From owner-freebsd-hackers Sat Sep 21 10:59:28 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id KAA11400 for hackers-outgoing; Sat, 21 Sep 1996 10:59:28 -0700 (PDT) Received: from DeepCore.dk (dial54.cybercity.dk [194.16.56.54]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id KAA11356 for ; Sat, 21 Sep 1996 10:59:18 -0700 (PDT) Received: (from sos@localhost) by DeepCore.dk (8.7.6/8.7.3) id QAA00459; Sat, 21 Sep 1996 16:36:57 +0200 (MET DST) Message-Id: <199609211436.QAA00459@DeepCore.dk> Subject: Re: SVGATextMode under FBSD To: proff@suburbia.net (Julian Assange) Date: Sat, 21 Sep 1996 16:36:57 +0200 (MET DST) Cc: hackers@FreeBSD.org In-Reply-To: <199609210731.RAA15662@suburbia.net> from Julian Assange at "Sep 21, 96 05:31:35 pm" From: sos@FreeBSD.org Reply-to: sos@FreeBSD.org X-Mailer: ELM [version 2.4ME+ PL15 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-hackers@FreeBSD.org X-Loop: FreeBSD.org Precedence: bulk In reply to Julian Assange who wrote: > > I have a fetish for large text screens (the one I'm writing this email on > is 180x60 (8x16 font) 10800 chars on screen at once, that's 5 x as much > information before my brain). FreeBSD's inability to do this is the > biggest problem I have with it, so I've started porting linux's SVGATextMode > package. Oh, someone down that road again :) > Which are really just backends to KDADDIO|KDDISABIO and KDENABIO|KDDELIO > respectively. > > Linux impliments KDADDIO/ioperm as a 128 byte bitmap for each process. > This is the way I'd like to see it in FreeBSD, but I do not understand > the how inb/outb interacts with the mm system well enough to do this > confidently (comments on this?). As a crude hack, I intend to wrap > KDADDIO|ioperm in linux_compat to KDENABIO. iopl wraps cleanly. This > should be enough to run svgalib programs, which is a big win, IMEO. We have no prvision for doing this (yet) it _might_ be done in not so long time.. > ioctl's used for VT resizing etc, most of which do not have syscon/pcvt > counterparts, but all of which should be relatively easy to impliment. > In terms of the linux emmulation code, should these be implimented in > syscons() using the linux defines/values, or as KD* extensions, with > FreeBSD values, and a linux_compat ioctl() wrapper? I'm in favour of > VT_* in order to avoid polluting the USL name space. Syscons uses the winsize structure, and reports back when there is a resolution change :), the extra VT defines is a linux bogosity :) All, in all, the way this should be done, is writing a lkm module that can be loaded and adds the prober ioctl's to the console driver (well at least syscons). The provisions for this is in the next update I have for syscons (most of it is allready there). If you want to do it this way, I'd help where I can, I might know a trick or two in this area :) -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Søren Schmidt (sos@FreeBSD.org) FreeBSD Core Team Even more code to hack -- will it ever end ..