From owner-freebsd-current@FreeBSD.ORG Wed Jan 14 07:18:02 2004 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 023CA16A4CE for ; Wed, 14 Jan 2004 07:18:02 -0800 (PST) Received: from smtp3.andrew.cmu.edu (SMTP3.andrew.cmu.edu [128.2.10.83]) by mx1.FreeBSD.org (Postfix) with ESMTP id 6138143D67 for ; Wed, 14 Jan 2004 07:17:53 -0800 (PST) (envelope-from dpapasia@andrew.cmu.edu) Received: from DAMNLEFTIST.RES.cmu.edu (DAMNLEFTIST.RES.cmu.edu [128.2.164.41]) (user=dpapasia mech=GSSAPI (0 bits))i0EFHqi8003060 for ; Wed, 14 Jan 2004 10:17:52 -0500 Date: Wed, 14 Jan 2004 10:18:25 -0500 (EST) From: Daniel Papasian X-X-Sender: dpapasia@damnleftist.res.cmu.edu To: freebsd-current@freebsd.org Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Subject: 5.2-RELEASE: Possible bug in filesystem code? X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 14 Jan 2004 15:18:02 -0000 I really hate to potentially create some distress over something that may be very difficult to repeat, but... I was working on a 5.2-RELEASE system that had been improperly shut down, and while the background process was checking/fixing the consistency of the filesystem, I did a relatively disk-intensive process (installed a port) and the system crashed in an apparent panic. I then attempted to see if this repeated, and sure enough it did, twice. So I built a kernel with debugging options and prepared to get a crash dump. To see if I hadn't tweaked inadvertently fixed the problem, I repeated the crash with the debugging kernel but before I had set it to save core by calling dumpon, and sure enough the system did crash again. In retrospect, this was a huge mistake, because the next time, when dumpon was called and I was ready to get a crash dump, the crash did not repeat. A case of ridiculously bad luck? It sounds it. When I get some time this weekend, I'll try a bit harder to repeat the problem and get a crashdump, but considering the nature of what I suspect the problem to be, I'm guessing it was caused by a specific inconsistency in the filesystem from the improper shutdown, something that has since been corrected, and would therefore be hard to reproduce. The filesystem used softupdates. I'm wondering if anyone has had any similar experiences, or if this is somehow a known problem in any way. -Daniel Papasian