[linux-sunxi] Re: [PATCH v3 2/3] dt-bindings: input: Add R_LRADC support for A83T

[alexandre.belloni@free-electrons.com (Alexandre Belloni)]

On 27/06/2017 at 19:36:31 +0200, Maxime Ripard wrote:
> Hi,
> 
> On Tue, Jun 27, 2017 at 11:29:10PM +0800, icenowy at aosc.io wrote:
> > Maxime, here's another problem: if we have already a GP LRADC driver,
> > how can we tell the kernel to use it as IIO ADC rather than keys?
> 
> The GPADC IIO driver is not for the LRADC driver, but the GPADC /
> temperature sensor.
> 
> We used to have an LRADC IIO driver in the CHIP BSP written by Alex
> (in CC):
> https://github.com/NextThingCo/CHIP-linux/commit/8675b761c54be73dc7cc0113209f02e10cc63a27
> 
> But he never mainlined it.
> 
> > Should we introduce a new property for this once ready?
> 
> We need to keep the current binding. We can just check for the
> presence or not of child nodes to see if it has some keys, and we'd
> need an IIO-to-input driver that is yet to be written.
> 

The whole submission is here and can already replace the existing driver
but it will be polling instead of using interrupts:

http://lists.infradead.org/pipermail/linux-arm-kernel/2016-July/440734.html

It is not upstream because it seems the full replacement was required at
once instead of doing it incrementally and there was (is) no API for
in-kernel events consumers.

Also, the DT ABI stuff would have to be solved.

-- 
Alexandre Belloni, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com