Re: [PATCH v1 1/3] dt-bindings: Document MIPID02 bindings

[Mickael GUENE]

Hi Sakari,

Thanks for your review. Find my comments below.

On 3/16/19 22:46, Sakari Ailus wrote:
> Hi Mickael,
> 
> Thanks for the patchset.
> 
> On Tue, Mar 12, 2019 at 07:44:03AM +0100, Mickael Guene wrote:
>> This adds documentation of device tree for MIPID02 CSI-2 to PARALLEL
>> bridge.
>>
>> Signed-off-by: Mickael Guene <mickael.guene@xxxxxx>
>> ---
>>
>>  .../bindings/media/i2c/st,st-mipid02.txt           | 69 ++++++++++++++++++++++
>>  1 file changed, 69 insertions(+)
>>  create mode 100644 Documentation/devicetree/bindings/media/i2c/st,st-mipid02.txt
>>
>> diff --git a/Documentation/devicetree/bindings/media/i2c/st,st-mipid02.txt b/Documentation/devicetree/bindings/media/i2c/st,st-mipid02.txt
>> new file mode 100644
>> index 0000000..a1855da
>> --- /dev/null
>> +++ b/Documentation/devicetree/bindings/media/i2c/st,st-mipid02.txt
>> @@ -0,0 +1,69 @@
>> +STMicroelectronics MIPID02 CSI-2 to PARALLEL bridge
>> +
>> +MIPID02 has two CSI-2 input ports, only one of those ports can be active at a
>> +time. Active port input stream will be de-serialized and its content outputted
>> +through PARALLEL output port.
>> +CSI-2 first input port is a dual lane 800Mbps whereas CSI-2 second input port is
> 
> 800 Mbps per lane (or total)?
> 
800 Mbps per lane. I will document it.
>> +a single lane 800Mbps. Both ports support clock and data lane polarity swap.
>> +First port also supports data lane swap.
>> +PARALLEL output port has a maximum width of 12 bits.
>> +Supported formats are RAW6, RAW7, RAW8, RAW10, RAW12, RGB565, RGB888, RGB444,
>> +YUV420 8-bit, YUV422 8-bit and YUV420 10-bit.
>> +
>> +Required Properties:
>> +- compatible: should be "st,st-mipid02"
>> +- clocks: reference to the xclk input clock.
>> +- clock-names: should be "xclk".
>> +- VDDE-supply: sensor digital IO supply. Must be 1.8 volts.
>> +- VDDIN-supply: sensor internal regulator supply. Must be 1.8 volts.
> 
> Perhaps Rob can confirm, but AFAIR the custom is to use lower case letters.
> 
 It seems there is a 50-50 ratio between upper and lower case usage in
Documentation/devicetree/bindings/media/i2. I will wait Rob's answer to change
it or not.
>> +
>> +Optional Properties:
>> +- reset-gpios: reference to the GPIO connected to the xsdn pin, if any.
>> +	       This is an active low signal to the mipid02.
>> +
>> +Required subnodes:
>> +  - ports: A ports node with one port child node per device input and output
>> +	   port, in accordance with the video interface bindings defined in
>> +	   Documentation/devicetree/bindings/media/video-interfaces.txt. The
>> +	   port nodes are numbered as follows:
>> +
>> +	   Port Description
>> +	   -----------------------------
>> +	   0    CSI-2 first input port
>> +	   1    CSI-2 second input port
>> +	   2    PARALLEL output
> 
> Please document which endpoint properties are relevant. From the above
> description I'd presume this to be at least clock-lanes (1st input),
> data-lanes, lane-polarities (for CSI-2) as well as bus-width for the
> parallel bus.
> 
ok. I will add documentation.
>> +
>> +Example:
>> +
>> +mipid02: mipid02@14 {
> 
> The node should be a generic name. "csi2rx" is used by a few devices now.
>
 If I understand you well, you would prefer:
csi2rx: mipid02@14 {
 I show no usage of csi2rx node naming except for MIPI-CSI2 RX controller.
>> +	compatible = "st,st-mipid02";
>> +	reg = <0x14>;
>> +	status = "okay";
>> +	clocks = <&clk_ext_camera_12>;
>> +	clock-names = "xclk";
>> +	VDDE-supply = <&vdd>;
>> +	VDDIN-supply = <&vdd>;
>> +	ports {
>> +		#address-cells = <1>;
>> +		#size-cells = <0>;
>> +		port@0 {
>> +			reg = <0>;
>> +
>> +			ep0: endpoint {
>> +				clock-lanes = <0>;
>> +				data-lanes = <1 2>;
>> +				remote-endpoint = <&mipi_csi2_in>;
>> +			};
>> +		};
>> +		port@2 {
>> +			reg = <2>;
>> +
>> +			ep2: endpoint {
>> +				bus-width = <8>;
>> +				hsync-active = <0>;
>> +				vsync-active = <0>;
>> +				remote-endpoint = <&parallel_out>;
>> +			};
>> +		};
>> +	};
>> +};
>