Re: [PATCH v8 1/2] dt-bindings: phy: qcom: Add CSI2 C-PHY/DPHY schema

[Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org>]

On 7/2/26 01:54, Bryan O'Donoghue wrote:
> On 02/06/2026 22:59, Vladimir Zapolskiy wrote:
>>> +
>>> +  "#phy-cells":
>>> +    const: 1
>>> +    description:
>>> +      The single cell specifies the PHY operating mode.
>> #phy-cells should be 0, because the PHY operating mode is well defined
>> by 'bus-type' property of an endpoint on the sensor side, the opposite
>> side of CAMSS/CSID as a CSIPHY "consumer" should not dictate the PHY type.
> 
> So going through the list I don't believe this is correct.

Be specific what is incorrect.

> 
> Inserting bus-type into the PHY layer is borrowing from one domain
> linux-media and pushing the concept into PHYs.

It's the other way around, "bus-type" is an endpoint property of media
devices, reference to video-interfaces.yaml to get more details about it.

> phy-cells = 1 with CPHY/DPHY specified in the consumer, as was Rob's
> suggestion will specify the mode.

This cell is just not needed, and unneeded complexity should not be added.

> Then to understand if we are in regular or combo mode, we parse the
> ports/endspoints in the PHY node as we've agreed already.
> 

That's correct, and while it's unavoidable to parse endpoint properties,
a phy cell value will not bring any additional information, in the best
case it's useless, or it opens a door to a clash, if different phy types
are occasionally specified twice.

-- 
Best wishes,
Vladimir