On Wed, 5 Feb 2020 19:18:44 +0100
Christian Borntraeger <borntraeger@xxxxxxxxxx> wrote:
> On 05.02.20 14:47, Cornelia Huck wrote:
> [..]
> >> --- a/arch/s390/kvm/interrupt.c
> >> +++ b/arch/s390/kvm/interrupt.c
> >> @@ -571,6 +571,14 @@ static int __write_machine_check(struct kvm_vcpu *vcpu,
> >> union mci mci;
> >> int rc;
> >>
> >> + if (kvm_s390_pv_is_protected(vcpu->kvm)) {
> >> + vcpu->arch.sie_block->iictl = IICTL_CODE_MCHK;
> >> + vcpu->arch.sie_block->mcic = mchk->mcic;
> >> + vcpu->arch.sie_block->faddr = mchk->failing_storage_address;
> >> + vcpu->arch.sie_block->edc = mchk->ext_damage_code;
> >
> > Maybe add a comment that we don't need with other machine-check related data?
>
> Not sure I get this point. Can you make a proposal?
/*
* All other possible payload for a machine check will
* not be handled by the hypervisor, as it does not have
* the needed information for protected guests.
*/
Something like that?
>
> >
> >> + return 0;
> >> + }
> >> +
> >> mci.val = mchk->mcic;
> >> /* take care of lazy register loading */
> >> save_fpu_regs();
> >
> > Anyway,
> > Reviewed-by: Cornelia Huck <cohuck@xxxxxxxxxx>
> >
>
