From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1030209AbWBYLHt (ORCPT ); Sat, 25 Feb 2006 06:07:49 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1030212AbWBYLHt (ORCPT ); Sat, 25 Feb 2006 06:07:49 -0500 Received: from liaag1af.mx.compuserve.com ([149.174.40.32]:17623 "EHLO liaag1af.mx.compuserve.com") by vger.kernel.org with ESMTP id S1030209AbWBYLHt (ORCPT ); Sat, 25 Feb 2006 06:07:49 -0500 Date: Sat, 25 Feb 2006 06:03:41 -0500 From: Chuck Ebbert <76306.1226@compuserve.com> Subject: Re: [-mm patch] x86: start early_printk at sensible screen row To: Andrew Morton Cc: linux-kernel@vger.kernel.org, stsp@aknet.ru, ak@suse.de Message-ID: <200602250607_MC3-1-B93B-782@compuserve.com> MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org In-Reply-To: <20060225001533.3b06bbe0.akpm@osdl.org> On Sat, 25 Feb 2006 at 00:15:33 -0800, Andrew Morton wrote: > > > > Use boot info to start early_printk() at the proper row on VGA console. > > > > Define "proper". > > > > > - current_ypos = max_ypos; > > + current_ypos = SCREEN_INFO.orig_y; > > Is that the place at which the boot loader left the cursor? Yes, that's what I should have said instead of "proper". So far I can't break it; after booting with "vga=ask" and doing some things at the prompt, the kernel output still starts at the right place. But the emergency fallback early_printk still starts at a fixed place and the x86_64 'direct mapping' message will screw things up when there's no earlyprintk boot parameter. -- Chuck "Equations are the Devil's sentences." --Stephen Colbert