From owner-freebsd-hackers Tue Mar 21 05:53:48 1995 Return-Path: hackers-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id FAA20516 for hackers-outgoing; Tue, 21 Mar 1995 05:53:48 -0800 Received: from deacon.cogsci.ed.ac.uk (deacon.cogsci.ed.ac.uk [129.215.144.7]) by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id FAA20491 for ; Tue, 21 Mar 1995 05:52:01 -0800 Received: (from richard@localhost) by deacon.cogsci.ed.ac.uk (8.6.10/8.6.9) id NAA04769; Tue, 21 Mar 1995 13:47:20 GMT Date: Tue, 21 Mar 1995 13:47:20 GMT Message-Id: <199503211347.NAA04769@deacon.cogsci.ed.ac.uk> From: Richard Tobin Subject: Re: NMI Error success story To: Nate Williams , hackers@FreeBSD.org In-Reply-To: Nate Williams's message of Sat, 18 Mar 1995 23:31:35 -0700 Organization: just say no Sender: hackers-owner@FreeBSD.org Precedence: bulk My home machine is a 386DX33 with an EISA bus, about 5 years old now. Since the first release of 386BSD I've had problems with it getting NMIs and then panicing, apparently during heavy disk (IDE) i/o (often during fsck). The problem vanishes completely if I remove the "halt" instruction from the idle loop (incidentally, this means I had no problem with NetBSD 1.0, which doesn't use the halt instruction). I don't understand why this is, but it's something that others having a similar problem might try. -- Richard