On 23/03/2024 18:44, Lothar Rubusch wrote: > On Sat, Mar 23, 2024 at 3:27 PM Krzysztof Kozlowski > <krzysztof.kozlowski@xxxxxxxxxx> wrote: >> >> On 23/03/2024 13:04, Lothar Rubusch wrote: >>> On Fri, Mar 22, 2024 at 3:17 AM Rob Herring <robh@xxxxxxxxxx> wrote: >>>> >>>> On Fri, Mar 22, 2024 at 12:37:13AM +0000, Lothar Rubusch wrote: >>>>> Provide the optional spi-3wire in the example. >>>> >>>> That doesn't match the diff as you don't touch the example. But really, >>>> this should say why you need spi-3wire. >>> >>> I understand. The change does not add anything to the example. which >>> is definitely wrong. >>> Anyway I'm unsure about this change in particular. I know the spi-3wire >>> binding exists and can be implemented. Not all spi devices offer it. Not all >>> drivers implement it. My patch set tries to implement spi-3wire for the >>> particular accelerometer. >>> Do I need to add something here to dt-bindings documentation of the >>> adxl345? Or, as an optional spi feature, is it covered anyway by >>> documentation of optional spi bindings? So, should I refrase this particular >>> patch or may I drop it entirely? Could you please clarify. >> >> Whether you need to change bindings or not, dtbs_check will tell you. >> Just run dtbs_check on your DTS. >> > > I'm not changing upstream DTS. At most, the documentation should > mention something. Nothing should stop you testing from downstream DTS... Best regards, Krzysztof