On Sat, May 06, 2023 at 09:45:07AM +0800, Changhuang Liang wrote: > > > On 2023/5/5 20:38, Conor Dooley wrote: > > On Fri, May 05, 2023 at 09:29:15AM +0800, Changhuang Liang wrote: > > > >> But if keep this "starfive,jh7110-aon-syscon" compatible. Which .yaml match to > >> it? Use this series dt-bindings or syscon series dt-bindings. > > > > There is no syscon series anymore, it's part of the PLL series now: > > https://lore.kernel.org/linux-clk/20230414024157.53203-1-xingyu.wu@xxxxxxxxxxxxxxxx/ > > > > I don't really care what you, Walker & Xingyu decide to do, but add the > > binding in one series in a complete form. It's far less confusing to > > have only have one version of the binding on the go at once. > > > > Due to the current aon pmu needs to be adjusted, it affects the syscon in PLL series. > So It's inevitable to change syscon in PLL series. > > My current idea is PLL series don't add the aon_syscon node. I will add it in my > aon pmu series in next version That's fine. Rob was happy with the clock related parts, which was the original source of confusion there. > like this: > > aon_syscon: syscon@17010000 { > compatible = "starfive,jh7110-aon-pmu", "syscon"; The syscon does a bunch of things of which one is a pmu. I don't see a reason to name this other than "starfive,jh100-aon-syscon". > reg = <0x0 0x17010000 0x0 0x1000>; > #power-domain-cells = <1>; > }; > > In my opinion, the first we add "starfive,jh7110-aon-syscon" because "syscon" can > not appear alone in the compatible. If we have "starfive,jh7110-aon-pmu", this > "starfive,jh7110-aon-syscon" is not a must-be need. > > Do you agree with doing so. > > Thanks, > Changhuang
Attachment:
signature.asc
Description: PGP signature
