Re: [PATCH v5 1/2] dt-bindings: adc: provide max34408/9 device tree binding document
From: Jonathan Cameron
Date: Fri Oct 13 2023 - 14:12:09 EST
On Fri, 13 Oct 2023 11:53:33 +0200
Krzysztof Kozlowski <krzysztof.kozlowski@xxxxxxxxxx> wrote:
> On 13/10/2023 11:09, Jonathan Cameron wrote:
> >>>>>>>> + shtdn-enable-gpios:
> >>>>>>>
> >>>>>>> I guess the review crossed with you sending v5. There is some
> >>>>>>> feedback on v4 you need
> >>>>>>> to address here.
> >>>>>>
> >>>>>> Jonathan, I thought I did, I've changed ena to powerdown-gpios from
> >>>>>> Krzysztof's comments but about this one pin I'm still not sure, it
> >>>>>> looks like *-enable-gpios (like in *-enable-gpios pins in
> >>>>>> iio/frequency/adi,adf4377.yaml) pin or is it not? Or maybe any
> >>>>>> other
> >>>>>> suggestions about naming of this one?
> >>>>>>
> >>>>>> Thanks.
> >>>>>
> >>>>> shutdown-gpios and make the sense (active high / low) such that
> >>>>> setting
> >>>>> it results in teh device being shut down.
> >>>>> Or treat it as an enable and enable-gpios
> >>>>>
> >>>>> Something that indicates both shutdown and enable is confusing ;)
> >>>>>
> >>>>> Jonathan
> >>>>
> >>>>
> >>>> Jonathan, then I make these changes:
> >>>>
> >>>> powerdown-gpios: -> output-enable:
> >>> Needs to retain the gpios bit as we want the standard gpio stuff to pick
> >>> them up. I'm not that keen on output-enable-gpios though. The activity
> >>> here is very much 'shutdown because of error or not enabled' I think.
> >>> So perhaps we flip the sense and document that it needs to be active low?
> >>>
> >>>> shtdn-enable-gpios: -> enable-gpios:
> >>>>
> >>>> Is it ok?
> >>>
> >>> Conor, Rob, Krzysztof - you probably have a better insight into this than
> >>> I do.
> >>>
> >>
> >> "enable-gpios" are for turning on a specific feature, not powering
> >> on/off entire device. For example to enable regulator output.
> >>
> >> "powerdown-gpios" are for turning device on/off.
> >>
> >> I don't know what do you have in your device.
> > Ok. Sounds like that what is enable-gpios above should be shutdown-gpios.
>
> shutdown-gpios sounds exactly the same as powerdown-gpios and it is
> already used in exactly same context.
Oops. Yup. powerdown-gpios seems appropriate.
>
> > The other case is a device output indicating whether the device is
> > shutdown. That can happen because it was told to do so (via the other gpio),
> > or because it is in an error state. What's a good naming convention for that?
>
> There is no convention and I did not see such case so far.
> powerdown-status-gpios? powerdown-state-gpios?
Either seems reasonable.
Thanks,
J
>
>
>
> Best regards,
> Krzysztof
>