From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751485AbWFJLN6 (ORCPT ); Sat, 10 Jun 2006 07:13:58 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751488AbWFJLN6 (ORCPT ); Sat, 10 Jun 2006 07:13:58 -0400 Received: from smtprelay01.ispgateway.de ([80.67.18.13]:7565 "EHLO smtprelay01.ispgateway.de") by vger.kernel.org with ESMTP id S1751485AbWFJLN5 (ORCPT ); Sat, 10 Jun 2006 07:13:57 -0400 From: Ingo Oeser To: "Antonino A. Daplas" Subject: Re: [PATCH 3/5] VT binding: Update fbcon to support binding Date: Sat, 10 Jun 2006 13:10:02 +0200 User-Agent: KMail/1.9.3 Cc: Andrew Morton , Linux Fbdev development list , Linux Kernel Development References: <448933EB.3070003@gmail.com> In-Reply-To: <448933EB.3070003@gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200606101310.04174.ioe-lkml@rameria.de> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Hi Antonio, On Friday, 9. June 2006 10:40, Antonino A. Daplas wrote: > 5. When fbcon is completely unbound from the console layer, fbcon will > also release (iow, decrement module reference counts to zero) all fbdev > drivers. In other words, a bind or unbind request from the console layer > will propagate down to the framebuffer drivers. > > 6. If fbcon is not bound to the console, it will ignore all notifier > events (except driver registration and unregistration) and all sysfs > requests. Wow! Now one can: - implement different framebuffer drivers for a chip. - try a stable and development version without rebooting, - have probing with user interaction ("if you see me please press enter else I will try the previous driver in 15 seconds.") instead of deciding on "failsafe" (vga/vesa) and "fast" (special driver) at boot. I love it! Just the take_over_console() as alternative API looks strange. It's ok as an intermeditate step (likely, if I remember the discussions correctly). If it should stay, it should require the console to be registered first. Can be changed later, once as you make progress and don't reach a release kernel. This is NO show stopper for me. Progress is the most important thing for development in this areas until it reaches a release kernel. Regards Ingo Oeser