From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S271189AbTG1W6o (ORCPT ); Mon, 28 Jul 2003 18:58:44 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S271191AbTG1W6o (ORCPT ); Mon, 28 Jul 2003 18:58:44 -0400 Received: from mailhost.tue.nl ([131.155.2.7]:36882 "EHLO mailhost.tue.nl") by vger.kernel.org with ESMTP id S271189AbTG1W6k (ORCPT ); Mon, 28 Jul 2003 18:58:40 -0400 Date: Tue, 29 Jul 2003 00:58:38 +0200 From: Andries Brouwer To: Claas Langbehn Cc: dean gaudet , linux-kernel@vger.kernel.org Subject: Re: 2.6.0-test2 has i8042 mux problems Message-ID: <20030728225838.GA1845@win.tue.nl> References: <20030728052614.GA5022@rootdir.de> <20030728113626.GA1706@win.tue.nl> <20030728210910.GA832@rootdir.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20030728210910.GA832@rootdir.de> User-Agent: Mutt/1.3.25i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Mon, Jul 28, 2003 at 11:09:10PM +0200, Claas Langbehn wrote: > Hello! > > its strange. After switching debug on, it suddenly works. Maybe the timing changed? [It would also be nice to see a log where things failed. Probably syslog would still work.] > Below my debug info. Thanks! Let me comment it. ------------------------------------------------------------- check_mux: First we probe for a Synaptics mux. Find none. check_aux: Then we probe for a PS/2 mouse port. Find one. Good. Found something. Call serio_register_port(). This searches for attached devices. We know already that it is the PS/2 mouse port, but who knows what might be connected. Let us try all devices we know about. Is it perhaps a keyboard? > drivers/input/serio/i8042.c: d4 -> i8042 (command) [0] > drivers/input/serio/i8042.c: f2 -> i8042 (parameter) [0] We wait for 99 jiffies, but no keyboard ID shows up. Then try to set LEDs. > drivers/input/serio/i8042.c: d4 -> i8042 (command) [99] > drivers/input/serio/i8042.c: ed -> i8042 (parameter) [99] > drivers/input/serio/i8042.c: fe <- i8042 (interrupt, aux, 0, timeout) [99] Get an error back. This mouse doesnt want to set its LEDs. Is it perhaps a mouse? > drivers/input/serio/i8042.c: d4 -> i8042 (command) [113] > drivers/input/serio/i8042.c: f2 -> i8042 (parameter) [113] No mouse ID either. No devices connected to this port, but at least we have a port: > serio: i8042 AUX port at 0x60,0x64 irq 12 Then we continue, and check the keyboard side of the i8042 controller. Things work satisfactorily, we detect a keyboard. Silly enough we set it to scancode Set 2, but nevertheless give commands that are only meaningful for scancode Set 3. Keyboards that know only about Set 2 will be unhappy. > input: AT Set 2 keyboard on isa0060/serio0 > serio: i8042 KBD port at 0x60,0x64 irq 1 OK, found keyboard. ------------------------------------------------------------- So far the story of your debug file. What is your actual hardware? Is there a mouse? Andries