Is the MIPS kernel buildable from HEAD?

[Patrick Mahan]

> -----Original Message-----
> From: juli at clockworksquid.com [mailto:juli at clockworksquid.com] On Behalf Of
> Juli Mallett
> Sent: Wednesday, August 25, 2010 3:15 PM
> To: M. Warner Losh
> Cc: Patrick Mahan; freebsd-mips at freebsd.org
> Subject: Re: Is the MIPS kernel buildable from HEAD?
>
> On Wed, Aug 25, 2010 at 15:02, M. Warner Losh <imp at bsdimp.com> wrote:
> > In message:
> <AANLkTinhMdBVKEr2dNETKF=aBAu+1yTCq_2dk97LGVng at mail.gmail.com>
> >            Juli Mallett <jmallett at FreeBSD.org> writes:
> > : Hi Patrick,
> > :
> > : Those aren't errors per se; if nothing is happening after that,
> > : something else is wrong.  If you aren't using a recent U-Boot, I
> > : suggest installing a recent U-Boot.  If you are using a recent U-Boot
> > : and know how to recompile it, I can send you patches that might help.
> > : There is an unused data structure, linux_app_boot_info_t, which
> > : clobbers some of the memory used for launching multiple processors in
> > : U-Boot 1.9.0.  I have patches to remove it if you would like to give
> > : that a try.
> >
> > I've also noticed that each of the boards have a different load
> > address.  Maybe 200000000 isn't right for your hardware?
>
> If it were wrong you'd be (very unlikely) to see it recognized as an ELF
> file.
>
> I don't understand you silly people who use addresses by hand instead
> of 0 which uses $(loadaddr) (only with Octeon U-Boot, though) which is
> usually correct for the board, anyway :)

Because we are hoary old silly kernel people who still remember using punch-tape
and still don't trust these new fangled serial consoles... VBEG

I've used 20000000, 18000000 and 21000000.  It seems dependent upon the Octeon
eval board.

Patrick