Bug 112266
Summary: | [GTK][WK2][stable] Logging into Facebook hits an unimplemented virtual method | ||
---|---|---|---|
Product: | WebKit | Reporter: | Claudio Saavedra <csaavedra> |
Component: | WebKitGTK | Assignee: | Nobody <webkit-unassigned> |
Status: | RESOLVED FIXED | ||
Severity: | Blocker | CC: | cgarcia, mrobinson, sergio, spenap, xan.lopez, zan |
Priority: | P2 | ||
Version: | 528+ (Nightly build) | ||
Hardware: | Unspecified | ||
OS: | Unspecified |
Claudio Saavedra
This can be reproduced with MiniBrowser, built from the stable branch, and also Epiphany. Log into Facebook and boom!
#0 0x0000003af8c35ba5 in __GI_raise (sig=sig@entry=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:63
#1 0x0000003af8c37358 in __GI_abort () at abort.c:90
#2 0x0000003afb460e1d in __gnu_cxx::__verbose_terminate_handler () at ../../../../libstdc++-v3/libsupc++/vterminate.cc:95
#3 0x0000003afb45eeb6 in __cxxabiv1::__terminate (handler=<optimized out>) at ../../../../libstdc++-v3/libsupc++/eh_terminate.cc:40
#4 0x0000003afb45eee3 in std::terminate () at ../../../../libstdc++-v3/libsupc++/eh_terminate.cc:50
#5 0x0000003afb45f9ff in __cxxabiv1::__cxa_pure_virtual () at ../../../../libstdc++-v3/libsupc++/pure.cc:50
#6 0x00007f23178782ce in WebCore::JSNodeOwner::isReachableFromOpaqueRoots(JSC::Handle<JSC::Unknown>, void*, JSC::SlotVisitor&) () from /opt/gnome-3.0/lib64/libwebkit2gtk-3.0.so.22
#7 0x00007f2319455935 in JSC::WeakBlock::visit(JSC::HeapRootVisitor&) () from /opt/gnome-3.0/lib64/libjavascriptcoregtk-3.0.so.0
#8 0x00007f2319452bcb in JSC::MarkedSpace::visitWeakSets(JSC::HeapRootVisitor&) () from /opt/gnome-3.0/lib64/libjavascriptcoregtk-3.0.so.0
#9 0x00007f2319449415 in JSC::Heap::markRoots() () from /opt/gnome-3.0/lib64/libjavascriptcoregtk-3.0.so.0
#10 0x00007f231944ad66 in JSC::Heap::collect(JSC::Heap::SweepToggle) () from /opt/gnome-3.0/lib64/libjavascriptcoregtk-3.0.so.0
#11 0x00007f231781e102 in WebCore::collect(void*) () from /opt/gnome-3.0/lib64/libwebkit2gtk-3.0.so.22
#12 0x00007f23186cfad2 in WebCore::ThreadTimers::sharedTimerFiredInternal() () from /opt/gnome-3.0/lib64/libwebkit2gtk-3.0.so.22
#13 0x00007f2318b80de2 in WebCore::timeout_cb(void*) () from /opt/gnome-3.0/lib64/libwebkit2gtk-3.0.so.22
#14 0x00007f2313b60f9a in g_timeout_dispatch (source=0x3b16380, callback=0x7f2318b80dd0 <WebCore::timeout_cb(void*)>, user_data=0x0) at gmain.c:4413
#15 0x00007f2313b5f23d in g_main_dispatch (context=0x20f4d80) at gmain.c:3054
#16 0x00007f2313b5ffa2 in g_main_context_dispatch (context=0x20f4d80) at gmain.c:3630
#17 0x00007f2313b60192 in g_main_context_iterate (context=0x20f4d80, block=1, dispatch=1, self=0x2134630) at gmain.c:3701
#18 0x00007f2313b605c2 in g_main_loop_run (loop=0x21c2740) at gmain.c:3895
#19 0x00007f2317753074 in WebProcessMainGtk () from /opt/gnome-3.0/lib64/libwebkit2gtk-3.0.so.22
#20 0x0000003af8c21a05 in __libc_start_main (main=0x400b80 <main>, argc=2, ubp_av=0x7fff15e578a8, init=<optimized out>, fini=<optimized out>, rtld_fini=<optimized out>, stack_end=0x7fff15e57898) at libc-start.c:225
#21 0x0000000000400ab9 in _start ()
Attachments | ||
---|---|---|
Add attachment proposed patch, testcase, etc. |
Claudio Saavedra
This can be also reproduced, at least from:
- goodreads.com
- plus.google.com
- gmail's chat
Martin Robinson
(In reply to comment #1)
> This can be also reproduced, at least from:
>
> - goodreads.com
> - plus.google.com
> - gmail's chat
Perhaps we are missing a file in the source list.
Xan Lopez
(In reply to comment #2)
> (In reply to comment #1)
> > This can be also reproduced, at least from:
> >
> > - goodreads.com
> > - plus.google.com
> > - gmail's chat
>
> Perhaps we are missing a file in the source list.
Also, shouldn't pure virtual methods give you compiler warnings if you try to make an instance of a class that has one of those?
Zan Dobersek
Still reproducible?
Claudio Saavedra
Can't reproduce it with the stable build in F19.
Zan Dobersek
I can see through dumping the symbols that on both ToT and the 2.0.2 release, the JSNodeOwner::isReachableFromOpaqueRoots method is defined.
I can also confirm that the crashes don't occur on ToT. Closing.