Bug 135876

Summary: [GTK] WebKitGtk+ 2.5.2 crashes using Facebook
Product: WebKit Reporter: Andres Gomez Garcia <agomez>
Component: WebKitGTKAssignee: Nobody <webkit-unassigned>
Status: RESOLVED FIXED    
Severity: Normal CC: berto, cgarcia, mcatanzaro, mrobinson, zan
Priority: P2    
Version: 528+ (Nightly build)   
Hardware: Unspecified   
OS: Linux   

Description Andres Gomez Garcia 2014-08-13 00:25:10 PDT
I'm using WebKitGtk+ with my own JHBuild setting:
https://github.com/tanty/jhbuild-epiphany/tree/wkgtk-devel

Today I updated from WebKitGtk+ 2.5.1 to 2.5.2 and consistently get crashes when browsing Facebook.

This is a BT.

Program received signal SIGSEGV, Segmentation fault.
0x00007ffff536534d in webkit_web_view_get_tls_info () from /opt/gnome-os/tanty/epiphany/install/lib/libwebkit2gtk-4.0.so.37
(gdb) bt
#0  0x00007ffff536534d in webkit_web_view_get_tls_info () from /opt/gnome-os/tanty/epiphany/install/lib/libwebkit2gtk-4.0.so.37
#1  0x0000000000466495 in load_changed_cb (web_view=0x29b7870, load_event=<optimized out>, user_data=<optimized out>) at ephy-web-view.c:1619
#2  0x00007ffff2114e48 in g_closure_invoke (closure=0x29b7fa0, return_value=0x0, n_param_values=2, param_values=0x7fffffffc2a0, invocation_hint=0x7fffffffc240) at gclosure.c:768
#3  0x00007ffff2126add in signal_emit_unlocked_R (node=node@entry=0xa8e740, detail=detail@entry=0, instance=instance@entry=0x29b7870, emission_return=emission_return@entry=0x0, 
    instance_and_params=instance_and_params@entry=0x7fffffffc2a0) at gsignal.c:3553
#4  0x00007ffff212ead8 in g_signal_emit_valist (instance=<optimized out>, signal_id=<optimized out>, detail=<optimized out>, var_args=var_args@entry=0x7fffffffc438) at gsignal.c:3309
#5  0x00007ffff212ed92 in g_signal_emit (instance=<optimized out>, signal_id=<optimized out>, detail=<optimized out>) at gsignal.c:3365
#6  0x00007ffff536134c in webkitWebViewLoadChanged(_WebKitWebView*, WebKitLoadEvent) () from /opt/gnome-os/tanty/epiphany/install/lib/libwebkit2gtk-4.0.so.37
#7  0x00007ffff523f7cc in WebKit::WebPageProxy::didStartProvisionalLoadForFrame(unsigned long, unsigned long, WTF::String const&, WTF::String const&, IPC::MessageDecoder&) ()
   from /opt/gnome-os/tanty/epiphany/install/lib/libwebkit2gtk-4.0.so.37
#8  0x00007ffff53a6bf6 in void IPC::handleMessageVariadic<Messages::WebPageProxy::DidStartProvisionalLoadForFrame, WebKit::WebPageProxy, void (WebKit::WebPageProxy::*)(unsigned long, unsigned long, WTF::String const&, WTF::String const&, IPC::MessageDecoder&)>(IPC::MessageDecoder&, WebKit::WebPageProxy*, void (WebKit::WebPageProxy::*)(unsigned long, unsigned long, WTF::String const&, WTF::String const&, IPC::MessageDecoder&)) ()
   from /opt/gnome-os/tanty/epiphany/install/lib/libwebkit2gtk-4.0.so.37
#9  0x00007ffff53a4f96 in WebKit::WebPageProxy::didReceiveMessage(IPC::Connection*, IPC::MessageDecoder&) () from /opt/gnome-os/tanty/epiphany/install/lib/libwebkit2gtk-4.0.so.37
#10 0x00007ffff51906b0 in IPC::MessageReceiverMap::dispatchMessage(IPC::Connection*, IPC::MessageDecoder&) () from /opt/gnome-os/tanty/epiphany/install/lib/libwebkit2gtk-4.0.so.37
#11 0x00007ffff52506eb in WebKit::WebProcessProxy::didReceiveMessage(IPC::Connection*, IPC::MessageDecoder&) () from /opt/gnome-os/tanty/epiphany/install/lib/libwebkit2gtk-4.0.so.37
#12 0x00007ffff518acf4 in IPC::Connection::dispatchMessage(std::unique_ptr<IPC::MessageDecoder, std::default_delete<IPC::MessageDecoder> >) () from /opt/gnome-os/tanty/epiphany/install/lib/libwebkit2gtk-4.0.so.37
#13 0x00007ffff518ae6c in IPC::Connection::dispatchOneMessage() () from /opt/gnome-os/tanty/epiphany/install/lib/libwebkit2gtk-4.0.so.37
#14 0x00007ffff657af93 in WTF::RunLoop::performWork() () from /opt/gnome-os/tanty/epiphany/install/lib/libwebkit2gtk-4.0.so.37
#15 0x00007ffff0422df8 in WTF::GMainLoopSource::voidCallback() () from /opt/gnome-os/tanty/epiphany/install/lib/libjavascriptcoregtk-4.0.so.18
#16 0x00007ffff0422e29 in WTF::GMainLoopSource::voidSourceCallback(WTF::GMainLoopSource*) () from /opt/gnome-os/tanty/epiphany/install/lib/libjavascriptcoregtk-4.0.so.18
#17 0x00007ffff1e179e5 in g_main_dispatch (context=0x702a70) at gmain.c:3064
#18 g_main_context_dispatch (context=context@entry=0x702a70) at gmain.c:3663
#19 0x00007ffff1e17d48 in g_main_context_iterate (context=context@entry=0x702a70, block=block@entry=1, dispatch=dispatch@entry=1, self=<optimized out>) at gmain.c:3734
#20 0x00007ffff1e17dec in g_main_context_iteration (context=0x702a70, context@entry=0x0, may_block=may_block@entry=1) at gmain.c:3795
#21 0x00007ffff23ff18c in g_application_run (application=0x765220, argc=argc@entry=1
Comment 1 Alberto Garcia 2014-08-13 01:17:39 PDT
This was not happening in 2.5.1 then?
Comment 2 Zan Dobersek 2014-08-18 02:15:11 PDT
I haven't checked the time frame of his work and whether it would affect 2.5.2, but Edu's crypto patches that landed over the past few weeks required bumping the version of the libgnutls library. Might that be related?
Comment 3 Andres Gomez Garcia 2014-12-03 03:09:41 PST
(In reply to comment #1)
> This was not happening in 2.5.1 then?

