From mboxrd@z Thu Jan 1 00:00:00 1970 From: alexandre.belloni@free-electrons.com (Alexandre Belloni) Date: Thu, 14 Dec 2017 20:21:31 +0100 Subject: DT dtc warnings In-Reply-To: References: <20171214183642.GF2559@piout.net> Message-ID: <20171214192131.GH2559@piout.net> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On 14/12/2017 at 13:00:07 -0600, Rob Herring wrote: > On Thu, Dec 14, 2017 at 12:36 PM, Alexandre Belloni > wrote: > > Hi Rob, > > > > On 14/12/2017 at 12:21:06 -0600, Rob Herring wrote: > >> Hi all, > >> > >> Below is a current list of ARM boards with more than 40 dtc warnings > >> when building with W=1. There's a treewide patch in flight to fix some > >> unit-address warnings[1], so those aren't included here. The list is > >> grouped by maintainer. AT91, Exynos, and Allwinner continue to be at > >> the top of the list and have been for some time (though having > >> multiple boards for an SoC can cause lots of duplicated warnings). > >> > >> Please help fix these. More checks are being added[2]. > >> > > > > For the at91 ones, IIRC, they are coming from the clock binding and we > > will have to break the ABI to fix it. This is not something we wanted to > > do before 4.14 but it will happen at some point. > > Looks like they are just missing unit-address. How does adding a > unit-address break the ABI? Though, aren't you planning to change from > a node per clock to clock controller node(s)? If so, then fixing when > doing that is fine. Adding the unit-address breaks the lookup of the clocks unless you want to have nodes named prog0 at 0, prog1 at 1, pck0 at 8, etc... I don't think it is worth the hassle going through all the dtsi to do that. -- Alexandre Belloni, Free Electrons Embedded Linux and Kernel engineering http://free-electrons.com