Subject: Re: CVS commit: src
To: matthew green <>
From: Christian Limpach <>
List: source-changes
Date: 05/13/2004 18:15:34

On Fri, May 14, 2004 at 01:14:00AM +1000, matthew green wrote:
>    From: "matthew green" <>
>    >    Log Message:
>    >    `` -m xen-i386 release'' now builds a release for NetBSD/xen
>    >    for i386.  The resulting release consists of:
>    >    - NetBSD/xen for i386 kernel, loader and docuemntation
>    >    - NetBSD/i386 userland sets
>    >
>    > if the userland is the same, is there any particular reason that
>    > i386 can't just include a "xen" kernel?  eg, the way that sparc
>    > includes a 32bit "sparc64" kernel.  see sparc/conf/GENERIC_SUN4U.
>    I'd actually prefer this since you install NetBSD/i386 first and then switch
>    to the NetBSD/xen kernel.
>    There might be 2 problems with this though:
>    - if arch/i386/conf/GENERIC_XEN doesn't compile, release for i386 will fail
> i don't think this is a problem.

It has the definite advantage that anybody making changes to arch/i386
which break the xen kernel build will notice sooner.

>    - nothing will build the NetBSD/xen loader and documentation
> that can be changed... in the sparc* case, both platforms build
> the sparc64 loaders (bootblk, which is in forth, and ofwboot,
> which is largely in C.  the 32 bit ofwboot can load 32 bit or
> 64 bit kernels.)  docs is a SMOD(? :-)

The loader build could be added to src/etc/etc.i386/'s
snap_md_post target and src/sys/arch/Makefile would probably have
to enter arch/xen when building for i386.  Alternatively the loader
could be moved under arch/i386/stand since it's i386 specific

I'm struggling with the docs anyway since the structure of INSTALL.*
doesn't really fit.

>    > this would avoid having to have another port to "release".
>    I was thinking that we could maybe add a knob which prevents building the
>    sets at all for -m xen-*.
> hmmm something along these lines would be ok if we go this route.

I have a slight preference for "release with i386".