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 409961 - rhythmbox crash to gst_mpeg_demux_send_subbuffer()
rhythmbox crash to gst_mpeg_demux_send_subbuffer()
Status: RESOLVED INCOMPLETE
Product: GStreamer
Classification: Platform
Component: gst-plugins-ugly
0.10.4
Other Linux
: Normal critical
: NONE
Assigned To: GStreamer Maintainers
GStreamer Maintainers
Depends on:
Blocks:
 
 
Reported: 2007-02-20 10:30 UTC by Sebastien Bacher
Modified: 2007-03-29 17:00 UTC
See Also:
GNOME target: ---
GNOME version: 2.15/2.16



Description Sebastien Bacher 2007-02-20 10:30:35 UTC
That bug has described on https://launchpad.net/bugs/86157

"Binary package hint: rhythmbox

I told music player to import my ~50 gig music collection.* At about 15 gigs, I got a pop up message that The music player had encountered an error report and that I could submit a bug report if I wanted. I could still see the music player status bar counting up the gigs,so I just left it sit there. after music player was done, I told it to go ahead and submit the crash report, intending to give it this message.
...
http://librarian.launchpad.net/6478257/_usr_lib_rhythmbox_rhythmbox-metadata.1000.crash
the bigg version of the bug report

It's ubuntu 6.10, fully up to date. Here's the big crash file attached. The reason it wasn't attached to begin with is because of an apparent bug in the crash reporter."


Debug backtrace for the crash:

Using host libthread_db library "/lib/tls/i686/cmov/libthread_db.so.1".
Core was generated by `/usr/lib/rhythmbox/rhythmbox-metadata unix:tmpdir=/tmp'.
Program terminated with signal 11, Segmentation fault.
  • #0 gst_mpeg_demux_send_subbuffer
    at gstmpegdemux.c line 925
  • #0 gst_mpeg_demux_send_subbuffer
    at gstmpegdemux.c line 925
  • #1 gst_dvd_demux_send_subbuffer
    at gstdvddemux.c line 948
  • #2 gst_mpeg_demux_parse_pes
    at gstmpegdemux.c line 875
  • #3 gst_mpeg_parse_chain
    at gstmpegparse.c line 737
  • #4 gst_pad_chain_unchecked
    at gstpad.c line 3344
  • #5 gst_pad_push
    at gstpad.c line 3510
  • #6 gst_id3demux_get_type
    from /usr/lib/gstreamer-0.10/libgstid3demux.so
  • #7 gst_pad_chain_unchecked
    at gstpad.c line 3344
  • #8 gst_pad_push
    at gstpad.c line 3510
  • #9 gst_type_find_element_chain
    at gsttypefindelement.c line 570
  • #10 gst_pad_chain_unchecked
    at gstpad.c line 3344
  • #11 gst_pad_push
    at gstpad.c line 3510
  • #12 gst_proxy_pad_do_chain
    at gstghostpad.c line 188
  • #13 gst_pad_chain_unchecked
    at gstpad.c line 3344
  • #14 gst_pad_push
    at gstpad.c line 3510
  • #15 gst_base_src_loop
    at gstbasesrc.c line 1529
  • #16 gst_task_func
    at gsttask.c line 192
  • #17 g_thread_pool_thread_proxy
    at gthreadpool.c line 265
  • #18 g_thread_create_proxy
    at gthread.c line 553
  • #19 start_thread
    from /lib/tls/i686/cmov/libpthread.so.0

Comment 1 Tim-Philipp Müller 2007-02-20 11:51:44 UTC
Do you have the file that causes the crash available somewhere?

