From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dmitry Torokhov Subject: Re: Touchpad toggle mess Date: Thu, 4 Nov 2010 19:20:05 -0700 Message-ID: <20101105022004.GA26173@core.coreip.homeip.net> References: <1288915209.5418.72.camel@novo.hadess.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from mail-gw0-f46.google.com ([74.125.83.46]:45376 "EHLO mail-gw0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752087Ab0KECUQ (ORCPT ); Thu, 4 Nov 2010 22:20:16 -0400 Content-Disposition: inline In-Reply-To: <1288915209.5418.72.camel@novo.hadess.net> Sender: linux-input-owner@vger.kernel.org List-Id: linux-input@vger.kernel.org To: Bastien Nocera Cc: linux-input , linux-hotplug@vger.kernel.org, xorg@lists.freedesktop.org, Peter Hutterer , Matthew Garrett , Martin Pitt Hi Bastien, On Fri, Nov 05, 2010 at 12:00:09AM +0000, Bastien Nocera wrote: > Heya, > > In GNOME 2.32, we added support for touchpad toggle buttons, by > capturing the XF86TouchpadToggle button in gnome-settings-daemon, > showing a nice popup, and disabling the touchpad in software. > > That works great for a number of laptops, but completely breaks a number > of others where the hardware handles the request, but still sends a key > event (usually separate ones for on and off). In those cases, we'd need > 2 more key types, in the kernel, and in X.org. Rfkill story all over I guess ;( > > The patches in https://bugs.freedesktop.org/show_bug.cgi?id=31300 > implement this for the kernel, and for X.org. > > Then we have the problem that udev's keymaps seem to use different > function keys depending on the hardware [1], when X.org (because of the > limitations of XKB) standardised on F22 for "XF86TouchpadToggle". > > So we'll need to standardise on the keys used. I selected F21 for > XF86TouchpadToggle, F22 for XF86TouchpadOn and F23 for XF86TouchpadOff. > See the patch in: > https://bugs.freedesktop.org/show_bug.cgi?id=31333 > > The remaining fixes would need to be in udev's keymaps, to set hardware > handled keys to F22 and F23, and software ones to F21. Patch is attached > for that. > > Questions, comments? I still believe we need dedicated keycodes in kernel for that and make sure that in-kernel maps use proper keycodes so that programs that do not have the same limitations that X has could differentiate between touchpad events and real F2* events. X can still continue to use f21, 22 and 23 (at least while it can't handle higher keycodes), and common setups may rely on udev's facilities to remap the keys. Thanks. -- Dmitry