On 11/04/2023 20:03, Konrad Dybcio wrote: >>> - qcom,sdm660-rpmpd >>> diff --git a/include/dt-bindings/power/qcom-rpmpd.h b/include/dt-bindings/power/qcom-rpmpd.h >>> index 1bf8e87ecd7e..867b18e041ea 100644 >>> --- a/include/dt-bindings/power/qcom-rpmpd.h >>> +++ b/include/dt-bindings/power/qcom-rpmpd.h >>> @@ -90,6 +90,15 @@ >>> #define SM8150_MMCX 9 >>> #define SM8150_MMCX_AO 10 >>> >>> +/* SA8155P is a special case, kept for backwards compatibility */ >> >> This is a new binding, thus I don't understand what is here backwards >> compatible? > Check the DT structure, 8155-adp (so, de facto 8155) included > 8150 before, but that was not exactly the correct approach.. > It bit us after we accidentally discovered 8155 advertises MMCX, > LCX and LMX PDs in cmd-db but triggers a bite on access attempts.. > > By mimicking the indices to match the ones of 8150, we only have > to fix up the domains that were axed from 8155 (in comparison to > 8150 which has the full fat setup). > > Konrad >> >>> +#define SA8155P_CX SM8150_CX >>> +#define SA8155P_CX_AO SM8150_CX_AO The DTS diff (the patch here) does not show it, I would need to check the context by opening other DTS in linux-next. Therefore it would be great if commit msg explained this. BTW, using here directly numbers matching SM8150 values, would also be backwards compatible... Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@xxxxxxxxxx> Best regards, Krzysztof
