Bug 771872 – _cairo_surface_has_snapshots (surface)' failed

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 771872 - _cairo_surface_has_snapshots (surface)' failed


Summary:	_cairo_surface_has_snapshots (surface)' failed


Status:	RESOLVED NOTGNOME

Product:	epiphany
Classification:	Core
Component:	General
Version:	3.22.x (obsolete)
Hardware:	Other Linux

Importance:	High critical
Target Milestone:	---
Assigned To:	Epiphany Maintainers
QA Contact:	Epiphany Maintainers

URL:
Whiteboard:

Depends on:
Blocks:

Reported:	2016-09-23 10:50 UTC by john.frankish@outlook.com
Modified:	2016-09-24 07:55 UTC

See Also:	https://bugs.webkit.org/show_bug.cgi?id=162497
GNOME target:	---
GNOME version:	---

Description john.frankish@outlook.com 2016-09-23 10:50:30 UTC

Using epiphany-3.22.0 and webkitgtk-2.14.0

fails with:

epiphany: cairo-surface.c:1652: cairo_surface_mark_dirty_rectangle: Assertion `! _cairo_surface_has_snapshots (surface)' failed.
Aborted

epiphany-3.21.92 and webkitgtk-2.13.2 worked without problems

Comment 1 Michael Catanzaro 2016-09-23 13:27:06 UTC

I can't reproduce it. We need a stacktrace please, using 'bt full' in gdb. Canned reply:

Thanks for taking the time to report this.
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 https://wiki.gnome.org/Community/GettingInTouch/Bugzilla/GettingTraces for more information on how to do so. When pasting a stack trace in this bug report, please reset the status of this bug report from NEEDINFO to its previous status. Thanks in advance!

Comment 2 john.frankish@outlook.com 2016-09-23 14:13:51 UTC

To use glibc with debug symbols would be awkward.

If I recompile cairo, epiphany glib2, gtk3 and webkitgtk with debug symbols is that enough?

(gdb) bt full

+ Trace 236690

#0 raise
from /lib/libc.so.6
#1 abort
from /lib/libc.so.6
#2 ??
from /lib/libc.so.6
#3 __assert_fail
from /lib/libc.so.6
#4 cairo_surface_mark_dirty_rectangle
from /usr/local/lib/libcairo.so.2
#5 cairo_surface_mark_dirty
from /usr/local/lib/libcairo.so.2
#6 ??
from /usr/local/lib/libwebkit2gtk-4.0.so.37
#7 ??
from /usr/local/lib/libwebkit2gtk-4.0.so.37
#8 ??
from /usr/local/lib/libgtk-3.so.0
#9 gtk_container_propagate_draw
from /usr/local/lib/libgtk-3.so.0
#10 ??
from /usr/local/lib/libgtk-3.so.0
#11 ??
from /usr/local/lib/libgtk-3.so.0
#12 gtk_container_propagate_draw
from /usr/local/lib/libgtk-3.so.0
#13 ??
from /usr/local/lib/libgtk-3.so.0
#14 ??
from /usr/local/lib/libgtk-3.so.0
#15 ??
from /usr/local/lib/libgtk-3.so.0
#16 ??
from /usr/local/lib/libgtk-3.so.0
#17 ??
from /usr/local/lib/libgtk-3.so.0
#18 gtk_container_propagate_draw
from /usr/local/lib/libgtk-3.so.0
#19 ??
from /usr/local/lib/libgtk-3.so.0
#20 ??
from /usr/local/lib/libgtk-3.so.0
#21 ??
from /usr/local/lib/libgtk-3.so.0
#22 ??
from /usr/local/lib/libgtk-3.so.0
#23 ??
from /usr/local/lib/libgtk-3.so.0
#24 ??
from /usr/local/lib/libgtk-3.so.0
#25 gtk_container_propagate_draw
from /usr/local/lib/libgtk-3.so.0
#26 ??
from /usr/local/lib/libgtk-3.so.0
#27 ??
from /usr/local/lib/libgtk-3.so.0
#28 ??
from /usr/local/lib/libgtk-3.so.0
#29 ??
from /usr/local/lib/libgtk-3.so.0
#30 ??
from /usr/local/lib/libgtk-3.so.0
#31 ??
from /usr/local/lib/libgtk-3.so.0
#32 ??
from /usr/local/lib/libgtk-3.so.0
#33 gtk_container_propagate_draw
from /usr/local/lib/libgtk-3.so.0
#34 ??
from /usr/local/lib/libgtk-3.so.0
#35 ??
from /usr/local/lib/libgtk-3.so.0
#36 ??
from /usr/local/lib/libgtk-3.so.0
#37 ??
from /usr/local/lib/libgtk-3.so.0
#38 gtk_main_do_event
from /usr/local/lib/libgtk-3.so.0
#39 ??
from /usr/local/lib/libgdk-3.so.0
#40 ??
from /usr/local/lib/libgdk-3.so.0
#41 ??
from /usr/local/lib/libgdk-3.so.0
#42 ??
from /usr/local/lib/libgdk-3.so.0
#43 g_closure_invoke
from /usr/local/lib/libgobject-2.0.so.0
#44 ??
from /usr/local/lib/libgobject-2.0.so.0
#45 g_signal_emit_valist
from /usr/local/lib/libgobject-2.0.so.0
#46 g_signal_emit
from /usr/local/lib/libgobject-2.0.so.0
#47 ??
from /usr/local/lib/libgdk-3.so.0
#48 ??
from /usr/local/lib/libgdk-3.so.0
#49 ??
from /usr/local/lib/libglib-2.0.so.0
#50 g_main_context_dispatch
from /usr/local/lib/libglib-2.0.so.0
#51 ??
from /usr/local/lib/libglib-2.0.so.0
#52 g_main_context_iteration
from /usr/local/lib/libglib-2.0.so.0
#53 g_application_run
from /usr/local/lib/libgio-2.0.so.0
#54 main

