From owner-freebsd-acpi@FreeBSD.ORG Tue May 31 17:14:58 2005 Return-Path: X-Original-To: freebsd-acpi@freebsd.org Delivered-To: freebsd-acpi@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 00AB916A41C for ; Tue, 31 May 2005 17:14:57 +0000 (GMT) (envelope-from nate@root.org) Received: from www.cryptography.com (li-22.members.linode.com [64.5.53.22]) by mx1.FreeBSD.org (Postfix) with ESMTP id B2AAD43D4C for ; Tue, 31 May 2005 17:14:57 +0000 (GMT) (envelope-from nate@root.org) Received: from [10.0.0.34] (adsl-67-119-74-222.dsl.sntc01.pacbell.net [67.119.74.222]) by www.cryptography.com (8.12.8/8.12.8) with ESMTP id j4VHEtZ6014155 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NOT); Tue, 31 May 2005 10:14:56 -0700 Message-ID: <429C9B56.2000802@root.org> Date: Tue, 31 May 2005 10:13:58 -0700 From: Nate Lawson User-Agent: Mozilla Thunderbird 1.0.2 (Windows/20050317) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Takashi Inoue References: <429ADD9D.90506@sophia.ac.jp> <429B6A8E.1070906@root.org> <429BCC04.3080207@sophia.ac.jp> In-Reply-To: <429BCC04.3080207@sophia.ac.jp> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: freebsd-acpi@freebsd.org Subject: Re: cpufreq on 5.4R X-BeenThere: freebsd-acpi@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: ACPI and power management development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 31 May 2005 17:14:58 -0000 Takashi Inoue wrote: > Hi Nate, > > Thanks for your post. > >> Only acpi_throttle is supported in the 5.4 version of cpufreq. I'm >> hoping to get an MFC of the -current version in soon although there >> are a couple nits I'd like to fix there first. > > I see. I'll try the throttle. I hope your excelent works are MFCed to 5 > soon. acpi_throttle is compiled into acpi.ko so if you are using ACPI, you already have acpi_throttle available if your system supports it. >> I'm not sure the port works properly with cpufreq. A better thing is >> to use the est.c from -current. Or just wait for MFC. > > I see. I'll also try cvsup current cpufreq when I have time. > By the way, I cannot see the difference between est and ichss. > Should I use est insted of ichss? ThinkPad X40 has an Intel i855GME > chipset. > I thought I can use either est or ichss and ichss is better. > Maybe I'm confused. There were three versions of SpeedStep that are supported in -current. The oldest, SMI-based SS is present in 440bx parts and is supported by the smist driver. The next oldest is system IO based SS, which is an evolution of smist and is supported by the ichss driver. Both offer only two speeds, full speed and some lower rate (typically ~70% of full speed). Both are supported by the chipset alone with no CPU interaction. The newest, Pentium-M type SpeedStep, is accessed via MSRs and sometimes chipset-assisted system IO. It offers multiple speeds and steps both voltage and clocks, saving much more power. It is supported by the est driver. Both ichss and est capabilities are sometimes exported by the BIOS through ACPI Px states, supported by the acpi_perf driver. Enabling ACPI provides support for acpi_throttle and acpi_perf. Enabling cpufreq enables support for smist, ichss, and est. The OS should autoprobe and enable the best one supported by your system, in this order: acpi_perf > est > ichss > smist -- Nate