From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261471AbTJMG05 (ORCPT ); Mon, 13 Oct 2003 02:26:57 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261492AbTJMG05 (ORCPT ); Mon, 13 Oct 2003 02:26:57 -0400 Received: from user-118bgnr.cable.mindspring.com ([66.133.194.251]:42629 "EHLO BL4ST") by vger.kernel.org with ESMTP id S261471AbTJMG04 (ORCPT ); Mon, 13 Oct 2003 02:26:56 -0400 Date: Sun, 12 Oct 2003 23:27:27 -0700 From: Eric Wong To: linux-kernel , Greg Norris Subject: Re: retaining use of the mouse wheel with "psmouse_noext=1"? Message-ID: <20031013062727.GB6958@BL4ST> References: <20031011202411.GA1397@glitch.localdomain> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20031011202411.GA1397@glitch.localdomain> User-Agent: Mutt/1.5.4i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Greg Norris wrote: > I have to use the "psmouse_noext=1" option to psmouse.ko, in order to > keep the pointer from going stark-raving mad[1] when I use my KVM > switch. Unfortunately, this seems to squash the use of the mouse wheel > for scrolling (it still works as a 3rd button), which is a feature I'd > really like to retain. Is there any way to keep the wheel fully > functional when using this option (or otherwise fix the KVM issue)? > It's a Belkin SOHO 4-port switch, if that makes a difference. A stab in the dark (I don't have a KVM handy, or even a mouse at the moment to test on) but you could try editing the psmouse_extensions() function in drivers/input/mouse/psmouse-base.c : if (psmouse_noext) return PSMOUSE_PS2; replace PSMOUSE_PS2 with another one of the defined variables found near the bottom of: drivers/input/mouse/psmouse.h You should be able to figure out which name most closely matches you mouse by looking at the psmouse_extensions() function. -- Eric Wong