From mboxrd@z Thu Jan 1 00:00:00 1970 From: Benjamin Herrenschmidt Subject: Re: fb_console_init fix. Date: Sat, 21 Feb 2004 09:56:56 +1100 Sender: linux-fbdev-devel-admin@lists.sourceforge.net Message-ID: <1077317816.9623.20.camel@gaston> References: Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Return-path: Received: from sc8-sf-mx1-b.sourceforge.net ([10.3.1.11] helo=sc8-sf-mx1.sourceforge.net) by sc8-sf-list1.sourceforge.net with esmtp (Exim 4.30) id 1AuJlI-0004K1-FU for linux-fbdev-devel@lists.sourceforge.net; Fri, 20 Feb 2004 15:08:48 -0800 Received: from gate.crashing.org ([63.228.1.57] ident=root) by sc8-sf-mx1.sourceforge.net with esmtp (TLSv1:AES256-SHA:256) (Exim 4.30) id 1AuJfC-0002GW-JX for linux-fbdev-devel@lists.sourceforge.net; Fri, 20 Feb 2004 15:02:30 -0800 In-Reply-To: Errors-To: linux-fbdev-devel-admin@lists.sourceforge.net List-Unsubscribe: , List-Id: List-Post: List-Help: List-Subscribe: , List-Archive: Content-Type: text/plain; charset="us-ascii" To: James Simmons Cc: Linus Torvalds , Linux Fbdev development list , Linux Kernel Mailing List On Sat, 2004-02-21 at 08:58, James Simmons wrote: > Hi! > > This patch fixes fb_console_init from being called twice. I still need > to fix set_con2fb but this helps but this is still important to get in. I hate those #ifdef MODULE. They are plaguing the drivers too, because of our crazy initialization scheme... (Linus: that patch should still go in, doing better is probably a 2.7 target in this case). If we want the fb stuff to be initialized before other drivers, we probably want to create an init step before devices and after PCI probe, but is that really necessary ? The core fbdev should just need a subsys initcall. fbcon could use that too and register a notifier. Then we could use the notifier mecanism to notify fbcon when fbdev's are added (from register_framebuffer) and get rid of all of the init crap The only remaining junk is to deal with the setup= stuff, but then, it could be done by having the drivers provide their "name" string & setup function in the ops structure passed on register, and have us call back the driver straight away, though that's a big change of the way drivers are initialized. Or we could add an fb_check_setup("name", my_driver_setup) call to be done by the drivers before register_framebuffer is called Ben . ------------------------------------------------------- SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps & Web services for Linux with a free DVD software kit from IBM. Click Now! http://ads.osdn.com/?ad_id=1356&alloc_id=3438&op=click From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261159AbUBTXCi (ORCPT ); Fri, 20 Feb 2004 18:02:38 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261419AbUBTXCf (ORCPT ); Fri, 20 Feb 2004 18:02:35 -0500 Received: from gate.crashing.org ([63.228.1.57]:46763 "EHLO gate.crashing.org") by vger.kernel.org with ESMTP id S261159AbUBTXCc (ORCPT ); Fri, 20 Feb 2004 18:02:32 -0500 Subject: Re: fb_console_init fix. From: Benjamin Herrenschmidt To: James Simmons Cc: Linus Torvalds , Linux Fbdev development list , Linux Kernel Mailing List In-Reply-To: References: Content-Type: text/plain Message-Id: <1077317816.9623.20.camel@gaston> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.4.5 Date: Sat, 21 Feb 2004 09:56:56 +1100 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Sat, 2004-02-21 at 08:58, James Simmons wrote: > Hi! > > This patch fixes fb_console_init from being called twice. I still need > to fix set_con2fb but this helps but this is still important to get in. I hate those #ifdef MODULE. They are plaguing the drivers too, because of our crazy initialization scheme... (Linus: that patch should still go in, doing better is probably a 2.7 target in this case). If we want the fb stuff to be initialized before other drivers, we probably want to create an init step before devices and after PCI probe, but is that really necessary ? The core fbdev should just need a subsys initcall. fbcon could use that too and register a notifier. Then we could use the notifier mecanism to notify fbcon when fbdev's are added (from register_framebuffer) and get rid of all of the init crap The only remaining junk is to deal with the setup= stuff, but then, it could be done by having the drivers provide their "name" string & setup function in the ops structure passed on register, and have us call back the driver straight away, though that's a big change of the way drivers are initialized. Or we could add an fb_check_setup("name", my_driver_setup) call to be done by the drivers before register_framebuffer is called Ben .