Re: [PATCH] KVM: s390x: add debug statement for diag 318 CPNC data

[Janosch Frank]

On 11/8/21 13:50, Christian Borntraeger wrote:
> Am 08.11.21 um 13:48 schrieb Janosch Frank:
> > On 11/8/21 13:04, Christian Borntraeger wrote:
> > > Am 08.11.21 um 12:12 schrieb Janosch Frank:
> > > > On 10/27/21 04:54, Collin Walling wrote:
> > > > > The diag 318 data contains values that denote information regarding the
> > > > > guest's environment. Currently, it is unecessarily difficult to observe
> > > > > this value (either manually-inserted debug statements, gdb stepping, mem
> > > > > dumping etc). It's useful to observe this information to obtain an
> > > > > at-a-glance view of the guest's environment, so lets add a simple VCPU
> > > > > event that prints the CPNC to the s390dbf logs.
> > > > >
> > > > > Signed-off-by: Collin Walling <walling@xxxxxxxxxxxxx>
> > > > > ---
> > > > >     arch/s390/kvm/kvm-s390.c | 1 +
> > > > >     1 file changed, 1 insertion(+)
> > > > >
> > > > > diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> > > > > index 6a6dd5e1daf6..da3ff24eabd0 100644
> > > > > --- a/arch/s390/kvm/kvm-s390.c
> > > > > +++ b/arch/s390/kvm/kvm-s390.c
> > > > > @@ -4254,6 +4254,7 @@ static void sync_regs_fmt2(struct kvm_vcpu *vcpu)
> > > > >         if (kvm_run->kvm_dirty_regs & KVM_SYNC_DIAG318) {
> > > > >             vcpu->arch.diag318_info.val = kvm_run->s.regs.diag318;
> > > > >             vcpu->arch.sie_block->cpnc = vcpu->arch.diag318_info.cpnc;
> > > > > +        VCPU_EVENT(vcpu, 2, "setting cpnc to %d", vcpu->arch.diag318_info.cpnc);
> > > > >         }
> > > > >         /*
> > > > >          * If userspace sets the riccb (e.g. after migration) to a valid state,
> > > >
> > > > Won't that turn up for every vcpu and spam the log?
> > >
> > > only if the userspace always sets the dirty bit (which it should not).
> >
> > But that's exactly what it does, no?
> > We do a loop over all vcpus and call kvm_s390_set_diag318() which sets the info in kvm_run and sets the diag318 bit in the kvm_dirty_regs.
>
> Yes, ONCE per CPU. And this is exactly what I want to see. (and it did show a bug in qemu that we only set it for one cpu to the correct value).

Ok.
I didn't really want to have n entries in the log hence I asked.

318 is a bit weird as it's a per VM value we need to put into all sie 
blocks.

> > @Collin: Could you check that please?