Re: [Qemu-devel] [PATCH v4 0/6] save/restore on Xen

[Avi Kivity]

On 01/24/2012 02:00 PM, Stefano Stabellini wrote:
> On Tue, 24 Jan 2012, Avi Kivity wrote:
> > On 01/24/2012 12:21 PM, Gerd Hoffmann wrote:
> > > >>>
> > > >>> But viewing RAM as just another device, having Xen only restore a
> > > >>> subset of
> > > >>> devices should be a reasonable thing to do moving forward.
> > >
> > > I don't think modeling device memory (i.e. vga vram) as something
> > > independent from the device (vga) is a good idea.  Because it isn't.
> > 
> > Right.  We should have VMSTATE_RAM() to express the dependency.
>
> So if I understand correctly you are planning on removing the call to
> vmstate_register_ram_global in hw/vga.c?

I can't say I'm planning to do so, but that's the direction I think we
need to go.

> In that case it might be better if I wait for your "VMSTATE_RAM" patch
> before I can send a new patch to save only non-RAM devices, because I
> guess they will clash with one another.

I don't understand what sense a patch to save non-RAM devices makes (or
even, what non-RAM devices are).

We have a special case here where Xen manages the RAM even though qemu
thinks it does.  IMO an if (xen) is the best way to express this
specialness.

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