From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dmitry Torokhov Subject: Re: [4.12 regression] Thinkpad X250 Touchpad and Trackpoint not recognized anymore; commit e839ffa: "Input: synaptics - add support for Intertouch devices" Date: Sat, 20 May 2017 10:35:26 -0700 Message-ID: <20170520173526.GB38800@dtor-ws> References: <41de7f92-8d98-3266-5d28-28ba48c5da38@pa-w.de> <20170519202303.GA19281@dtor-ws> <20170519204914.GD19281@dtor-ws> <59adce69-743d-98f1-ece7-1197c49e1b5e@pa-w.de> <20170520095950.GB6808@mail.corp.redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from mail-pg0-f47.google.com ([74.125.83.47]:35805 "EHLO mail-pg0-f47.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752475AbdETRf3 (ORCPT ); Sat, 20 May 2017 13:35:29 -0400 Content-Disposition: inline In-Reply-To: <20170520095950.GB6808@mail.corp.redhat.com> Sender: linux-input-owner@vger.kernel.org List-Id: linux-input@vger.kernel.org To: Benjamin Tissoires Cc: Pascal Wichmann , linux-input@vger.kernel.org, linux-kernel@vger.kernel.org On Sat, May 20, 2017 at 11:59:50AM +0200, Benjamin Tissoires wrote: > Hi, > > On May 20 2017 or thereabouts, Pascal Wichmann wrote: > > > Looks like you running your patched kernel? > > That's right. > > > > > > >>> CONFIG_RMI4_CORE=m > > >>> CONFIG_RMI4_I2C=m > > >>> CONFIG_RMI4_SPI=m > > >>> # CONFIG_RMI4_SMB is not set > > > > > > This is your issue I believe. > > > > Indeed, enabling that configuration solves that issue. > > > > However, I think it is quite unintuitive that a module (psmouse) chooses > > a default mode which requires another driver which is not necessarily > > included; though it would probably be not a very clean solution to > > explicitly check that as well. > > > > Is this behaviour, that one module requires another without > > communicating that clearly, wanted? > > > > I can see 3 solutions: > 1. Have PS2_SMBUS depending on RMI_SMBUS (and ELAN_I2C, and others when > required) > 2. Have PS2_SMBUS selecting RMI_SMBUS (and the others when time comes) > 3. Changing the default value of synaptics_intertouch to > SYNAPTICS_INTERTOUCH_OFF when RMI_SMBUS is not set > > Solution 3. might be interesting because it doesn't prevent users to > compile the module on the side and is Synaptics only. > > Dmitry, any comments? I like #3. We might also want to stick a warning into synaptics.c when we see a device that has intertouch, but RMI_SMBUS is disabled, so we could nudge users to switch over to RMI. Thanks. -- Dmitry