Re: [PATCH 29/50] pckbd: more vmstate_register() from i8042_mm_init() to i8042_mmio_realize()

[Mark Cave-Ayland]

On 23/05/2022 07:34, Hervé Poussineau wrote:

> Hello,
>
> If you want, you can break migration compatibility. I don't see it as a problem.
>
> Hervé

Thanks for suggesting this. Given that migration of the magnum machine was broken for 
a long time until I fixed it recently, I don't think it would be a problem either.

I was considering a similar proposal for lasips2 but was planning to do the migration 
break in the part 2 series as part of the QOM modelling improvements rather than in 
this initial series. I'll update the I8042_MMIO accordingly when my current working 
branch for part 2.

> Le 22/05/2022 à 20:18, Mark Cave-Ayland a écrit :
> > Note in this case it is not possible to register a (new) VMStateDescription in
> > the DeviceClass without breaking migration compatibility for the MIPS magnum
> > machine.
> >
> > Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
> > ---
> >   hw/input/pckbd.c | 5 +++--
> >   1 file changed, 3 insertions(+), 2 deletions(-)
> >
> > diff --git a/hw/input/pckbd.c b/hw/input/pckbd.c
> > index eb77ad193e..7f3578aa4d 100644
> > --- a/hw/input/pckbd.c
> > +++ b/hw/input/pckbd.c
> > @@ -681,6 +681,9 @@ static void i8042_mmio_realize(DeviceState *dev, Error 
> > **errp)
> >       memory_region_init_io(&s->region, OBJECT(dev), &i8042_mmio_ops, ks,
> >                             "i8042", s->size);
> > +
> > +    /* Note we can't use dc->vmsd without breaking migration compatibility */
> > +    vmstate_register(NULL, 0, &vmstate_kbd, ks);
> >   }
> >   static void i8042_mmio_init(Object *obj)
> > @@ -721,8 +724,6 @@ MMIOKBDState *i8042_mm_init(qemu_irq kbd_irq, qemu_irq 
> > mouse_irq,
> >       s->irq_kbd = kbd_irq;
> >       s->irq_mouse = mouse_irq;
> > -    vmstate_register(NULL, 0, &vmstate_kbd, s);
> > -
> >       s->kbd = ps2_kbd_init(kbd_update_kbd_irq, s);
> >       s->mouse = ps2_mouse_init(kbd_update_aux_irq, s);


ATB,

Mark.