Re: [PATCH 1/3] dt-bindings: pwm: Add binding for RPi firmware PWM bus

[Nicolas Saenz Julienne <nsaenzjulienne@suse.de>]

[-- Attachment #1: Type: text/plain, Size: 3085 bytes --]

On Tue, 2020-10-13 at 14:08 +0200, Uwe Kleine-König wrote:
> On Tue, Oct 13, 2020 at 12:35:38PM +0200, Nicolas Saenz Julienne wrote:
> > Hi Uwe, thanks for having a look at this.
> > 
> > On Mon, 2020-10-12 at 09:01 +0200, Uwe Kleine-König wrote:
> > > On Fri, Oct 09, 2020 at 05:30:28PM +0200, Nicolas Saenz Julienne wrote:
> > > > The PWM bus controlling the fan in RPi's official PoE hat can only be
> > > > controlled by the board's co-processor.
> > > > 
> > > > Signed-off-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
> > > > ---
> > > >  .../arm/bcm/raspberrypi,bcm2835-firmware.yaml | 21 +++++++++++++++++++
> > > >  .../pwm/raspberrypi,firmware-pwm.h            | 13 ++++++++++++
> > > >  2 files changed, 34 insertions(+)
> > > >  create mode 100644 include/dt-bindings/pwm/raspberrypi,firmware-pwm.h
> > > > 
> > > > diff --git a/Documentation/devicetree/bindings/arm/bcm/raspberrypi,bcm2835-firmware.yaml b/Documentation/devicetree/bindings/arm/bcm/raspberrypi,bcm2835-firmware.yaml
> > > > index a2c63c8b1d10..dcaf00e8602e 100644
> > > > --- a/Documentation/devicetree/bindings/arm/bcm/raspberrypi,bcm2835-firmware.yaml
> > > > +++ b/Documentation/devicetree/bindings/arm/bcm/raspberrypi,bcm2835-firmware.yaml
> > > > @@ -64,6 +64,22 @@ properties:
> > > >        - compatible
> > > >        - "#reset-cells"
> > > >  
> > > > +  pwm:
> > > > +    type: object
> > > > +
> > > > +    properties:
> > > > +      compatible:
> > > > +        const: raspberrypi,firmware-pwm
> > > > +
> > > > +      "#pwm-cells":
> > > > +        const: 1
> > > > +        description: >
> > > > +          The argument is the PWM bus number.
> > > 
> > > This is wrong. #pwm-cells specifies the number of "arguments" for
> > > phandles pointing to this node. And I would prefer this being 2 to match
> > > the stuff described in the generic pwm binding.
> > 
> > I saw buses out there with the same limitation as this one (unable to change
> > frequency) that used a single cell, so I whent with it. That said I'll be happy
> > to change it and drop the custom *_xlate() function in benefit of the default
> > one.
> 
> As the first cell after the phandle is for the period and only the
> second if for flags, this is a poor argument.

In that case aren't these bindings wrong (and associated xlate() functions)?

google,cros-ec-pwm.yaml:
[...]
	properties:
	  compatible:
	    const: google,cros-ec-pwm
	  "#pwm-cells":
	    description: The cell specifies the PWM index.
	    const: 1
[...]

cirrus,clps711x-pwm.txt:
[...]
	- #pwm-cells: Should be 1. The cell specifies the index of the channel.
[...]

Note that pxa-pwm.txt behaves as you comment.

Ultimately note that in of_pwm_simple_xlate() the second argument is used to
assign the pwm period, the first one is passed as an index to
pwm_request_from_chip().

> So yes, use #pwm-cells = <2> and drop the custom xlate() function please.

I'll still go this way nontheless. Just want to make sure I understand things
correctly.

Regards,
Nicolas


[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 488 bytes --]