Bug 674574 – playbin about-to-finish signal Segfault with pygi

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 674574 - playbin about-to-finish signal Segfault with pygi


Summary:	playbin about-to-finish signal Segfault with pygi


Status:	RESOLVED OBSOLETE

Product:	GStreamer
Classification:	Platform
Component:	gst-python
Version:	0.10.x
Hardware:	Other Linux

Importance:	High critical
Target Milestone:	git master
Assigned To:	GStreamer Maintainers
QA Contact:	GStreamer Maintainers

URL:
Whiteboard:

Depends on:
Blocks:

Reported:	2012-04-22 17:08 UTC by Jonas H.
Modified:	2012-12-17 11:43 UTC

See Also:
GNOME target:	---
GNOME version:	---

Attachments
The sound used to reproduce the bug (8.78 KB, video/ogg) 2012-04-22 17:08 UTC, Jonas H.	Details

Description Jonas H. 2012-04-22 17:08:32 UTC

Created attachment 212550 [details]
The sound used to reproduce the bug

Running this code http://paste.pocoo.org/show/585630/ results in this crash http://paste.pocoo.org/show/585628/ after the sound has been played.

python-gst 0.10.22

Comment 1 André Klapper 2012-04-22 19:51:53 UTC

Thanks!Thanks for taking the time to report this.
Unfortunately, that stacktrace misses debug symbols for glib2 and gstreamer. Please install them first and paste an updated stacktrace as a comment in this bug report. See http://live.gnome.org/GettingTraces for more information on how to do so. Thanks in advance!

Comment 2 Jonas H. 2012-04-22 20:41:55 UTC

I can do that but maybe you can reproduce the crash yourself so I don't have to go through all the compilation hassle?

Comment 3 André Klapper 2012-04-23 07:49:23 UTC

Here you go on Fedora16:


Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0xb70ffb40 (LWP 3301)]
0x44d05189 in PyObject_Call (func=<function at remote 0xb7c4dfb4>, arg=(<__main__.GstPlayBin2 at remote 0xb7c61d24>,), kw=0x0)
    at /usr/src/debug/Python-2.7.2/Objects/abstract.c:2527
2527	        if (Py_EnterRecursiveCall(" while calling a Python object"))
(gdb) thread apply all bt

+ Trace 230121

Thread 3 (Thread 0xb70ffb40 (LWP 3301))

#0 PyObject_Call
#1 PyEval_CallObjectWithKeywords
#2 PyObject_CallObject
#3 pyg_closure_marshal
at pygtype.c line 1217
#4 g_closure_invoke
at gclosure.c line 774
#5 signal_emit_unlocked_R
at gsignal.c line 3272
#6 g_signal_emit_valist
at gsignal.c line 3003
#7 g_signal_emit
at gsignal.c line 3060
#8 drained_cb
at gstplaybin2.c line 2895
#9 g_cclosure_marshal_VOID__VOID
at gmarshal.c line 85
#10 g_closure_invoke
at gclosure.c line 774
#11 signal_emit_unlocked_R
at gsignal.c line 3272
#12 g_signal_emit_valist
at gsignal.c line 3003
#13 g_signal_emit
at gsignal.c line 3060
#14 proxy_drained_signal
at gsturidecodebin.c line 1518
#15 g_cclosure_marshal_VOID__VOID
at gmarshal.c line 85
#16 g_closure_invoke
at gclosure.c line 774
#17 signal_emit_unlocked_R
at gsignal.c line 3272
#18 g_signal_emit_valist
at gsignal.c line 3003
#19 g_signal_emit
at gsignal.c line 3060
#20 gst_decode_chain_handle_eos
at gstdecodebin2.c line 2927
#21 gst_decode_pad_handle_eos
at gstdecodebin2.c line 2868
#22 source_pad_event_probe
at gstdecodebin2.c line 3442
#23 source_pad_event_probe
at gstdecodebin2.c line 3432
#24 gst_marshal_BOOLEAN__POINTER
at gstmarshal.c line 586
#25 g_closure_invoke
at gclosure.c line 774
#26 signal_emit_unlocked_R
at gsignal.c line 3272
#27 g_signal_emitv
at gsignal.c line 2907
#28 gst_pad_emit_have_data_signal
at gstpad.c line 4143
#29 gst_pad_push_event
at gstpad.c line 5236
#30 gst_proxy_pad_do_event
at gstghostpad.c line 114
#31 gst_pad_send_event
at gstpad.c line 5398
#32 gst_pad_push_event
at gstpad.c line 5250
#33 vorbis_dec_sink_event
at gstvorbisdec.c line 544
#34 gst_pad_send_event
at gstpad.c line 5398
#35 gst_pad_push_event
at gstpad.c line 5250
#36 gst_single_queue_push_one
at gstmultiqueue.c line 1004
#37 gst_multi_queue_loop
at gstmultiqueue.c line 1179
#38 gst_task_func
at gsttask.c line 318
#39 default_func
at gsttaskpool.c line 70
#40 g_thread_pool_thread_proxy
at gthreadpool.c line 319
#41 g_thread_create_proxy
at gthread.c line 1962
#42 start_thread
at pthread_create.c line 309
#43 clone
at ../sysdeps/unix/sysv/linux/i386/clone.S line 133

Comment 4 Thiago Sousa Santos 2012-04-30 14:49:41 UTC

The paste site you used has shut down, could you attach the files here or use a different service?

Comment 5 Jonas H. 2012-04-30 16:18:41 UTC

André has already posted a much better traceback for the same bug.

Comment 6 Thiago Sousa Santos 2012-04-30 16:45:18 UTC

I was looking for the code that you also pasted there.

Comment 7 Jonas H. 2012-04-30 16:57:20 UTC

Right, here you go


from gi.repository import Gst

Gst.init(None)

def on_about_to_finish(playbin):
    pass

playbin = Gst.ElementFactory.make('playbin2', 'playbin')
playbin.set_property('uri', 'file:///jonas/stuff/noise.ogg')
playbin.set_state(Gst.State.PLAYING)
playbin.connect('about-to-finish', on_about_to_finish)
# ausprobieren mit audio-sinks und dann bugreport filen

import time
while 1:
    time.sleep(.1)

Comment 8 Tim-Philipp Müller 2012-09-23 18:33:59 UTC

> from gi.repository import Gst

Not a gst-python bug then. Just to be sure, py-gi will never work right with GStreamer 0.10.x.

The crash is also reproducable with GStreamer 0.11/1.0 though. Don't see anything wrong with our signal registration, so I would suspect a pygobject bug.

Comment 9 Sebastian Dröge (slomo) 2012-12-17 11:43:00 UTC

Closing this bug now, gst-python is only an extension module to pygi now and this bug doesn't make much sense anymore in this context.