Re: [Qemu-devel] [PATCH RFC 0/3] qom: Generalize qdev init to "realize"

[Anthony Liguori]

On 03/27/2012 11:19 AM, Paolo Bonzini wrote:
> Il 26/03/2012 15:46, Andreas Färber ha scritto:
> > Hello Anthony,
> >
> > Here's a mini series introducing ObjectClass::realize(Object *) and
> > forwarding it to existing DeviceClass::init(DeviceState *).
>
> I think a major difference between realize and init should be that the
> realize property also propagates down the whole composition tree (in
> pre-order for setting to true, and post-order for setting to false).

Yes, I haven't reviewed this series yet, but my expectation would be that 
realize propagates and that the default implementation of DeviceClass::realize() 
would explicitly *not* propagate and instead just call ::init.

Likewise, unrealize() should do the same with reset().

Regards,

Anthony Liguori

> An important, related point is whether the composition tree should be
> created before or after realize/init.  Right now the few examples in the
> tree do it after, but this sounds wrong to me and would block the above
> auto-propagation.
>
> Paolo