Re: [PATCH] KVM: x86/mmu: Remove KVM MMU write lock when accessing indirect_shadow_pages

[Jim Mattson]

On Mon, Jun 5, 2023 at 10:42 AM Mingwei Zhang <mizhang@xxxxxxxxxx> wrote:
>
> On Mon, Jun 5, 2023 at 9:55 AM Jim Mattson <jmattson@xxxxxxxxxx> wrote:
> >
> > On Sun, Jun 4, 2023 at 5:43 PM Mingwei Zhang <mizhang@xxxxxxxxxx> wrote:
> > >
> > > Remove KVM MMU write lock when accessing indirect_shadow_pages counter when
> > > page role is direct because this counter value is used as a coarse-grained
> > > heuristics to check if there is nested guest active. Racing with this
> > > heuristics without mmu lock will be harmless because the corresponding
> > > indirect shadow sptes for the GPA will either be zapped by this thread or
> > > some other thread who has previously zapped all indirect shadow pages and
> > > makes the value to 0.
> > >
> > > Because of that, remove the KVM MMU write lock pair to potentially reduce
> > > the lock contension and improve the performance of nested VM. In addition
> > > opportunistically change the comment of 'direct mmu' to make the
> > > description consistent with other places.
> > >
> > > Reported-by: Jim Mattson <jmattson@xxxxxxxxxx>
> > > Signed-off-by: Mingwei Zhang <mizhang@xxxxxxxxxx>
> > > ---
> > >  arch/x86/kvm/x86.c | 10 ++--------
> > >  1 file changed, 2 insertions(+), 8 deletions(-)
> > >
> > > diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> > > index 5ad55ef71433..97cfa5a00ff2 100644
> > > --- a/arch/x86/kvm/x86.c
> > > +++ b/arch/x86/kvm/x86.c
> > > @@ -8585,15 +8585,9 @@ static bool reexecute_instruction(struct kvm_vcpu *vcpu, gpa_t cr2_or_gpa,
> > >
> > >         kvm_release_pfn_clean(pfn);
> > >
> > > -       /* The instructions are well-emulated on direct mmu. */
> > > +       /* The instructions are well-emulated on Direct MMUs. */
> > >         if (vcpu->arch.mmu->root_role.direct) {
> > > -               unsigned int indirect_shadow_pages;
> > > -
> > > -               write_lock(&vcpu->kvm->mmu_lock);
> > > -               indirect_shadow_pages = vcpu->kvm->arch.indirect_shadow_pages;
> > > -               write_unlock(&vcpu->kvm->mmu_lock);
> > > -
> > > -               if (indirect_shadow_pages)
> > > +               if (READ_ONCE(vcpu->kvm->arch.indirect_shadow_pages))
> >
> > I don't understand the need for READ_ONCE() here. That implies that
> > there is something tricky going on, and I don't think that's the case.
>
> READ_ONCE() is just telling the compiler not to remove the read. Since
> this is reading a global variable,  the compiler might just read a
> previous copy if the value has already been read into a local
> variable. But that is not the case here...

Not a global variable, actually, but that's not relevant. What would
be wrong with using a previously read copy?

We don't always wrap reads in READ_ONCE(). It's actually pretty rare.
So, there should be an explicit and meaningful reason.

> Note I see there is another READ_ONCE for
> kvm->arch.indirect_shadow_pages, so I am reusing the same thing.

That's not a good reason. "If all of your friends jumped off a cliff,
would you?"

> I did check the reordering issue but it should be fine because when
> 'we' see indirect_shadow_pages as 0, the shadow pages must have
> already been zapped. Not only because of the locking, but also the
> program order in __kvm_mmu_prepare_zap_page() shows that it will zap
> shadow pages first before updating the stats.