Re: [Xen-devel] [PATCH 3/4] x86/e820: assume memmap provided when booted as a Xen guest is correct

[Roger Pau Monné <roger.pau@xxxxxxxxxx>]

On Thu, Dec 27, 2018 at 09:03:00PM +0000, Andrew Cooper wrote:
> On 27/12/2018 15:56, Roger Pau Monne wrote:
> > This implies there's no need to forcefully reserve the VGA MMIO
> > region, since the memory map provided will be correct.
> >
> > Reported-by: Andrew Cooper <andrew.cooper3@xxxxxxxxxx>
> > Signed-off-by: Roger Pau Monné <roger.pau@xxxxxxxxx>
> > ---
> > Cc: Jan Beulich <jbeulich@xxxxxxxx>
> > Cc: Andrew Cooper <andrew.cooper3@xxxxxxxxxx>
> > Cc: Wei Liu <wei.liu2@xxxxxxxxxx>
> > ---
> >  xen/arch/x86/e820.c | 11 ++++++-----
> >  1 file changed, 6 insertions(+), 5 deletions(-)
> >
> > diff --git a/xen/arch/x86/e820.c b/xen/arch/x86/e820.c
> > index 7c5b85118f..5e46d4ec53 100644
> > --- a/xen/arch/x86/e820.c
> > +++ b/xen/arch/x86/e820.c
> > @@ -537,11 +537,12 @@ static void __init 
> > machine_specific_memory_setup(struct e820map *raw)
> >  
> >      reserve_dmi_region();
> >  
> > -    /*
> > -     * Some BIOSes claim RAM in the 640k - 1M region.
> > -     * Not right. Fix it up.
> > -     */
> > -    reserve_vga_region();
> > +    if ( !xen_guest )
> 
> Having thought on this some more, how about using !cpu_has_hypervisor
> here?  Booting PVH under Xen isn't the only case where we won't have a
> legacy VGA range.  HyperV gen-2 VMs are the next most obvious candidate.

I think it's OK to assume that the memory map provided by an
hypervisor must be correct, let me send v2.

Note that booting the Xen kernel under HyperV gen-2 is quite
trivial, the main problem is getting the dom0 kernel to work in Xen PV
mode while using the HyperV extensions.

Thanks, Roger.

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