Re: [PATCH v7 07/12] KVM: arm64: PMU: Set PMCR_EL0.N for vCPU based on the associated PMU

[Raghavendra Rao Ananta]

On Mon, Oct 16, 2023 at 12:16 PM Oliver Upton <oliver.upton@xxxxxxxxx> wrote:
>
> On Mon, Oct 16, 2023 at 12:02:27PM -0700, Raghavendra Rao Ananta wrote:
> > On Mon, Oct 16, 2023 at 6:35 AM Sebastian Ott <sebott@xxxxxxxxxx> wrote:
> > >
> > > On Mon, 9 Oct 2023, Raghavendra Rao Ananta wrote:
> > > > u64 kvm_vcpu_read_pmcr(struct kvm_vcpu *vcpu)
> > > > {
> > > > -     return __vcpu_sys_reg(vcpu, PMCR_EL0);
> > > > +     u64 pmcr = __vcpu_sys_reg(vcpu, PMCR_EL0) &
> > > > +                     ~(ARMV8_PMU_PMCR_N_MASK << ARMV8_PMU_PMCR_N_SHIFT);
> > > > +
> > > > +     return pmcr | ((u64)vcpu->kvm->arch.pmcr_n << ARMV8_PMU_PMCR_N_SHIFT);
> > > > }
> > > > diff --git a/arch/arm64/kvm/sys_regs.c b/arch/arm64/kvm/sys_regs.c
> > > > index ff0f7095eaca..c750722fbe4a 100644
> > > > --- a/arch/arm64/kvm/sys_regs.c
> > > > +++ b/arch/arm64/kvm/sys_regs.c
> > > > @@ -745,12 +745,8 @@ static u64 reset_pmcr(struct kvm_vcpu *vcpu, const struct sys_reg_desc *r)
> > > > {
> > > >       u64 pmcr;
> > > >
> > > > -     /* No PMU available, PMCR_EL0 may UNDEF... */
> > > > -     if (!kvm_arm_support_pmu_v3())
> > > > -             return 0;
> > > > -
> > > >       /* Only preserve PMCR_EL0.N, and reset the rest to 0 */
> > > > -     pmcr = read_sysreg(pmcr_el0) & (ARMV8_PMU_PMCR_N_MASK << ARMV8_PMU_PMCR_N_SHIFT);
> > > > +     pmcr = kvm_vcpu_read_pmcr(vcpu) & (ARMV8_PMU_PMCR_N_MASK << ARMV8_PMU_PMCR_N_SHIFT);
> > >
> > > pmcr = ((u64)vcpu->kvm->arch.pmcr_n << ARMV8_PMU_PMCR_N_SHIFT);
> > > Would that maybe make it more clear what is done here?
> > >
> > Since we require the entire PMCR register, and not just the PMCR.N
> > field, I think using kvm_vcpu_read_pmcr() would be technically
> > correct, don't you think?
>
> No, this isn't using the entire PMCR value, it is just grabbing
> PMCR_EL0.N.
>
Oh sorry, my bad.
> What's the point of doing this in the first place? The implementation of
> kvm_vcpu_read_pmcr() is populating PMCR_EL0.N using the VM-scoped value.
>
I guess originally the change replaced read_sysreg(pmcr_el0) with
kvm_vcpu_read_pmcr(vcpu) to maintain consistency with others.
But if you and Sebastian feel that it's an overkill and directly
getting the value via vcpu->kvm->arch.pmcr_n is more readable, I'm
happy to make the change.

Thank you.
Raghavendra
> --
> Thanks,
> Oliver