On Wed, Jul 21, 2010 at 2:32 PM, Jon Jackson <[email protected]>wrote:
> > > On Mon, Jul 19, 2010 at 12:29 PM, Josselin Mouette <[email protected]>wrote: > >> Le mercredi 14 juillet 2010 à 15:55 -0400, Jonathan Jackson a écrit : >> > On a Debian Lenny installation that has performed well for 6+ months, >> gnome sessions now hang during logon for all >> > system users. This is after the desktop background and panels appear, >> but before any desktop icons or panel >> > icons/menus are visible. KDE sessions are fine. >> > >> > the last entry in ~/.xsession errors is "Initializing gnome-mount >> extension" >> > >> > /var/log/messages indicates the following errors coinciding with the >> session hang: >> > >> > Jul 14 15:27:39 genova kernel: [ 661.024456] gnome-vfs-daemo[15293]: >> segfault at 0 ip 00007fc8584ceaf2 sp 00007fffc43cc3f8 error 4 in >> libc-2.7.so[7fc858454000+14a000] >> [etc.] >> >> Thanks for your report. >> >> Unfortunately we need a backtrace of the crash to understand what is >> happening. See http://wiki.debian.org/HowToGetABacktrace >> >> > Ok, so I've generated a trace following the instructions on the wiki page, > thanks. > > $ gdb /usr/lib/gnome-vfs-2.0/gnome-vfs-daemon > GNU gdb 6.8-debian > Copyright (C) 2008 Free Software Foundation, Inc. > License GPLv3+: GNU GPL version 3 or later < > http://gnu.org/licenses/gpl.html> > This is free software: you are free to change and redistribute it. > There is NO WARRANTY, to the extent permitted by law. Type "show copying" > and "show warranty" for details. > This GDB was configured as "x86_64-linux-gnu"... > (gdb) set pagination 0 > (gdb) run > Starting program: /usr/lib/gnome-vfs-2.0/gnome-vfs-daemon > [Thread debugging using libthread_db enabled] > [New Thread 0x7fa61029f6e0 (LWP 22354)] > > (process:22354): libgnomevfs-CRITICAL **: > gnome_vfs_get_uri_from_local_path: assertion `g_path_is_absolute > (local_full_path)' failed > > (process:22354): libgnomevfs-CRITICAL **: > gnome_vfs_get_uri_from_local_path: assertion `g_path_is_absolute > (local_full_path)' failed > > Program received signal SIGSEGV, Segmentation fault. > [Switching to Thread 0x7fa61029f6e0 (LWP 22354)] > 0x00007fa60ea8caf2 in strcmp () from /lib/libc.so.6 > (gdb) bt > #0 0x00007fa60ea8caf2 in strcmp () from /lib/libc.so.6 > #1 0x000000000040e530 in > _gnome_vfs_volume_monitor_find_mtab_volume_by_activation_uri > (volume_monitor=<value optimized out>, activation_uri=0x0) at > gnome-vfs-volume-monitor.c:673 > #2 0x000000000040cc2c in update_fstab_drives > (volume_monitor_daemon=0x19f3280) at gnome-vfs-volume-monitor-daemon.c:708 > #3 0x000000000040df2e in gnome_vfs_volume_monitor_daemon_init > (volume_monitor_daemon=0x19f3280) at gnome-vfs-volume-monitor-daemon.c:191 > #4 0x00007fa60f275fa5 in g_type_create_instance () from > /usr/lib/libgobject-2.0.so.0 > #5 0x00007fa60f25a7ab in ?? () from /usr/lib/libgobject-2.0.so.0 > #6 0x00007fa60f25b5d2 in g_object_newv () from > /usr/lib/libgobject-2.0.so.0 > #7 0x00007fa60f25c327 in g_object_new_valist () from > /usr/lib/libgobject-2.0.so.0 > #8 0x00007fa60f25c46c in g_object_new () from /usr/lib/libgobject-2.0.so.0 > #9 0x000000000040f071 in _gnome_vfs_get_volume_monitor_internal (create=1) > at gnome-vfs-volume-monitor.c:446 > #10 0x0000000000407b6e in main (argc=<value optimized out>, argv=<value > optimized out>) at vfs-daemon.c:945 > > > I've followed this up based on 'clue' given by (_gnome_vfs_volume_monitor_find_mtab_volume_by_activation_uri ) and found this has been also been reported in Ubuntu https://bugs.launchpad.net/ubuntu/+source/gnome-vfs2/+bug/85477 and upstream https://bugzilla.gnome.org/show_bug.cgi?id=408553 And it appears to be related to parsing /etc/fstab Looking at my /etc/fstab for anything unusual, I had an entry UUID=A86EF81A6EF7DECE windows ntfs noauto,user,umask=227 0 0 which is valid in the sense that I mounted the file system without problems cd /mnt sudo mount windows however the mount point is not a proper full path. Changing the mount point to /mnt/windows fixes the segfault in gnome-vfs-daemon and gnome sessions are working again. I'll post a comment also upstream Thanks for recommending a backtrace as this was key to finding a solution,
