Re: [PATCH v1 3/8]: PVH startup changes (enlighten.c)

[Mukesh Rathor <mukesh.rathor@oracle.com>]

On Thu, 4 Oct 2012 09:38:40 +0100
Ian Campbell <Ian.Campbell@citrix.com> wrote:

> On Wed, 2012-10-03 at 23:37 +0100, Mukesh Rathor wrote:
> > On Wed, 3 Oct 2012 12:58:22 +0100
> > Ok, finally, focussing on this, the issue with pfn in dom0 is that
> > I need pfn allocated in construct_dom0() and be mapped so that the
> > guest can just do :
> > 
> > HYPERVISOR_shared_info=(struct shared_info
> > *)__va(xen_start_info->shared_info);
> > 
> > How about following I am experimenting with right now:
> > 
> > in construct_dom0():
> > 
> >     vstartinfo_end   = (vstartinfo_start +
> >                         sizeof(struct start_info) +
> >                         sizeof(struct dom0_vga_console_info));
> > 
> >     if ( is_hybrid_domain(d) ) {
> >         start_info_pfn_addr = round_pgup(vstartinfo_end) - v_start;
> >         vstartinfo_end   += PAGE_SIZE;
> >     }
> > 
> > I can then put (PFN: start_info_pfn_addr)->(MFN:
> > virt_to_maddr(d->shared_info)) in the p2m, and dom0 just has to do
> > __va(), like domU does now.  I wont' need to special case dom0 then.
> > 
> > Do you foresee any problems with this approach?
> 
> Hard to say without all the surrounding context but it seems plausible
> to me.


Ok, above works. So no dom0 special case in linux now to map shared_page.

thanks
Mukesh