Re: [PATCH v11 10/11] KVM: x86/pmu: Check guest LBR availability in case host reclaims them

[Xu, Like]

On 2020/5/19 19:15, Peter Zijlstra wrote:
> On Thu, May 14, 2020 at 04:30:53PM +0800, Like Xu wrote:
>
> > diff --git a/arch/x86/kvm/vmx/pmu_intel.c b/arch/x86/kvm/vmx/pmu_intel.c
> > index ea4faae56473..db185dca903d 100644
> > --- a/arch/x86/kvm/vmx/pmu_intel.c
> > +++ b/arch/x86/kvm/vmx/pmu_intel.c
> > @@ -646,6 +646,43 @@ static void intel_pmu_lbr_cleanup(struct kvm_vcpu *vcpu)
> >   		intel_pmu_free_lbr_event(vcpu);
> >   }
> >   
> > +static bool intel_pmu_lbr_is_availabile(struct kvm_vcpu *vcpu)
> > +{
> > +	struct kvm_pmu *pmu = vcpu_to_pmu(vcpu);
> > +
> > +	if (!pmu->lbr_event)
> > +		return false;
> > +
> > +	if (event_is_oncpu(pmu->lbr_event)) {
> > +		intel_pmu_intercept_lbr_msrs(vcpu, false);
> > +	} else {
> > +		intel_pmu_intercept_lbr_msrs(vcpu, true);
> > +		return false;
> > +	}
> > +
> > +	return true;
> > +}
> This is unreadable gunk, what?

Abstractly, it is saying "KVM would passthrough the LBR satck MSRs if
event_is_oncpu() is true, otherwise cancel the passthrough state if any."

I'm using 'event->oncpu != -1' to represent the guest LBR event
is scheduled on rather than 'event->state == PERF_EVENT_STATE_ERROR'.

For intel_pmu_intercept_lbr_msrs(), false means to passthrough the LBR stack
MSRs to the vCPU, and true means to cancel the passthrough state and make
LBR MSR accesses trapped by the KVM.

> > +/*
> > + * Higher priority host perf events (e.g. cpu pinned) could reclaim the
> > + * pmu resources (e.g. LBR) that were assigned to the guest. This is
> > + * usually done via ipi calls (more details in perf_install_in_context).
> > + *
> > + * Before entering the non-root mode (with irq disabled here), double
> > + * confirm that the pmu features enabled to the guest are not reclaimed
> > + * by higher priority host events. Otherwise, disallow vcpu's access to
> > + * the reclaimed features.
> > + */
> > +static void intel_pmu_availability_check(struct kvm_vcpu *vcpu)
> > +{
> > +	lockdep_assert_irqs_disabled();
> > +
> > +	if (lbr_is_enabled(vcpu) && !intel_pmu_lbr_is_availabile(vcpu) &&
> > +		(vmcs_read64(GUEST_IA32_DEBUGCTL) & DEBUGCTLMSR_LBR))
> > +		pr_warn_ratelimited("kvm: vcpu-%d: LBR is temporarily unavailable.\n",
> > +			vcpu->vcpu_id);
> More unreadable nonsense; when the events go into ERROR state, it's a
> permanent fail, they'll not come back.
It's not true. The guest LBR event with 'ERROR state' or 'oncpu != -1' 
would be
lazy released and re-created in the next time the 
intel_pmu_create_lbr_event() is
called and it's supposed to be re-scheduled and re-do availability_check() 
as well.

From the perspective of the guest user, the guest LBR is only temporarily 
unavailable
until the host no longer reclaims the LBR.
> > +}
> > +
> >   struct kvm_pmu_ops intel_pmu_ops = {
> >   	.find_arch_event = intel_find_arch_event,
> >   	.find_fixed_event = intel_find_fixed_event,
> > @@ -662,4 +699,5 @@ struct kvm_pmu_ops intel_pmu_ops = {
> >   	.reset = intel_pmu_reset,
> >   	.deliver_pmi = intel_pmu_deliver_pmi,
> >   	.lbr_cleanup = intel_pmu_lbr_cleanup,
> > +	.availability_check = intel_pmu_availability_check,
> >   };
> > diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c
> > index 9969d663826a..80d036c5f64a 100644
> > --- a/arch/x86/kvm/vmx/vmx.c
> > +++ b/arch/x86/kvm/vmx/vmx.c
> > @@ -6696,8 +6696,10 @@ static fastpath_t vmx_vcpu_run(struct kvm_vcpu *vcpu)
> >   
> >   	pt_guest_enter(vmx);
> >   
> > -	if (vcpu_to_pmu(vcpu)->version)
> > +	if (vcpu_to_pmu(vcpu)->version) {
> >   		atomic_switch_perf_msrs(vmx);
> > +		kvm_x86_ops.pmu_ops->availability_check(vcpu);
> > +	}
> AFAICT you just did a call out to the kvm_pmu crud in
> atomic_switch_perf_msrs(), why do another call?
In fact, availability_check() is only called here for just one time.

The callchain looks like:
- vmx_vcpu_run()
ÂÂÂ - kvm_x86_ops.pmu_ops->availability_check();
ÂÂÂ ÂÂÂ - intel_pmu_availability_check()
ÂÂÂ ÂÂÂ ÂÂÂ - intel_pmu_lbr_is_availabile()
ÂÂÂ ÂÂÂ ÂÂÂ ÂÂÂ - event_is_oncpu() ...

Thanks,
Like Xu
>