bug#7728: 24.0.50; GDB backtrace from abort

[Drew Adams]

> > There's still one thing I don't understand: why do we call
> > Fselect_frame?  AFAICT, Fset_window_configuration has no reason to
> > select a new frame, it all works within the selected-frame.
> 
> Probably because of minibuffer-only frames or something.
> 
> Perhaps Drew could publish the relevant parts of the window
> configuration that was being restored in that case (or any other
> similar case).

Sorry, I don't follow this at all at the C level, and even at the Lisp level I'm
not sure I can be much help here.

In the last GDB backtrace I sent (yesterday), I did the following, in my version
of Emacs, in Icicle mode:

C-h f  g r a p h i c  S-TAB

Then click `mouse-2' on completion candidate `display-graphic-p'.  That should
end completion and show *Help* (in a separate frame, in my setup) with the
output of `describe-function'.  Instead, Emacs crashes (or whatever you call it
- no error raised).

That crash (or whatever it is) is reproducible in my setup.  But as there are so
many things going on (I use a standalone minibuffer frame, etc.) it's unclear
how much help that description is, except that it is probably not very helpful.