Bug 1105663 - [abrt] mate-window-manager: __gconv_open(): marco killed by SIGSEGV
Summary: [abrt] mate-window-manager: __gconv_open(): marco killed by SIGSEGV
Keywords:
Status: CLOSED WONTFIX
Alias: None
Product: Fedora
Classification: Fedora
Component: mate-window-manager
Version: 20
Hardware: x86_64
OS: Unspecified
unspecified
unspecified
Target Milestone: ---
Assignee: Dan Mashal
QA Contact: Fedora Extras Quality Assurance
URL: https://retrace.fedoraproject.org/faf...
Whiteboard: abrt_hash:8147266aa386c221e2b125bbd14...
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2014-06-06 16:25 UTC by Davide Repetto
Modified: 2014-11-21 10:59 UTC (History)
4 users (show)

Fixed In Version:
Clone Of:
Environment:
Last Closed: 2014-11-21 10:59:46 UTC
Type: ---
Embargoed:


Attachments (Terms of Use)
File: backtrace (54.91 KB, text/plain)
2014-06-06 16:25 UTC, Davide Repetto
no flags Details
File: cgroup (172 bytes, text/plain)
2014-06-06 16:25 UTC, Davide Repetto
no flags Details
File: core_backtrace (17.00 KB, text/plain)
2014-06-06 16:25 UTC, Davide Repetto
no flags Details
File: dso_list (8.75 KB, text/plain)
2014-06-06 16:25 UTC, Davide Repetto
no flags Details
File: environ (2.49 KB, text/plain)
2014-06-06 16:25 UTC, Davide Repetto
no flags Details
File: exploitable (93 bytes, text/plain)
2014-06-06 16:25 UTC, Davide Repetto
no flags Details
File: limits (1.29 KB, text/plain)
2014-06-06 16:25 UTC, Davide Repetto
no flags Details
File: maps (44.04 KB, text/plain)
2014-06-06 16:25 UTC, Davide Repetto
no flags Details
File: open_fds (711 bytes, text/plain)
2014-06-06 16:25 UTC, Davide Repetto
no flags Details
File: proc_pid_status (962 bytes, text/plain)
2014-06-06 16:25 UTC, Davide Repetto
no flags Details
File: var_log_messages (344 bytes, text/plain)
2014-06-06 16:25 UTC, Davide Repetto
no flags Details

Description Davide Repetto 2014-06-06 16:25:29 UTC
Description of problem:
I was adding fonts to local .fonts folder

Version-Release number of selected component:
mate-window-manager-1.6.2-6.fc20

Additional info:
reporter:       libreport-2.2.2
backtrace_rating: 4
cmdline:        marco
crash_function: __gconv_open
executable:     /usr/bin/marco
kernel:         3.14.5-200.fc20.x86_64
runlevel:       N 5
type:           CCpp
uid:            1000

Truncated backtrace:
Thread no. 1 (10 frames)
 #3 __gconv_open at gconv_open.c:282
 #4 iconv_open at iconv_open.c:71
 #5 try_conversion at gconvert.c:199
 #6 g_iconv_open at gconvert.c:251
 #7 open_converter at gconvert.c:338
 #8 g_convert at gconvert.c:575
 #9 make_list at gdkselection-x11.c:537
 #10 gdk_text_property_to_utf8_list_for_display at gdkselection-x11.c:623
 #11 gdk_text_property_to_utf8_list at gdkselection.c:96
 #12 meta_text_property_to_utf8 at ui/ui.c:813

Comment 1 Davide Repetto 2014-06-06 16:25:34 UTC
Created attachment 902974 [details]
File: backtrace

Comment 2 Davide Repetto 2014-06-06 16:25:35 UTC
Created attachment 902975 [details]
File: cgroup

Comment 3 Davide Repetto 2014-06-06 16:25:37 UTC
Created attachment 902976 [details]
File: core_backtrace

Comment 4 Davide Repetto 2014-06-06 16:25:38 UTC
Created attachment 902977 [details]
File: dso_list

Comment 5 Davide Repetto 2014-06-06 16:25:39 UTC
Created attachment 902978 [details]
File: environ

Comment 6 Davide Repetto 2014-06-06 16:25:40 UTC
Created attachment 902979 [details]
File: exploitable

Comment 7 Davide Repetto 2014-06-06 16:25:42 UTC
Created attachment 902980 [details]
File: limits

Comment 8 Davide Repetto 2014-06-06 16:25:44 UTC
Created attachment 902981 [details]
File: maps

Comment 9 Davide Repetto 2014-06-06 16:25:45 UTC
Created attachment 902982 [details]
File: open_fds

Comment 10 Davide Repetto 2014-06-06 16:25:46 UTC
Created attachment 902983 [details]
File: proc_pid_status

Comment 11 Davide Repetto 2014-06-06 16:25:48 UTC
Created attachment 902984 [details]
File: var_log_messages

Comment 12 Wolfgang Ulbrich 2014-06-21 15:14:34 UTC
Does it happend everytime and marco is unusable now or was it only a one shot abrt alarm?
Did marco crashed or was it only a abrt alarm?

Comment 13 Davide Repetto 2014-06-21 21:54:47 UTC
It happened on more than one occasion for sure.
I was updating fonts with rsync on a number of workstations and while all of the first hundred or so experienced weird and long -but temporary- lockups of the display (I suppose it was marco "hanging" while waiting for the fonts subsystem), a few had this marco crash.
Since I had not planned any downtime and to avoid further problems, I opted to proceed upgrading fonts in a separate folder that would later be moved in place, once the sync was fully done. This way the remaining 600-something upgrades went smoothly.
We're talking about the refresh of few hundred megs of already-existing fonts.
The equivalent of the fonts normally shipped in windows7 or in office 2013 for the mac. Which where copied over older versions of themselves.

Marco was still usable after any of those issues. Even when it outright crashed.
In fact I believe the crash itself was due to the continuous "reloading or re-caching" of the fonts that happened every time a single font-file was touched by rsync during the "hot update" of the fonts.

Comment 14 Wolfgang Ulbrich 2014-06-21 23:20:05 UTC
forward to upstream
https://github.com/mate-desktop/marco/issues/111
Pls answer there if upstream have a question.

Comment 15 Davide Repetto 2014-06-22 10:05:57 UTC
will do. :)

Comment 16 Wolfgang Ulbrich 2014-11-21 10:59:46 UTC
closed, package are renamed to marco


Note You need to log in before you can comment on or make changes to this bug.