On 27/11/2023 15:19, Geert Uytterhoeven wrote: >> --- /dev/null >> +++ b/Documentation/devicetree/bindings/dts-coding-style.rst > >> + /* SoC DTSI */ >> + >> + / { >> + cpus { >> + /* ... */ >> + }; >> + >> + psci { >> + /* ... */ >> + }; >> + >> + soc@ { > > "soc@" is invalid, that should be "soc". soc@0 is valid. > > As the "soc" node is special, you may want to elaborate: > > compatible = "simple-bus"; > #address-cells = <1>; > #size-cells = <1>; > ranges; but then we go to missing address/size cells in root node. Your comment is in general correct, but what you propose here is not a coding style, but DTS correctness and I only wanted to show the order of nodes. dtc already enforces the proper unit addresses, ranges and cells. > >> + dma: dma-controller@10000 { >> + /* ... */ >> + }; >> + >> + clk: clock-controller@80000 { >> + /* ... */ >> + }; >> + }; >> + }; >> + >> + /* Board DTS - alphabetical order */ >> + >> + &clk { >> + /* ... */ >> + }; >> + >> + &dma { >> + /* ... */ >> + }; >> + >> + /* Board DTS - alternative order, keep as DTSI */ >> + >> + &dma { >> + /* ... */ >> + }; >> + >> + &clk { >> + /* ... */ >> + }; > > IMO that alternative order is hard to review: you need to have multiple > files open. It will also make validation hard, as you can only validate > the end result, not individual files. Rob commented on this - tools (will) solve the issue. :) > > Anyway, this is already quite usable so > Reviewed-by: Geert Uytterhoeven <geert+renesas@xxxxxxxxx> > > Gr{oetje,eeting}s, > > Geert > Best regards, Krzysztof