Re: [PATCH v2 3/4] dt-bindings: update google,cros-ec-pwm documentation

[Rob Herring <robh@kernel.org>]

On Thu, Apr 07, 2022 at 02:38:55PM +0000, Fabio Baltieri wrote:
> Hi Rob,
> 
> On Wed, Apr 06, 2022 at 10:41:45AM -0500, Rob Herring wrote:
> > On Thu, Mar 31, 2022 at 12:58:17PM +0000, Fabio Baltieri wrote:
> > > Update google,cros-ec-pwm node documentation to mention the
> > > google,use_pwm_type property.
> > > 
> > > Signed-off-by: Fabio Baltieri <fabiobaltieri@chromium.org>
> > > ---
> > >  .../devicetree/bindings/pwm/google,cros-ec-pwm.yaml         | 6 ++++++
> > >  1 file changed, 6 insertions(+)
> > > 
> > > diff --git a/Documentation/devicetree/bindings/pwm/google,cros-ec-pwm.yaml b/Documentation/devicetree/bindings/pwm/google,cros-ec-pwm.yaml
> > > index 4cfbffd8414a..9c895c990ed8 100644
> > > --- a/Documentation/devicetree/bindings/pwm/google,cros-ec-pwm.yaml
> > > +++ b/Documentation/devicetree/bindings/pwm/google,cros-ec-pwm.yaml
> > > @@ -19,6 +19,12 @@ description: |
> > >  properties:
> > >    compatible:
> > >      const: google,cros-ec-pwm
> > > +
> > > +  google,use-pwm-type:
> > > +    description:
> > > +      Use PWM types (CROS_EC_PWM_DT_<...>) instead of generic channels.
> > > +    type: boolean
> > 
> > Either do a new compatible string if the cell interpretation is mutually 
> > exclusive (channel number vs. type) or split the number space for the 
> > 1st cell between type and channel number. IOW, set a bit (31?) to 
> > signify the number is a type, not a channel.
> 
> Split the number space was my first (tentative) implementation as well,
> but it turns out that the PWM subsystem really wants channels to be
> zero-based[1], so I don't think flags or bitmasks are really an option.

Fix the PWM subsystem then...

> New compatible sounds good though, I'll rework a v3 with that change.
> 
> Thanks!
> Fabio
> 
> [1] https://elixir.bootlin.com/linux/v5.17/source/drivers/pwm/core.c#L423
> 
> -- 
> Fabio Baltieri