Bug 645228 – segfault on "Move on Toolbar"

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 645228 - segfault on "Move on Toolbar"


Summary:	segfault on "Move on Toolbar"


Status:	RESOLVED FIXED

Product:	libegg
Classification:	Other
Component:	toolbar
Version:	unspecified
Hardware:	Other Linux

Importance:	High critical
Target Milestone:	---
Assigned To:	Soren Sandmann Pedersen
QA Contact:	Libegg maintenance

URL:
Whiteboard:

Depends on:
Blocks:

Reported:	2011-03-19 15:54 UTC by Steve
Modified:	2011-04-03 15:41 UTC

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

Attachments
stack trace (4.16 KB, application/octet-stream) 2011-04-02 17:46 UTC, Adam Dingle		Details
dnddevice.diff (930 bytes, patch) 2011-04-02 22:06 UTC, Xan Lopez	none	Details \| Review
dndfix.diff (931 bytes, patch) 2011-04-03 00:15 UTC, Xan Lopez	none	Details \| Review
dndfix.diff (1.85 KB, patch) 2011-04-03 15:37 UTC, Xan Lopez	none	Details \| Review

Description Steve 2011-03-19 15:54:06 UTC

Epiphany 2.91.91.1 built on Slackware-current.

Right-click on bookmark on Toolbar and select "Move on Toolbar" causes Epiphany to segfault.

Comment 1 Akhil Laddha 2011-03-20 01:18:32 UTC

Thanks for taking the time to report this bug.
Without a stack trace from the crash it's very hard to determine what caused it.
Can you get us a stack trace? Please see http://live.gnome.org/GettingTraces for more information on how to do so. Thanks in advance!

Comment 2 Adam Dingle 2011-04-02 17:46:07 UTC

I'm seeing this too.  I'm using Epiphany trunk on Fedora 15 alpha.  I'll attach a stack trace.

Comment 3 Adam Dingle 2011-04-02 17:46:30 UTC

Created attachment 184957 [details]
stack trace

Comment 4 Xan Lopez 2011-04-02 21:55:20 UTC

The bt:

+ Trace 226551

Thread 1 (Thread 0xb7fd3ab0 (LWP 18659))

#0 g_type_check_instance_cast
at gtype.c line 3994
#1 gdk_window_cache_new
at gdkdnd-x11.c line 506
#2 gdk_window_cache_get
at gdkdnd-x11.c line 641
#3 drag_context_find_window_cache
at gdkdnd-x11.c line 3275
#4 gdk_x11_drag_context_find_window
at gdkdnd-x11.c line 3297
#5 gdk_drag_find_window_for_screen
at gdkdnd.c line 278
#6 gtk_drag_update_idle
at gtkdnd.c line 3963
#7 gdk_threads_dispatch
at gdk.c line 754
#8 g_idle_dispatch
at gmain.c line 4545
#9 g_main_dispatch
at gmain.c line 2440
#10 g_main_context_dispatch
at gmain.c line 3013
#11 g_main_context_iterate
at gmain.c line 3091
#12 g_main_loop_run
at gmain.c line 3299
#13 gtk_main
at gtkmain.c line 1358
#14 main
at ../../src/ephy-main.c line 747


Seems like a GTK+ bug to me?

Comment 5 Xan Lopez 2011-04-02 21:58:42 UTC

Running with fatal warnings:


Gdk-CRITICAL **: gdk_device_get_source: assertion `GDK_IS_DEVICE (device)' failed
aborting...

Program received signal SIGTRAP, Trace/breakpoint trap.
0x022255fb in g_logv (log_domain=0x36e037 "Gdk", log_level=G_LOG_LEVEL_CRITICAL, format=0x22a2ddc "%s: assertion `%s' failed", 
    args1=0xbfffcb7c "\254\345\066") at gmessages.c:553
553			G_BREAKPOINT ();
(gdb) bt

+ Trace 226552

#0 g_logv
at gmessages.c line 553
#1 g_log
at gmessages.c line 577
#2 g_return_if_fail_warning
#3 gdk_device_get_source
at gdkdevice.c line 645
#4 gtk_drag_begin_internal
at gtkdnd.c line 2325
#5 gtk_drag_begin
at gtkdnd.c line 2548
#6 move_item_cb
at ../../../lib/egg/egg-editable-toolbar.c line 309
#7 g_cclosure_marshal_VOID__VOID
at gmarshal.c line 79
#8 g_closure_invoke
at gclosure.c line 767
#9 signal_emit_unlocked_R
at gsignal.c line 3256
#10 g_signal_emit_valist
at gsignal.c line 2987
#11 g_signal_emit
at gsignal.c line 3044
#12 _gtk_action_emit_activate
at gtkaction.c line 799
#13 gtk_action_activate
at gtkaction.c line 829
#14 gtk_real_menu_item_activate
at gtkmenuitem.c line 1734
#15 g_cclosure_marshal_VOID__VOID
at gmarshal.c line 79
#16 g_type_class_meta_marshal
at gclosure.c line 878
#17 g_closure_invoke
at gclosure.c line 767
#18 signal_emit_unlocked_R
at gsignal.c line 3186
#19 g_signal_emit_valist
at gsignal.c line 2987
#20 g_signal_emit
at gsignal.c line 3044
#21 gtk_widget_activate
at gtkwidget.c line 6109
#22 gtk_menu_shell_activate_item
at gtkmenushell.c line 1403
#23 gtk_menu_shell_button_release
at gtkmenushell.c line 803
#24 gtk_menu_button_release
at gtkmenu.c line 3489
#25 _gtk_marshal_BOOLEAN__BOXED
#26 g_type_class_meta_marshal
at gclosure.c line 878
#27 g_closure_invoke
at gclosure.c line 767
#28 signal_emit_unlocked_R
at gsignal.c line 3294
#29 g_signal_emit_valist
at gsignal.c line 2997
#30 g_signal_emit
at gsignal.c line 3044
#31 gtk_widget_event_internal
at gtkwidget.c line 6078
#32 gtk_widget_event
at gtkwidget.c line 5794
#33 gtk_propagate_event
at gtkmain.c line 2597
#34 gtk_main_do_event
at gtkmain.c line 1872
#35 _gdk_event_emit
at gdkevents.c line 71
#36 gdk_event_source_dispatch
at gdkeventsource.c line 318
#37 g_main_dispatch
at gmain.c line 2440
#38 g_main_context_dispatch
at gmain.c line 3013
#39 g_main_context_iterate
at gmain.c line 3091
#40 g_main_loop_run
at gmain.c line 3299
#41 gtk_main
at gtkmain.c line 1358
#42 main
at ../../src/ephy-main.c line 747

Comment 6 Xan Lopez 2011-04-02 22:06:42 UTC

Created attachment 184994 [details] [review]
dnddevice.diff

This seems to fix the bug, the problem was in libegg.

Comment 7 Xan Lopez 2011-04-03 00:15:13 UTC

Created attachment 184999 [details] [review]
dndfix.diff

Matthias suggests to use gdk_event_get_device since it's a bit more in line with the surrounding code.

Comment 8 Xan Lopez 2011-04-03 15:37:36 UTC

Created attachment 185034 [details] [review]
dndfix.diff

Use gdk_event_set_device.

Comment 9 Xan Lopez 2011-04-03 15:41:22 UTC

Fixed in master (both of ephy and libegg).