Re: [PATCH v4 2/2] media: dt-bindings: media: i2c: Add IMX300 CMOS sensor binding

[Sakari Ailus]

On Sun, Jan 17, 2021 at 06:51:04PM +0100, AngeloGioacchino Del Regno wrote:
> Il 17/01/21 00:44, Sakari Ailus ha scritto:
> > Hi AngeloGioacchino,
> > 
> > On Wed, Jan 13, 2021 at 07:29:34PM +0100, AngeloGioacchino Del Regno wrote:
> > > Add YAML device tree binding for IMX300 CMOS image sensor, and
> > > the relevant MAINTAINERS entries.
> > > 
> > > Signed-off-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@xxxxxxxxxxxxxx>
> > > ---
> > >   .../bindings/media/i2c/sony,imx300.yaml       | 112 ++++++++++++++++++
> > >   MAINTAINERS                                   |   7 ++
> > >   2 files changed, 119 insertions(+)
> > >   create mode 100644 Documentation/devicetree/bindings/media/i2c/sony,imx300.yaml
> > > 
> > > diff --git a/Documentation/devicetree/bindings/media/i2c/sony,imx300.yaml b/Documentation/devicetree/bindings/media/i2c/sony,imx300.yaml
> > > new file mode 100644
> > > index 000000000000..4fa767feea80
> > > --- /dev/null
> > > +++ b/Documentation/devicetree/bindings/media/i2c/sony,imx300.yaml
> > > @@ -0,0 +1,112 @@
> > > +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
> > > +%YAML 1.2
> > > +---
> > > +$id: http://devicetree.org/schemas/media/i2c/sony,imx300.yaml#
> > > +$schema: http://devicetree.org/meta-schemas/core.yaml#
> > > +
> > > +title: Sony 1/2.3-Inch 25Mpixel Stacked CMOS Digital Image Sensor
> > > +
> > > +maintainers:
> > > +  - AngeloGioacchino Del Regno <angelogioacchino.delregno@xxxxxxxxxxxxxx>
> > > +
> > > +description: |-
> > > +  The Sony IMX300 is a 1/2.3-inch Stacked CMOS (Exmor-RS) digital image
> > > +  sensor with a pixel size of 1.08um and an active array size of
> > > +  5948H x 4140V. It is programmable through I2C interface at address 0x10.
> > > +  Image data is sent through MIPI CSI-2, which is configured as either 2 or
> > > +  4 data lanes.
> > > +
> > > +properties:
> > > +  compatible:
> > > +    const: sony,imx300
> > > +
> > > +  reg:
> > > +    maxItems: 1
> > > +
> > > +  clocks:
> > > +    maxItems: 1
> > 
> > Please add assigned clock related properties; see
> > Documentation/driver-api/media/camera-sensor.rst .
> > 
> Will do!
> 
> > > +
> > > +  vdig-supply:
> > > +    description:
> > > +      Digital I/O voltage supply, 1.15-1.20 volts
> > > +
> > > +  vana-supply:
> > > +    description:
> > > +      Analog voltage supply, 2.2 volts
> > > +
> > > +  vddl-supply:
> > > +    description:
> > > +      Digital core voltage supply, 1.8 volts
> > > +
> > > +  reset-gpios:
> > > +    description: |-
> > > +      Reference to the GPIO connected to the xclr pin, if any.
> > > +      Must be released (set high) after all supplies are applied.
> > > +
> > > +  # See ../video-interfaces.txt for more details
> > > +  port:
> > > +    type: object
> > > +    properties:
> > > +      endpoint:
> > > +        type: object
> > > +
> > > +        properties:
> > > +          data-lanes:
> > > +            description: |-
> > > +              The driver only supports four-lane operation.
> > 
> > This can be removed as bindings describe hardware, not driver operation.
> > 
> Ack.
> 
> > > +            items:
> > > +              - const: 0
> > > +              - const: 1
> > > +              - const: 2
> > > +              - const: 3
> > 
> > Two lanes here, too?
> > 
> 
> The driver only supports four-lane operation.
> I am 100% sure that this sensor can also work with two lanes, but it needs
> special configuration which I'm not able to produce, nor test.
> 
> As you may imagine (and as you can read in the driver itself), all of this
> was reverse-engineering work, as Sony has never released any datasheet for
> this sensor - and I have a hunch - they never will (but that's another
> story).

That's all fine. The bindings describe the hardware, not the driver's
capabilities.

