Re: [Qemu-block] [Qemu-devel] [PATCH v2 1/1] virtio-scsi-ccw: use ioeventfd even when KVM is disabled

[Cornelia Huck]

On Mon, 3 Jul 2017 12:20:03 +0200
Christian Borntraeger <address@hidden> wrote:

> On 07/03/2017 10:51 AM, QingFeng Hao wrote:
>
> > diff --git a/target/s390x/kvm.c b/target/s390x/kvm.c
> > index a3d00196f4..c37f9c3b9e 100644
> > --- a/target/s390x/kvm.c
> > +++ b/target/s390x/kvm.c
> > @@ -2220,6 +2220,9 @@ int kvm_s390_assign_subch_ioeventfd(EventNotifier 
> > *notifier, uint32_t sch,
> >          .addr = sch,
> >          .len = 8,
> >      };
> > +    if (!kvm_enabled()) {
> > +        return 0;
> > +    }  
> 
> thinking more about it. wouldnt it be better to do something like this instead
> 
> diff --git a/target/s390x/cpu.h b/target/s390x/cpu.h
> index 058ddad..cc47831 100644
> --- a/target/s390x/cpu.h
> +++ b/target/s390x/cpu.h
> @@ -1240,7 +1240,11 @@ static inline int 
> s390_assign_subch_ioeventfd(EventNotifier *notifier,
>                                                uint32_t sch_id, int vq,
>                                                bool assign)
>  {
> -    return kvm_s390_assign_subch_ioeventfd(notifier, sch_id, vq, assign);
> +    if (kvm_enabled()) {
> +        return kvm_s390_assign_subch_ioeventfd(notifier, sch_id, vq, assign);
> +    } else {
> +        return 0;
> +    }
>  }
>  
>  static inline void s390_crypto_reset(void)

Agreed.

> 
> 
> FWIW, it seems that we (s390) do not have a functional equivalent function as 
> commit
> 8c56c1a592b5092d91da8d8943c17777d6462a6f ("memory: emulate ioeventfd") , so 
> we will
> not use the iothreads. 

Should not be a problem, I think. We may want to cook up an equivalent,
though.

> 
> 
> >      if (!kvm_check_extension(kvm_state, KVM_CAP_IOEVENTFD)) {
> >          return -ENOSYS;
> >      }
> >   
> 
>