bryan o'connor <bryan@lunch.org> writes:

> > There have been more changes at 06/29 and 06/30.  Your search.c is
> > newer than that?
> > If it is, could you run Emacs under gdb and give a backtrace when it
> > bugs out?
> 
> i just tested a cvs version dated 6/28 and it did abort().  the
> backtrace
> is attached below.
> 
> my keystrokes were:
> 	M-x slime [return]
> 	M-. prin1 [return]
> 	M-. prin1 [return]
> 	M-. prin1 [return]

I have repeatedly stated that I had no clue what I was doing there, as
I have no idea how Lisp data structures are implemented in C.  I have
begged in desperation that someone take a look at the patch before I
check it in.

No reaction.  So that's where we are now.

Does the following patch (where I, on a hunch, try using XSETBUFFER
whenever assigning a buffer, without a clue about whether this has any
relevance or not) change anything?