Re: [PATCH v8 3/6] software node: implement reference properties

[Rafael J. Wysocki]

On Tue, Nov 10, 2020 at 1:39 PM Heikki Krogerus
<heikki.krogerus@xxxxxxxxxxxxxxx> wrote:
>
> On Mon, Nov 09, 2020 at 09:05:51PM +0200, Andy Shevchenko wrote:
> > On Mon, Nov 09, 2020 at 10:53:05AM -0800, Dmitry Torokhov wrote:
> > > On Mon, Nov 09, 2020 at 07:18:37PM +0100, Lukasz Stelmach wrote:
> > > > It was <2020-11-09 pon 19:24>, when Andy Shevchenko wrote:
> >
> > ...
> >
> > > > Probably I have missed something and I will be greatful, if you tell me
> > > > where I can find more information about software nodes. There are few
> > > > users in the kernel and it isn't obvious for me how to use software
> > > > nodes properly.
> > >
> > > Yeah, I disagree with Andy here. The lookup tables are a crutch that we
> > > have until GPIO and PWM a taught to support software nodes (I need to
> > > resurrect my patch series for GPIO, if you have time to test that would
> > > be awesome).
> >
> > We don't have support for list of fwnodes, this will probably break things
> > where swnode is already exist.
> >
> > I think Heikki may send a documentation patch to clarify when swnodes can and
> > can't be used. Maybe I'm mistaken and above is a good use case by design.
>
> Yeah, the problem is that I'm not sure that we can limit the swnodes
> for any specific purpose, or dictate very strictly how they are used
> with other possible fwnodes.

Generally agreed, but if there are known problems, they need to be
documented at least for the time being until they are addressed.

> Right now I'm thinking we should simply not talk about the
> relationship a software node should have or can have with other
> fwnodes (regardless of their type) in the swnode documentation.

This sounds reasonable to me, with the above exception.