Re: [EXT] Re: [PATCH v4 2/5] dt-bindings: i3c: Add NXP P3H2x4x i3c-hub support

[Frank Li]

On Thu, Jan 15, 2026 at 10:01:03PM +0100, Krzysztof Kozlowski wrote:
> On 15/01/2026 13:21, Aman Kumar Pandey wrote:
> >
> > On Tue, Jan 13, 2026 at 01:45:26PM +0200, Aman Kumar Pandey wrote:
> >> +https://eur01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fdevi
> >>> +cetree.org%2Fschemas%2Fmfd%2Fnxp%2Cp3h2840-i3c-
> >> hub.yaml%23&data=05%7C
> >>>
> >> +02%7Caman.kumarpandey%40nxp.com%7C75a7d6f1f8b74fd3ac5808de5346
> >> 3133%7C
> >>>
> >> +686ea1d3bc2b4c6fa92cd99c5c301635%7C0%7C0%7C639039758104017624%7
> >> CUnkno
> >>>
> >> +wn%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwM
> >> CIsIlAiOiJX
> >>>
> >> +aW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=Ze
> >> GSgy3cF
> >>> +whQbcO%2BserynVrrrqYGOIwBMgVjgdY25Lg%3D&reserved=0
> >>
> >> This is not a MFD device. Place it in appropriate place, e.g. i2c or i3c
> >> controllers directory.
> >>
> >
> > Thanks for the review and the comment.
> >
> > I would like to clarify why I modeled the P3H2840 as an MFD.
>
> I do not see any model as MFD. You placed it in MFD directory and this
> is what I contest.
>
> Otherwise please point me to any hardware description what "MFD" is,
> because for sure your I3C hub is not this:
> https://en.wikipedia.org/wiki/MFD - scroll down to "Multifunction
> device" in "Computing".
>
> > While the primary function of the chip is an I3C hub,
>
> Wrap your emails so it will be possible to reply and read it.
>

Aman Kumar Pandey:

	Don't use outlook reply community's email. You can use b4 + mutt
https://josefbacik.github.io/kernel/2021/10/18/lei-and-b4.html

	Company's outlook always mess up email.

Frank


> So there you have your primary function.
>
> > this device also contains a programmable regulator which is exposed externally and can be used to power downstream devices on the board, not only the internal hub logic. This regulator has:
> > - Independent enable/disable control
> > - Independent voltage configuration
> >
> > From the board and hardware point of view, this block is not only a "hub-internal" power control, but a board-level power resource that other devices depend on.
> > hub logic and the regulator block have distinct roles and different Linux subsystems (I3C hub vs regulator framework), I modeled this as an MFD with two child devices.
>
> Linux subsystems do not matter. We speak about bindings, not drivers.
>
> You ignored some of my other comments. Your email is completely
> unreadable, so answer is still NAK. Read my comments fully, fix your
> email setup and then answer without all this messed quotes.
>
> Best regards,
> Krzysztof