Re: [PATCH v7 00/11] KVM: xen: update shared_info and vcpu_info handling

[Paul Durrant]

On 09/11/2023 10:02, David Woodhouse wrote:
> On Thu, 2023-10-05 at 09:36 +0100, Paul Durrant wrote:
> > On 05/10/2023 07:41, David Woodhouse wrote:
> > > On Mon, 2023-10-02 at 09:57 +0000, Paul Durrant wrote:
> > > > From: Paul Durrant <pdurrant@xxxxxxxxxx>
> > > >
> > > > The following text from the original cover letter still serves as an
> > > > introduction to the series:
> > > >
> > > > "Currently we treat the shared_info page as guest memory and the VMM
> > > > informs KVM of its location using a GFN. However it is not guest memory as
> > > > such; it's an overlay page. So we pointlessly invalidate and re-cache a
> > > > mapping to the *same page* of memory every time the guest requests that
> > > > shared_info be mapped into its address space. Let's avoid doing that by
> > > > modifying the pfncache code to allow activation using a fixed userspace HVA
> > > > as well as a GPA."
> > > >
> > > > This version of the series is functionally the same as version 6. I have
> > > > simply added David Woodhouse's R-b to patch 11 to indicate that he has
> > > > now fully reviewed the series.
> > >
> > > Thanks. I believe Sean is probably waiting for us to stop going back
> > > and forth, and for the dust to settle. So for the record: I think I'm
> > > done heckling and this is ready to go in.
> > >
> > > Are you doing the QEMU patches or am I?
> >
> > I'll do the QEMU changes, once the patches hit kvm/next.
>
> Note that I disabled migration support in QEMU for emulated Xen
> guests. You might want that for testing, since the reason for this work
> is to enable pause/serialize workflows.
>
> Migration does work all the way up to XenStore itself, and
> https://gitlab.com/qemu-project/qemu/-/commit/766804b101d *was* tested
> with migration enabled. There are also unit tests for XenStore
> serialize/deserialize.
>
> I disabled it because the PV backends on the XenBus don't have
> suspend/resume support. But a guest using other emulated net/disk
> devices should still be able to suspend/resume OK if we just remove the
> 'unmigratable' flag from xen_xenstore, I believe.

Ok. Enabling suspend/resume for backends really ought not to be that 
hard. The main reason for this series was to enable pause 
for-for-memory-reconfiguration but I can look into 
suspend/resume/migrate once I've done the necessary re-work.