Description of problem: I ran an update yesterday (yum update) after about 1 month of not updating, started fresh today (new boot, not from hibernate), logged in into X and immediately saw abrtd messages about caja crashing. The crashes continued repeatedly in a loop. Version-Release number of selected component: caja-1.8.1-5.fc20 Additional info: reporter: libreport-2.2.3 backtrace_rating: 4 cmdline: caja --sm-client-id 10de6699d426b79e7d140032711117028200000016820022 --sm-client-state-file /home/jin/.config/session-state/caja-1412366238.desktop crash_function: _g_log_abort executable: /usr/bin/caja kernel: 3.16.3-200.fc20.x86_64 runlevel: N 5 type: CCpp uid: 1000 Truncated backtrace: Thread no. 1 (10 frames) #2 _g_log_abort at gmessages.c:255 #5 caja_file_new_from_filename at caja-file.c:505 #6 caja_file_get_internal at caja-file.c:696 #7 caja_file_get_by_uri at caja-file.c:743 #8 caja_directory_get_corresponding_file at caja-directory.c:463 #9 begin_location_change at caja-window-manage-views.c:990 #10 caja_window_slot_open_location_full at caja-window-manage-views.c:698 #11 caja_window_go_to at caja-window.c:241 #12 caja_application_load_session at caja-application.c:2141 #13 caja_application_startup at caja-application.c:919
Created attachment 943829 [details] File: backtrace
Created attachment 943830 [details] File: cgroup
Created attachment 943831 [details] File: core_backtrace
Created attachment 943832 [details] File: dso_list
Created attachment 943833 [details] File: environ
Created attachment 943834 [details] File: limits
Created attachment 943835 [details] File: maps
Created attachment 943836 [details] File: open_fds
Created attachment 943837 [details] File: proc_pid_status
Created attachment 943838 [details] File: var_log_messages
Originally I was not able to submit the trace because the abrt tool was complaining about keyring unlocking and no password that I entered worked. I then tried to start caja from command line, to see how it crashes: killall caja && caja --sm-client-id 10de6699d426b79e7d140032711117028200000016820022 --sm-client-state-file /home/jin/.config/session-state/caja-1412366238.desktop ** ERROR:caja-file.c:505:caja_file_new_from_filename: assertion failed: (filename[0] != '\0') Aborted (core dumped) Afterwards I looked at the .desktop file, it had the following [Caja] section in the end: [Caja] documents=<?xml version="1.0" encoding="UTF-8"?>\n<session>\n <history>\n <bookmark name="Unnamed Device" icon="folder" uri="mtp://[usb:002,061]/"/>\n <bookmark name="tmp" icon="inode-directory:inode-x-generic:folder" uri="file:///tmp"/>\n </history>\n <window type="spatial" location="">\n <slot location="" active="TRUE"/>\n </window>\n <window type="spatial" location="">\n <slot location="" active="TRUE"/>\n </window>\n</session>\n I then deleted this section from the .desktop file and the problem disappeared, only then I was able to submit the backtrace/create this bug via the abrt application. Hope that helps, let me know if you need more info.
could you please post a list of updated packages? Run 'yum history' and 'yum history info <id>'. And does caja work now after the update or do you get permanent crashs and abrt alarms?
These are the two updates that I did, one yesterday and one today. Loaded plugins: langpacks ID | Command line | Date and time | Action(s) | Altered ------------------------------------------------------------------------------- 445 | update | 2014-10-04 09:50 | Update | 54 444 | update | 2014-10-03 20:04 | E, I, O, U | 304 EE 443 | install vlc-plugin-jack | 2014-09-13 01:07 | Install | 1 442 | update | 2014-08-28 23:48 | Update | 43 .... The problem started after update id 444, update 445 did not have any influence (problem persisted). I attaching the update info logs. Caja stopped working (permanent crashes) after I rebooted after yesterdays update. However, it's possible that the problem was triggered by me rebooting (and thus terminating the X session) and not by the update itself, before that I was only hibernating for more than a month and did not log out of X. Caja started to work again after I manually edited the .desktop file as mentioned in my first post, only then the crashes stopped, so caja working again is not related to the update.
Created attachment 943839 [details] yum update id 444
Created attachment 943840 [details] yum update id 445
In the update log i found not abnormals. But i don't have such a file /home/jin/.config/session-state/caja-1412366238.desktop because i never use the save session funktion. I would try to delete or save the file to get a clear configuration. This should be save to do because than the seesion starts without a caja window in session-state. Could it be possible that this uri doesn't exists anymore? uri="mtp://[usb:002,061]/
I am pretty sure that this uri does not exist anymore - it was a connected smartphone at some point which is not connected anymore. The problem is that caja seems to crash when it attempts to open uris that are not longer valid, ideally it should purge the invalid entries from the .desktop file or at least ignore them. As I mentioned earlier, as soon as I removed that part of the configuration from the .desktop file, caja stopped crashing. I can't tell you why this configuration remained in the .desktop file after the mtp device was disconnected, must be a different bug/problem elsewhere. Aand yes, I do use the "save session" functionality.
Yes, i agree that in a ideally world we only use Mate as desktop for linux and we have tons of develover who fixes issues ;) But reality is that only me maintain Mate in fedora, and mate upstream is a small team, in result i'm doubt that such a issue is on top their list. Don't forget that those devs comes from debian and issues from fedora aren't reproducable for them because of other versions od base packages, and lot of bugs in fedora are happend in the background but abrt will record them. Debian doesn't have such a tool so they will never get noticed by them. I hope you agree with me that this issue is easy to fix by a user (me), and we can close this report. If you think this is important enough please open a report directly by upstream. https://github.com/mate-desktop/caja/issues Ps: if the issue ocours more than acceptable i will re-open the report for shure. Thank you