From mboxrd@z Thu Jan 1 00:00:00 1970 From: Wei Ni Subject: Re: [PATCH v2 3/3] ARM: tegra: dalmore: add thermal zones for nct1008 Date: Fri, 7 Mar 2014 11:45:09 +0800 Message-ID: <531940C5.3040102@nvidia.com> References: <1393932664-13319-1-git-send-email-wni@nvidia.com> <1393932664-13319-4-git-send-email-wni@nvidia.com> <5318B790.8090808@wwwdotorg.org> Mime-Version: 1.0 Content-Type: text/plain; charset="ISO-8859-1" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <5318B790.8090808-3lzwWm7+Weoh9ZMKESR00Q@public.gmane.org> Sender: linux-tegra-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org To: Stephen Warren , Eduardo Valentin , "linux-0h96xk9xTtrk1uMJSBkQmQ@public.gmane.org" Cc: "linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org" , "linux-tegra-u79uwXL29TY76Z2rM5mHXA@public.gmane.org" , "lm-sensors-GZX6beZjE8VD60Wz+7aTrA@public.gmane.org" List-Id: linux-tegra@vger.kernel.org On 03/07/2014 01:59 AM, Stephen Warren wrote: > On 03/04/2014 04:31 AM, Wei Ni wrote: >> Add dt node to describe the thermal zone for the nct1008. > >> diff --git a/arch/arm/boot/dts/tegra114-dalmore.dts b/arch/arm/boot/dts/tegra114-dalmore.dts > >> + thermal-zones { >> + nct1008-local { >> + polling-delay-passive = <2000>; /* milliseconds */ >> + polling-delay = <0>; /* milliseconds */ > > A delay of 0 implies that software continually polls this zone, thus > consuming an entire CPU. Is this really what you want? No, if this delay is 0, the driver will not run polling queue, but just cancel it. > >> + >> + thermal-sensors = <&nct1008 0>; >> + }; > > This zone (and the other) is missing the required (per the binding > documentation) trips and cooling-maps sub-nodes. Eduardo, Are the trips and cooling-maps required properties ? I think if miss these sub-nodes, the thermal zone device also can work, and in the of-thermal framework, it doesn't handle them as required property either. > > I've also sent a query in response to the thermal-zones binding so see > whether or not it can be fixed so that software doesn't bind against the > node name, but rather against a compatible property like almost > everything else. Let's see what the response is there before reposting > this patch. Sure. Thanks. Wei. >