Re: [PATCH 2/7] kvm-vfio: Decouple only when we match a group

[Paolo Bonzini]

On 09/06/2017 23:59, Alex Williamson wrote:
> Unset-KVM and decrement-assignment only when we find the group in our
> list.  Otherwise we can get out of sync if the user triggers this for
> groups that aren't currently on our list.
> 
> Signed-off-by: Alex Williamson <alex.williamson@xxxxxxxxxx>
> Cc: Paolo Bonzini <pbonzini@xxxxxxxxxx>
> ---
>  virt/kvm/vfio.c |    6 ++----
>  1 file changed, 2 insertions(+), 4 deletions(-)
> 
> diff --git a/virt/kvm/vfio.c b/virt/kvm/vfio.c
> index d32f239eb471..db9036ef8c73 100644
> --- a/virt/kvm/vfio.c
> +++ b/virt/kvm/vfio.c
> @@ -201,18 +201,16 @@ static int kvm_vfio_set_group(struct kvm_device *dev, long attr, u64 arg)
>  				continue;
>  
>  			list_del(&kvg->node);
> +			kvm_arch_end_assignment(dev->kvm);
> +			kvm_vfio_group_set_kvm(kvg->vfio_group, NULL);
>  			kvm_vfio_group_put_external_user(kvg->vfio_group);
>  			kfree(kvg);
>  			ret = 0;
>  			break;
>  		}
>  
> -		kvm_arch_end_assignment(dev->kvm);
> -
>  		mutex_unlock(&kv->lock);
>  
> -		kvm_vfio_group_set_kvm(vfio_group, NULL);
> -
>  		kvm_vfio_group_put_external_user(vfio_group);
>  
>  		kvm_vfio_update_coherency(dev);
> 

Reviewed-by: Paolo Bonzini <pbonzini@xxxxxxxxxx>

Should these two patches be applied to stable kernels too?  In any case,
please take care your self of getting these to Linus!

Thanks,

Paolo