From owner-freebsd-current Wed Jan 7 16:15:27 1998 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id QAA27892 for current-outgoing; Wed, 7 Jan 1998 16:15:27 -0800 (PST) (envelope-from owner-freebsd-current) Received: from d198-232.uoregon.edu (d198-232.uoregon.edu [128.223.198.232]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id QAA27867; Wed, 7 Jan 1998 16:14:42 -0800 (PST) (envelope-from mini@d198-232.uoregon.edu) Received: (from mini@localhost) by d198-232.uoregon.edu (8.8.5/8.8.5) id QAA11337; Wed, 7 Jan 1998 16:13:54 -0800 (PST) Message-ID: <19980107161353.55587@micron.mini.net> Date: Wed, 7 Jan 1998 16:13:53 -0800 From: Jonathan Mini To: Julian Elischer Cc: current@FreeBSD.ORG Subject: Re: DEVFS-11 Reply-To: Jonathan Mini References: <199801071404.GAA11334@freefall.freebsd.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.88e In-Reply-To: <199801071404.GAA11334@freefall.freebsd.org>; from Julian Elischer on Wed, Jan 07, 1998 at 06:04:38AM -0800 X-files: The Truth is Out There Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Cool. :) Currently, I am implementing access to the display buffer of the video card via block device. (The VBE windowing system uses the device pager to handle efficient as well as coordinated synching) This means you can newfs and mount your video card. Hmmm... I think that wins my award for "coolest ram drive." (followed closely by a DOS ramdrive driver that uses the memory on a GUS. :) This one wins becauseyou get to "watch" the disk access and storage.) Seriously now, I have been using the block device system to deal with the video card, and I wondered whether or not this was "proper." I've noticed that the only block devices are disk drives of some sort or another. Is this merely coincidental, or are block devices "restricted" to being filesystem-type devices? *waits fearfully for a big explosion... * Julian Elischer stands accused of saying: > One little change and a fix > > change.. > You can now newfs /dev/rsd0 and mount it (/dev/sd0) without ever > disklabelling it. (its just a pile of blocks, why not have a FS on it?) > This is a byproduct of the fact that all slices appear identical, > whether or not they are subslices of subslices of subslices, or just a > whole raw disk. More interesting is the capacity to newfs and mount an > MBR slice (e.g. wd0s2) without going to the extra step of adding the > disklabel. Of course you need to disklabel to make a BOOTABLE part. :) > > This is "Even more dangeroulsy dedicated mode" > It is more useful for the floppy where you can now do: > newfs /dev/rfd0 floppy3 > > Without the disklabel step. and still mount it. > This wan't the AIM but it's a byproduct of the aim of making all > disk abstractions identical. > > > fix: > Can now mount filesystems from floppies again without crashing the system. > (Thanks to Mike Smith for falling into that pit..) > > julian > > This may be the version that gets checked in (Branch JULIAN_SLICES). -- Jonathan Mini Ingenious Productions Software Development P.O. Box 5693, Eugene, Or. 97405 "A child of five could understand this! Quick -- Fetch me a child of five."