Subject: Re: Proposed targets to obsolete for gcc 3.5
To: Ben Elliston <bje@au.ibm.com>
From: Ian Dall <ian@beware.dropbear.id.au>
List: port-pc532
Date: 06/29/2004 23:20:45
Ben Elliston <bje@au.ibm.com> writes:

> David Brownlee <abs@absd.org> writes:
> 
> > ns32k-*
> >    Nobody is manufacturing these old beasts.  And again, the only
> >    maintainance done on this port is making sure that it still builds.
> >    There are no test results from the past year, the latest results posted
> >    are from December 2002[8].  In addition, this is another cc0 target,
> >    it is unknown if a free simulator exists, the port uses assembly
> >    prologues/epilogues, it does not use define_constants.  In short, old
> >    stuff that few still use/test, if anyone at all.
> > 
> > http://gcc.gnu.org/ml/gcc/2004-06/msg01113.html
> > 
> >  	Maybe someone should reply to Steven Bosscher indicating ns32k
> >  	support is still alive in one area of the world?
> 
> Ian Dall <ian at beware.dropbear.id.au> is listed as the ns32k GCC
> maintainer.  I wonder if he has any comments on its current state of
> maintenance?

Sombody called? The last time I built this it was OK! However, I admit
to not having been active lately. My pc532 is currently turned off and
longer in routine use.

My take is that unless someone gets the energy to do the elf port of
binutils, the pc532 port is going to die and then there is no
motivation for the gcc ns32k port either. I still have the interest,
but not the time!

I'm happy to continue with the gcc port if someone picks up the elf
port of binutils. I suspect it is not really that hard, but there is
an overhead in maintaining familiarity with the body of code and I
just don't have the time to do that with gcc *and* binutils.

For example, I am aware that "best practice" in gcc ports is to treat
the condition codes as a proper register, instead of as a special case
(cc0 target). I am also aware that best practice is to generate RTL
for the function prologue/epilogue instead of assembly language. Both
of these would be reasonably significant efforts. I have no idea what
define_constant is about, so that is something to come up to speed on.
Just scanning the mailing lists is a significant effort!

Similar comments about the gas and binutils ports. We have a
functioning a.out port (albeit without dynamic linking), there are
plenty of examples of elf ports (vax is probably a good starting
place). But someone has to get there head around it all, understand
elf and the organization of binutils (the way macros are used is not
transparent) do the coding test it debug it etc.

Oh, and then there is the NetBSD kernel changes needed to load ns32k
elf binaries, and changes to the build process to build elf libraries
etc.

So, if someone can step up to to the plate and take on the elf work,
then I'm happy to continue with the gcc maintenance. Any volunteers?

Ian