From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S265397AbTLHNSK (ORCPT ); Mon, 8 Dec 2003 08:18:10 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S265398AbTLHNSK (ORCPT ); Mon, 8 Dec 2003 08:18:10 -0500 Received: from hastur.physics.ox.ac.uk ([163.1.243.65]:22694 "EHLO janus.physics.ox.ac.uk") by vger.kernel.org with ESMTP id S265397AbTLHNSH (ORCPT ); Mon, 8 Dec 2003 08:18:07 -0500 From: Tim Stadelmann Organization: University of Oxford To: linux-kernel@vger.kernel.org Date: Mon, 8 Dec 2003 13:12:58 +0000 User-Agent: KMail/1.5.4 MIME-Version: 1.0 Content-Disposition: inline Message-Id: <200312081312.58472.stadelmann@physics.ox.ac.uk> Subject: Dell Laptop APM problems/noise Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit X-SA-Exim-Mail-From: stadelmann@physics.ox.ac.uk Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org I observe the same power supply (?) noise as Jean-Marc on my Dell Latitude C600 with 2.6.0-test10 and -test11, if APM is enabled and the kernel is allowed to tell the BIOS that the CPU is idle. Some remarks, which I hope may be helpful: - in 2.6, APM idle calls produce the noise, but do not conserve power. This could be either because the call fails to have the desired effect, or because whatever causes the oscillation uses so much power that the effect is offset. Disabling idle calls avoids the oscillation, but reduces the battery life dramaticall if compared with 2.4 kernels or 2.6 using ACPI. - unlike Jean-Marc, I do not see the problem when using ACPI. Battery life is comparable to 2.4 with APM enabled. (of course, suspend does not work, but that's a different story...) - with 2.4 kernels, APM *does* work as expected. No oscillation happens, and the battery life is prolonged significantly when allowing idle calls. Something relevant must have changed. Regards, Tim Stadelmann PS: please CC questions and remarks to my personal address