On 26/09/2022 17:05, Luca Weiss wrote: > Hi Krzysztof, > > On Montag, 26. September 2022 10:54:23 CEST Krzysztof Kozlowski wrote: >> On 25/09/2022 18:18, Luca Weiss wrote: >>> The iadc node name is supposed to be just 'adc' and the compatible is >>> only supposed to be qcom,spmi-iadc according to the bindings. >>> >>> Adjust the node to match that. >>> >>> Signed-off-by: Luca Weiss <luca@xxxxxxxxx> >>> --- >>> >>> arch/arm/boot/dts/qcom-pm8941.dtsi | 4 ++-- >>> 1 file changed, 2 insertions(+), 2 deletions(-) >>> >>> diff --git a/arch/arm/boot/dts/qcom-pm8941.dtsi >>> b/arch/arm/boot/dts/qcom-pm8941.dtsi index 3c15eecf2f21..33517cccee01 >>> 100644 >>> --- a/arch/arm/boot/dts/qcom-pm8941.dtsi >>> +++ b/arch/arm/boot/dts/qcom-pm8941.dtsi >>> @@ -131,8 +131,8 @@ adc-chan@48 { >>> >>> }; >>> >>> }; >>> >>> - pm8941_iadc: iadc@3600 { >>> - compatible = "qcom,pm8941-iadc", > "qcom,spmi-iadc"; >>> + pm8941_iadc: adc@3600 { >>> + compatible = "qcom,spmi-iadc"; >> >> I am not sure this is correct. Usually specific compatibles are encouraged. > > I'm happy to change it the other way also. > > But the sibling of this compatible, qcom,spmi-vadc also only has that single > compatible so it'd align it with that. Ugh, there is a mess around them. Some other ADCs have specific compatibles, some not, so there is no consistency. I propose to have device specific compatible with qcom,spmi-iadc fallback, so basically document the DTS in bindings. Maybe other IADC will need some quirks, so specific compatible helps here. Best regards, Krzysztof