From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from gate.crashing.org (gate.crashing.org [63.228.1.57]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (Client did not present a certificate) by ozlabs.org (Postfix) with ESMTP id 6571A67B5C for ; Mon, 15 May 2006 16:23:38 +1000 (EST) Subject: Re: Load average 1.0 on idle machine From: Benjamin Herrenschmidt To: Dan Malek In-Reply-To: References: <87wtcplejw.fsf@blackdown.de> <17510.55656.607530.491585@zot.electricrain.com> <1147657558.21291.65.camel@localhost.localdomain> Content-Type: text/plain Date: Mon, 15 May 2006 16:23:27 +1000 Message-Id: <1147674207.21291.101.camel@localhost.localdomain> Mime-Version: 1.0 Cc: linuxppc-dev@ozlabs.org, "Brian D. Carlstrom" List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On Mon, 2006-05-15 at 02:08 -0400, Dan Malek wrote: > On May 14, 2006, at 9:45 PM, Benjamin Herrenschmidt wrote: > > > .... For now, consider > > the load avergage value bogus :) > > It's not bogus :-) If you happen to schedule periodic > tasks that are sampled with load average computes, > or sleep them "uninterrupible", they are counted in > the load average. The thermal control runs every second and sleeps with "interruptible"... so it should not have such a laod average imho :) Now it's possible that what it does during that "second" is long enough to end up not sleeping that much (though it's also sleeping while doing i2c accesses, but that is uninterruptible for obvious reasons at the i2c driver level). Ben