on this bug report. Since this bug hasn't been experienced with the ubuntu version of pidgin yet, I set the status of the bug report to invalid. To help Pidgin developpers to fix this issue, you can login and comment in their own bug report system here : http://developer.pidgin.im/ticket/1857

If somebody can reproduce this bug with pidgin from ubuntu repositories, the status of this bug should be set back to new.

[Alejandro R. Mosteo (mosteo)]

I'm seeing again this bug in 9.04, Pidgin version 2.5.5. It doesn't happen 100% of the time, it goes and comes intermitently.

$ apt-cache policy pidgin
pidgin:
  Installed: 1:2.5.5-1ubuntu8.1
  Candidate: 1:2.5.5-1ubuntu8.1

Procedure: click on the global buddy icon besides the status drop-down, the selector opens, and on double-clicking the icon or clicking OK, pidgin crashes. (Identical to that in pidgin-bugs #1857, which was supposedly fixed in 2.4.0)

It happens now both online and offline, having a previous icon or without one. I have enabled MSN, jabber, AIM, yahoo accounts.

The buddy icon I'm trying to select is a png. I got a backtrace after installing debug symbols with gdb:

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0xb70a6750 (LWP 25893)]
0xb769f51f in memcpy () from /lib/libc.so.6
(gdb) bt
#0 0xb769f51f in memcpy () from /lib/libc.so.6
#1 0xb79aab33 in gdk_pixbuf_loader_write () from /usr/lib/libgdk_pixbuf-2.0.so.0
#2 0x080ea3c4 in pidgin_status_box_redisplay_buddy_icon (status_box=0x8770270) at ../../pidgin/gtkstatusbox.c:2262
#3 0x080ecf4b in buddy_icon_set_cb (filename=0x895e338 "/home/user/buddy icons/me/me.png", box=0x8770270)
    at ../../pidgin/gtkstatusbox.c:1536
#4 0xb77e9866 in do_callbacks (name=0x80fc0d4 "/pidgin/accounts/buddyicon", pref=0x8643c38) at ../../libpurple/prefs.c:759
#5 0x080ed0d0 in icon_choose_cb (filename=0x400 <Address 0x400 out of bounds>, data=0x8770000) at ../../pidgin/gtkstatusbox.c:1559
#6 0x080f1f6e in icon_filesel_choose_cb (widget=0x895d020, response=-3, dialog=0x8f86860) at ../../pidgin/gtkutils.c:2375
#7 0xb79370ec in g_cclosure_marshal_VOID () from /usr/lib/libgobject-2.0.so.0
#8 0xb7929c7b in g_closure_invoke () from /usr/lib/libgobject-2.0.so.0
#9 0xb793fe57 in ?? () from /usr/lib/libgobject-2.0.so.0
#10 0xb79414b9 in g_signal_emit_valist () from /usr/lib/libgobject-2.0.so.0
#11 0xb7941936 in g_signal_emit () from /usr/lib/libgobject-2.0.so.0
#12 0xb7b10b31 in gtk_dialog_response () from /usr/lib/libgtk-x11-2.0.so.0
#13 0xb79373a4 in g_cclosure_marshal_VOID__VOID () from /usr/lib/libgobject-2.0.so.0
#14 0xb7929c7b in g_closure_invoke () from /usr/lib/libgobject-2.0.so.0
#15 0xb793fe57 in ?? () from /usr/lib/libgobject-2.0.so.0
#16 0xb79414b9 in g_signal_emit_valist () from /usr/lib/libgobject-2.0.so.0
#17 0xb7941936 in g_signal_emit () from /usr/lib/libgobject-2.0.so.0
#18 0xb7adebda in gtk_button_clicked () from /usr/lib/libgtk-x11-2.0.so.0
#19 0xb7ae00f5 in ?? () from /usr/lib/libgtk-x11-2.0.so.0
#20 0xb7ae0143 in ?? () from /usr/lib/libgtk-x11-2.0.so.0
#21 0xb79eb8fb in ?? () from /usr/lib/libgdk-x11-2.0.so.0
#22 0xb78a12b6 in ?? () from /usr/lib/libglib-2.0.so.0
#23 0xb78a0b88 in g_main_context_dispatch () from /usr/lib/libglib-2.0.so.0
#24 0xb78a40eb in ?? () from /usr/lib/libglib-2.0.so.0
#25 0xb78a45ba in g_main_loop_run () from /usr/lib/libglib-2.0.so.0
#26 0xb7b937d9 in gtk_main () from /usr/lib/libgtk-x11-2.0.so.0
#27 0x080c31da in main (argc=0, argv=0xbfe71a04) at ../../pidgin/gtkmain.c:880

[Saivann Carignan (oxmosys)]

mosteo : I can't reproduce the bug here, can you attach a png that crash pidgin? I tried with many png with XMPP, Yahoo, MSN and IRC enabled and I was not able to reproduce the problem. When you say "Jabber", you probably mean XMPP right? Are you able to reproduce the bug if your remove your AIM account from your configuration?