> 
> > > +
> > > +          clock-noncontinuous: true
> > > +
> > > +          link-frequencies:
> > > +            $ref: /schemas/types.yaml#/definitions/uint64-array
> > > +            description:
> > > +              Allowed data bus frequencies. The driver currently needs
> > > +              to switch between 780000000 and 480000000 Hz in order to
> > > +              guarantee functionality of all modes.
> > 
> > You can omit this description, too.
> > 
> 
> The intention here was to be clear and provide as much information as I
> could gather during the very time-consuming reverse engineering process that
> took place in the making of this driver.
> 
> But okay, I will remove this.

Again, this is about the hardware, not the driver. That information is also
part of the driver.

> 
> > > +
> > > +        required:
> > > +          - data-lanes
> > > +          - link-frequencies
> > > +
> > > +required:
> > > +  - compatible
> > > +  - reg
> > > +  - clocks
> > > +  - vana-supply
> > > +  - vdig-supply
> > > +  - vddl-supply
> > 
> > Are the regulators really required? I'm not quite sure about the
> > established practices; still the common case is that one or two of these
> > are hard-wired.
> > 
> 
> On all the Sony phones that I have (....many), with MSM8956, MSM8996,
> SDM630, equipped with the IMX300 camera assy, none of these three are
> hard-wired: sometimes they're wired to the LDOs of the PMIC, sometimes
> they're wired to fixed LDOs, enabled through GPIOs (fixed-regulator binding
> in this case).
> 
> So.. yeah, they're really required.

As noted, that depends on the board. You just happen to have some where
they are not hard-wired.

> 
> > > +  - port
> > > +
> > > +additionalProperties: false
> > > +
> > > +examples:
> > > +  - |
> > > +    i2c0 {
> > > +        #address-cells = <1>;
> > > +        #size-cells = <0>;
> > > +
> > > +        imx300: sensor@10 {
> > > +            compatible = "sony,imx300";
> > > +            reg = <0x10>;
> > > +            clocks = <&imx300_xclk>;
> > > +            vana-supply = <&imx300_vana>;   /* 2.2v */
> > > +            vdig-supply = <&imx300_vdig>;   /* 1.2v */
> > > +            vddl-supply = <&imx300_vddl>;   /* 1.8v */
> > > +
> > > +            port {
> > > +                imx300_0: endpoint {
> > > +                    remote-endpoint = <&csi1_ep>;
> > > +                    data-lanes = <0 1 2 3>;
> > > +                    clock-noncontinuous;
> > > +                    link-frequencies = /bits/ 64 <780000000 480000000>;
> > > +                };
> > > +            };
> > > +        };
> > > +    };
> > > +
> > > +...
> > > diff --git a/MAINTAINERS b/MAINTAINERS
> > > index ad9abb42f852..5e0f08f48d48 100644
> > > --- a/MAINTAINERS
> > > +++ b/MAINTAINERS
> > > @@ -16633,6 +16633,13 @@ T:	git git://linuxtv.org/media_tree.git
> > >   F:	Documentation/devicetree/bindings/media/i2c/imx290.txt
> > >   F:	drivers/media/i2c/imx290.c
> > > +SONY IMX300 SENSOR DRIVER
> > > +M:	AngeloGioacchino Del Regno <angelogioacchino.delregno@xxxxxxxxxxxxxx>
> > > +L:	linux-media@xxxxxxxxxxxxxxx
> > 
> > Please also add the git tree.
> > 
> > Ideally also the MAINTAINERS change comes with the first patch adding the
> > files, which should be the DT bindings. I.e. just reverse the order of the
> > patches.
> > 
> 
> I haven't added it because last time I did that I got reviews saying that if
> I'm not the owner of the git tree I shall not put it in.
> Though, if that's a requirement for media, then I didn't know that...

The documentation in MAINTAINERS doesn't say that at least.

I think it'd be useful to have it.

> 
> > > +S:	Maintained
> > > +F:	Documentation/devicetree/bindings/media/i2c/sony,imx300.yaml
> > > +F:	drivers/media/i2c/imx300.c
> > > +
> > >   SONY IMX319 SENSOR DRIVER
> > >   M:	Bingbu Cao <bingbu.cao@xxxxxxxxx>
> > >   L:	linux-media@xxxxxxxxxxxxxxx
> > 
> 
> Thank you for your review!

You're welcome!

-- 
Sakari Ailus