(I also think this has been fixed in CVS, but I might be wrong).
Comment 2 Nathan Friedly 2007-02-20 23:38:20 UTC
(In reply to comment #1)
> Do you have the file that causes the crash available somewhere?
> 
> (I also think this has been fixed in CVS, but I might be wrong).
> 
I just posted this on the other bug site, https://launchpad.net/ubuntu/+source/gst-plugins-ugly0.10/+bug/86157 but heres a copy for this one:


Ok, I reimported my collection today (I did not first delete it though) there were 221 errors:

I got the error "The MIME type of the file could not be identified" for various files including thumbs.db, some .doc files, some .nfo files, an itunes 4 music library.itl file, and a few .mp3 files (which played just fine in xmms)

I got some "The GStreamer plugin to decode "MP3" files cannot be found" double clicking that particular file causes the system to tell me that it's a web page. it's 28.4 megs. I renamed it to .html, however, and opened it and sure enough, it show a web page with a bunch of links at the top then a _whole_ lot of jiberish - did the mp3 get appended to a web page maybe..?

another one with that error turned out to be a password protected .rar file with a .mp3 extension.

there was a number of "The GStreamer plugin to decode "DOS/Windows executable" files cannot be found" errors on cd's that tried to install shit on your computer

Another one, this is maybe the one you're looking for, got this error: "Problem occured without error being set. This is a bug in Rhythombox or GStreamer." That file was in the same folder as the html file and played just fine in XMMS. I also got this on an .nfo file which opened in gedit when I double clicked it.

I got an "The GStreamer plugin to decode "unknown" files cannot be found" error for a .rmmj file. double clicking on that one brings up movie player with an error message that I don't have the right codec to play it. (real player maybe..?)

I got one "The file is encrypted and cannot be played" on a wma file. fscking drm.

I think that was about it. I can attach any of these files or ftp them to my site, just let me know what you're looking for.


Comment 3 Tim-Philipp Müller 2007-02-21 00:09:34 UTC
> Ok, I reimported my collection today (I did not first delete it though) there
> were 221 errors:

Thanks for doing this!

> Another one, this is maybe the one you're looking for, got this error: 
> "Problem occured without error being set. This is a bug in Rhythombox
> or GStreamer." > That file was in the same folder as the html file and
> played just fine in XMMS.
> ...
> I think that was about it. I can attach any of these files or ftp them to my
> site, just let me know what you're looking for.

I'm looking for a file that crashes ("Segmentation Fault") when you do

 $ gst-launch-0.10 filesrc location=/path/to/file.foo ! decodebin ! fakesink

:)

(as opposed to erroring out in some way or silently decoding to the end)
Comment 4 Nathan Friedly 2007-02-22 13:17:35 UTC
ok, its going to take me a few (more) days to get to this.
Comment 5 Tim-Philipp Müller 2007-03-02 10:44:28 UTC
Any updates yet?
Comment 6 Nathan Friedly 2007-03-02 14:39:04 UTC
Sorry, finals.. 

I did spend a couple hours the other day but I didn't find the file yet. 
 
Think you could do me a script that loopps through every file in /media/500g/music/ (including sub directories) and saves the output? 

I tried and it went for about 2 hours but it didn't get all of the files (like the one with the errors)
Comment 7 Tim-Philipp Müller 2007-03-08 10:51:14 UTC
> I did spend a couple hours the other day but I didn't find the file yet. 
> 
> Think you could do me a script that loopps through every file in
> /media/500g/music/ (including sub directories) and saves the output? 

The media test suite in GStreamer CVS should be able to do that for you. You can check it out with:

  cvs -d:pserver:anoncvs@anoncvs.freedesktop.org:/cvs/gstreamer co gst-media-test

Comment 8 Nathan Friedly 2007-03-29 00:46:52 UTC
I don't know.. I've been working on this for a good number of hours and I'm not finding what file's causing the segmentation fault. 

I'm about to nuke this system and do a fresh install of ubuntu, so if I come up with any useful information I'll email it to you. 

For now I'm going to have to say forget about this bug. If I'm the only one that's getting it then it's probably a fluke specific to my collection.

Thanks and God Bless you,
-Nathan
Comment 9 Tim-Philipp Müller 2007-03-29 17:00:43 UTC
> I don't know.. I've been working on this for a good number of hours and I'm not
> finding what file's causing the segmentation fault. 
> 
> I'm about to nuke this system and do a fresh install of ubuntu, so if I come up
> with any useful information I'll email it to you. 
> 
> For now I'm going to have to say forget about this bug. If I'm the only one
> that's getting it then it's probably a fluke specific to my collection.

Thanks for spending all this time trying to find the problematic file. Will close as INCOMPLETE for now then. Please re-open or file a new bug if you ever come across this again, thanks!