Re: [PATCH v2 3/3] dt-bindings: phy: realtek: Add the doc about the Realtek SoC USB 2.0/3.0 PHY

[Krzysztof Kozlowski]

On 02/06/2023 05:20, Stanley Chang[昌育德] wrote:
> Hi Krzysztof,
> 
>>>
>>> Most of the properties are about the phy parameters.
>>> Is the phy parameter data suitable to be placed in DTS?
>>> I referenced other phy drivers.
>>> These parameters should not be defined in dts.
>>> I would move the parameters to the driver.
>>
>> If these can be in the driver, why would ever they be in DTS in the first place?
>>
> Our platforms have 3 xhci controllers which map to 3 different phy ports.

You mean on the same SoC?

> And the three phy ports use the same driver, but the parameters are different.
> So I put the parameter settings in DTS, we have 3 usb-phy nodes representing 3 phy ports.
> Also, some parameters have to be adjusted for different boards.
> Therefore, it is more applicable in DTS than in driver.

Then it looks justified in DT, so please write proper descriptions for
proper properties. Underscores are not allowed in node names. No fake
nodes. Properties should usually describe physical/hardware effect not
the register value.

qcom,usb-snps-femto-v2.yaml is nice example. Few Mediatek bindings also
would work.
'

Best regards,
Krzysztof