Re: [PATCH v5 3/7] KVM: x86: Allow to use bitmap in ring-based dirty page tracking

[Peter Xu <peterx@redhat.com>]

On Wed, Oct 05, 2022 at 08:41:50AM +0800, Gavin Shan wrote:
> -8.29 KVM_CAP_DIRTY_LOG_RING/KVM_CAP_DIRTY_LOG_RING_ACQ_REL
> -----------------------------------------------------------
> +8.29 KVM_CAP_DIRTY_LOG_{RING, RING_ACQ_REL, RING_ALLOW_BITMAP}
> +--------------------------------------------------------------

Shall we make it a standalone cap, just to rely on DIRTY_RING[_ACQ_REL]
being enabled first, instead of making the three caps at the same level?

E.g. we can skip creating bitmap for DIRTY_RING[_ACQ_REL] && !_ALLOW_BITMAP
(x86).

> @@ -2060,10 +2060,6 @@ int kvm_get_dirty_log(struct kvm *kvm, struct kvm_dirty_log *log,
>  	unsigned long n;
>  	unsigned long any = 0;
>  
> -	/* Dirty ring tracking is exclusive to dirty log tracking */
> -	if (kvm->dirty_ring_size)
> -		return -ENXIO;

Then we can also have one dirty_ring_exclusive(), with something like:

bool dirty_ring_exclusive(struct kvm *kvm)
{
        return kvm->dirty_ring_size && !kvm->dirty_ring_allow_bitmap;
}

Does it make sense?  Thanks,

-- 
Peter Xu