(gdb)

Comment 3 Michael Catanzaro 2016-09-23 15:03:35 UTC

(In reply to john.frankish@outlook.com from comment #2)
> To use glibc with debug symbols would be awkward.
> 
> If I recompile cairo, epiphany glib2, gtk3 and webkitgtk with debug symbols
> is that enough?

Yeah, that'd be fine

Comment 4 john.frankish@outlook.com 2016-09-23 15:16:07 UTC

Here it is:

Program received signal SIGABRT, Aborted.
0x00007fffeb26b7c1 in raise () from /lib/libc.so.6
(gdb) bt full

+ Trace 236691

#0 raise
#1 abort
#2 __assert_fail
#3 __assert_perror_fail
#4 cairo_surface_mark_dirty_rectangle
#5 cairo_surface_mark_dirty
#6 WebKit::AcceleratedBackingStoreX11::paint(_cairo*, WebCore::IntRect const&)
#7 webkitWebViewBaseDraw(_GtkWidget*, _cairo*)
#8 gtk_widget_draw_internal
#9 gtk_container_propagate_draw
#10 gtk_container_draw
#11 gtk_widget_draw_internal
#12 gtk_container_propagate_draw
#13 gtk_paned_render
#14 gtk_css_custom_gadget_draw
#15 gtk_css_gadget_draw
#16 gtk_paned_draw
#17 gtk_widget_draw_internal
#18 gtk_container_propagate_draw
#19 gtk_container_draw
#20 gtk_box_draw_contents
#21 gtk_css_custom_gadget_draw
#22 gtk_css_gadget_draw
#23 gtk_box_draw
#24 gtk_widget_draw_internal
#25 gtk_container_propagate_draw
#26 gtk_notebook_draw_stack
#27 gtk_css_custom_gadget_draw
#28 gtk_css_gadget_draw
#29 gtk_box_gadget_draw
#30 gtk_css_gadget_draw
#31 gtk_notebook_draw
#32 gtk_widget_draw_internal
#33 gtk_container_propagate_draw
#34 gtk_container_draw
#35 gtk_window_draw.lto_priv
#36 gtk_widget_draw_internal
#37 gtk_widget_render
#38 gtk_main_do_event
#39 _gdk_event_emit
#40 _gdk_window_process_updates_recurse_helper.lto_priv.323
#41 gdk_window_process_updates_internal
#42 gdk_window_process_updates_with_mode.lto_priv
#43 g_closure_invoke
#44 signal_emit_unlocked_R.lto_priv.242
#45 g_signal_emit_valist
#46 g_signal_emit
#47 gdk_frame_clock_paint_idle
#48 gdk_threads_dispatch
#49 g_timeout_dispatch
#50 g_main_context_dispatch
#51 g_main_context_iterate.isra.42.lto_priv
#52 g_main_context_iteration
#53 g_application_run
#54 main

Comment 5 Michael Catanzaro 2016-09-23 17:06:46 UTC

It looks like a regression in WebKitGTK+ 2.14. I'll ask you to report this to bugs.webkit.org, but first it's important to get a better backtrace; we need to see local variables. Can you please take the backtrace with 'bt full' instead of just 'bt'? Once you have a backtrace showing local variables, then we can file a WebKit bug. Canned reply:

Thanks for taking the time to report this.
This issue is most likely a bug in WebKit. Please report the bug to https://bugs.webkit.org/ including a link to this bug report and noting the version of WebKitGTK+ that you have installed. When reporting the WebKit bug, be sure to include the prefix '[GTK]' in the bug summary and select the 'WebKit Gtk' component.

Comment 6 john.frankish@outlook.com 2016-09-23 17:25:38 UTC

that was "bt full"...

Comment 7 john.frankish@outlook.com 2016-09-23 17:32:16 UTC

https://bugs.webkit.org/show_bug.cgi?id=162497

Comment 8 john.frankish@outlook.com 2016-09-24 07:55:16 UTC

just for the record, both epiphany 3.21.92 and 3.22.0 work with webkitgtk-2.13.2, so it does seem to be a webkitgtk issue

note however that at least one of the gnome-*-3.22.0 packages requires webkitgtk-2.23.92, which creates a problem to revert to webkitgtk-2.13.2