From owner-freebsd-stable Thu Mar 8 14:40:11 2001 Delivered-To: freebsd-stable@freebsd.org Received: from bitbucket.extern.uniface.nl (bitbucket.extern.uniface.nl [193.78.88.139]) by hub.freebsd.org (Postfix) with ESMTP id F315F37B719 for ; Thu, 8 Mar 2001 14:40:04 -0800 (PST) (envelope-from driehuis@playbeing.org) Received: from bh2.nl.compuware.com (unknown [172.16.17.82]) by bitbucket.extern.uniface.nl (Postfix) with ESMTP id 5F0298287 for ; Thu, 8 Mar 2001 23:40:03 +0100 (CET) Received: from trashcan.nl.compuware.com ([172.16.16.52]) by bh2.nl.compuware.com with SMTP (Microsoft Exchange Internet Mail Service Version 5.5.2650.21) id GRMZR2X9; Thu, 8 Mar 2001 23:40:03 +0100 Received: from c1111.nl.compuware.com (c1111.nl.compuware.com [172.16.16.36]) by trashcan.nl.compuware.com (Postfix) with ESMTP id EDDCE145A4; Thu, 8 Mar 2001 23:40:02 +0100 (CET) Date: Thu, 8 Mar 2001 23:40:02 +0100 (CET) From: Bert Driehuis X-Sender: bertd@c1111.nl.compuware.com To: stable@FreeBSD.ORG Subject: Re: ABIT KT7 and temp monitoring In-Reply-To: <200103082209.f28M9hU04940@gollum.esys.ca> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On Thu, 8 Mar 2001, Lyndon Nerenberg wrote: > > healthd works fine on systems that use supported monitoring chips. > > Unfortunately the number of chips that healthd supports is pretty small > > compared to monitoring tools on other platforms, e.g. Motherboard Monitor on > > Windows or LM Sensors on Linux. > > Which says to me that the interface should be abstracted out. > Something like a device driver that presents a /dev device that > you can read or ioctl to get the information in a somewhat chip- > independent fashion. I've toyed with this idea (actually, I've implemented it for BSD/OS 4.x using the sysctl interface). However, there are two problems with it that made me drop the project for the time being. First of all, putting it in the kernel means a rigid interface, not to mention the associated bloat and near-constant kernel updates as new chips get glued to motherboards (and wired in new an innovative ways). Second, the user needs to define what sensor means what anyway, because of the ease of miswiring, and because a broken fan is indistinguisable from a nonexistant fan, which means a config file must be used anyway. Finally, none of the monitoring chips I've seen can be detected reliably. The ones attached to the ISA bus can be checked for sanity with some confidence (the W83781D has a one byte id), but scanning the I2C bus can easily lock a motherboard (I think that that's what's causing my Intel board to act up when the ichsmb driver gets added to the kernel). Cheers, -- Bert -- Bert Driehuis -- driehuis@playbeing.org -- +31-20-3116119 If the only tool you've got is an axe, every problem looks like fun! To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message