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

[Dmitry Baryshkov]

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?


-- 
With best wishes
Dmitry