On Mon, Oct 16, 2023, Dongli Zhang wrote:
> Hi Sean,
>
> On 10/16/23 11:49, Sean Christopherson wrote:
> > Compile tested only, but the below should fix the vCPU hotplug case. Then
> > someone (not me) just needs to figure out why kvm_xen_shared_info_init() forces
> > a masterclock update.
> >
> > I still think we should clean up the periodic sync code, but I don't think we
> > need to periodically sync the masterclock.
>
> This looks good to me. The core idea is to not update master clock for the
> synchronized cases.
>
>
> How about the negative value case? I see in the linux code it is still there?
See below.
> (It is out of the scope of my expectation as I do not need to run vCPUs in
> different tsc freq as host)
>
> Thank you very much!
>
> Dongli Zhang
>
> >
> > ---
> > arch/x86/kvm/x86.c | 29 ++++++++++++++++-------------
> > 1 file changed, 16 insertions(+), 13 deletions(-)
> >
> > diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> > index c54e1133e0d3..f0a607b6fc31 100644
> > --- a/arch/x86/kvm/x86.c
> > +++ b/arch/x86/kvm/x86.c
> > @@ -2510,26 +2510,29 @@ static inline int gtod_is_based_on_tsc(int mode)
> > }
> > #endif
> >
> > -static void kvm_track_tsc_matching(struct kvm_vcpu *vcpu)
> > +static void kvm_track_tsc_matching(struct kvm_vcpu *vcpu, bool new_generation)
> > {
> > #ifdef CONFIG_X86_64
> > - bool vcpus_matched;
> > struct kvm_arch *ka = &vcpu->kvm->arch;
> > struct pvclock_gtod_data *gtod = &pvclock_gtod_data;
> >
> > - vcpus_matched = (ka->nr_vcpus_matched_tsc + 1 ==
> > - atomic_read(&vcpu->kvm->online_vcpus));
> > + /*
> > + * To use the masterclock, the host clocksource must be based on TSC
> > + * and all vCPUs must have matching TSCs. Note, the count for matching
> > + * vCPUs doesn't include the reference vCPU, hence "+1".
> > + */
> > + bool use_master_clock = (ka->nr_vcpus_matched_tsc + 1 ==
> > + atomic_read(&vcpu->kvm->online_vcpus)) &&
> > + gtod_is_based_on_tsc(gtod->clock.vclock_mode);
> >
> > /*
> > - * Once the masterclock is enabled, always perform request in
> > - * order to update it.
> > - *
> > - * In order to enable masterclock, the host clocksource must be TSC
> > - * and the vcpus need to have matched TSCs. When that happens,
> > - * perform request to enable masterclock.
> > + * Request a masterclock update if the masterclock needs to be toggled
> > + * on/off, or when starting a new generation and the masterclock is
> > + * enabled (compute_guest_tsc() requires the masterclock snaphot to be
> > + * taken _after_ the new generation is created).
> > */
> > - if (ka->use_master_clock ||
> > - (gtod_is_based_on_tsc(gtod->clock.vclock_mode) && vcpus_matched))
> > + if ((ka->use_master_clock && new_generation) ||
> > + (ka->use_master_clock != use_master_clock))
> > kvm_make_request(KVM_REQ_MASTERCLOCK_UPDATE, vcpu);
> >
> > trace_kvm_track_tsc(vcpu->vcpu_id, ka->nr_vcpus_matched_tsc,
> > @@ -2706,7 +2709,7 @@ static void __kvm_synchronize_tsc(struct kvm_vcpu *vcpu, u64 offset, u64 tsc,
> > vcpu->arch.this_tsc_nsec = kvm->arch.cur_tsc_nsec;
> > vcpu->arch.this_tsc_write = kvm->arch.cur_tsc_write;
> >
> > - kvm_track_tsc_matching(vcpu);
> > + kvm_track_tsc_matching(vcpu, !matched);
If my analysis of how the negative timestamp occurred is correct, the problematic
scenario was if cur_tsc_nsec/cur_tsc_write were updated without a masterclock update.
Passing !matched for @new_generation means that KVM will force a masterclock update
if cur_tsc_nsec/cur_tsc_write are changed, i.e. prevent the negative timestamp bug.
