Re: [PATCH 1/3] dt-bindings: auxdisplay: hit, hd44780: drop redundant GPIO node

From: Ralf Schlatterbeck
Date: Tue Feb 13 2024 - 04:08:13 EST

On Mon, Feb 12, 2024 at 02:38:27PM +0100, Krzysztof Kozlowski wrote:
> On 12/02/2024 12:58, Ralf Schlatterbeck wrote:
> > On Mon, Feb 12, 2024 at 12:25:48PM +0100, Krzysztof Kozlowski wrote:
> >>
> >> Hm, I don't understand how exactly it helps. The GPIO expander has its
> >> own example and as you pointed below, this is basically the same code,
> >> except rw and backlight GPIOs.
> >
> > The hd44780 is a display that is very often used.
> GPIO expanders and their usage is nothing specific to this device -
> other devices also might benefit of them. Or the SoCs which have enough
> of GPIOs... I really do not understand why do we need expander here and
> how does it help

The hd44780 is most often sold together with that specific I/O expander.
The idea was to help people with that combination how to get their
device working.

> Anyway, binding examples should not be collection of unrelated
> solutions, because then we should accept for each device schema several
> other variations and combinations.

The solutions in that case are not unrelated because they document the
most-often-used hw combo.

I also didn't find any documentation of how to actually *use* the
pcf8575 I/O expander. Even
has only docs on how to instantiate the device on the i2c bus but not
how then to use the I/Os of the chip for something else.

So I'd ask again to not remove that piece of useful documentation.

And to get somehow philosophic:
I think that docs should be didactic, not optimized to the least

Thanks and kind regards,
Dr. Ralf Schlatterbeck Tel: +43/2243/26465-16
Open Source Consulting www:
Reichergasse 131, A-3411 Weidling email: office@xxxxxxxxxx