On 8/16/22 4:59 AM, Krzysztof Kozlowski wrote: > On 15/08/2022 07:34, Samuel Holland wrote: >> Some sunxi SoCs have in-package regulators controlled by a register in >> the system control MMIO block. Allow a child node for these regulators >> in addition to SRAM child nodes. >> >> Signed-off-by: Samuel Holland <samuel@xxxxxxxxxxxx> >> --- >> >> Changes in v3: >> - Require the regulators node to have a unit address >> - Reference the regulator schema from the SRAM controller schema >> - Move the system LDOs example to the SRAM controller schema >> - Reorder the patches so the example passes validation >> >> Changes in v2: >> - New patch for v2 >> >> .../allwinner,sun4i-a10-system-control.yaml | 29 +++++++++++++++++++ >> 1 file changed, 29 insertions(+) >> >> diff --git a/Documentation/devicetree/bindings/sram/allwinner,sun4i-a10-system-control.yaml b/Documentation/devicetree/bindings/sram/allwinner,sun4i-a10-system-control.yaml >> index d64c1b28fb61..915ca85c3f10 100644 >> --- a/Documentation/devicetree/bindings/sram/allwinner,sun4i-a10-system-control.yaml >> +++ b/Documentation/devicetree/bindings/sram/allwinner,sun4i-a10-system-control.yaml >> @@ -56,6 +56,10 @@ properties: >> ranges: true >> >> patternProperties: >> + "^regulators@[0-9a-f]+$": >> + $ref: /schemas/regulator/allwinner,sun20i-d1-system-ldos.yaml# >> + unevaluatedProperties: false > > unevaluatedProperties is not needed. Your other schema does not allow > anything else here. I can remove it. I added it because it looks like the dt-schema tools use it as an indicator that the matched properties are child nodes[1]. Maybe that is not relevant here? Regards, Samuel [1]: https://github.com/devicetree-org/dt-schema/commit/b12b3737cabc