Re: [Qemu-devel] [PATCH 00/10] xen: pv domain support.

[Avi Kivity]

Alexander Graf wrote:
> > > 1) Migration path. If you could already use KVM on a Xen host, you 
> > > could have Xen PV guests and KVM guests in parallel, easing 
> > > migration to KVM for customers.
> >
> > I like this, of course, but we have a path through Xenner.  Maybe 
> > this (kvm-on-xen) path will be easier to take.
>
> Oh, I thought Xenner was semi on hold? Is it still actively developed? 
> I thought Gerd's main goal for now was to get qemu-dm into upstream 
> qemu for now.

I understood that Xenner is the next thing on the list.

> > With npt/ept pv performance might be higher running under kvm+xenner 
> > than with software-only Xen by letting the guest kernel access 
> > pagetables directly.  Though Gerd had some issues with 64-bit guests 
> > IIRC, which is a pity since it's there that the pv performance hit is 
> > greatest.
>
> Well that might be true, but I'm not fully convinced yet :-). Either 
> way with "current" hardware (pre-Nehalem, pre-Barcelona) PV is still 
> faster.

Right, but every day this is becoming less important.

> Also, FWIW IO performance on PV Xen is still way superior to KVM. 
> Please correct me if I stand wrong there.

We still have some work some work to do on virtio, but bandwidth should 
be comparable.  And it has nothing to do with pv-vs-fv.

--
error compiling committee.c: too many arguments to function