On Sun, Jun 14, 2026 at 12:50:14PM +0300, Avihai Horon wrote:
>
> On 6/12/2026 7:16 PM, Peter Xu wrote:
> > External email: Use caution opening links or attachments
> >
> >
> > On Tue, Jun 09, 2026 at 10:58:05AM +0300, Avihai Horon wrote:
> > > diff --git a/hw/core/machine.c b/hw/core/machine.c
> > > index 4d8b15d99e..8219f13779 100644
> > > --- a/hw/core/machine.c
> > > +++ b/hw/core/machine.c
> > > @@ -43,6 +43,7 @@ GlobalProperty hw_compat_11_0[] = {
> > > { "chardev-vc", "encoding", "cp437" },
> > > { "tpm-crb", "cap-chunk", "off" },
> > > { "tpm-crb", "x-allow-chunk-migration", "off" },
> > > + { "migration", "switchover-ack-legacy", "on" },
> > > };
> > This line needs to be moved to the last patch, or people will get very
> > confused on last patch flipping default without caring about old behavior..
>
> This patch Implements the compatibility flows for legacy switchover-ack
> which use this property, so moving this line to last patch sounds even more
> confusing IMHO.
My comment only applies to this one line to change hw_compat_11_0[], not
the property itself. It won't affect this patch. It's common pattern to
do this in one patch to "update default value, but let's keep the old
machines unaffected".
Btw, since this value still has a default value ON in this patch, moving
this line to the last patch won't change anything except making last patch
clearer.
Thanks,
>
> The last patch should not be viewed standalone, it's tightly coupled with
> this one. Plus, it doesn't make sense to have the switchover-ack-legacy
> property if it's always off.
> So I personally don't find it confusing, the last patch simply enables the
> feature once everything is in place.
>
> If that helps, I can extend the last patch's commit message to something
> like:
>
> Now that VFIO has implemented new switchover-ack, enable it for new
> machines. Note that legacy switchover-ack is still used for
> compatibility with older machines via a compatibility entry for
> switchover-ack-legacy property.
>
> >
> > After move, feel free to take:
> >
> > Reviewed-by: Peter Xu <[email protected]>
> >
> > PS: If Cédric feels comfortable updating that while queuing it'll be fine
> > too.
>
> Thanks.
>
--
Peter Xu
