evolution crashed with SIGSEGV in e_cal_menu_target_new_select()

Bug #247480 reported by Laurent Bonnaud
82
Affects Status Importance Assigned to Milestone
Evolution
Fix Released
Critical
evolution (Ubuntu)
Fix Released
Medium
Ubuntu Desktop Bugs
Hardy
Won't Fix
Medium
Ubuntu Desktop Bugs

Bug Description

Binary package hint: evolution

Evolution crashed while my laptop was left unattended. Therefore I cannot tell which action lead to this crash.

In the terminal where I started evolution, the following messages were displayed before the crash:

(evolution:2814): Gtk-CRITICAL **: gtk_status_icon_set_visible: assertion `GTK_IS_STATUS_ICON (status_icon)' failed

(evolution:2814): GLib-GObject-CRITICAL **: g_object_unref: assertion `G_IS_OBJECT (object)' failed
camel-Message: --
Making error
camel-Message: --

(evolution:2814): calendar-gui-CRITICAL **: e_cal_model_get_component_at: assertion `row >= 0 && row < priv->objects->len' failed

(evolution:2814): calendar-gui-CRITICAL **: e_cal_model_copy_component_data: assertion `comp_data != NULL' failed

[1]+ Segmentation fault (core dumped) evolution (wd: ~)

ProblemType: Crash
Architecture: i386
Date: Thu Jul 10 19:53:49 2008
DistroRelease: Ubuntu 8.04
ExecutablePath: /usr/bin/evolution
NonfreeKernelModules: fglrx
Package: evolution 2.22.3.1-0ubuntu1
PackageArchitecture: i386
ProcCmdline: evolution
Signal: 11
SourcePackage: evolution
StacktraceTop:
 e_cal_menu_target_new_select ()
 tasks_control_sensitize_commands ()
 ?? () from /usr/lib/evolution/2.22/components/libevolution-calendar.so
 g_cclosure_marshal_VOID () from /usr/lib/libgobject-2.0.so.0
 g_closure_invoke () from /usr/lib/libgobject-2.0.so.0
Title: evolution crashed with SIGSEGV in e_cal_menu_target_new_select()
Uname: Linux 2.6.24-19-generic i686
UserGroups: adm admin audio cdrom dialout dip floppy fuse kvm lpadmin plugdev pulse-rt scanner staff video

Revision history for this message
Laurent Bonnaud (laurent-bonnaud) wrote :
Revision history for this message
Laurent Bonnaud (laurent-bonnaud) wrote :

Let me add that I do not use evolution's calendar component.

Revision history for this message
Apport retracing service (apport) wrote : Symbolic stack trace

StacktraceTop:e_cal_menu_target_new_select ()
tasks_control_sensitize_commands ()
?? () from /usr/lib/evolution/2.22/components/libevolution-calendar.so
IA__g_cclosure_marshal_VOID__INT (closure=0x92a0670, return_value=0x0, n_param_values=2,
IA__g_closure_invoke (closure=0x92a0670, return_value=0x0, n_param_values=2,

Revision history for this message
Apport retracing service (apport) wrote : Symbolic threaded stack trace
Revision history for this message
Sebastien Bacher (seb128) wrote :

Thank you for your bug report. The crash is similar to http://bugzilla.gnome.org/show_bug.cgi?id=335618

Changed in evolution:
assignee: nobody → desktop-bugs
importance: Undecided → Medium
status: New → Triaged
Changed in evolution:
status: Unknown → Confirmed
Changed in evolution:
status: Confirmed → Fix Released
Revision history for this message
CTenorman (ctenorman) wrote :

Does anyone know when this will be updated in Hardy? This particular bug has made the tasks application almost unusable for my type of work, and it is the primary means of organizing all my information. A quick fix on the Ubuntu end would be greatly appreciated, thanks!

Revision history for this message
Sebastien Bacher (seb128) wrote :

the bug has not been fixed upstream yet so there is nothing to update in hardy

Revision history for this message
TJ (tj) wrote :

I've begun experiencing this with v2.22.3.1 and Tasks.

$ evolution --component=tasks

evolution-shell-Message: Killing old version of evolution-data-server...
(evolution:1628): e-table-CRITICAL **: e_table_model_column_count: assertion `e_table_model != NULL' failed
(evolution:1628): e-table-CRITICAL **: e_table_model_column_count: assertion `e_table_model != NULL' failed
(evolution:1628): e-table-CRITICAL **: e_table_model_column_count: assertion `e_table_model != NULL' failed
(evolution:1628): calendar-gui-CRITICAL **: e_cal_model_get_component_at: assertion `row >= 0 && row < priv->objects->len' failed
(evolution:1628): calendar-gui-CRITICAL **: e_cal_model_copy_component_data: assertion `comp_data != NULL' failed
Segmentation fault

It is reproducable when there are a number of open and completed tasks in the task view, sorted by "%complete" (lowest to highest).

