On Wed May 8, 2024 at 10:36 PM AEST, Michael Ellerman wrote:
> Gautam Menghani <gautam@xxxxxxxxxxxxx> writes:
> > PAPR hypervisor has introduced three new counters in the VPA area of
> > LPAR CPUs for KVM L2 guest (see [1] for terminology) observability - 2
> > for context switches from host to guest and vice versa, and 1 counter
> > for getting the total time spent inside the KVM guest. Add a tracepoint
> > that enables reading the counters for use by ftrace/perf. Note that this
> > tracepoint is only available for nestedv2 API (i.e, KVM on PowerVM).
> ...
> > diff --git a/arch/powerpc/kvm/book3s_hv.c b/arch/powerpc/kvm/book3s_hv.c
> > index 8e86eb577eb8..ed69ad58bd02 100644
> > --- a/arch/powerpc/kvm/book3s_hv.c
> > +++ b/arch/powerpc/kvm/book3s_hv.c
> > @@ -4108,6 +4108,54 @@ static void vcpu_vpa_increment_dispatch(struct kvm_vcpu *vcpu)
> > }
> > }
> >
> > +static inline int kvmhv_get_l2_counters_status(void)
> > +{
> > + return get_lppaca()->l2_counters_enable;
> > +}
>
> This is breaking the powernv build:
[...]
All the nested KVM code should really go under CONFIG_PSERIES.
Possibly even moved out to its own file.
For now maybe you could just ifdef these few functions and
replace with noop variants for !PSERIES.
Thanks,
Nick
