Re: [PATCH] iio: st-sensors: Update ST Sensor bindings

From: Linus Walleij <linus.walleij@linaro.org>
To: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Jonathan Cameron <jic23@kernel.org>,
	linux-iio@vger.kernel.org, Hartmut Knaack <knaack.h@gmx.de>,
	Lars-Peter Clausen <lars@metafoo.de>,
	Peter Meerwald-Stadler <pmeerw@pmeerw.net>,
	"open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS" 
	<devicetree@vger.kernel.org>, Rob Herring <robh@kernel.org>
Subject: Re: [PATCH] iio: st-sensors: Update ST Sensor bindings
Date: Fri, 28 Jan 2022 16:57:04 +0100	[thread overview]
Message-ID: <CACRpkdYGYQNQ27DD5HaUt8ERgoiNjxGE95W8EHbZRFNaXsH8VQ@mail.gmail.com> (raw)
In-Reply-To: <CACRpkdbrX-JiPJh2pnVPt39mNgNNrAFkOfS8Q9kO002+oVagyg@mail.gmail.com>

On Fri, Jan 28, 2022 at 4:51 PM Linus Walleij <linus.walleij@linaro.org> wrote:
> On Wed, Jan 26, 2022 at 3:59 PM Geert Uytterhoeven <geert@linux-m68k.org> wrote:
> > On Mon, Apr 12, 2021 at 2:24 PM Linus Walleij <linus.walleij@linaro.org> wrote:
>
> > >    interrupts:
> > > +    description: interrupt line(s) connected to the DRDY line(s) and/or the
> > > +      Intertial interrupt lines INT1 and INT2 if these exist. This means up to
> > > +      three interrupts, and the DRDY must be the first one if it exists on
> >
> > So this says three (the LSM9DS0 datasheet agrees)...
> >
> > > +      the package. The trigger edge of the interrupts is sometimes software
> > > +      configurable in the hardware so the operating system should parse this
> > > +      flag and set up the trigger edge as indicated in the device tree.
> > >      minItems: 1
> > > +    maxItems: 2
> >
> > ... while this says two?
>
> Looks like a bug, could you send a patch? (I'm a bit preoccupied right now.)

Oh wait a minute, LSM9DS0 is one of those with more than one component
inside it isn't it?

While it is a bit awkward, we do bindings per-subcomponent on these, so
for example lsm330dlc registers as "st,lsm330dlc-accel" and "st,lsm330dlc-gyro"
and it makes a bit of sense because they each have different I2C addresses
as well.

I see it as two components just sharing a physical package rather than one
component in a package.

So the IRQs are per-subcomponent, not for the entire package.

Does this influence the situation you have with LSM9DS0?

Yours,
Linus Walleij