From owner-freebsd-hackers Wed Oct 7 15:28:39 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id PAA09699 for freebsd-hackers-outgoing; Wed, 7 Oct 1998 15:28:39 -0700 (PDT) (envelope-from owner-freebsd-hackers@FreeBSD.ORG) Received: from lariat.lariat.org (lariat.lariat.org [206.100.185.2]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id PAA09660 for ; Wed, 7 Oct 1998 15:28:34 -0700 (PDT) (envelope-from brett@lariat.org) Received: (from brett@localhost) by lariat.lariat.org (8.8.8/8.8.6) id QAA27135; Wed, 7 Oct 1998 16:28:28 -0600 (MDT) Message-Id: <4.1.19981007162637.04032380@mail.lariat.org> X-Sender: brett@mail.lariat.org X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Date: Wed, 07 Oct 1998 16:27:17 -0600 To: hackers@FreeBSD.ORG From: Brett Glass Subject: Another instance of the crash I was seeing Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Just a few days ago, Jim Dennis of Starshine encountered exactly the same crash I've been experiencing here (with FreeBSD 2.2.7). As you may recall, our backup script, which piped dump through gzip to ftp, was causing the machine to crash with a seg fault. The debugger said that the seg fault occurred in the kernel while it was in an idle loop.... Possibly waiting for completion of an IDE command. (I suspect this is indeed where the problem occurred, because the machine actually had a lot of work to do in userland at the time; CPU usage was pegged. But the IDE driver blocks and busy-waits for command completion in the kernel, so we would see it enter the idle loop while it was waiting.) My SWAG is that the problem is most likely in the IDE/ATAPI or ed0 drivers, and/or in interrupt handling during the busy wait. But I have no idea how to figure out where the problem is. Jim's crashes, like the ones I've been seeing, occurred while dumping disk data to and from the network under relatively high CPU loads. He posted a description to -questions recently. Has anyone else seen this? It's beginning to look like a pattern, and we need to nail it so that we can reinstate our backup procedure. (The one we're using now is a stopgap.) --Brett To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message