RE: [PATCH 1/2] dt-bindings: iio: imu: Add bindings for ADI ADIS16607

[Sabau, Radu bogdan]

> -----Original Message-----
> From: Conor Dooley <conor@xxxxxxxxxx>
> Sent: Thursday, October 1, 2026 10:05 PM

...

> > +properties:
> > +  compatible:
> > +    enum:
> > +      - adi,adis16607-2
> > +      - adi,adis16607-3
> 
> Yet they are not compatible. Why?
> pw-bot: changes-requested

Yep, Sashiko also pointed this, it was an earlier version which won't be released
anymore and I left it mentioned in the commit message by accident, I will
remove that, sorry for the confusion, again.

> 
> > +
> > +  reg:

...

> > +  adi,fifo-enable:
> > +    type: boolean
> > +    description:
> > +      Enable hardware FIFO mode for buffered data acquisition. When
> enabled,
> > +      the driver uses the device FIFO with configurable watermark threshold
> > +      instead of single-sample burst reads.
> 
> What aspect of the usecase does setting this depend on?
> Why not just always enable this?

This one has given me some thinking, and I think you have a really great point
here, FIFO should always be enabled. At first I thought that the advantage of
direct single burst readings is that you get the diagnose at each point, but
after some consideration I think the FIFO way is also correct enough since it
gets that diagnose too if things go wrong so no point in keeping that direct
burst mode. Thanks for this!