[Webkit-unassigned] [Bug 238494] REGRESSION(r290360): [GLX] Crash on process exit

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Wed Mar 30 00:32:41 PDT 2022


https://bugs.webkit.org/show_bug.cgi?id=238494

--- Comment #7 from Carlos Garcia Campos <cgarcia at igalia.com> ---
(In reply to robert.nagy from comment #5)
> New backtrace with the patch:
> 
> Core was generated by `WebKitWebProcess'.
> Program terminated with signal SIGSEGV, Segmentation fault.
> #0  native () at
> /home/pobj/webkitgtk4-2.36.0/webkitgtk-2.36.0/Source/WebCore/platform/
> graphics/x11/PlatformDisplayX11.h:50
> 50          ::Display* native() const { return m_display; }
> [Current thread is 1 (process 281893)]
> (gdb) bt
> #0  native () at
> /home/pobj/webkitgtk4-2.36.0/webkitgtk-2.36.0/Source/WebCore/platform/
> graphics/x11/PlatformDisplayX11.h:50
> #1  sharedDisplay () at
> /home/pobj/webkitgtk4-2.36.0/webkitgtk-2.36.0/Source/WebCore/platform/
> graphics/x11/XUniqueResource.cpp:46
> #2  deleteXResource () at
> /home/pobj/webkitgtk4-2.36.0/webkitgtk-2.36.0/Source/WebCore/platform/
> graphics/x11/XUniqueResource.cpp:79
> #3  0x000005cf5a2b40dc in reset () at
> /home/pobj/webkitgtk4-2.36.0/webkitgtk-2.36.0/Source/WebCore/platform/
> graphics/x11/XUniqueResource.h:86
> #4  ~XUniqueResource () at
> /home/pobj/webkitgtk4-2.36.0/webkitgtk-2.36.0/Source/WebCore/platform/
> graphics/x11/XUniqueResource.h:77
> #5  ~GLContextGLX () at
> /home/pobj/webkitgtk4-2.36.0/webkitgtk-2.36.0/Source/WebCore/platform/
> graphics/glx/GLContextGLX.cpp:341
> #6  0x000005cf5a2b418f in ~GLContextGLX () at
> /home/pobj/webkitgtk4-2.36.0/webkitgtk-2.36.0/Source/WebCore/platform/
> graphics/glx/GLContextGLX.cpp:323
> #7  0x000005cf5a2c3441 in operator() () at
> /usr/include/c++/v1/__memory/unique_ptr.h:57
> #8  reset () at /usr/include/c++/v1/__memory/unique_ptr.h:318
> #9  operator= () at /usr/include/c++/v1/__memory/unique_ptr.h:276
> #10 ~PlatformDisplayX11 () at
> /home/pobj/webkitgtk4-2.36.0/webkitgtk-2.36.0/Source/WebCore/platform/
> graphics/x11/PlatformDisplayX11.cpp:90
> #11 ~PlatformDisplayX11 () at
> /home/pobj/webkitgtk4-2.36.0/webkitgtk-2.36.0/Source/WebCore/platform/
> graphics/x11/PlatformDisplayX11.cpp:87
> #12 0x000005d02ae80041 in _libc___cxa_finalize (dso=0x0) at
> /usr/src/lib/libc/stdlib/atexit.c:177
> #13 0x000005d02ae5d401 in _libc_exit (status=0) at
> /usr/src/lib/libc/stdlib/exit.c:54
> #14 0x000005cd39f65a09 in _start ()
> (gdb)

Ok, that confirms it's not that gtk closed the display. So, I guess the problem is that for whatever reason it's not ok to call PlatformDisplay::shared() from its destructor. The patch is still good, I think, so I'll try to find a solution on top of it.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20220330/119dc16f/attachment-0001.htm>


More information about the webkit-unassigned mailing list