Re: [PATCH v2 1/1] arm64: dts: qcom: monaco-evk: Add Interface Plus Mezzanine

[Umang Chheda]

Hello Krzysztof,

On 2/24/2026 3:39 AM, Dmitry Baryshkov wrote:
> On Mon, Feb 23, 2026 at 09:37:53PM +0100, Krzysztof Kozlowski wrote:
>> On 23/02/2026 20:02, Dmitry Baryshkov wrote:
>>>>> So I presume what you're saying is that we should at most declare one
>>>>> level of non-controlled fixed regulators?
>>>> In general, non-controller fixed regulators should not be there at all,
>>>> except when they serve certain purpose, like fulfill the binding
>>>> requirement. It's their only point.
>>>>
>>>> And a chain of:
>>>>
>>>> A -> B -> C -> device
>>>>
>>>> is completely redundant if all A+B+C are non-controlled.
>>> I think that came from me. I don't consider that to be completely
>>> redundant. It helps in reviews and in some understanding of the board
>>> logic. I'm not asking to implement all the intermediate regulators, but
>>> to implement the meaningful relationship between end-user regulators.
>> These are not end-user regulators. These are fixed things which no one
>> touches and no one needs. There is no single purpose for user-space to
>> see them.
>>
>> Why do you not insist on defining all of such external oscilators, rest
>> of regulators, all possible little ICs?
> So, where is the boundary from you point of view? Do we define fixed
> regulators powering DRM bridges / USB hubs and other similar devices?
> Or do we do it only if the bindings require us to do it?

Can you help share your point of view on the above query from Dmitry ? In this case to adhere to  bindings requirements

Is it okay if we define fixed regulators like A - > B - > device ? Instead of defining all the intermediate regulators.

>

Thanks,
Umang