From owner-freebsd-acpi@FreeBSD.ORG Wed Apr 28 11:45:42 2004 Return-Path: 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 BCFE616A4CE for ; Wed, 28 Apr 2004 11:45:42 -0700 (PDT) Received: from root.org (root.org [67.118.192.226]) by mx1.FreeBSD.org (Postfix) with SMTP id BC95743D2F for ; Wed, 28 Apr 2004 11:45:41 -0700 (PDT) (envelope-from nate@root.org) Received: (qmail 88278 invoked by uid 1000); 28 Apr 2004 18:45:42 -0000 Date: Wed, 28 Apr 2004 11:45:42 -0700 (PDT) From: Nate Lawson To: Jeffrey Katcher In-Reply-To: <20040428160504.36451.qmail@web41114.mail.yahoo.com> Message-ID: <20040428114036.R88145@root.org> References: <20040428160504.36451.qmail@web41114.mail.yahoo.com> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII cc: freebsd-acpi@freebsd.org Subject: Re: Fan Control Success on IBM T40? (another quick Q) X-BeenThere: freebsd-acpi@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: ACPI and power management development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 28 Apr 2004 18:45:42 -0000 On Wed, 28 Apr 2004, Jeffrey Katcher wrote: > >From everything I've seen (on the linux-acpi lists), the T40 doesn't have _any_ > fan control in the DSDT. Apparently it's explicitly controlled via SMBIOS. > There's a SMAPI interface available from FreeBSD, but I don't have enough info > yet to exploit it. That's similar to my T23. It actually handles fans automatically with SMM code and does not export the info through ACPI. > Second question, still on power/clock issues, but diverging a little from ACPI: > What's the interaction between SpeedStep clock and ACPI clock percentage? > If I start on batteries, apparently the clock is at 600Mhz (even though dmesg > reports it as 1600Mhz no matter what). ACPI CPU then can reduce this base by > eighths? So if I start on AC, it's eighths of 1600 and DC, it's eighths of > 600? SpeedStep and ACPI performance states are clock/voltage. Throttling is a % of the current clock. So they are additive. I'm working on a cpufreq driver for the former and acpi_cpu supports the latter currently. Thinkpads automatically switch SpeedStep based on AC line at boot time so if I boot with the AC line unplugged, I get the lowest clock/voltage combo. Throttling, if activated, would be in addition to that. So if your chipset, like mine, offers 1 Ghz and 733 Mhz clock/voltage settings, and 100% and 50% throttling settings, you have four actual settings: 1000 mhz 733 mhz 500 mhz (1 ghz + 50% throttle) 366 mhz (733 mhz + 50% throttle) You won't be able to adjust SpeedStep-like settings until the cpufreq driver is finished. It will automatically combine the two additive settings and export a unified interface that just allows you to select the clock rate. -Nate