After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 711385 - Crash when recoding video file
Crash when recoding video file
Status: RESOLVED OBSOLETE
Product: GStreamer
Classification: Platform
Component: dont know
1.2.1
Other Linux
: Normal normal
: NONE
Assigned To: GStreamer Maintainers
GStreamer Maintainers
Depends on:
Blocks:
 
 
Reported: 2013-11-04 10:19 UTC by Jan ONDREJ (SAL)
Modified: 2015-06-14 19:45 UTC
See Also:
GNOME target: ---
GNOME version: 3.7/3.8

Attachments
thread all backtrace (22.10 KB, text/plain)
2013-11-04 10:37 UTC, Jan ONDREJ (SAL) 		Details
thread apply all bt with more debuginfo (24.34 KB, text/plain)
2013-11-12 11:57 UTC, Jan ONDREJ (SAL) 		Details

Description Jan ONDREJ (SAL) 2013-11-04 10:19:49 UTC 
gst-launch-1.0 filesrc location=20131104-091054-test.mpg '!' decodebin name=demux '{' matroskamux name=mux '!' filesink location=20131104-091054-test.mkv '}' '{' demux. '!' queue '!' audioconvert '!' avenc_aac bitrate=96000 '!' queue '!' mux. '}' '{' demux. '!' deinterlace '!' queue '!' x264enc bitrate=1200 '!' mux. '}'
Setting pipeline to PAUSED ...
Pipeline is PREROLLING ...
Missing element: Teletext decoder
Redistribute latency...
Caught SIGSEGV
Redistribute latency...

+ Trace 232706

  • #0 poll
    from /lib64/libc.so.6
  • #1 g_main_context_iterate.isra.22
  • #2 g_main_loop_run
    from /lib64/libglib-2.0.so.0
  • #3 gst_bus_poll
    from /lib64/libgstreamer-1.0.so.0
  • #4 event_loop
  • #5 main
  • #0 poll
    at ../sysdeps/unix/syscall-template.S line 81
  • #1 g_main_context_poll
    at gmain.c line 3995
  • #2 g_main_context_iterate
    at gmain.c line 3696
  • #3 g_main_loop_run
    at gmain.c line 3895
  • #4 gst_bus_poll
    at gstbus.c line 1082
  • #5 event_loop
    at gst-launch.c line 507
  • #6 main
    at gst-launch.c line 1042 

I can provide this 67 MB test file, if required, or may be can create any other file.
Comment 1 Sebastian Dröge (slomo) 2013-11-04 10:23:51 UTC 
Can you provide a backtrace of all threads? You can get that with
> thread apply all backtrace

Which version of GStreamer are you using? 1.2.0?
Comment 2 Jan ONDREJ (SAL) 2013-11-04 10:37:24 UTC 
Created attachment 258908 [details]
thread all backtrace

Attaching backtrace for all threads. Sure, it's 1.2.0 (filled in bugzilla request). I use 1.2.0 compiled from Fedora rawhide to Fedora 19.
Comment 3 Jan ONDREJ (SAL) 2013-11-04 12:28:20 UTC 
Switching avenc_aac to lamemp3enc helped me to workaround this problem.
Comment 4 Sebastian Dröge (slomo) 2013-11-11 12:36:57 UTC 
Could you get debug symbols for gstreamer-libav and the other GStreamer modules and get a new backtrace? The important information is missing :)

Also note that avenc_aac is marked as experimental encoder by libav, and does not work at all in gstreamer-libav 1.2.x. Not sure how you got it working like that, it fails to initialize unless you pass specific flags to libav, which gstreamer-libav only does in git master. Use vo-aacenc or faac instead.
Comment 5 Jan ONDREJ (SAL) 2013-11-12 11:57:16 UTC 
Created attachment 259652 [details]
thread apply all bt with more debuginfo

OK, attached new backtrace.
Comment 6 Tim-Philipp Müller 2015-06-14 16:40:38 UTC 
I know it's been a while, but is this still a problem with 1.4 or 1.5/git master?
Comment 7 Jan ONDREJ (SAL) 2015-06-14 19:30:46 UTC 
Looks like syntax a bit changed. After removal of {}, on:

1.4.5 - don't crash, but does nothing, just waits, no error, output file is zero length, no CPU usage

1.5.1 - works well, output is generated

git/master - I don't know.

I think you can close this as solved in 1.5.1.
Comment 8 Tim-Philipp Müller 2015-06-14 19:45:51 UTC 
Good to hear, thanks for re-testing.

(Closing as OBSOLETE since I have no idea what actually fixed it where).