From mboxrd@z Thu Jan 1 00:00:00 1970 From: Maxime Ripard Subject: Re: [PATCH] arm64: dts: allwinner: a64: Re-add "syscon" compatible Date: Wed, 26 Sep 2018 11:52:02 +0200 Message-ID: <20180926095202.tdibor3x7fsn6afz@flea> References: <20180917152857.162662-1-andre.przywara@arm.com> <20180921153522.0413c280@donnerap.Emea.Arm.com> <20180921151625.wkaypyueyi2pguur@flea> <20180924112230.73e3009b@donnerap.Emea.Arm.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Content-Disposition: inline In-Reply-To: <20180924112230.73e3009b@donnerap.Emea.Arm.com> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+linux-arm-kernel=m.gmane.org@lists.infradead.org To: Andre Przywara Cc: Mark Rutland , Rob Herring , devicetree , Chen-Yu Tsai , Icenowy Zheng , linux-arm-kernel , Corentin Labbe List-Id: devicetree@vger.kernel.org On Mon, Sep 24, 2018 at 11:22:30AM +0100, Andre Przywara wrote: > > The fundamental difference is that we're mostly just a bunch of spare > > time programmers working on this platform, with a partial > > documentation for the controllers, at best. > > > > You forced me to ask these developpers to work on their weekends and > > evenings on some crazy corner cases to maintain the backward > > compatibility. And honestly, both from a technical and human > > standpoint, I definitely understand if some of them are just leaving > > and don't want to work on it anymore. I would probably do the same in > > their position. > > > > And having to ask that for companies like ARM or SUSE just makes it > > more frustrating to be honest. So there's simply no way you have > > forward compatibility while I'm there. Or you manage to convince all > > the ARM maintainers and enforce that compatibility for all the > > platforms. > > I understand that from your point of view there is no way of investing > huge efforts in staying forward compatible, but I am not asking for > that (and by no way forcing this!). Instead this suggestion is a small > tweak to achieve this. We're trying to remain compatible but if there's any technical reason, then we won't be. I don't want anyone to assume we will, and to rely on the fact that we are actually guaranteeing this. > So I am sorry if those things frustrate you (which I can understand > very well), but I believe fixing the DT in a proper way is > much more user friendly in the long term (actually this issue was > brought forward by a user[1]). Given the current state of the industry, I don't really see how the DT can allow you to do what you are trying to achieve. Maxime -- Maxime Ripard, Bootlin Embedded Linux and Kernel engineering https://bootlin.com