Re: [PATCH 2/3] kvm: fix potentially corrupt mmio cache

[Paolo Bonzini <pbonzini@redhat.com>]

Il 02/09/2014 18:44, David Matlack ha scritto:
> >
> > -#define MMIO_GEN_SHIFT                 19
> > -#define MMIO_GEN_LOW_SHIFT             9
> > -#define MMIO_GEN_LOW_MASK              ((1 << MMIO_GEN_LOW_SHIFT) - 1)
> > +#define MMIO_GEN_SHIFT                 20
> > +#define MMIO_GEN_LOW_SHIFT             10
> > +#define MMIO_GEN_LOW_MASK              ((1 << MMIO_GEN_LOW_SHIFT) - 2)
> >  #define MMIO_GEN_MASK                  ((1 << MMIO_GEN_SHIFT) - 1)
> >  #define MMIO_MAX_GEN                   ((1 << MMIO_GEN_SHIFT) - 1)
> >
> > @@ -4428,7 +4432,7 @@ void kvm_mmu_invalidate_mmio_sptes(struct kvm *kvm)
> >          * The very rare case: if the generation-number is round,
> >          * zap all shadow pages.
> >          */
> > -       if (unlikely(kvm_current_mmio_generation(kvm) >= MMIO_MAX_GEN)) {
> > +       if (unlikely(kvm_current_mmio_generation(kvm) == 0)) {
> 
> This should be in patch 1/3.

I don't think so.  This change is not due to the removal of biasing in
x86.c, but rather due to removing bit 0 from MMIO_GEN_LOW_MASK.

I placed it here, because the previous test works just fine until bit 0
is removed from MMIO_GEN_LOW_MASK.

Paolo