Re: [PATCH 4/4] ppc/pnv: Add xscom- prefix to pervasive-control region name

[Philippe Mathieu-Daudé]

On 26/11/24 04:54, Nicholas Piggin wrote:
> On Tue Nov 26, 2024 at 5:28 AM AEST, Philippe Mathieu-Daudé wrote:
> > Hi,
> >
> > On 25/11/24 14:20, Nicholas Piggin wrote:
> > > By convention, xscom regions get a xscom- prefix.
> > >
> > > Fixes: 1adf24708bf7 ("hw/ppc: Add pnv nest pervasive common chiplet model")
> > > Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
> > > ---
> > >    hw/ppc/pnv_nest_pervasive.c | 2 +-
> > >    1 file changed, 1 insertion(+), 1 deletion(-)
> > >
> > > diff --git a/hw/ppc/pnv_nest_pervasive.c b/hw/ppc/pnv_nest_pervasive.c
> > > index 77476753a4..780fa69dde 100644
> > > --- a/hw/ppc/pnv_nest_pervasive.c
> > > +++ b/hw/ppc/pnv_nest_pervasive.c
> > > @@ -177,7 +177,7 @@ static void pnv_nest_pervasive_realize(DeviceState *dev, 
> > > Error **errp)
> > >        pnv_xscom_region_init(&nest_pervasive->xscom_ctrl_regs_mr,
> > >                              OBJECT(nest_pervasive),
> > >                              &pnv_nest_pervasive_control_xscom_ops,
> > > -                          nest_pervasive, "pervasive-control",
> > > +                          nest_pervasive, "xscom-pervasive-control",
> >
> > Could this break migration stream? Or only RAM regions need to
> > have a stable name? I don't remember, but try be be cautions with
> > such cosmetic change just before the release ;)
>
> Hey Phil,
>
> Thanks for always somehow being across everything :)

;)

Answering myself, I/O regions are migrated within the device state,
per field, via DeviceClass::vmsd structure.

IIRC RAM regions are the ones tied to their name, which is built
using the object owner, which is why we can't change migrated RAM
MR owners.

> For the powernv machine we are okay since we don't support migration
> at all. It's on the wishlist, but at the moment we have lots of
> hardware models that don't provide vmstate. So I think we are okay.

Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>