On Fri, Nov 06, 2026 at 11:27:38PM +0100, Carlo Caione wrote:
> On Fri, Nov 06, 2026 at 20:48:01 +0100, Conor Dooley wrote:
> > On Fri, Nov 06, 2026 at 08:34:37PM +0100, Carlo Caione wrote:
> > > On Fri, Nov 06, 2026 at 17:34:14 +0100, Conor Dooley wrote:
> > > 
> > > > > +    #include <dt-bindings/clock/amlogic,s7-pll-clkc.h>
> > > > > +    #include <dt-bindings/clock/amlogic,s7-scmi-clkc.h>
> > > > > +
> > > > > +    soc {
> > > > > +        #address-cells = <2>;
> > > > > +        #size-cells = <2>;
> > > > > +
> > > > > +        clock-controller@0 {
> > > > > +            compatible = "amlogic,s7-peripherals-clkc";
> > > > > +            reg = <0x0 0x0 0x0 0x49c>;
> > > > 
> > > > What's with the weird numbers in this example?
> > > > 
> > > > Examples don't have to be real, but when it looks like this the
> > > > immediate thought is that the modelling of the device might be
> > > > incorrec.t
> > > 
> > > The numbers are actually real, but the example simply lost their context
> > > because I copied the node from the S7 dtsi (not yet submitted), where
> > > the controller sits under the APB bus node (bus@fe000000) and the bus
> > > ranges make the addresses relative to it.
> > > 
> > > 		apb: bus@fe000000 {
> > > 			compatible = "simple-bus";
> > > 			reg = <0x0 0xfe000000 0x0 0x480000>;
> > > 			#address-cells = <2>;
> > > 			#size-cells = <2>;
> > > 			ranges = <0x0 0x0 0x0 0xfe000000 0x0 0x480000>;
> > > 
> > > 			clkc_periphs: clock-controller@0 {
> > > 				compatible = "amlogic,s7-peripherals-clkc";
> > > 				reg = <0x0 0x0 0x0 0x49c>;
> > > 				#clock-cells = <1>;
> > > 				...
> > > 
> > > In V2 I'll use an absolute address for the sake of clarity (PLL has the
> > > same issue).
> > 
> > Before you do that, can you please share with me the s7 dtsi? Cos surely
> > you've got it in some git repo online somewhere.
> 
> The S7 dtsi is already upstream at [0], the peripheral clock controller
> and PLL will be enabled in that DTSI with a following series after the
> bindings and the drivers in this set are acked, following the diffs that
> will be submitted:

Right. The device does seem to have nodes of odd sizes and start
addresses, so this at least is not making anything worse.

Acked-by: Conor Dooley <conor.dooley@microchip.com>
pw-bot: not-applicable