Re: [Xen-devel] [PATCH 2/2] xen/mm: Introduce PG_state_uninitialised

[Jan Beulich <jbeulich@xxxxxxxx>]

On 18.03.2020 11:41, Paul Durrant wrote:
> > -----Original Message-----
> > From: Jan Beulich <jbeulich@xxxxxxxx>
> > Sent: 18 March 2020 10:10
> > To: paul@xxxxxxx
> > Cc: 'David Woodhouse' <dwmw2@xxxxxxxxxxxxx>; sstabellini@xxxxxxxxxx; 
> > julien@xxxxxxx; wl@xxxxxxx;
> > konrad.wilk@xxxxxxxxxx; george.dunlap@xxxxxxxxxxxxx; andrew.cooper3@xxxxxxxxxx;
> > ian.jackson@xxxxxxxxxxxxx; george.dunlap@xxxxxxxxxx; 
> > jeff.kubascik@xxxxxxxxxxxxxxx; 'Xia, Hongyan'
> > <hongyxia@xxxxxxxxxx>; stewart.hildebrand@xxxxxxxxxxxxxxx; 
> > xen-devel@xxxxxxxxxxxxxxxxxxxx
> > Subject: Re: [PATCH 2/2] xen/mm: Introduce PG_state_uninitialised
> >
> > On 18.03.2020 09:53, Paul Durrant wrote:
> > > > -----Original Message-----
> > > > From: Xen-devel <xen-devel-bounces@xxxxxxxxxxxxxxxxxxxx> On Behalf Of David 
> > > > Woodhouse
> > > > Sent: 17 March 2020 22:15
> > > >
> > > > On Thu, 2020-02-20 at 12:59 +0100, Jan Beulich wrote:
> > > > > On 07.02.2020 19:04, David Woodhouse wrote:
> > > > > > --- a/xen/arch/x86/mm.c
> > > > > > +++ b/xen/arch/x86/mm.c
> > > > > > @@ -488,7 +488,8 @@ void share_xen_page_with_guest(struct page_info *page, 
> > > > > > struct domain *d,
> > > > > >
> > > > > >        page_set_owner(page, d);
> > > > > >        smp_wmb(); /* install valid domain ptr before updating refcnt. */
> > > > > > -    ASSERT((page->count_info & ~PGC_xen_heap) == 0);
> > > > > > +    ASSERT((page->count_info & ~PGC_xen_heap) == PGC_state_inuse ||
> > > > > > +           (page->count_info & ~PGC_xen_heap) == PGC_state_uninitialised);
> > > > > Can uninitialized pages really make it here?
> > > > Yep, we share the low 1MiB with dom_io.
> > > OOI anyone know why we do this? Is it actually necessary?
> > Yes, for Dom0 to be able to access things like EBDA, IBFT, or data
> > found in BIOS space.
> Ok. I am still wondering why dom0's low 1MiB of pfn space is not
> simply mapped 1:1 though. Just historical?
Well, in a way perhaps. Using the DomIO approach is less of a special
case than mapping some arbitrary range 1:1. Furthermore Dom0 being PV
wouldn't necessarily expect any BIOS in its PFN range there, but
rather views it as normal RAM.

Jan

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel