On Mon, Mar 03, 2025 at 05:11:31PM +0100, Paolo Bonzini wrote:
> On 3/3/25 02:23, Yan Zhao wrote:
> > On Sat, Mar 01, 2025 at 02:34:26AM -0500, Paolo Bonzini wrote:
> > > From: Yan Zhao <yan.y.zhao@xxxxxxxxx>
> > >
> > > Introduce supported_quirks in kvm_caps to store platform-specific force-enabled
> > > quirks. Any quirk removed from kvm_caps.supported_quirks will never be
> > > included in kvm->arch.disabled_quirks, and will cause the ioctl to fail if
> > > passed to KVM_ENABLE_CAP(KVM_CAP_DISABLE_QUIRKS2).
> > >
> > > Signed-off-by: Yan Zhao <yan.y.zhao@xxxxxxxxx>
> > > Message-ID: <20250224070832.31394-1-yan.y.zhao@xxxxxxxxx>
> > > Signed-off-by: Paolo Bonzini <pbonzini@xxxxxxxxxx>
> > > ---
> > > arch/x86/kvm/x86.c | 7 ++++---
> > > arch/x86/kvm/x86.h | 2 ++
> > > 2 files changed, 6 insertions(+), 3 deletions(-)
> > >
> > > diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> > > index fd0a44e59314..a97e58916b6a 100644
> > > --- a/arch/x86/kvm/x86.c
> > > +++ b/arch/x86/kvm/x86.c
> > > @@ -4782,7 +4782,7 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext)
> > > r = enable_pmu ? KVM_CAP_PMU_VALID_MASK : 0;
> > > break;
> > > case KVM_CAP_DISABLE_QUIRKS2:
> > > - r = KVM_X86_VALID_QUIRKS;
> > > + r = kvm_caps.supported_quirks;
> >
> > As the concern raised in [1], it's confusing for
> > KVM_X86_QUIRK_EPT_IGNORE_GUEST_PAT to be present on AMD's platforms while not
> > present on Intel's non-self-snoop platforms.
>
> To make it less confusing, let's rename it to
> KVM_X86_QUIRK_IGNORE_GUEST_PAT. So if userspace wants to say "I don't want
Hmm, it looks that quirk IGNORE_GUEST_PAT is effectively always enabled on Intel
platforms without enabling EPT.
And kvm_mmu_may_ignore_guest_pat() does not care quirk IGNORE_GUEST_PAT on AMD
or on Intel when enable_ept is 0.
bool kvm_mmu_may_ignore_guest_pat(struct kvm *kvm)
{
return shadow_memtype_mask &&
kvm_check_has_quirk(kvm, KVM_X86_QUIRK_EPT_IGNORE_GUEST_PAT);
}
After renaming the quirk, should we also hide quirk IGNORE_GUEST_PAT from being
exposed in KVM_CAP_DISABLE_QUIRKS2 when enable_ept is 0?
However, doing so will complicate kvm_noncoherent_dma_assignment_start_or_stop().
> KVM to ignore guest's PAT!", it can do so easily:
>
> - it can check unconditionally that the quirk is included in
> KVM_CAP_DISABLE_QUIRKS2, and it will pass on both Intel with self-snoop with
> AMD;
So, when userspace finds the quirk IGNORE_GUEST_PAT is exposed in
KVM_CAP_DISABLE_QUIRKS2, it cannot treat this as an implication that KVM is
currently ignoring guest PAT, but it only means that this is a new KVM with
quirk IGNORE_GUEST_PAT taken into account.
> - it can pass it unconditionally to KVM_X86_QUIRK_IGNORE_GUEST_PAT, knowing
> that PAT will be honored.
For AMD, since userspace does not explicitly disable quirk IGNORE_GUEST_PAT, you
introduced kvm_caps.inapplicable_quirks to allow IGNORE_GUEST_PAT to be listed
in KVM_CAP_DISABLED_QUIRKS.
