From owner-freebsd-current Fri Apr 5 15:14:12 1996 Return-Path: owner-current Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id PAA20430 for current-outgoing; Fri, 5 Apr 1996 15:14:12 -0800 (PST) Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.211]) by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id PAA20424 for ; Fri, 5 Apr 1996 15:14:09 -0800 (PST) Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id QAA25090; Fri, 5 Apr 1996 16:08:31 -0700 From: Terry Lambert Message-Id: <199604052308.QAA25090@phaeton.artisoft.com> Subject: Re: 2.2-960323 Panic in mount_msdos To: franky@pinewood.nl (Frank ten Wolde) Date: Fri, 5 Apr 1996 16:08:31 -0700 (MST) Cc: terry@lambert.org, freebsd-current@FreeBSD.org In-Reply-To: <9604050935.ZM9063@pwood1.pinewood.nl> from "Frank ten Wolde" at Apr 5, 96 09:35:25 am X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-current@FreeBSD.org X-Loop: FreeBSD.org Precedence: bulk > > Not bounds-checking dereferences isn't an error; it an optimization, > > and it's allowable because mount is not a user accessable command > > (you have to be root). > > I slightly disagree. Even as root I could make a typo and by mistake > specifiy the wrong partition/slice to mount, causing the entire system > to die. > > It would have been nice if some sanity checking would have been performed, > so that the kernel simply would abort the mount(2) system call with an > appropriate error (wrong FS type, or something similar). For instance, if you accidently type "/dev/kmem" instead of a disk device? Typo's are unlikely, since the "cannonically correct" way to do transient mounts that you expect to make is to put them as not mounted by default entries in the /etc/fstab. Then you mount them by device name, and typos get an error because they would need two device names. > I simply pointed out the panic. Maybe the code maintainer of the DOSFS > can use this info to make the system even more stable. Appreciated; but from what I understand, it is a *total* rewrite, so unless you get the developement code before it is released, any bug reports on the old system are bit-bucket fodder. I was explicit (possibly to the point of looking annoyed when I'm not) in the posting because of this. Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers.