On Fri, Oct 9, 2020 at 12:48 AM Ricardo Cañuelo <ricardo.canuelo@xxxxxxxxxxxxx> wrote: > > Hi Rob, > > Thanks for taking the time to review this. Find my answers below: > > On jue 08-10-2020 13:38:18, Rob Herring wrote: > > > + codecs: > > > > Doesn't moving this require a driver change? > > I studied the driver as thoroughly as I could (what > sound/soc/codecs/cros_ec_codec.c:cros_ec_codec_platform_probe does) and > I think the driver should still be able to handle this. Unfortunately, I > can't test it. Can any of the driver maintainers share their > input about this? I'm adding Cheng-Yi Chiang to cc as well. The question is not what cros_ec_codec_platform_probe does, but how the platform device is created by the mfd driver. I think that's just a call to of_platform_populate which will only create immediate child devices unless there's a simple-bus or simple-mfd compatible. So I guess you could also add 'simple-mfd' rather than a driver change. However, there could be some expectation in the codec driver that the immediate parent is the EC node. > > If you only need 1 codec, then just drop the unit-address. > > Thank you. Yes, as far as I can tell there's only this codec (so far). I would probably go this route. You could add this level if there's ever more than one codec. However, I'm still not clear what the address represents for the codec. Is it needed? The address space/format is defined by the parent node. So is this defined by the EC? If so, other components don't have an address? Rob
