From: Takashi Iwai Subject: Re: 100.0% power usage of Device Audio codec (HDA) on wake Date: Wed, 21 Nov 2012 15:26:00 +0100 Message-ID: References: <33f3a5d64a489d40499192a0659dd76c@sourcepole.ch> <50AB3A39.1090604@canonical.com> <50AB45B3.6060904@canonical.com> <752ffd5c0c257045c7d8b23d95ea9ffa@sourcepole.ch> Mime-Version: 1.0 (generated by SEMI 1.14.6 - "Maruoka") Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Received: from mx2.suse.de (cantor2.suse.de [195.135.220.15]) by alsa0.perex.cz (Postfix) with ESMTP id 0DA42265D4E for ; Wed, 21 Nov 2012 15:26:01 +0100 (CET) In-Reply-To: <752ffd5c0c257045c7d8b23d95ea9ffa@sourcepole.ch> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: alsa-devel-bounces@alsa-project.org Sender: alsa-devel-bounces@alsa-project.org To: Tomas Pospisek Cc: Alsa-devel@alsa-project.org, David Henningsson List-Id: alsa-devel@alsa-project.org At Wed, 21 Nov 2012 15:15:09 +0100, Tomas Pospisek wrote: > > On Tue, 20 Nov 2012 10:01:18 +0100, Takashi Iwai wrote: > > At Tue, 20 Nov 2012 09:56:19 +0100, > > David Henningsson wrote: > >> > >> On 11/20/2012 09:32 AM, Takashi Iwai wrote: > >> > At Tue, 20 Nov 2012 09:07:21 +0100, > >> > David Henningsson wrote: > >> >> > >> >> It does put another question - from where does powertop get its > >> >> values, > >> >> and do we have to do anything on our side to make sure / aid them > >> >> being > >> >> correct? > >> > > >> > powertop checks /sys/class/sound/hwC*D*/power_{off|on}_acct files. > >> > They contain the time (in ms) how long the codec is in the power-save > >> > mode. > >> > >> Thanks. But I have also recently seen a computer where powertop showed > a > >> Watt number, essentially saying that the codec consumed 1.4 Watts or so > > >> (I don't remember the exact number). Have you seen that too? > > > > When 1.4W is constantly consumed even without PCM, it's a bit too > > high. But with PCM playback, about 1W usage can be measured normally, > > I guess. > > man powertop says: > > --calibrate runs powertop in calibration mode. When running on battery, > powertop can track power consumption as well as system > activity. When there are enough measurements, powertop > can start to report power estimates. One can get more > accurate estimates by using this option to enable a > calibration cycle. This will cycle through various display > levesl and USB device activities and workloads. > > So that's what powertop does I think: it estimates. But 100% things from hwC0D0 has nothing to do with the real power usage. It's just the time of HD-audio power-saving mode on and off. Takashi