Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <20220723012325.1715714-1-seanjc@google.com> <20220723012325.1715714-3-seanjc@google.com>
 <Yt8hu/+I8YzVckvU@google.com>
In-Reply-To: <Yt8hu/+I8YzVckvU@google.com>
From:   David Matlack <dmatlack@google.com>
Date:   Mon, 25 Jul 2022 16:08:42 -0700
Message-ID: <CALzav=djkQWrxxXR5qik8A=GK5fpg+RN3Qfrmgwxoh1NVu7+aQ@mail.gmail.com>
Subject: Re: [PATCH v2 2/6] KVM: x86/mmu: Properly account NX huge page
 workaround for nonpaging MMUs
To:     Sean Christopherson <seanjc@google.com>
Cc:     Paolo Bonzini <pbonzini@redhat.com>,
        kvm list <kvm@vger.kernel.org>,
        LKML <linux-kernel@vger.kernel.org>,
        Yosry Ahmed <yosryahmed@google.com>,
        Mingwei Zhang <mizhang@google.com>,
        Ben Gardon <bgardon@google.com>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Mon, Jul 25, 2022 at 4:05 PM David Matlack <dmatlack@google.com> wrote:
>
> On Sat, Jul 23, 2022 at 01:23:21AM +0000, Sean Christopherson wrote:
> > Account and track NX huge pages for nonpaging MMUs so that a future
> > enhancement to precisely check if shadow page cannot be replaced by a NX
> > huge page doesn't get false positives.  Without correct tracking, KVM can
> > get stuck in a loop if an instruction is fetching and writing data on the
> > same huge page, e.g. KVM installs a small executable page on the fetch
> > fault, replaces it with an NX huge page on the write fault, and faults
> > again on the fetch.
> >
> > Alternatively, and perhaps ideally, KVM would simply not enforce the
> > workaround for nonpaging MMUs.  The guest has no page tables to abuse
> > and KVM is guaranteed to switch to a different MMU on CR0.PG being
> > toggled so there's no security or performance concerns.  However, getting
> > make_spte() to play nice now and in the future is unnecessarily complex.
> >
> > In the current code base, make_spte() can enforce the mitigation if TDP
> > is enabled or the MMU is indirect, but make_spte() may not always have a
> > vCPU/MMU to work with, e.g. if KVM were to support in-line huge page
> > promotion when disabling dirty logging.
> >
> > Without a vCPU/MMU, KVM could either pass in the correct information
> > and/or derive it from the shadow page, but the former is ugly and the
> > latter subtly non-trivial due to the possitibility of direct shadow pages
> > in indirect MMUs.  Given that using shadow paging with an unpaged guest
> > is far from top priority _and_ has been subjected to the workaround since
> > its inception, keep it simple and just fix the accounting glitch.
> >
> > Signed-off-by: Sean Christopherson <seanjc@google.com>
>
> It's odd that KVM enforced NX Huge Pages but just skipped the accounting.
> In retrospect, that was bound to cause some issue.
>
> Aside from the comment suggestion below,
>
> Reviewed-by: David Matlack <dmatlack@google.com>
>
> > ---
> >  arch/x86/kvm/mmu/mmu.c          |  2 +-
> >  arch/x86/kvm/mmu/mmu_internal.h |  8 ++++++++
> >  arch/x86/kvm/mmu/spte.c         | 11 +++++++++++
> >  3 files changed, 20 insertions(+), 1 deletion(-)
> >
> > diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
> > index 1112e3a4cf3e..493cdf1c29ff 100644
> > --- a/arch/x86/kvm/mmu/mmu.c
> > +++ b/arch/x86/kvm/mmu/mmu.c
> > @@ -3135,7 +3135,7 @@ static int __direct_map(struct kvm_vcpu *vcpu, struct kvm_page_fault *fault)
> >                       continue;
> >
> >               link_shadow_page(vcpu, it.sptep, sp);
> > -             if (fault->is_tdp && fault->huge_page_disallowed)
> > +             if (fault->huge_page_disallowed)
> >                       account_nx_huge_page(vcpu->kvm, sp,
> >                                            fault->req_level >= it.level);
> >       }
> > diff --git a/arch/x86/kvm/mmu/mmu_internal.h b/arch/x86/kvm/mmu/mmu_internal.h
> > index ff4ca54b9dda..83644a0167ab 100644
> > --- a/arch/x86/kvm/mmu/mmu_internal.h
> > +++ b/arch/x86/kvm/mmu/mmu_internal.h
> > @@ -201,6 +201,14 @@ struct kvm_page_fault {
> >
> >       /* Derived from mmu and global state.  */
> >       const bool is_tdp;
> > +
> > +     /*
> > +      * Note, enforcing the NX huge page mitigation for nonpaging MMUs
> > +      * (shadow paging, CR0.PG=0 in the guest) is completely unnecessary.
> > +      * The guest doesn't have any page tables to abuse and is guaranteed
> > +      * to switch to a different MMU when CR0.PG is toggled on (may not
> > +      * always be guaranteed when KVM is using TDP).  See also make_spte().
> > +      */
> >       const bool nx_huge_page_workaround_enabled;
> >
> >       /*
> > diff --git a/arch/x86/kvm/mmu/spte.c b/arch/x86/kvm/mmu/spte.c
> > index 7314d27d57a4..9f3e5af088a5 100644
> > --- a/arch/x86/kvm/mmu/spte.c
> > +++ b/arch/x86/kvm/mmu/spte.c
> > @@ -147,6 +147,17 @@ bool make_spte(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp,
> >       if (!prefetch)
> >               spte |= spte_shadow_accessed_mask(spte);
> >
> > +     /*
> > +      * For simplicity, enforce the NX huge page mitigation even if not
> > +      * strictly necessary.  KVM could ignore if the mitigation if paging is
> > +      * disabled in the guest, but KVM would then have to ensure a new MMU
> > +      * is loaded (or all shadow pages zapped) when CR0.PG is toggled on,
> > +      * and that's a net negative for performance when TDP is enabled.  KVM
> > +      * could ignore the mitigation if TDP is disabled and CR0.PG=0, as KVM
> > +      * will always switch to a new MMU if paging is enabled in the guest,
> > +      * but that adds complexity just to optimize a mode that is anything
> > +      * but performance critical.
> > +      */
>
> I had some trouble parsing the last sentence. How about this for slightly
> better flow:
>
>         /*
>          * For simplicity, enforce the NX huge page mitigation even if not
>          * strictly necessary.  KVM could ignore if the mitigation if paging is
>          * disabled in the guest, but KVM would then have to ensure a new MMU
>          * is loaded (or all shadow pages zapped) when CR0.PG is toggled on,
>          * and that's a net negative for performance when TDP is enabled.  When
>          * TDP is disabled, KVM will always switch to a new MMU when CR0.PG is
>          * toggled, but that would tie make_spte() further to vCPU/MMU state
>          * and add complexity just to optimize a mode that is anything but
>          * performance critical.

Blegh. Should be:

"... but leveraging that to ignore the mitigation would tie
make_spte() further..."

>          */
>
> >       if (level > PG_LEVEL_4K && (pte_access & ACC_EXEC_MASK) &&
> >           is_nx_huge_page_enabled(vcpu->kvm)) {
> >               pte_access &= ~ACC_EXEC_MASK;
> > --
> > 2.37.1.359.gd136c6c3e2-goog
> >