[PATCH] RFC: input: Extend matrix-keypad device tree binding

[Rob Herring]

On 12/20/2012 03:41 PM, Simon Glass wrote:
> Hi Dmitry,
> 
> On Thu, Dec 20, 2012 at 1:13 PM, Dmitry Torokhov
> <dmitry.torokhov at gmail.com> wrote:
>> Hi Simon,
>>
>> On Thu, Dec 20, 2012 at 12:23:58PM -0800, Simon Glass wrote:
>>>
>>> For the key repeat feature, we need to set this after the input device
>>> is registered. So we would need to add a matrix_keypad_setup_input() or
>>> similar to be called by the driver after input_register_device(). I am
>>> less keen on that idea, and less again on the alternative of perhaps
>>> matrix_keypad_register_device() which does input_register_device() and
>>> then sets up the key repeat. Thoughts?
>>
>> No, we already have default rate and delay. Unless you can prove that
>> random firmware writer's idea of appropriate delay and rate is better
>> then current default - for everyone - and then can successfully argue
>> that that obviously best delay/rate combo should not replace the current
>> one but stay only in DT bindings, let's keep relying on users adjusting
>> their own preferences from respective desktop environments/console/etc.
> 
> Seems reasonable. My only comment on this is that the device tree
> comes from kernel, not firmware. This lets us configure an embedded
> system easily (where the user may not have access to repeat rate
> preferences).

Maybe for your case, but in general the DT is not tied to the kernel.

> Grant are you OK with me just dropping the repeat settings, and
> keeping the other two? If so I will respin the patch.
> 

One comment is these properties are already defined for omap. You should
delete them from omap-keypad.txt and refer to this.

Rob

> Regards,
> Simon
> 
>>
>> Thanks.
>>
>> --
>> Dmitry
> _______________________________________________
> devicetree-discuss mailing list
> devicetree-discuss at lists.ozlabs.org
> https://lists.ozlabs.org/listinfo/devicetree-discuss
>