Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Thu, 15 Apr 2021 08:54:20 +0200
From:   Jacopo Mondi <jacopo@jmondi.org>
To:     Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Cc:     Jacopo Mondi <jacopo+renesas@jmondi.org>,
        Geert Uytterhoeven <geert+renesas@glider.be>,
        Magnus Damm <magnus.damm@gmail.com>,
        Kieran Bingham <kieran.bingham@ideasonboard.com>,
        Rob Herring <robh+dt@kernel.org>,
        linux-renesas-soc@vger.kernel.org, devicetree@vger.kernel.org,
        linux-kernel@vger.kernel.org
Subject: Re: [PATCH v3 1/5] dt-bindings: media: max9286: Define
 'maxim,gpio-poc'
Message-ID: <20210415065420.lbba2hig4p2ynx6c@uno.localdomain>
References: <20210414135128.180980-1-jacopo+renesas@jmondi.org>
 <20210414135128.180980-2-jacopo+renesas@jmondi.org>
 <YHd/AKIfxIHhTjWO@pendragon.ideasonboard.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
In-Reply-To: <YHd/AKIfxIHhTjWO@pendragon.ideasonboard.com>
Precedence: bulk

Hi Laurent,

On Thu, Apr 15, 2021 at 02:47:12AM +0300, Laurent Pinchart wrote:
> Hi Jacopo,
>
> Thank you for the patch.
>
> On Wed, Apr 14, 2021 at 03:51:24PM +0200, Jacopo Mondi wrote:
> > Define a new vendor property in the maxim,max9286 binding schema.
> >
> > The new property allows to declare that the remote camera
> > power-over-coax is controlled by one of the MAX9286 gpio lines.
> >
> > As it is currently not possible to establish a regulator as consumer
> > of the MAX9286 gpio controller for this purpose, the property allows to
> > declare that the camera power is controlled by the MAX9286 directly.
> >
> > The property accepts a gpio-index (0 or 1) and one line polarity
> > flag as defined by dt-bindings/gpio/gpio.h.
> >
> > Signed-off-by: Jacopo Mondi <jacopo+renesas@jmondi.org>
> > ---
> >  .../bindings/media/i2c/maxim,max9286.yaml     | 53 ++++++++++++++++++-
> >  1 file changed, 52 insertions(+), 1 deletion(-)
> >
> > diff --git a/Documentation/devicetree/bindings/media/i2c/maxim,max9286.yaml b/Documentation/devicetree/bindings/media/i2c/maxim,max9286.yaml
> > index ee16102fdfe7..480a491f3744 100644
> > --- a/Documentation/devicetree/bindings/media/i2c/maxim,max9286.yaml
> > +++ b/Documentation/devicetree/bindings/media/i2c/maxim,max9286.yaml
> > @@ -70,6 +70,24 @@ properties:
> >        a remote serializer whose high-threshold noise immunity is not enabled
> >        is 100000 micro volts
> >
> > +  maxim,gpio-poc:
>
> I would have written poc-gpio to match the order of the GPIO bindings
> syntax.
>

That's what I had :) but then the property gets matched against the
gpio schema and I get complains because it expects a phandle as first
argument... Maybe there's a way I've missed to prevent the property to
be matched with *-gpio ?

> > +    $ref: '/schemas/types.yaml#/definitions/uint32-array'
> > +    minItems: 2
> > +    maxItems: 2
> > +    description: |
> > +      Identifier of gpio line that controls Power over Coax to the cameras and
>
> I'd write "Index of the MAX9286 GPIO output that ..." to make it clear
> that this is not a generic GPIO.
>

Ack

> > +      the associated polarity flag (GPIO_ACTIVE_HIGH and GPIO_ACTIVE_LOW)
> > +      as defined in <include/dt-bindings/gpio/gpio.h>.
> > +
> > +      When the remote cameras power is controlled by one of the MAX9286 gpio
> > +      lines, this property has to be used to specify which line among the two
> > +      available ones controls the remote camera power enablement.
> > +
> > +      When this property is used it is not possible to register a gpio
> > +      controller as the gpio lines are controlled directly by the MAX9286 and
> > +      not available for consumers, nor the 'poc-supply' property should be
> > +      specified.
>
> Only one of the two lines would be controlled directly. Shouldn't we
> still register a GPIO controller for the other line ?

I considered that and thought it was a bit of an overkill (and I also
had a bit of troubles identifying how to register only gpio #1, as it
would be identified as gpio #0 if the actual #0 is not registered)

>
> Could you also mention somewhere that the first item in the array should
> be 0 or 1 ? It may be hard to express in a YAML schema, so I'm fine just
> documenting it in the description.
>

Sure, I tried identifying how to express that with yaml and failed :)

> I've been wondering whether this would be a common enough issue that it
> could justify support in the GPIO core to handle consumer-provider
> loops, but even if that happens at some point in the future, I think the
> proposal here is good enough and we won't need to switch.
>

Please note that with the suggestion offline from rob I will add to
the next version:

# If 'maxim,gpio-poc' is present, then 'poc-supply' and 'gpio-controller'
# are not allowed.
if:
  required:
    - maxim,gpio-poc
then:
  allOf:
    - not:
        required:
          - poc-supply
    - not:
        required:
          - gpio-controller

> > +
> >    ports:
> >      $ref: /schemas/graph.yaml#/properties/ports
> >
> > @@ -182,7 +200,6 @@ required:
> >    - reg
> >    - ports
> >    - i2c-mux
> > -  - gpio-controller
> >
> >  additionalProperties: false
> >
> > @@ -327,4 +344,38 @@ examples:
> >            };
> >          };
> >        };
> > +
> > +      /*
> > +       * Example of a deserializer that controls the camera Power over Coax
> > +       * through one of its gpio lines.
> > +       */
> > +      gmsl-deserializer@6c {
> > +        compatible = "maxim,max9286";
> > +        reg = <0x6c>;
> > +        enable-gpios = <&gpio 14 GPIO_ACTIVE_HIGH>;
> > +
> > +        /*
> > +         * The remote camera power is controlled by MAX9286 GPIO line #0.
> > +         * No 'poc-supply' nor 'gpio-controller' are specified.
> > +         */
> > +        maxim,gpio-poc = <0 GPIO_ACTIVE_LOW>;
> > +
> > +        /*
> > +         * Do not describe connections as they're the same as in the previous
> > +         * example.
> > +         */
> > +        ports {
> > +          #address-cells = <1>;
> > +          #size-cells = <0>;
> > +
> > +          port@4 {
> > +            reg = <4>;
> > +          };
> > +        };
> > +
> > +        i2c-mux {
> > +          #address-cells = <1>;
> > +          #size-cells = <0>;
> > +        };
> > +      };
> >      };
>
> It's customary to indent DT examples with 4 spaces. The existing
> examples use two spaces, so maybe a patch on top of this would be useful
> to increase readability ?
>

Ah weird! I can add a patch before this one!

> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
>

Thanks
   j

> --
> Regards,
>
> Laurent Pinchart