banshee weird behavior and random freezes

[arpunk]

Binary package hint: banshee

Banshee keeps freezing forcing me to kill the application, this is the output of one of the freezes (the other ones didnt output nothing related to the possible crash):

Warning: [5/3/2006 10:35:59 PM] (Cannot connect to NetworkManager) - An available, working network connection will be assumed
Debug: [5/3/2006 10:36:00 PM] (Default player engine) - GStreamer 0.10
Debug: [5/3/2006 10:36:00 PM] (Audio CD Core Initialised) -
Audioscrobbler starting protocol engine
Scanning library for tracks to update
Done scanning library
Processing track queue for pending queries
Setting MusicBrainz proxy to www.musicbrainz.org:80
Done processing track queue
*** glibc detected *** double free or corruption (!prev): 0x08b69270 ***

Another thing ive seen is that in some of the songs (depending on how it was encoded), the song duration keeps changing while the song is playing. Could this be a banshee or a gstreamer issue?

[Sebastian Dröge (slomo)]

The duration changing is known and upstream calls it a feature until a better solution is found.

For the double free... could you please run banshee with gdb?

gdb --args mono /usr/lib/banshee/banshee.exe
handle SIGXCPU SIG33 SIGPWR nostop noprint
run
[wait for the double free]
backtrace

then call "p mono_print_method_from_ip ($address)" on any address that isn't resolved and paste the complete output here.
For more informations about (gdb) debugging of mono applications look here:
http://www.mono-project.com/Debugging

[arpunk (ricardo-lanziano)]

Here is the ouput of gdb:

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread -1212876480 (LWP 26155)]
0xb7ce99de in g_hash_table_insert () from /usr/lib/libglib-2.0.so.0
(gdb) backtrace
#0 0xb7ce99de in g_hash_table_insert () from /usr/lib/libglib-2.0.so.0
#1 0xb6b8e405 in gdk_event_new () from /usr/lib/libgdk-x11-2.0.so.0
#2 0xb6c9f648 in gtk_container_propagate_expose () from /usr/lib/libgtk-x11-2.0.so.0
#3 0xb6c9f71e in gtk_container_propagate_expose () from /usr/lib/libgtk-x11-2.0.so.0
#4 0xb6c64fd8 in gtk_box_reorder_child () from /usr/lib/libgtk-x11-2.0.so.0
#5 0xb6c9dc56 in gtk_container_forall () from /usr/lib/libgtk-x11-2.0.so.0
#6 0xb6c9f492 in gtk_container_get_focus_hadjustment () from /usr/lib/libgtk-x11-2.0.so.0
#7 0xb6d27720 in _gtk_marshal_BOOLEAN__BOXED () from /usr/lib/libgtk-x11-2.0.so.0
#8 0xb6a0f16f in g_cclosure_new_swap () from /usr/lib/libgobject-2.0.so.0
#9 0xb6a0f79f in g_closure_invoke () from /usr/lib/libgobject-2.0.so.0
#10 0xb6a1e9ce in g_signal_stop_emission () from /usr/lib/libgobject-2.0.so.0
#11 0xb6a1f886 in g_signal_emit_valist () from /usr/lib/libgobject-2.0.so.0
#12 0xb6a1fe89 in g_signal_emit () from /usr/lib/libgobject-2.0.so.0
#13 0xb6e09b1f in gtk_widget_activate () from /usr/lib/libgtk-x11-2.0.so.0
#14 0xb6c9f6c7 in gtk_container_propagate_expose () from /usr/lib/libgtk-x11-2.0.so.0
#15 0xb6c9f71e in gtk_container_propagate_expose () from /usr/lib/libgtk-x11-2.0.so.0
#16 0xb6c618ab in gtk_bin_get_type () from /usr/lib/libgtk-x11-2.0.so.0
#17 0xb6c9dc56 in gtk_container_forall () from /usr/lib/libgtk-x11-2.0.so.0
#18 0xb6c9f492 in gtk_container_get_focus_hadjustment () from /usr/lib/libgtk-x11-2.0.so.0
#19 0xb6e1b37d in gtk_window_get_position () from /usr/lib/libgtk-x11-2.0.so.0
#20 0xb6d27720 in _gtk_marshal_BOOLEAN__BOXED () from /usr/lib/libgtk-x11-2.0.so.0
#21 0xb6a0f16f in g_cclosure_new_swap () from /usr/lib/libgobject-2.0.so.0
#22 0xb6a0f79f in g_closure_invoke () from /usr/lib/libgobject-2.0.so.0
#23 0xb6a1e9ce in g_signal_stop_emission () from /usr/lib/libgobject-2.0.so.0
#24 0xb6a1f886 in g_signal_emit_valist () from /usr/lib/libgobject-2.0.so.0
#25 0xb6a1fe89 in g_signal_emit () from /usr/lib/libgobject-2.0.so.0
#26 0xb6e09b1f in gtk_widget_activate () from /usr/lib/libgtk-x11-2.0.so.0
#27 0xb6d26466 in gtk_main_do_event () from /usr/lib/libgtk-x11-2.0.so.0
#28 0xb6ba2dd7 in gdk_window_clear_area_e () from /usr/lib/libgdk-x11-2.0.so.0
#29 0xb6ba2eab in gdk_window_process_all_updates () from /usr/lib/libgdk-x11-2.0.so.0
#30 0xb6c9d8d5 in gtk_container_check_resize () from /usr/lib/libgtk-x11-2.0.so.0
#31 0xb7cf6b92 in g_child_watch_add () from /usr/lib/libglib-2.0.so.0
#32 0xb7cf4876 in g_main_context_dispatch () from /usr/lib/libglib-2.0.so.0
#33 0xb7cf7936 in g_main_context_check () from /usr/lib/libglib-2.0.so.0
#34 0xb7cf7c58 in g_main_loop_run () from /usr/lib/libglib-2.0.so.0
#35 0xb6d255a5 in gtk_main () from /usr/lib/libgtk-x11-2.0.so.0
#36 0xb36222c1 in ?? ()
#37 0x00847c3c in ?? ()
#38 0xb3622280 in ?? ()
#39 0x080674c8 in ?? ()
#40 0x080674b0 in ?? ()
#41 0x088e8a90 in ?? ()
#42 0xb7f122ec in ?? () ...

