On Fri, 2006-07-28 at 10:57 +0100, Keir Fraser wrote:
> On 27 Jul 2006, at 21:40, Hollis Blanchard wrote:
> > Keir, you mentioned that start_info probably doesn't make much sense
> > for
> > PowerPC. However, if for no other reason than linux/drivers/xen has
> > many
> > references to it, we'd like to fix it for now and figure out how to
> > replace it later.
> I think that structure is quite x86-specific. I'd be happy to see
> direct references to start_info replaced in drivers/xen. Obvious (and
> ugly) low-hanging fruit is test for SIF_INITDOMAIN. Replacing with
> arch-specific is_initial_domain() or similar would be a great patch to
> A couple of the others (e.g., console_mfn, store_mfn) are a
> bit harder -- how do you communicate those on ppc?
We intend to communicate those via the Open Firmware device tree.
(There's a brief overview at
IBM Linux Technology Center
Xen-ppc-devel mailing list