Subject: Re: Binutil grip (fwd)
To: None <port-mac68k@NetBSD.ORG>
From: Benoit MARTEL <magus@cs.mcgill.ca>
List: port-mac68k
Date: 06/01/1997 16:53:23
Date: Sun, 1 Jun 1997 16:47:54 -0400 (EDT)
From: Benoit MARTEL <magus@cs.mcgill.ca>
To: Michael R Zucca <mrz5149@cs.rit.edu>
Cc: port-mac-68k@netbsd.org
Subject: Re: Binutil grip

On Sun, 1 Jun 1997, Michael R Zucca wrote:

> > >back to "sizeof(struct llc)". I'd have pushed for us leaving the
> > >8's in place until we got the compiler fixed (with big XXX's). Oh
> > >well.
> 
> This is probably a little off topic but I'm just wondering what in the heck
> is wrong with the standard GNU distribution of gcc and binutils?

That seem to be the main concern on tech-toolchain...
>From reading the archive of it, I'm left with the feeling that it's 
feasible but needs some people do dedicate time.

> 
> The fact that we use bastardized versions of this software is the reason
> a good cross-compiler setup is so hard to do. I made a serious effort at this
> but I gave up after I tried compiling gas from the source tree. It just
> isn't friendly to compile anywhere else but on NetBSD on the machine it's
> supposed to be working on.

I've also had some difficulties when trying to setup a cross compiler 
under Linux. I think off-the-shelf gcc/binutils should make it easier.
I hear (tech-toolchain) that stock gcc can be used as a cross compiler to 
compile kernels (on sparc for sun3) but that their are still problems 
with shared libs. It would be really good if we could just get rid of the 
basterdized versions once and for all.

> 
> Is there any effort under way to get the source to work with the standard
> GNU versions?
> 
> BTW, does anybody have a cross compile environment going for Power Mac
> linux? If so, I will be quiet :)
> 
> Another BTW: Somebody asked me if there was a SUP for linux. I found the
> source out on the net. I believe it's somewhere in the Debian Linux
> distribution (disk 7?).
> 

-----------------------------------------------------------------
"Because user errors often produce unpredictable results, the user
should try to avoid them."

IBM MVS/XA System Programming Library.