1. Add a new task (which appears at the top of the list).
2. Double-click the new task to edit it.
3. Do Options > Status Details and change values including "Percent complete".
4. When pressing the Close button Evolution segfaults.

as described here and in the upstream bug report.

Revision history for this message
dvhart (darren-dvhart) wrote :

Would it be possible to bump the priority on this? I just enabled apport on Hardy and realized I'm seeing the same bug. This crashes evolution everytime I edit the Categories of a task, I think this merits higher priority than "Medium" as it's pretty much a show stopper.

Revision history for this message
Sebastien Bacher (seb128) wrote :

the bug has been fixed upstream now

Changed in evolution:
status: Triaged → Fix Committed
Revision history for this message
dvhart (darren-dvhart) wrote :

OK, what does that mean for Hardy LTS users? Will Hardy see the fix or will we have to upgrade to Intrepid? I apologize if this is inherent in your statement, but "upstream" is a little ambiguous (from my perspective).

Revision history for this message
Sebastien Bacher (seb128) wrote :

it's not clear if the hardy and intrepid versions have the exact same issue but if there is a clear way to trigger the crash on hardy and somebody backport an upstream change to fix the issue that can be considered as a stable update

Revision history for this message
dvhart (darren-dvhart) wrote :

See my bug report for Bug 267933 for a trivial reproduction scenario on Hardy. I went through the trouble to enable apport on hardy and collected all the relevant crash data there as well. Hopefully that will encourage getting a fix backported to hardy. Without it, the evolution tasklist is quite unusable (unless you don't care about changing due dates or using categories...).

Revision history for this message
Alan Johnson (nilgiri) wrote :

I'm having the same problem with tasks. I'm using it to connect to an Exchange server via OWA, and every time I change the priority, Evolutions crashes. It starts right up again and my change has taken effect, so not the end of the world, but this is certainly the kind of thing that needs to be fixed before I can begin to suggest using Ubuntu across my entire organization (400+ users). Of course, it crashes once in a while when doing other things in tasks, but reliably crashes when changing priorities.

Anyway, it really does need to be backported into the main Hardy repos, but is there some way to get just the relevant packages to update from some repo for the time being? I don't anything in proposed updates or unsupported for this, and only tar balls from the Evolution site. I would be happy with dep packages from anywhere for now.

Revision history for this message
TJ (tj) wrote :

I've cherry-picked the patch from upstream and so far my testing seems to indicate the issue is resolved. I've uploaded a fixed package to my PPA so if you would like to test and confirm the fix we can see about getting it approved as a SRU.

evolution (2.22.3.1-0ubuntu2) hardy; urgency=low

  * debian/patches 64 Fix task list crash (Gnome #544187 [r36276], #551382 [r36282]) (LP: #247480).

 -- TJ <email address hidden> Mon, 29 Sep 2008 22:00:00 +0200

Revision history for this message
dvhart (darren-dvhart) wrote :

TJ, I tried a dozen combinations of things that all crashed before and they all worked as expected this time. Thanks for putting up the package on your PPA. In case others want to try and don't know what a PPA is - I didn't:

https://launchpad.net/~intuitivenipple/+archive

Just add the hardy deb source and then install evolution with aptitude (or whatever).

So this fixes the issue for me, thanks TJ!

Revision history for this message
Alan Johnson (nilgiri) wrote :

I just upgraded using the Evolution packages from TJ's PPA and I beat on the tasks pretty hard for a few minutes with no errors. I use this daily and will update immediately if I have any more crashes. So far, all is beautiful! I'm subscribed to this bug, so if anyone wants an update on my status before promoting the fix, just post a request here and I'll get right on it.

Thanks TJ!

Also, thanks dvhart! because I may never have figured out the PPA without your post. =)

Rock on!

Revision history for this message
TJ (tj) wrote :

SRU Justification:

Impact: When using Evolution Tasks list and editing tasks Evolution will crash immediately the task is saved.

Testcase:

It is reproducable when there are a number of open and completed tasks in the task view, sorted by "%complete" (lowest to highest). There are other scenarios too (see LP bug #267933).

Select the Tasks window, .
1. Add a new task (which appears at the top of the list).
2. Double-click the new task to edit it.
3. Do Options > Status Details and change values including "Percent complete".
4. When pressing the Close button Evolution segfaults.

Revision history for this message
Sebastien Bacher (seb128) wrote :

thank you for your work on the updates, some notes though:

- the upload target should be hardy-proposed and not hardy
- no need to list GNOME revision, or should you wrap the lines after 72 chars
- there is a .cproject in the diff which should probably not be there

could you fix those and update the bug?

Revision history for this message
Sebastien Bacher (seb128) wrote :

the new version is in intrepid now

Changed in evolution:
status: Fix Committed → Fix Released
assignee: nobody → desktop-bugs
importance: Undecided → Medium
status: New → Confirmed
Revision history for this message
Alan Johnson (nilgiri) wrote :

I just had a related crash that I occationally received before. I updated a due date, change the priority, and then tried to open that task to edit the notes, and I Evolution did not crash, but tasks did. The error message was something like "tasks have crashed. you will not be able to access your tasks again until evolution has been restarted." (sorry I didn't get the exact text before I closed the dialog.) Sure enough, restarting evolution did the trick just fine. Again, I has see this before updating with the PPA posted here, but this is the first time since.

Should this be a new bug? I was considering it a part of what was being listed here. Perhaps a more complete update from upstream is all that is needed? Maybe I am not saying that right, but the package from the PPA has a 2.2 version and was "cherry picked" from the upstream which has a 2.4 version in the latest stable release. What I mean to say is that perhaps more of the 2.4 needs to be pulled in. All would be ideal IMHO.

Revision history for this message
mercutio22 (macabro22) wrote :

This happened to me in intrepid ibex =P

Revision history for this message
Nick Jenkins (nickpj) wrote :

Can we please get an upload to hardy-proposed? I'm happy to test this update. I just did a "sudo aptitude install -t hardy-proposed evolution", but there weren't any packages; closest I could see were evolution-data-server and evolution-data-server-common, but there weren't any updates for the evolution front-end.

Revision history for this message
Svt (sveta-d) wrote :

Thanks for your update. I've tried it on hardy, but it solved my problem only _partially_

My problem differed a little from those mentioned above. I had problem with the field 'Start date' of the task. When I had a task with some start date assigned to it, and then I wanted just to clear start date info - evolution failed with the same error as mentioned.

After installing your update, it stopped to close unexpectedly, but now this field doesn't refreshed in the task list view. Inside the task I see that the date field is clear. But in task list the old one stays. The value in the list changes only after restart of evolution.

Revision history for this message
Alan Johnson (nilgiri) wrote :

I can confirm the exact behavior described by Svt. I still wonder if the latest version would fix this. The version I references before were off a bit. The current Ubuntu version 2.22 and the latest stable from the Evolution project is 2.24.

TJ, could you get the entire 2.24 release into your repo for testing? Or is that what you meant by "cherry picked". I get the feeling that "cherry picked" meant you only grabbed certain bits of the 2.24 code release in your package, but I would be very interested and happy to test the full 2.24 if it were available in a .deb somewhere.

Please advise. Thanks!

Revision history for this message
Alan Johnson (nilgiri) wrote :

I see that Evolution 2.24 is in Intrepid, but I really want to stick with the LTS. Is there any hope that 2.24 will be released into Hardy main, ever? Or even backports? I'll upgrade if I have to, but LTS is supposed to include massive bug fix releases like this one, isn't it? No big deal either way. An upgrade over xmas break will give me an excuse to hide from the family ;-) , but if some one could please relay the directly of this effort, I would greatly appreciate it. I just don't want to upgrade to Intrepid only to have 2.24 make it in to Hardy after all.

Thanks!

Revision history for this message
Martin Pitt (pitti) wrote : Re: [Bug 247480] Re: evolution crashed with SIGSEGV in e_cal_menu_target_new_select()

Alan Johnson [2008-12-05 15:02 -0000]:
> I see that Evolution 2.24 is in Intrepid, but I really want to stick
> with the LTS. Is there any hope that 2.24 will be released into Hardy
> main, ever?

No, not hardy proper. hardy-backports would be considerable, though.

Revision history for this message
Sebastien Bacher (seb128) wrote :

you should not try to upgrade to 2.24 if you want a stable version, they rewrote the storage code to use sqlite this cycle and the new code still has quite some issues

Changed in evolution:
importance: Unknown → Critical
Revision history for this message
Rolf Leggewie (r0lf) wrote :

Hardy has seen the end of its life and is no longer receiving any updates. Marking the Hardy task for this ticket as "Won't Fix".

Changed in evolution (Ubuntu Hardy):
status: Confirmed → Won't Fix
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.