From mboxrd@z Thu Jan 1 00:00:00 1970 From: Arnd Bergmann Subject: Re: [PATCH 2/3] ARM: dts: msm8974-hammerhead: Add regulator nodes for hammerhead Date: Mon, 18 Jul 2016 21:12:13 +0200 Message-ID: <4129314.Hc5WdF5TEZ@wuerfel> References: <20160717105208.9596-1-bshah@kde.org> <1957086.XgnIUz2uOJ@wuerfel> <20160718171155.GE13516@tuxbot> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7Bit Return-path: In-Reply-To: <20160718171155.GE13516@tuxbot> Sender: linux-arm-msm-owner@vger.kernel.org To: Bjorn Andersson Cc: Bhushan Shah , Andy Gross , David Brown , Rob Herring , Mark Rutland , Russell King , linux-arm-msm , "open list:ARM/QUALCOMM SUPPORT" , devicetree@vger.kernel.org List-Id: devicetree@vger.kernel.org On Monday, July 18, 2016 10:11:55 AM CEST Bjorn Andersson wrote: > On Mon 18 Jul 00:44 PDT 2016, Arnd Bergmann wrote: > > On Sunday, July 17, 2016 8:44:01 PM CEST Bjorn Andersson wrote: > > 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? Fair enough, let's do that for now. Arnd