Re: [Xen-devel] [PATCH RFC v12 08/21] pvh: vmx-specific changes

[George Dunlap <george.dunlap@xxxxxxxxxxxxx>]

On 16/09/13 08:37, Jan Beulich wrote:
> > > > On 13.09.13 at 18:38, George Dunlap <george.dunlap@xxxxxxxxxxxxx> wrote:
> > On 13/09/13 17:25, George Dunlap wrote:
> > > +    /* PVH: Why is ET not set? */
> > > +    v->arch.hvm_vcpu.guest_cr[0] = is_pvh_domain(d) ?
> > > +        ( X86_CR0_PG | X86_CR0_NE | X86_CR0_PE | X86_CR0_WP )
> > > +        : ( X86_CR0_PE | X86_CR0_ET );
> > I also couldn't figure out why CR0_ET wasn't set for PVH vms; is it not
> > set for PV vms?
> This one's certainly wrong - it shouldn't even be possible to clear
> CR0.ET in hardware on anything post-486 (i.e. trying to clear it
> would be silently ignored by the hardware, and not setting it in
> the code above is merely confusing, but not actively breaking
> anything).
Right -- so it would be better if this were something like:

v->arch.hvm_vcpu.guest_cr[0] = $SHARED_FLAGS;
if ( is_pvh_domain(d))
    v->arch.hvm_vcpu.guest_cr[0] |= $PVH_FLAGS.

 -George


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel