Re: [PATCH v3 1/8] dt-bindings: clock: renesas: Split out RZ/T2H CPG binding

Hi Geer,

On Fri, Aug 7, 2026 at 2:30 PM Geert Uytterhoeven <[email protected]> wrote:
>
> Hi Prabhakar,
>
> On Mon, 20 Jul 2026 at 13:14, Lad, Prabhakar <[email protected]> wrote:
> > On Mon, Jul 20, 2026 at 11:53 AM Krzysztof Kozlowski <[email protected]> wrote:
> > > On 20/07/2026 12:50, Lad, Prabhakar wrote:
> > > > On Mon, Jul 20, 2026 at 6:55 AM Krzysztof Kozlowski <[email protected]> wrote:
> > > >> On Thu, Jul 16, 2026 at 01:34:27PM +0100, Prabhakar wrote:
> > > >>> From: Lad Prabhakar <[email protected]>
> > > >>>
> > > >>> Move the RZ/T2H and RZ/N2H clock controller bindings from the generic
> > > >>> renesas,cpg-mssr.yaml schema into a dedicated binding.
> > > >>>
> > > >>> Unlike the traditional Renesas CPG/MSSR block, the RZ/T2H and RZ/N2H
> > > >>> system controller integrates additional functionality beyond clock and
> > > >>> module-stop control, including reset control, clock monitor support,
> > > >>> write-protection registers and other system management features.
> > > >>
> > > >> Nothing of above differences got reflected in the code, so I do not
> > > >> understand this reasoning. Binding is exactly the same as the old one.
> > > >>
> > > > A syscon regmap is implemented [0] as part of the RZ/T2H CPG driver.
> > > > Since the RZ/T2H clock driver uses this syscon wrapper, a function
> > > > pointer is implemented to handle the RZ/T2H-specific configuration.
> > > >
> > > > Also, as the initial implementation used separate device tree nodes,
> > > > Geert suggested splitting this binding schema off into a dedicated
> > > > file [1].
> > > >
> > >
> > > I still miss the difference. Can you point me (can be short diff) to
> > > actual difference in this binding?
> > >
> > Fair enough! You're right that the schema text itself doesn't differ.
> > I'll drop this patch in v4 and leave the bindings in
> > renesas,cpg-mssr.yaml (unless Geert has a different opinion).
>
> Given you no longer add a bunch of RZ/T2H-specific properties
> and subnodes, I no longer see a reason for the split.
>
Ok, I will drop this change in the nxt version.

Cheers,
Prabhakar