On Mon 18 Jul 00:44 PDT 2016, Arnd Bergmann wrote: > On Sunday, July 17, 2016 8:44:01 PM CEST Bjorn Andersson wrote: > > On Sun, Jul 17, 2016 at 8:34 PM, Bhushan Shah <bshah@xxxxxxx> wrote: > > > On Sun, Jul 17, 2016 at 09:21:48PM +0200, Arnd Bergmann wrote: > > >> On Sunday, July 17, 2016 4:22:07 PM CEST Bhushan Shah wrote: > > >> > + > > >> > + smd { > > >> > + rpm { > > >> > + rpm_requests { > > >> > + pm8841-regulators { > > >> > + s1 { > > >> > + regulator-min-microvolt = <675000>; > > >> > + regulator-max-microvolt = <1050000>; > > >> > + }; > > >> > > >> Maybe add a label at either the rpm_requests or the pm8841-regulators > > >> node so you can add properties in the leaf nodes withoutout having to > > >> specify the whole path? > > > > > > Sure, I will adjust patch. > > > > Please don't. After running into several cases where this would end us > > up in having a multitude of nodes each describing just a snippet of > > each level we decided not to do so in the general case for the > > Qualcomm boards. > > > > There are a few where we apparently ended up doing so anyways, but for > > all other cases of regulators we express the full tree in the dts, so > > please follow that so we don't mix the styles too much. > > Ok, then how about this instead: > > /smd/rpm/rpm_requests/pm8841-regulators { The problem I have with this is that in the dtsi we have properties and other nodes under each one of these, hence we end up with completely different overall structure depending on if I look in the dtsi or the dts. The problem I have with it in the dts is that we have properties and nodes under "smd" and "rpm_requests". So siblings are no longer grouped together. I have a hard time finding my way through flattened trees, often spread out over multiple files, that I need to puzzle together in my head. Perhaps there are better ways to keep this comprehensible, without maintaining the structure. > s1 { > regulator-min-microvolt = <675000>; > regulator-max-microvolt = <1050000>; > }; > > ... > }; > > That avoids the ridiculous intendation level but uses no labels. > I do share your dislike of the indentation level. I do have a few other concerns about style and scalability in other places. How about we follow how I've done this in the other files for now (i.e. keep the structure of the patch as is) and sit down at LAS16 to discuss what to do about this? Regards, Bjorn -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html