From owner-freebsd-stable@FreeBSD.ORG Sat May 6 16:29:58 2006 Return-Path: X-Original-To: freebsd-stable@freebsd.org Delivered-To: freebsd-stable@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 1DE5716A416 for ; Sat, 6 May 2006 16:29:58 +0000 (UTC) (envelope-from rwatson@FreeBSD.org) Received: from cyrus.watson.org (cyrus.watson.org [209.31.154.42]) by mx1.FreeBSD.org (Postfix) with ESMTP id 794F643D46 for ; Sat, 6 May 2006 16:29:57 +0000 (GMT) (envelope-from rwatson@FreeBSD.org) Received: from fledge.watson.org (fledge.watson.org [209.31.154.41]) by cyrus.watson.org (Postfix) with ESMTP id 9D43446C4A; Sat, 6 May 2006 12:29:56 -0400 (EDT) Date: Sat, 6 May 2006 17:29:56 +0100 (BST) From: Robert Watson X-X-Sender: robert@fledge.watson.org To: Kris Kennaway In-Reply-To: <20060506153154.GA43951@xor.obsecurity.org> Message-ID: <20060506171118.A17611@fledge.watson.org> References: <20060505011127.dadc75f8.rosti.bsd@gmail.com> <200605051001.22929.doconnor@gsoft.com.au> <20060506030333.b5c4bccf.rosti.bsd@gmail.com> <200605060936.32792.doconnor@gsoft.com.au> <20060506181859.dead1bd4.rosti.bsd@gmail.com> <20060506153154.GA43951@xor.obsecurity.org> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Cc: Rostislav Krasny , freebsd-stable@freebsd.org Subject: Re: system crash during file copy to a floppy with bad sectors X-BeenThere: freebsd-stable@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Production branch of FreeBSD source code List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 06 May 2006 16:29:58 -0000 On Sat, 6 May 2006, Kris Kennaway wrote: >>> On Saturday 06 May 2006 09:33, Rostislav Krasny wrote: >>>>> Can you get a back trace? ie enable crash dumps and do it again, or >>>>> transcribe, or photograph the screen as it panics if you are local. >>>> >>>> Unfortunately I cannot reproduce it now. Doesn't the old log help? >>>> There is one "Fatal trap 12". >>> >>> Only if you have the backtrace I think. >> >> In attempt to "emulate" bad sectors I got another crash, that looks similar >> (but not exactly same) to the original two crashes. I just pulled out the >> diskette before file copy is finished. I know, that must never be done, but >> sometimes people are in a great hurry or just do mistakes. IMHO that isn't >> a good reason for a system crash. > > It's a known and hard-to-fix problem, though. BTW, this is generally a question of resilience of individual file systems to on-disk corruption and failures. UFS is quite tricky to make robust against disk destruction, since it relies on strong consistenty guarantees for performance reasons. FAT, on the other hand, is a file system we should be able to make more robust quite a bit more easily. I think there would really be two phases to such a project: (1) Make FAT (msdosfs) MPSAFE, perform general cleanup. (2) Improve robustiness in the face of media failure. The reason FAT is particularly interesting, of course, is its widespread use on removable media such as USB sticks. Robert N M Watson