deja-dup-monitor: Program terminated with signal 11, Segmentation fault.

Bug #495589 reported by Michal Nowak
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Déjà Dup
Incomplete
Critical
Unassigned

Bug Description

UUID: 7c3612b1d3b8f1bf54607601798c69baee1ffd94
architecture: i686
cmdline: deja-dup-monitor
component: deja-dup
description: /var/cache/abrt/ccpp-1260554391-1949/description
executable: /usr/bin/deja-dup-monitor
kernel: 2.6.31.6-166.fc12.i686
package: deja-dup-13.3-1.fc12
rating: 4
reason: Process was terminated by signal 11 (Segmentation fault)
release: Fedora release 12 (Constantine)

How to reproduce
-----
1. Installed and configured nothing
2. restart and log into GNOME
3. spotted the crash at the very beginning of GNOME loading

backtrace
-----
Core was generated by `deja-dup-monitor'.
Program terminated with signal 11, Segmentation fault.
#0 0x0804d3cd in deja_dup_backend_is_ready (self=0x0, when=0xbfeaf084)
    at Backend.c:220
220 return DEJA_DUP_BACKEND_GET_CLASS (self)->is_ready (self, when);

Thread 1 (Thread 1949):
#0 0x0804d3cd in deja_dup_backend_is_ready (self=0x0, when=0xbfeaf084)
    at Backend.c:220
No locals.
#1 0x0804cb47 in monitor_is_ready (when=<value optimized out>)
    at monitor.c:156
        _tmp3_ = <value optimized out>
        _tmp0_ = 0x0
        _tmp4_ = <value optimized out>
        _tmp2_ = 0x0
        result = <value optimized out>
        _inner_error_ = 0x0
#2 monitor_kickoff (when=<value optimized out>) at monitor.c:468
        result = <value optimized out>
        _inner_error_ = 0x0
        wait_time = -68391
        _tmp0_ = <value optimized out>
        when = <value optimized out>
        _tmp3_ = <value optimized out>
        _tmp2_ = -1075122036
        _tmp1_ = 0x0
        _tmp4_ = 0
#3 0x0804ce58 in monitor_prepare_next_run () at monitor.c:582
        wait_time = -68390
#4 0x0804d140 in monitor_main (args_length1=1, args=0xbfeaf1d4)
    at monitor.c:704
        _inner_error_ = 0x0
        context = 0x8410998
        status = 0
        _tmp1_ = <value optimized out>
        result = <value optimized out>
        _tmp2_ = 0x0
#5 main (args_length1=1, args=0xbfeaf1d4) at monitor.c:715
No locals.

Revision history for this message
Michael Terry (mterry) wrote :

You're a segfault-finding machine, thanks! Again, will look this weekend. I suspect they're both the same underlying cause.

Changed in deja-dup:
importance: Undecided → Critical
Revision history for this message
Michael Terry (mterry) wrote :

The crash aspect of this bug is fixed in trunk by providing a default value for the backend gconf value. It appears you didn't install the gconf schema?

Even with this fix, something doesn't add up. You say you did no configuration, but the monitor shouldn't have gotten to that section of code unless a certain gconf value (/apps/deja-dup/periodic) was set to true... Can I see your gconf values [1] just to make sure?

[1] gconftool-2 --dump /apps/deja-dup > /tmp/deja-dup.gconf

Changed in deja-dup:
status: New → Incomplete
Revision history for this message
Michal Nowak (newman) wrote :

Sending what files I have installed wrt deja-dup in Fedora. If there's something missing or wrong, note it to fedora Package Review too, please.

Will investigate that gconf value once I am back home, as I spotted it there.

Revision history for this message
Michal Nowak (newman) wrote :

newman ~ $ gconftool-2 --dump /apps/deja-dup
<gconfentryfile>
  <entrylist base="/apps/deja-dup">
    <entry>
      <key>delete-after</key>
      <value>
        <int>0</int>
      </value>
    </entry>
    <entry>
      <key>encrypt</key>
      <value>
        <bool>true</bool>
      </value>
    </entry>
    <entry>
      <key>periodic</key>
      <value>
        <bool>true</bool>
      </value>
    </entry>
    <entry>
      <key>periodic-period</key>
      <value>
        <int>1</int>
      </value>
    </entry>
  </entrylist>
</gconfentryfile>

Revision history for this message
Michael Terry (mterry) wrote :

I marked this as a dup of bug 495248 because it is the same underlying bug. It will be fixed in 13.4, out in a week. The remaining detail where I thought something else might be wrong (because I didn't expect to hit that block of code if periodic=false) was a red herring since periodic=true for you.

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.