[Alejandro R. Mosteo (mosteo)]

I suspect that the particular PNG is not at fault, since I've used it other times and in other machines... anyway, attached it is.

You're right, by jabber I meant XMPP gmail accounts.

Removing the AIM account (deleted entirely) made no change.

Want a strace?

[xteejx (xteejx)]

Unfortunately, I cannot reproduce this either in Jaunty, pidgin version 2.5.5-1ubuntu8.1
I think the easiest way for you to provide the info we need is to force apport to see it. You can do this by running "sudo force_start=1 /etc/init.d/apport restart" without quotes. This will enable apport for one session only, so can you do that for us, and force pidgin to crash how you did before, and let it report a new bug, giving it full access to Launchpad.
If you can also put a quick comment on this bug report for the bug number you get, we can mark this as a duplicate so developers can see both sets of info.
Thank you in advance!

[Alejandro R. Mosteo (mosteo)]

I see that the apport info includes a core dump, won't that contain my passwords?

Anyway, I've got good news: I've pinpointed it to the 'mystatusbox' plugin. With it disabled, no crash. Enabled, crash. Its configuration is with both checkboxes disabled.

I hope with this info you can reproduce it. I see however that the plugin comes from universe.

[xteejx (xteejx)]

A core dump is only included if apport cannot retrace itself, and as a matter of policy, only the Ubuntu Bug Control team and yourself can see any info contained within these files, which will be edited if there is any such info.
Again I have tried to reproduce the crash how you mention, and still cannot do it. Can you provide the information asked above with apport? Without this info, and without being able to reproduce it, we will have to close this report. Thank you.

[Saivann Carignan (oxmosys)]

mosteo : I didn't find mystatusbox plugin in default pidgin plugins and on synaptic. Can you tell how you got that plugin?

[Alejandro R. Mosteo (mosteo)]

It's in the package pidgin-plugin-pack (version 2.2.0-1)

Gimme some time for the apport thing...

[xteejx (xteejx)]

No problem. When you get a bug number mark this as a duplicate if you
can, or give us the bug number here. Thank you.

[Alejandro R. Mosteo (mosteo)]

The duplicate with the apport info is bug 383986.

[Saivann Carignan (oxmosys)]

I confirm that I was able to crash pidgin with segmentation fault with mystatusbox plugin enabled. Like mosteo said, this is not reproducible 100% of the time, however I think that since this bug is not the same as the initial bug report (which affected pidgin without plugin), this bug report should be set back to "fix released" and the new apport bug report should get a nice description, including all steps to reproduce.

[Saivann Carignan (oxmosys)]

mosteo : I get a Forbidden page when trying to access the apport bug, probably because it is filed for upstream pidgin project rather than pidgin in ubuntu. If not already set, can you click on "Also Affects distribution" and choose "Ubuntu" as the distribution, and "pidgin" as the source package for the apport bug? That will probably allow access to ubuntu bug control team to that bug while keeping it private.

[Alejandro R. Mosteo (mosteo)]

@Saivann: I tried what you suggested but the bug was already tagged so. In the end, I've made the duplicate 383986 public, since I tested with a temporary password in a throwaway account, so there should be nothing compromising in the dump.

[xteejx (xteejx)]

I see the confusion above has been sorted out. Launchpad does not allow you to mark a private bug as a duplicate or vice-versa. But it's done now, so no need to worry. The upstream bug has been fixed, will check versions, etc and see when it will be released into Ubuntu.

[xteejx (xteejx)]

Actually THIS bug should have been marked as the duplicate, will fix that.

[Saivann Carignan (oxmosys)]

Teej : This is not the end of the world, but this bug should probably not be a duplicate since mosteo bug a similar, but different bug (which require a plugin to segfault, that was not the case for bug #139081 which is fixed).