From owner-freebsd-current Mon Nov 25 01:23:43 1996 Return-Path: owner-current Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id BAA19706 for current-outgoing; Mon, 25 Nov 1996 01:23:43 -0800 (PST) Received: from irz301.inf.tu-dresden.de (irz301.inf.tu-dresden.de [141.76.1.11]) by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id BAA19692 for ; Mon, 25 Nov 1996 01:23:35 -0800 (PST) Received: from sax.sax.de (sax.sax.de [193.175.26.33]) by irz301.inf.tu-dresden.de (8.6.12/8.6.12-s1) with ESMTP id KAA22538; Mon, 25 Nov 1996 10:22:59 +0100 Received: (from uucp@localhost) by sax.sax.de (8.6.12/8.6.12-s1) with UUCP id KAA10550; Mon, 25 Nov 1996 10:22:58 +0100 Received: (from j@localhost) by uriah.heep.sax.de (8.8.2/8.6.9) id KAA16329; Mon, 25 Nov 1996 10:04:54 +0100 (MET) From: J Wunsch Message-Id: <199611250904.KAA16329@uriah.heep.sax.de> Subject: Re: again, fixit floppy isn't useable... To: freebsd-current@FreeBSD.org (FreeBSD-current users) Date: Mon, 25 Nov 1996 10:04:54 +0100 (MET) Cc: andreas@klemm.gtn.com (Andreas Klemm) In-Reply-To: from Andreas Klemm at "Nov 24, 96 11:11:45 pm" X-Phone: +49-351-2012 669 X-PGP-Fingerprint: DC 47 E6 E4 FF A6 E9 8F 93 21 E0 7D F9 12 D6 4E X-Mailer: ELM [version 2.4ME+ PL17 (25)] 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 As Andreas Klemm wrote: > I saw in a commit log from Joerg, that he does some fixes to the > fixit floppy generation. Don't know exactly what fixes. But today That's your problem. You should probably watch the commit logs more closely, then you knew what has been fixed, and what not. > But both, the one from 2.1.6 and from 2.2-ALPHA are still broken. Nobody cares about 2.1.6. People ``buying'' this system most likely do already have their own recovery strategy, since it's targeted to ISPs and such. Even if they don't have, they wouldn't stumple across a defunct MAKEDEV script since they know how to mknod manually. Namely the regexp fix that finally fixed the `expr' on the fixit floppy wasn't yet in 2.2-ALPHA either. I wasn't 101.1 % sure about whether all people would like the way i've fixed it, and hence imposed a `grace period' before pulling the fix into the 2.2 branch. It's in now, so this one should work in 2.2-BETA once this is ready. > fixit# mt status > /dev/nrst0: no such file or directory Quite possible. There's no space for any and all device nodes, in particular not on the root file system (which has _nothing_ to do with the fixit floppy). Maybe export TAPE /mnt2/dev/nrst0 mt status might have worked better. :) (If i don't forget it, i can take the export statement into the fixit's .profile.) > The device nrst0 is only present on the fixit floppy, under /mnt2. Yes, so whaddayawant? There's only limited space in the MFS root f/s, and we don't have DEVFS ready either (which will be the way out of all that shit). > fixit# sh MAKEDEV sd0s3e > expr: not found see above > fixit# sh MAKEDEV sd0s3e > /mnt2: write failed, file system full > Memory fault > /mnt2: write failed, file system full > Memory fault > bad unit for disk in: sd0s3e(unit=,slice=,part=) > > No space ???? Yes. Read the message, PLEEZE! Memory fault -- _that's_ what caused the ENOSPC. There wasn't enough space for a 600 KB coredump neither on the root MFS, nor on the fixit floppy. Did you expect that much space there? :) > The only difference is, that the 2.2-ALPHA floppy had an vi and > an ed. The 2.1.6 boot/fixit floppy had only and ed, vi was missing. RTFcommitmessages. vi has been left out due to space constraints. > Well, I see no possibility to restore a crashed system :-( Andreas, you are far more than a Unix' beginner. I still don't get your whining, in particular not if the reported problems are being worked on. I expect from you that you could figure out how to use `mknod' manually to create the desired device nodes, at least after reading the MAKEDEV script. Again: most of this _is_ fixed by now. 2.2-ALPHA ain't `now'. Things have been going on, and your comments were welcomed and recognized. What we can't fix with the current scheme is the lack of a multitude of device nodes, either in the root MFS or on the fixit floppy. This has to be left for 3.0 which will hopefully and finally have a working DEVFS. -- cheers, J"org joerg_wunsch@uriah.heep.sax.de -- http://www.sax.de/~joerg/ -- NIC: JW11-RIPE Never trust an operating system you don't have sources for. ;-)