GNOME Bugzilla – Bug 739997
Crash when selecting contact
Last modified: 2018-05-28 11:28:41 UTC
gnome-contacts starts correctly and lists all the contacts. Upon clicking on one of them the application crashes with this backtrace: ]2;gdb /usr/local/bin/gnome-contacts]1;gdbGNU gdb 6.3 Copyright 2004 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 "amd64-unknown-openbsd5.6"... (gdb) run Starting program: /usr/local/bin/gnome-contacts Program received signal SIGSEGV, Segmentation fault. 0x00000b3a1aa01cc3 in _folks_potential_match_contains (self=0xb3990adb910, haystack=0xb39adbdede0, haystack_length1=7, c=3755991007, pos=136, max_dist=15) at potential-match.c:2233 2233 _tmp3_ = _tmp1_[_tmp2_]; (gdb) bt
+ Trace 234324
(gdb) Is there any further information I could provide to fix this issue which currently renders it unusable?
Yes, you can provide some info about your folks library, evolution-data-server module, mostly versions and architecture I guess. Now I wonder, why do you launch your gnome-contacts instance from '/usr/local/bin'
The libraries: folks-0.10.0 evolution-data-server-3.12.7.1 architecture: OpenBSD/amd64 5.6 I launch it from /usr/local/bin because that's where it's installed. It's no different than from running it from the gnome-shell application grid.
Hi. Any news about this issue? I keep running into it and that makes gnome-contacts totally unusable :-( Thanks.
(In reply to comment #3) > Hi. Any news about this issue? I keep running into it and that makes > gnome-contacts totally unusable :-( > Thanks. It would be useful having some data about the contact. Use folks-inspect to gather the information about the contact causing the crash. Also I would like to know if this happens every time time, or if this is a matter of a corrupt environment. Plain, does this happens on a fresh install?
> It would be useful having some data about the contact. Use folks-inspect to > gather the information about the contact causing the crash. Ok I will try to play with that tool and report back. Note that it happens with any contact (local or remote). > Also I would like to know if this happens every time time, or if this is a > matter of a corrupt environment. Plain, does this happens on a fresh install? Everytime. Fresh installation as well. It is 100% reproducible for me.
Hi Eric. So, hopefully that's the kind of info you are looking for: (the "Can't convert from type 'GeeSet' to 'gchararray'" message looks dubious) Persona 'eds:1399738798.8502.6@peck.home.lan:http\://www.google.com/m8/feeds/contacts/foo.bar%40gmail.com/full/3b6836450cbc95f2': iid 1399738798.8502.6@peck.home.lan:http://www.google.com/m8/feeds/contacts/foo.bar%40gmail.com/full/3b6836450cbc95f2 uid eds:1399738798.8502.6@peck.home.lan:http\://www.google.com/m8/feeds/contacts/foo.bar%40gmail.com/full/3b6836450cbc95f2 display-id 1399738798.8502.6@peck.home.lan:http://www.google.com/m8/feeds/contacts/foo.bar%40gmail.com/full/3b6836450cbc95f2 is-user FALSE store 0x4e7720c3b90: eds, 1399738798.8502.6@peck.home.lan (Contacts) individual 0x4e77b4df5a0 contact 0x4e770e30c50 web-service-addresses { } local-ids { '1399738798.8502.6@peck.home.lan:http://www.google.com/m8/feeds/contacts/foo.bar%40gmail.com/full/3b6836450cbc95f2' } location 0x0 postal-addresses { } phone-numbers { +33123456789 } email-addresses { } notes { } linkable-properties { 'im-addresses', 'email-addresses', 'local-ids', 'web-service-addresses' } writeable-properties { 'local-ids', 'web-service-addresses', 'gender', 'is-favourite', 'anti-links', 'full-name', 'email-addresses', 'postal-addresses', 'im-addresses', 'structured-name', 'p avatar (null) structured-name John contact-id http://www.google.com/m8/feeds/contacts/foo.bar%40gmail.com/full/3b6836450cbc95f2 full-name John nickname gender FOLKS_GENDER_UNSPECIFIED urls { } im-addresses { } groups { 'Imported 10/30/13', 'Personal' } calendar-event-id (null) birthday roles { } is-favourite FALSE anti-links { } system-groups Can't convert from type 'GeeSet' to 'gchararray' in-google-personal-group TRUE
Hi. It'd be really nice to be able to find the core of this crash before 3.16 is out. gnome-contacts is completely unusable on OpenBSD currently because of that. It really seems the bug is coming from folks, so maybe the Product field of this ticket should be changed.
Hey Antoine, are you still experiencing this issue? I haven’t been able to reproduce this in the latest version (neither stable or unstable).
Hi Niels. Unfortunately yes, the issue is still 100% reproducible for me (3.26).
Okay. I did some digging and I think I found the problem, which would be a bug in folks. I made bug 792185 over there with a patch that should hopefully fix this. Is there anyone here who could possibly test the patch over there and tell whether it fixed the problem for them?
Also possible by another bug in folks, for which I made bug 792196 (with a patch that should solve it).
Hi Niels. I applied both patches. They do fix a segfault (which I didn't reported before because I though it was the same issue as this ticket) but not the one described here :-/ gnome-contacts still crashes at : 0x0000184d2dddac34 in _folks_potential_match_jaro_dist () Anyway, thanks a lot for looking into this.
Hey Antoine! Thanks for applying the patches and giving the feedback. Would you mind posting a full backtrace of the crash? I'll be able to give this a better look then :-)
Hey Niels. There was one posted by Jasper when he opened the bug. Granted that was 3 years ago so I'll post a more recent one soonish :-)
Hey Antoine! Could you try the folks patch in bug 792238?
hey Niels. Just applied patch from bug 792238 but I'm still getting a segfault. Backtrace: Program received signal SIGSEGV, Segmentation fault. 0x00001c9f12dea2d1 in _folks_potential_match_matches (self=0x1c9efd746ba0, s1=0x1c9ed193cfc0, s1_length1=15, s2=0x1c9fc91d21e0, s2_length1=5, max_dist=6, t=0x7f7ffffece40) at potential-match.c:2168 2168 potential-match.c: No such file or directory. in potential-match.c Current language: auto; currently minimal (gdb) bt
+ Trace 238316
Ah, seems I forgot to check the length in the other string as well. Mind trying the second patch from bug 792238 as well?
Rock on! That finally fixed it :-) Thanks a lot Niels, that's awesome.
Great! I'll set it as FIXED as soon as the patch has been reviewed and committed in folks.
Okay, patch has landed. For those experiencing the same issue: try updating to folks 0.11.5 once it's released (which shouldn't be too long). Thanks everyone for the feedback!
Any chance you can release folks 0.11.5, please?