lkml.org 
[lkml]   [2018]   [Aug]   [4]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectRe: [PATCH] gpio: Add driver for PC Engines APU2/APU3 GPIOs
Date
On Thursday, August 2, 2018 11:30:00 PM CEST Linus Walleij wrote:
> On Wed, Aug 1, 2018 at 1:12 PM Florian Eckert <fe@dev.tdt.de> wrote:
>
> > Add a new device driver "gpio-apu" which will now handle the GPIOs on
> > APU2 and APU3 devices from PC Engines.
> >
> > - APU2/APU3 -> front button reset support
> > - APU3 -> SIM switch support
> >
> > Signed-off-by: Florian Eckert <fe@dev.tdt.de>
>
> Hi Florian, thanks for the patch!

There have been multiple attempts at upstreaming a gpio/pinctrl driver for
the different APU SoC by now. The last attempt I remember took place on
this ML too: <https://lkml.org/lkml/2017/12/20/824>

But sadly most attempts stalled, because there is already a leds-apu driver
<https://github.com/torvalds/linux/blob/master/drivers/leds/leds-apu.c>
(since the gpio-driver was never accepted) that will do in a pinch for the
LEDs (which most people care about).

> I looped in Andy Schevchenko and Mika Westerberg who are authorities on
> x86 platform drivers in general and GPIO and pin control in particular
> so they can help out with the review.
Intel helping AMD. That's nice to see ;-)

> I'm a bit confused whether these things are really GPIOs or just
> switches but since they can change direction they seem to be GPIOs.

Yes, it's a real gpio/pinctrl in AMDs Tech docs.
<https://support.amd.com/TechDocs/52740_16h_Models_30h-3Fh_BKDG.pdf>

Some of the pins are muxed. I.e GPIO167-170 can either be a GPIO or
used for HD audio.
"The HD audio controller supports up to four codecs with one AZ_SDIN
pin from each codec. The four AZ_SDIN pins are multiplexed with
GPIO167-170 (GPIOxA7-GPIOxAA). If a particular pin is to be used for
HD audio functionality". There's a full table in "3.26.12.1 GPIO Registers".




As for the APUs. The vendor (PC Engines) happily provides
PDFs and schematics for their boards:
<https://www.pcengines.ch/pdf/apu1.pdf>
<https://www.pcengines.ch/schema/apu1c.pdf>
<https://www.pcengines.ch/pdf/apu2.pdf>
<http://pcengines.ch/schema/apu2c.pdf>

So, it's possible to repurpose several test points as additional
GPIOs and more.

Note2:
On both boards there is also a dedicated GPIO pin header J19, but
these pins are controlled by the SuperIO Nuvoton NCT5104D.

> > +config GPIO_APU
> > + tristate "PC Engines APU2/APU3 GPIO support"
> > + depends on X86
> > + select GPIO_GENERIC
> > + help
> > + Say Y here to support GPIO functionality on APU2/APU3 boards
> > + from PC Engines.
> > + - APU2/APU3 -> front button reset support
> > + - APU3 -> SIM switch support
Well, by design this driver will sort of clash with the leds-apu driver.

Regards,
Christian


\
 
 \ /
  Last update: 2018-08-04 20:23    [W:0.123 / U:0.136 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site