From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mike Looijmans Subject: Re: [PATCH 2/2] hwmon: (max6650) Add initial rpm target devicetree documentation Date: Tue, 30 Aug 2016 18:56:12 +0200 Message-ID: <57C5BAAC.1090002@topic.nl> References: <1472026407-3448-1-git-send-email-mike.looijmans@topic.nl> <1472026407-3448-2-git-send-email-mike.looijmans@topic.nl> <20160830163804.GA5954@rob-hp-laptop> Mime-Version: 1.0 Content-Type: text/plain; charset=windows-1252; format=flowed Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20160830163804.GA5954@rob-hp-laptop> Sender: linux-hwmon-owner@vger.kernel.org To: Rob Herring Cc: linux-hwmon@vger.kernel.org, devicetree@vger.kernel.org, linux@roeck-us.net, linux-kernel@vger.kernel.org List-Id: devicetree@vger.kernel.org On 30-08-16 18:38, Rob Herring wrote: > On Wed, Aug 24, 2016 at 10:13:27AM +0200, Mike Looijmans wrote: >> Add devicetree property for early initialization of the fan controller >> to prevent overheating, for example when resetting the board while the >> fan was completely turned off. >> >> Signed-off-by: Mike Looijmans >> --- >> Documentation/devicetree/bindings/hwmon/max6650.txt | 5 +++++ >> 1 file changed, 5 insertions(+) > > The kernel boot seems kind of late to ensure your fan is running, but > > Acked-by: Rob Herring Thanks! Well, "late's better than never," my grandma used to say when we visited :) To put your mind at ease, the particular platform I'm using this for is an xc7z045 (Zynq) where the ARMs running the Linux kernel will only account for 0.5W, but once the logic awakens the chip can dissipate up to 30W. The logic bitstream is usually on the rootfs, so kernel boot is early enough for the fan to start spinning. And with the logic asleep, it need not spin at all. -- Mike Looijmans