Re: [PATCH] input: synaptics-rmi4 - cleanup rmi_i2c_probe()

[Christopher Heiny <cheiny@synaptics.com>]

On 01/08/2014 02:46 PM, Dmitry Torokhov wrote:
> On Wed, Jan 08, 2014 at 02:11:32PM -0800, Christopher Heiny wrote:
>> Moves i2c_check_functionality to occur before the gpio_config() call.  This
>> can catch some issues that would otherwise result in mysterious problems
>> in gpio_config().
>>
>> Reduces debugging output; updates remaining output to be more accurate.
>
> What kind of gpio config is there? Did not we add proper configuration
> of attn_gpio to the core? Do you really need these callbacks? They will
> hurt you when you will try to move to devicetree-based setups.

This is a gpio_config() does the platform specific gpio setup (power 
configuration, voltage levels, hardware reset lines, and so on).  This 
is used on every production platform using the driver that I know of, 
and the setup is different on every platform (sometimes even between 
revs of that platform).  We'd go crazy trying to handle that in 
rmi_i2c.c or rmi_driver.c, so it's pushed to the platform.

And yeah, I figure it'll hurt when we move to devicetree.  Right now 
we're just trying to get the current state of the code in the 
synaptics-rmi4 branch from "horribly broken" to "it lives!".  We're 
willing to put off that devicetree pain in order to get the code on its 
feet.