On Fri, Jul 10, 2020 at 2:07 PM <ansuelsmth@xxxxxxxxx> wrote: > > > > > -----Messaggio originale----- > > Da: Rob Herring <robh@xxxxxxxxxx> > > Inviato: venerdì 10 luglio 2020 18:27 > > A: Ansuel Smith <ansuelsmth@xxxxxxxxx> > > Cc: Amit Kucheria <amit.kucheria@xxxxxxxxxx>; Andy Gross > > <agross@xxxxxxxxxx>; Bjorn Andersson <bjorn.andersson@xxxxxxxxxx>; > > Zhang Rui <rui.zhang@xxxxxxxxx>; Daniel Lezcano > > <daniel.lezcano@xxxxxxxxxx>; linux-pm@xxxxxxxxxxxxxxx; linux-arm- > > msm@xxxxxxxxxxxxxxx; devicetree@xxxxxxxxxxxxxxx; linux- > > kernel@xxxxxxxxxxxxxxx > > Oggetto: Re: [PATCH 3/6] dt-bindings: thermal: tsens: document ipq8064 > > bindings > > > > On Thu, Jul 09, 2020 at 11:51:33PM +0200, Ansuel Smith wrote: > > > Document the use of regmap phandle for ipq8064 SoCs > > > > > > Signed-off-by: Ansuel Smith <ansuelsmth@xxxxxxxxx> > > > --- > > > .../bindings/thermal/qcom-tsens.yaml | 51 ++++++++++++++++--- > > > 1 file changed, 44 insertions(+), 7 deletions(-) > > > > > > diff --git a/Documentation/devicetree/bindings/thermal/qcom-tsens.yaml > > b/Documentation/devicetree/bindings/thermal/qcom-tsens.yaml > > > index d7be931b42d2..5ceb5d720e16 100644 > > > --- a/Documentation/devicetree/bindings/thermal/qcom-tsens.yaml > > > +++ b/Documentation/devicetree/bindings/thermal/qcom-tsens.yaml > > > @@ -24,6 +24,7 @@ properties: > > > - enum: > > > - qcom,msm8916-tsens > > > - qcom,msm8974-tsens > > > + - qcom,ipq8064-tsens > > > - const: qcom,tsens-v0_1 > > > > > > - description: v1 of TSENS > > > @@ -47,6 +48,11 @@ properties: > > > - description: TM registers > > > - description: SROT registers > > > > > > + regmap: > > > + description: > > > + Phandle to the gcc. On ipq8064 SoCs gcc and tsense share the same > > regs. > > > + $ref: /schemas/types.yaml#/definitions/phandle > > > > Can't you make this a child of the gcc and drop this property? > > > > Make the thermal a child of the gcc would be a little confusing. Anyway > making this > a child of gcc cause the not probing of the thermal driver as it's ignored > any child of > gcc. I pushed v2 with the fixed problem. Structure the DT as the h/w is structured, not what's convenient for the OS. Either the GCC driver has to call of_platform_populate on the child nodes or you can perhaps use 'simple-mfd'. Rob
