Protoization: m32r-stub.c and stuff.c?

[Kevin Buettner <kevinb at cygnus dot com>]

I'm nearing the end of my protoization activities for the "pure" gdb
sources.  (I still have a bit of work to do in tui and with the gdbtk
sources.)

Anyway, there are still two files left to protoize, or not...  and I
would appreciate some guidance on whether to protoize these files or
leave them alone.

One of these files is m32r-stub.c.  We had a discussion regarding
this file earlier this year.  See:

    http://sources.redhat.com/ml/gdb-patches/2000-09/msg00137.html
    http://sources.redhat.com/ml/gdb-patches/2000-09/msg00138.html
    http://sources.redhat.com/ml/gdb-patches/2000-09/msg00139.html
    http://sources.redhat.com/ml/gdb-patches/2000-09/msg00140.html
    http://sources.redhat.com/ml/gdb-patches/2000-09/msg00141.html
    http://sources.redhat.com/ml/gdb-patches/2000-09/msg00142.html
    http://sources.redhat.com/ml/gdb-patches/2000-09/msg00143.html
    http://sources.redhat.com/ml/gdb-patches/2000-09/msg00144.html
    http://sources.redhat.com/ml/gdb-patches/2000-09/msg00145.html
    http://sources.redhat.com/ml/gdb-patches/2000-09/msg00146.html
    http://sources.redhat.com/ml/gdb-patches/2000-09/msg00150.html
    http://sources.redhat.com/ml/gdb-patches/2000-09/msg00213.html

After reviewing this discussion, it looks to me as though some of the
participants (Stan and J.T.) were strongly leaning towards protoizing
this file, but that no firm conclusion was reached.

I've looked at stuff.c and am wondering if this file is even in use
anymore.  According to the comments in the file, it was written in
1986 by David Bridgham and is used to "stuff files into a specially
prepared space in kdb."

The most recent ChangeLog entry for stuff.c is by me earlier this year
when I used my fix-decls script to protoize the majority of the gdb
sources.  Prior to that, J.T. Conklin had changed this file in the
fall of 1995 to use "gdb_string.h" instead of <string.h>.  And prior
to that, in 1992, Fred Fish had also performed some maintenance
activities on this file.  Even John Gilmore's changes from 1991 appear
to be maintenance related.  I.e, to the best of my knowledge, all of
the changes to this file in the recorded annals of gdb have been
maintenance related.  (I've also checked the ``cvs log'' record
from Red Hat's internal gdb repository.)

It is my opinion that stuff.c can simply be removed from the gdb
sources.  If not, it contains two functions which need to be
protoized, main() and err().  The first will be easy to do.  The
second will require that the old vararg va_list style of declaration
be converted to use ANSI facilities.  I don't think it's worth it.

Here are my recommendations:

1) m32r-stub.c should be protoized.
2) stuff.c should be marked obsolete.  (And later removed.)

Other opinions?

Kevin