Re: [RFC PATCH 14/17] input: RMI4 F30 GPIO/LED control

On 08/27/2012 03:58 PM, Linus Walleij wrote:

GPIO/LED, nice since I'm a GPIO maintainer I'll take a closer look.

If the bus will start doing a lot of non-input business it should live under
drivers/mfd but I think this is just one exception, right?

On Fri, Aug 17, 2012 at 3:17 PM, Christopher Heiny [off-list ref] wrote:

(...)

quoted

diff --git a/drivers/input/rmi4/rmi_f30.c b/drivers/input/rmi4/rmi_f30.c

quoted

+#include <linux/kernel.h>
+#include <linux/rmi.h>
+#include <linux/input.h>
+#include <linux/slab.h>
+#include "rmi_driver.h"

The non-existance of <linux/gpio.h> and <linux/leds.h> tells us that something
is very wrong.

You should not model these GPIOs and LEDs by a set of obscure
sysfs attributes, instead use the proper kernel subsystems that
already exist for handling this! LEDs and GPIOs already have their
own (standardized) userspace sysfs interfaces.

Reading the code I see that this is what happens here, so please rewrite
this to be a real GPIO+LED driver using struct gpio_chip and
the same for LEDs.

Be inspired by drivers/gpio/* and drivers/leds/*

Roger.  We'll rework this and resubmit at a later date.