No.
Comment 4 Eduardo Lima Mitev 2014-12-03 03:12:19 PST
Removed myself from CC.
Comment 5 Michael Catanzaro 2014-12-03 09:00:57 PST
I can't reproduce this. Is it still happening, Andres? If so, with what versions of WebKitGTK+, GnuTLS, and glib-networking?
Comment 6 Andres Gomez Garcia 2014-12-03 23:39:17 PST
(In reply to comment #5)
> I can't reproduce this. Is it still happening, Andres? If so, with what
> versions of WebKitGTK+, GnuTLS, and glib-networking?

I have no longer that version. I can re-try in some days, though, but it will take time. However, being a development version I assume it would make more sense to test with a more modern one. FWIW, I'm using 2.6.4 nowadays and I'm not having that problem any more.
Comment 7 Alberto Garcia 2014-12-04 02:30:47 PST
But 2.5.2 is from the unstable branch, isn't it? Do we have a scenario where people can't upgrade from 2.5.2 to some 2.6? If not, does it make sense to keep this bug open?
Comment 8 Andres Gomez Garcia 2014-12-04 04:38:28 PST
(In reply to comment #7)
> But 2.5.2 is from the unstable branch, isn't it? Do we have a scenario where
> people can't upgrade from 2.5.2 to some 2.6? If not, does it make sense to
> keep this bug open?

That's what I was saying ... but I didn't know if there was any interest on checking the actual bug or just assume that something fixed the problem without really knowing the cause.

Anyway, I suppose there is none ...