From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751916AbaHDIoX (ORCPT ); Mon, 4 Aug 2014 04:44:23 -0400 Received: from cantor2.suse.de ([195.135.220.15]:54405 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751119AbaHDIoW (ORCPT ); Mon, 4 Aug 2014 04:44:22 -0400 Subject: Re: [PATCH 4/4] Return the fan speed via sysfs From: Jean Delvare To: kreijack@inwind.it Cc: Benjamin Herrenschmidt , linux-kernel@vger.kernel.org, bryan@whatroute.net In-Reply-To: <53DE6667.8070901@gmail.com> References: <1406901650-20841-1-git-send-email-kreijack@inwind.it> <1406901650-20841-5-git-send-email-kreijack@inwind.it> <20140803161738.485f3f60@endymion.delvare> <53DE54D5.2050806@inwind.it> <20140803175901.27519306@endymion.delvare> <53DE6667.8070901@gmail.com> Content-Type: text/plain; charset="UTF-8" Organization: Suse Linux Date: Mon, 04 Aug 2014 10:44:12 +0200 Message-ID: <1407141852.4302.8.camel@chaos.site> Mime-Version: 1.0 X-Mailer: Evolution 2.28.2 Content-Transfer-Encoding: 8bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Le Sunday 03 August 2014 à 18:42 +0200, Goffredo Baroncelli a écrit : > On 08/03/2014 05:59 PM, Jean Delvare wrote: > > The temperature attributes in hwmon would need different names and > > units too (temp1_input and temp2_input, in millidegree C.) The > > advantage is that all monitoring applications out there would pick up > > these values automatically. > > Are you suggesting to add also a "temp1_input" attribute ? Yes, but as a hwmon class device attribute, not as an attribute in /sys/devices/temperature. > >> I even thought to allow to change the fan speed from user space.... > > > > Ben will never let you do that ;-) > > What would be the risk ?. When the CPU temperature goes behind the limit, > then the computer is switched off by an hardware protection (I am sure because > I had to changed a cpu board because a drift of the temperature sensor). And you are still asking what the risk is? ;-) > I am not suggesting to allow to change the fan speed, I am only asking which would > be the risk. If the user forces the fan too low, the system can overheat, and that can result in user injury or hardware damage (or reduced lifetime of hardware, at least.) Some people think that the user should always be given full power over his/her hardware. Others think that hardware should just work and users should never have to care about such low-level details. In this specific case I'd say don't change it unless you have a compelling reason to. -- Jean Delvare SUSE L3 Support