tracker-extract crashed with SIGABRT in g_malloc0()

Bug #1364780 reported by Karthikeyan Sivaraj
270
This bug affects 44 people
Affects Status Importance Assigned to Milestone
tracker (Ubuntu)
Confirmed
Medium
Unassigned

Bug Description

Logged into ubuntu-gnome guest vm installed in virtualbox. Encountered this crash.

ProblemType: Crash
DistroRelease: Ubuntu 14.10
Package: tracker-extract 1.0.4-0ubuntu1
ProcVersionSignature: Ubuntu 3.16.0-12.18-generic 3.16.1
Uname: Linux 3.16.0-12-generic x86_64
ApportVersion: 2.14.7-0ubuntu1
Architecture: amd64
CurrentDesktop: GNOME
Date: Tue Sep 2 23:32:05 2014
ExecutablePath: /usr/lib/tracker/tracker-extract
InstallationDate: Installed on 2014-08-10 (23 days ago)
InstallationMedia: Ubuntu-GNOME 14.10 "Utopic Unicorn" - Alpha amd64 (20140730)
ProcCmdline: /usr/lib/tracker/tracker-extract
ProcEnviron:
 PATH=(custom, no user)
 XDG_RUNTIME_DIR=<set>
 LANG=en_US.UTF-8
 SHELL=/bin/bash
Signal: 6
SourcePackage: tracker
StacktraceTop:
 g_malloc0 () from /lib/x86_64-linux-gnu/libglib-2.0.so.0
 g_slice_free1 () from /lib/x86_64-linux-gnu/libglib-2.0.so.0
 g_queue_pop_tail () from /lib/x86_64-linux-gnu/libglib-2.0.so.0
 ?? () from /lib/x86_64-linux-gnu/libglib-2.0.so.0
 ?? () from /lib/x86_64-linux-gnu/libglib-2.0.so.0
Title: tracker-extract crashed with SIGABRT in g_malloc0()
UpgradeStatus: No upgrade log present (probably fresh install)
UserGroups: adm cdrom dip lpadmin plugdev sambashare sudo

Revision history for this message
Karthikeyan Sivaraj (karthikeyan-sivaraj) wrote :
Revision history for this message
Apport retracing service (apport) wrote :

StacktraceTop:
 g_malloc0 (n_bytes=2032) at /build/buildd/glib2.0-2.41.3/./glib/gmem.c:132
 thread_memory_from_self () at /build/buildd/glib2.0-2.41.3/./glib/gslice.c:519
 g_slice_free1 (mem_size=<optimized out>, mem_block=0x23986c0) at /build/buildd/glib2.0-2.41.3/./glib/gslice.c:1088
 g_list_free_1 (list=<optimized out>) at /build/buildd/glib2.0-2.41.3/./glib/glist.c:200
 g_queue_pop_tail (queue=queue@entry=0x239d018) at /build/buildd/glib2.0-2.41.3/./glib/gqueue.c:625

Revision history for this message
Apport retracing service (apport) wrote : Stacktrace.txt
Revision history for this message
Apport retracing service (apport) wrote : StacktraceSource.txt
Revision history for this message
Apport retracing service (apport) wrote : ThreadStacktrace.txt
Changed in tracker (Ubuntu):
importance: Undecided → Medium
tags: removed: need-amd64-retrace
Revision history for this message
Launchpad Janitor (janitor) wrote :

Status changed to 'Confirmed' because the bug affects multiple users.

Changed in tracker (Ubuntu):
status: New → Confirmed
information type: Private → Public
Revision history for this message
Martyn Russell (martyn-lanedo) wrote :

Hi all,

I am one of the upstream maintainers. I think I've seen this bug recently but need a way to reliably reproduce it to fix it.
Does anyone have a way to do that?

Normally, if you know the file in question being extracted, you should be able to run this on the command line to reproduce the crash:

 $ /usr/libexec/tracker-extract -v 3 -f /path/to/file.mp3

Once you have that crash, a file uploaded here to test with would help hugely.

Thanks,

Revision history for this message
Martyn Russell (martyn-lanedo) wrote :

Just to add, my initial suspicion is either:

a) a rogue extractor
b) the kernel could be killing us due to too much memory use, based on our call to setrlimit() which we mainly only use in tracker-extract. The limit = CLAMP (total_halfed, MEM_LIMIT_MIN, G_MAXLONG); where MEM_LIMIT_MIN = 256 Mb. Bare in mind, the memory use in the logs above is more than 256MB.

Revision history for this message
Karthikeyan Sivaraj (karthikeyan-sivaraj) wrote :

Crash occurs sometimes immediately after login. Captured full stack-trace for a recent occurrence.

tags: added: vivid
tags: added: bugpattern-needed
Revision history for this message
Ian W Scott (iscott) wrote :

In answer to @martyn-lanedo: I'm not sure how to reproduce this reliably, except that it happens for me in *every* ubuntu-gnome session (15.04) within 15 minutes or so of login. But it seems to be triggered by a background tracker action, not by an attempt to extract a particular file directly. If you can suggest a way of identifying which file was being extracted at the time of the crash, I can try to identify it.

Revision history for this message
Martyn Russell (martyn-lanedo) wrote :

Hi Ian,

It may well be a rogue file being indexed.
We've just fixed a bug which was allocating way too much memory here:
  https://bugzilla.gnome.org/show_bug.cgi?id=748608

This started off as a Tracker bug in tracker-extract too. Turned out to be a library we depend on.

The bug also mentions a link which I've added recently to help people investigate these issues, here is the link again:
  https://wiki.gnome.org/Projects/Tracker/Documentation/Debugging#High_memory_use.3F

Are you able to share the output of:

  $ tracker-extract -v 3 -f /path/to/file

? That might shed some light on this issue.
Alternatively, if you could attach the file in question for testing here, that could help too.

Thanks :)

Revision history for this message
scott mowerson (smowerson) wrote :

i am getting this immediately after gnome signon 16.10 UBUNTU.

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.