[Sebastian Dröge (slomo)]

Do you still have this with the versions in dapper?

[Alessandro Ambrogi (alessandro-ambrogi-2001)]

I can confirm this bug on the Dapper final version. It's always reproducible, and make banshee to freeze with the same error after some minute.
The cause is the Metadata Searcher plugin... if I disable it, banshee became very stable. But it's a very important plugin... it's very sad have to disable it...

[Benjamin Woods (woodsb02-gmail)]

I can also confirm this bug in ubuntu dapper final.

I will try disabling metadata searcher plugin to stop it from happening, but as Alessandro said - i love that plugin! :(

[Sebastian Dröge (slomo)]

Is this problem still there with 0.11.0-0ubuntu1 in edgy?

[Benjamin Woods (woodsb02-gmail)]

I am running the latest edgy and banshee version 0.10.12-1ubuntu1.
Perhaps there is an update I don't have yet.

Either way - it's actually fixed in this version! (I imagine it is also fixed in 0.11 but cannot yet confirm this).

I do however still get an error at the command line with this version:
(Banshee:11155): GdkPixbuf-WARNING **: GdkPixbufLoader finalized without calling gdk_pixbuf_loader_close() - this is not allowed. You must explicitly end the data stream to the loader before dropping the last reference.

Also - my ipod doesnt seem to be readable in this version. But it works in rhythmbox.

Perhaps 0.11 will fix this... i might manually grab it from the repositories and give it a go.

I will report back on that version soon.

[Sebastian Dröge (slomo)]

Ok so this bug is fixed... for the ipod support and 0.11.0 just wait 20 minutes or something and it will be on the mirrors.

[Benjamin Woods (woodsb02-gmail)]

Does dapper use this 0.11.0 version?

If not, then its not fixed... just asking since i no longer use dapper.

[Sebastian Dröge (slomo)]

dapper does not and will never use 0.11.0. First of all it's completely frozen since the release except for major, not too intrusive bugfixes... and except backports. Unfortunately it can't be backported because of too old mono in dapper and backporting mono is not possible without backporting many other mono applications, etc...

[Benjamin Woods (woodsb02-gmail)]

Fair enough.

I have just tested banshee 0.11.0 using the edgy debs.

Below is my output as I used the program. I have noticed that it no longer writes to the console when it uses musicbrainz... but i know it was using it because some song metadata was automatically updated.

My ipod works perfectly now with it - so i am happy :)

All fixed for me!

bwoods@slick:~$ banshee
Debug: [09/20/2006 22:55:36] (Default player engine) - GStreamer 0.10
Debug: [09/20/2006 22:55:36] (Audio CD Core Initialized) -
Audioscrobbler starting protocol engine
Warning: [09/20/2006 22:55:37] (Could not initialize plugin `Daap') - Daemon not running
Setting MusicBrainz proxy to www.musicbrainz.org:80

(Banshee:10734): Gtk-CRITICAL **: gtk_tree_view_get_cell_area: assertion `GTK_WIDGET_REALIZED (tree_view)' failed

(Banshee:10734): Gtk-CRITICAL **: gtk_tree_view_get_cell_area: assertion `GTK_WIDGET_REALIZED (tree_view)' failed
Setting IO Backend to Banshee.IO.Unix.IOConfig (unix)
Audioscrobbler stopping protocol engine
bwoods@slick:~$

[Sebastian Dröge (slomo)]

So is there still something wrong with 0.11.0?

[Benjamin Woods (woodsb02-gmail)]

Well the output makes it appear so - but everything seems to work fine.

So I until I find a bug I wouldn't say anything is broken as such...

Banshee:10734): Gtk-CRITICAL **: gtk_tree_view_get_cell_area: assertion `GTK_WIDGET_REALIZED (tree_view)' failed

Whatever it means I don't notice any problems... :)

[Sebastian Dröge (slomo)]

Yes that's a known problem but fortunately doesn't break anything :)