bug#16207: 24.3.50; emacs_backtrace.txt

[Eli Zaretskii]

> Date: Sat, 21 Dec 2013 19:31:47 +0100
> From: martin rudalics <rudalics@gmx.at>
> Cc: 16207@debbugs.gnu.org
> 
>  >  > Breakpoint 1, terminate_due_to_signal (sig=22,
>  >  > backtrace_limit=2147483647) at emacs.c:351
>  >  > 351       signal (sig, SIG_DFL);
>  >  > (gdb) bt
>  >  > #0  terminate_due_to_signal (sig=22, backtrace_limit=2147483647) at
>  > emacs.c:351
>  >  > #1  0x011615dc in die (msg=0x148b198 "BUFFERP (a)", file=0x148b10c
>  >  > "lisp.h", line=874) at alloc.c:6742
>  >  > #2  0x010eb7fb in XBUFFER (a=58132514) at lisp.h:874
>  >  > #3  0x01089e14 in run_window_configuration_change_hook (f=0x3aa1a08
>  >  > <__register_frame_info+61479432>) at window.c:3257
>  > ...
>  >  > "set-window-configuration" (0x88db28)
> 
> Which is obviously bug#16051 which I've been now able to reproduce for
> the first time.  And it's definitively the toolbar-window which causes
> the crash.

But then the XBUFFER thing is bogus, because the tool-bar window never
has a buffer.  Its 'contents' field is always nil.

But how did that window get to become the "selected" window??  It
should never be that, I think.