2021-09-16 14:08:29

by David Hildenbrand

Subject: Re: [PATCH v4 1/1] s390x: KVM: accept STSI for CPU topology information

> struct kvm_vm_stat {
> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> index 51d1594bd6cd..f3887e13c5db 100644
> --- a/arch/s390/kvm/kvm-s390.c
> +++ b/arch/s390/kvm/kvm-s390.c
> @@ -608,6 +608,9 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext)
> case KVM_CAP_S390_PROTECTED:
> r = is_prot_virt_host();
> break;
> + case KVM_CAP_S390_CPU_TOPOLOGY:
> + r = test_facility(11);
> + break;
> default:
> r = 0;
> }
> @@ -819,6 +822,19 @@ int kvm_vm_ioctl_enable_cap(struct kvm *kvm, struct kvm_enable_cap *cap)
> icpt_operexc_on_all_vcpus(kvm);
> r = 0;
> break;
> + case KVM_CAP_S390_CPU_TOPOLOGY:

As given in my example, this should be

r = -EINVAL;
mutex_lock(&kvm->lock);
if (kvm->created_vcpus) {
r = -EBUSY;
} else if (test_facility(11)) {
...
}

Similar to how we handle KVM_CAP_S390_VECTOR_REGISTERS.

[...]

> +
> + /* PTF needs both host and guest facilities to enable interpretation */
> + if (test_kvm_facility(vcpu->kvm, 11) && test_facility(11))
> + vcpu->arch.sie_block->ecb |= ECB_PTF;

This should be simplified to

if (test_kvm_facility(vcpu->kvm, 11))

then. (vsie code below is correct)


--
Thanks,

David / dhildenb

2021-09-17 13:13:48

by Pierre Morel

Subject: Re: [PATCH v4 1/1] s390x: KVM: accept STSI for CPU topology information

On 9/16/21 4:03 PM, David Hildenbrand wrote:
>>   struct kvm_vm_stat {
>> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
>> index 51d1594bd6cd..f3887e13c5db 100644
>> --- a/arch/s390/kvm/kvm-s390.c
>> +++ b/arch/s390/kvm/kvm-s390.c
>> @@ -608,6 +608,9 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm,
>> long ext)
>>       case KVM_CAP_S390_PROTECTED:
>>           r = is_prot_virt_host();
>>           break;
>> +    case KVM_CAP_S390_CPU_TOPOLOGY:
>> +        r = test_facility(11);
>> +        break;
>>       default:
>>           r = 0;
>>       }
>> @@ -819,6 +822,19 @@ int kvm_vm_ioctl_enable_cap(struct kvm *kvm,
>> struct kvm_enable_cap *cap)
>>           icpt_operexc_on_all_vcpus(kvm);
>>           r = 0;
>>           break;
>> +    case KVM_CAP_S390_CPU_TOPOLOGY:
>
> As given in my example, this should be
>
> r = -EINVAL;
> mutex_lock(&kvm->lock);
> if (kvm->created_vcpus) {
>     r = -EBUSY;
> } else if (test_facility(11)) {
> ...
> }
>
> Similar to how we handle KVM_CAP_S390_VECTOR_REGISTERS.
>
> [...]
>
>> +
>> +    /* PTF needs both host and guest facilities to enable
>> interpretation */
>> +    if (test_kvm_facility(vcpu->kvm, 11) && test_facility(11))
>> +        vcpu->arch.sie_block->ecb |= ECB_PTF;
>
> This should be simplified to
>
> if (test_kvm_facility(vcpu->kvm, 11))
>
> then. (vsie code below is correct)
>
>

OK, the idea was to let the hypervisor the possibility to do userland
emulation if it wanted.
But I can modify as you want.

Regards,
Pierre

--
Pierre Morel
IBM Lab Boeblingen