> At 3:15 PM 9/13/95, Dave Huang wrote:
> >Is it possible to set up some sort of cross development environment
> >that'll generate NetBSD/mac68k executables on a NetBSD/i386 machine? I
> >figure it might be neat to see if I could make a kernel that'll do
> >something interesting on my machine :) (Quadra 630... doubtful, huh? :)
> >especially since I don't know all that much about the low-level hardware).
> >
> >If anyone has any ideas or instructions on how to do this, please let me
> >know. (BTW, I did manage to compile as, although I have no idea if it
> >works or not... need to find something to assemble first :)
> 
> I doubt that it would work, even if it was cross compiled from the any
> source code version, other than the MacBSD source.  I too have a Q630, and
> have been waiting for a version of MacBSD that _will_ run on it.  However,
> due to the hardware that Apple puts in these things, It won't work without
> custom OS code to handle the mac.

I'm not sure what you mean by "wouldn't work."  Brad and I were able
to compile working MacBSD kernels on a Linux/i386 box.  (The 486 runs
a lot faster than the Mac II, plus you don't have to reboot into
MacBSD every time you want to compile another kernel.)

Dave: Get GNU gcc and configure it with "--target=NetBSD/m68k" or
something.  (There's a list of targets with the distribution.)  You
might have to get some special versions of the GNU assembler from the
NetBSD tree.  Just to start, download gcc and binutils, and compile
them until the system stops complaining.  If you can't get past
something, ask me and I'll try to figure out what we did two months
ago.

If you do get it to compile, though, it still won't run on your Quadra
because of the SCSI problems.  At least you'll have a system where
you can hack and feel useful.  Ask Allen (briggs@acm.vt.edu) for a
pointer to where to start looking for the interrupt problem.

Lawrence