Hi,

On Thu, May 13, 2004 at 09:15:45AM -0700, Chuck Silvers wrote:
> it seems to me that it would make more sense to do builds based on
> $(MACHINE_ARCH) than on $(MACHINE), since the vast majority of the
> files in a release are the same for (eg.) all 11 or 12 of the m68k
> platforms.

I was surprised that this wasn't the case.  Only the final release
on ftp.netbsd.org seems to have (manually setup?) sharing for a subset
of the m68k sets.  FWIW I like your idea/model.

> there are a few other issues with this kind of change:
> 
>  - kernels
>    build them all and enable sysinst to select one that matches
>    the hardware that it's running on.

 - kernel modules
   these are $(MACHINE) dependent unless all relevant kernel structures
   and interfaces are identical on all $(MACHINES)s of the same
   $(MACHINE_ARCH).  It would probably be best to move these out of
   base.tgz even now, since they are more dependent on a matching
   kernel than on a matching userland.  Do the i386 lkm's even work
   with a GENERIC.MP kernel?


Btw, kernel modules might be slightly different in the i386 sets built
for NetBSD/xen but if that's so then I consider it a bug and it should
be remedied.  I don't think it should be a reason not to release the
NetBSD/xen kernel with i386.

    christian