From mboxrd@z Thu Jan 1 00:00:00 1970 From: Hans de Goede Subject: Re: [PATCH 0/7] Various touchscreens: Support axis inversion / swapping / extra gpios Date: Fri, 20 Nov 2015 12:33:20 +0100 Message-ID: <564F0500.6020606@redhat.com> References: <1448018233-27824-1-git-send-email-hdegoede@redhat.com> <20151120112245.GC8644@n2100.arm.linux.org.uk> Reply-To: hdegoede-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8; format=flowed Return-path: In-Reply-To: <20151120112245.GC8644-l+eeeJia6m9vn6HldHNs0ANdhmdF6hFW@public.gmane.org> List-Post: , List-Help: , List-Archive: , List-Unsubscribe: , To: Russell King - ARM Linux Cc: Dmitry Torokhov , Maxime Ripard , devicetree , linux-sunxi-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org, Sander Vermin , linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r@public.gmane.org, linux-input-u79uwXL29TY76Z2rM5mHXA@public.gmane.org List-Id: linux-input@vger.kernel.org Hi, On 20-11-15 12:22, Russell King - ARM Linux wrote: > On Fri, Nov 20, 2015 at 12:17:06PM +0100, Hans de Goede wrote: >> This patch-set extends the of_touchscreen.c helpers to (optionally) also >> deal with axis inversion / swapping, and patches a bunch of drivers to use >> this new functionality. > > What's wrong with using tslib? tslib supports axis transformations in > userspace via a set of filters. Doing this in the kernel when we know that the touchscreen is always mounted upside down (or rotated 90 degress) on a certain tablet makes things just work, where as tslib requires manual configuration by the user. Also most distributions are not even shipping tslib, some of this can be done with the xf86-input-evdev driver (which is the upstream xorg default driver for touchscreens) too, but again requires manual configuration and the same goes for the new libinput stack to which xorg is moving and which is the default for wayland. Also note that this patch-set is not adding new functionality. We already have devicetree-bindings to indicate to the driver that axis inversion / swapping should be done in the driver, and these are already used in various drivers. This patch-set just makes it easier to add support for these already existing (and documented) devicetree properties. Regards, Hans