Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S264585AbUF1BOt (ORCPT ); Sun, 27 Jun 2004 21:14:49 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S264609AbUF1BOt (ORCPT ); Sun, 27 Jun 2004 21:14:49 -0400 Received: from caramon.arm.linux.org.uk ([212.18.232.186]:57349 "EHLO caramon.arm.linux.org.uk") by vger.kernel.org with ESMTP id S264585AbUF1BOr (ORCPT ); Sun, 27 Jun 2004 21:14:47 -0400 Date: Mon, 28 Jun 2004 02:14:39 +0100 From: Russell King To: Chris Wedgwood Cc: Erik Jacobson , Christoph Hellwig , Jesse Barnes , Andrew Morton , Pat Gefre , linux-kernel@vger.kernel.org Subject: Re: [PATCH 2.6] Altix serial driver Message-ID: <20040628021439.A17654@flint.arm.linux.org.uk> Mail-Followup-To: Chris Wedgwood , Erik Jacobson , Christoph Hellwig , Jesse Barnes , Andrew Morton , Pat Gefre , linux-kernel@vger.kernel.org References: <20040623143801.74781235.akpm@osdl.org> <200406231754.56837.jbarnes@engr.sgi.com> <20040625083130.GA26557@infradead.org> <20040625124807.GA29937@infradead.org> <20040626235248.GC12761@taniwha.stupidest.org> <20040628003311.GA23017@taniwha.stupidest.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5.1i In-Reply-To: <20040628003311.GA23017@taniwha.stupidest.org>; from cw@f00f.org on Sun, Jun 27, 2004 at 05:33:11PM -0700 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1649 Lines: 37 On Sun, Jun 27, 2004 at 05:33:11PM -0700, Chris Wedgwood wrote: > On Sun, Jun 27, 2004 at 07:24:34PM -0500, Erik Jacobson wrote: > > Maybe you can help me clear it up then. When I feed serial core the > > name ttyS with TTY_MAJOR and minor 64, the registration fails. If I > > disable 8250 in the kernel config, the registration works for us. > > I'm not sure why the 8250 code reserves/registers ttyS0 on Altix when > do such hardware exists. I vaguely recall patching it to prevent this > in a hacky way. > > I would like to know why the 8250 code is registering a driver for > hardware that doesn't exist and see that fixed. It's the way its always been done, and the way the tty layer works. You register a range of ttys that you're going to be driving, and you own those ttys whether or not you actually have hardware for them. You can't say "ok, I have ttyS1 and ttyS3, I'll leave ttyS0 and ttyS2 available for someone else to use" because the tty layer just doesn't work like that. It has the notion of a single driver for a range of non-overlapping ttys. Yes, it would be nice to get rid of that limitation, but we're in a stable kernel series and the tty layer doesn't have a maintainer to do the work... -- Russell King Linux kernel 2.6 ARM Linux - http://www.arm.linux.org.uk/ maintainer of: 2.6 PCMCIA - http://pcmcia.arm.linux.org.uk/ 2.6 Serial core - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/