Re: [PATCH 2/2] gpio: sch: Add interrupt support

[Jan Kiszka]

On 26.04.19 19:46, Manivannan Sadhasivam wrote:
> On Fri, Apr 26, 2019 at 07:39:56PM +0200, Jan Kiszka wrote:
> > On 26.04.19 19:33, Manivannan Sadhasivam wrote:
> > > On Fri, Apr 26, 2019 at 08:20:19PM +0300, Andy Shevchenko wrote:
> > > > On Fri, Apr 26, 2019 at 7:05 PM Jan Kiszka <jan.kiszka@xxxxxxxxxxx> wrote:
> > > > > On 26.04.19 16:42, Enrico Weigelt, metux IT consult wrote:
> > > > > > On 26.04.19 15:36, Jan Kiszka wrote:
> > > > > >
> > > > > > > At the same time, there are no real alternatives - to my> knowledge - for the value it brings (various bindings) to simply
> > > > > > switch> the engine.
> > > > > > Which value exactly does that collection of crude wrappers and broken
> > > > > > attempts to buypass the kernel (driving gpios via /dev/mem *facepalm*)
> > > > > > provide ?
> > > > >
> > > > > Leaving that blunt hack aside:
> > > > >
> > > > > import mraa
> > > > >
> > > > > pin = mraa.Gpio(13)
> > > > > pin.dir(mraa.DIR_OUT)
> > > > > pin.write(1)
> > > > >
> > > > > And the same goes for nodejs, java and c++.
> > > > >
> > > > > Moreover, this allows you to abstract away where "Pin 13" actually came from on
> > > > > that board if the kernel changes (BSP -> upstream...) or the extension board or
> > > > > ...
> > > >
> > > > The problem here is opaque number. This has to be chip + *relative* pin number/
> > > > See this:
> > > > https://stackoverflow.com/questions/55532410/how-do-linux-gpio-numbers-get-their-values/55579640#55579640
> > >
> > > But for platform like 96Boards we don't need controller specific lookup, these
> > > are all handled by the platform code [1] so that the users can use the standard
> > > pinout number to access GPIOs. For instance, pin 23 on the Low Speed expansion
> > > header is the GPIO for all 96Boards platform, so the user can access that pin
> > > using 23 itself in the application and it will run across all supported
> > > 96Boards.
> >
> > Can you ensure stable numbering when probing order changes, e.g. due to
> > adding an extension board?
>
> Good point! For tackling this, I'm planning to introduce an API for accessing
> the GPIO by its line name. It will be tricky to implement but once done, it
> will serve.

Whatever is stable and handy. As cited in the other thread, I played with the 
device path as well, see [1][2]. Feel free to pick what what is useful, I will 
likely not have time to work on that soon again.

Jan

[1] https://github.com/siemens/mraa/commit/034d787eea1a5b201ea77a6549ffc0bdbc3b776d
[2] https://github.com/siemens/mraa/commit/6fd8a3c27764718c1e62a4938a6dc88819c3f65f

--
Siemens AG, Corporate Technology, CT RDA IOT SES-DE
Corporate Competence Center Embedded Linux