[PATCH 6/11] KVM/MMU: Flush tlb with range list in sync_page()

[Tianyu Lan]

On Sat, Jan 5, 2019 at 12:30 AM Sean Christopherson
<sean.j.christopherson at intel.com> wrote:
>
> On Fri, Jan 04, 2019 at 04:54:00PM +0800, lantianyu1986 at gmail.com wrote:
> > From: Lan Tianyu <Tianyu.Lan at microsoft.com>
> >
> > This patch is to flush tlb via flush list function.
>
> More explanation of why this is beneficial would be nice.  Without the
> context of the overall series it's not immediately obvious what
> kvm_flush_remote_tlbs_with_list() does without a bit of digging.
>
> >
> > Signed-off-by: Lan Tianyu <Tianyu.Lan at microsoft.com>
> > ---
> >  arch/x86/kvm/paging_tmpl.h | 16 ++++++++++++++--
> >  1 file changed, 14 insertions(+), 2 deletions(-)
> >
> > diff --git a/arch/x86/kvm/paging_tmpl.h b/arch/x86/kvm/paging_tmpl.h
> > index 833e8855bbc9..866ccdea762e 100644
> > --- a/arch/x86/kvm/paging_tmpl.h
> > +++ b/arch/x86/kvm/paging_tmpl.h
> > @@ -973,6 +973,7 @@ static int FNAME(sync_page)(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp)
> >       bool host_writable;
> >       gpa_t first_pte_gpa;
> >       int set_spte_ret = 0;
> > +     LIST_HEAD(flush_list);
> >
> >       /* direct kvm_mmu_page can not be unsync. */
> >       BUG_ON(sp->role.direct);
> > @@ -980,6 +981,7 @@ static int FNAME(sync_page)(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp)
> >       first_pte_gpa = FNAME(get_level1_sp_gpa)(sp);
> >
> >       for (i = 0; i < PT64_ENT_PER_PAGE; i++) {
> > +             int tmp_spte_ret = 0;
> >               unsigned pte_access;
> >               pt_element_t gpte;
> >               gpa_t pte_gpa;
> > @@ -1029,14 +1031,24 @@ static int FNAME(sync_page)(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp)
> >
> >               host_writable = sp->spt[i] & SPTE_HOST_WRITEABLE;
> >
> > -             set_spte_ret |= set_spte(vcpu, &sp->spt[i],
> > +             tmp_spte_ret = set_spte(vcpu, &sp->spt[i],
> >                                        pte_access, PT_PAGE_TABLE_LEVEL,
> >                                        gfn, spte_to_pfn(sp->spt[i]),
> >                                        true, false, host_writable);
> > +
> > +             if (kvm_available_flush_tlb_with_range()
> > +                 && (tmp_spte_ret & SET_SPTE_NEED_REMOTE_TLB_FLUSH)) {
> > +                     struct kvm_mmu_page *leaf_sp = page_header(sp->spt[i]
> > +                                     & PT64_BASE_ADDR_MASK);
> > +                     list_add(&leaf_sp->flush_link, &flush_list);
> > +             }
> > +
> > +             set_spte_ret |= tmp_spte_ret;
> > +
> >       }
> >
> >       if (set_spte_ret & SET_SPTE_NEED_REMOTE_TLB_FLUSH)
> > -             kvm_flush_remote_tlbs(vcpu->kvm);
> > +             kvm_flush_remote_tlbs_with_list(vcpu->kvm, &flush_list);
>
> This is a bit confusing and potentially fragile.  It's not obvious that
> kvm_flush_remote_tlbs_with_list() is guaranteed to call
> kvm_flush_remote_tlbs() when kvm_available_flush_tlb_with_range() is
> false, and you're relying on the kvm_flush_remote_tlbs_with_list() call
> chain to never optimize away the empty list case.  Rechecking
> kvm_available_flush_tlb_with_range() isn't expensive.

That makes sense. Will update. Thanks.

>
> >
> >       return nr_present;
> >  }
> > --
> > 2.14.4
> >



-- 
Best regards
Tianyu Lan