Re: [PATCH 1/2] dt-bindings: hwmon: Introduce ADS71x8

[Jonathan Cameron <jic23@kernel.org>]

On Mon, 2 Sep 2024 13:04:55 +0000
"Sperling, Tobias" <Tobias.Sperling@Softing.com> wrote:

> > On Fri, 30 Aug 2024 07:30:16 -0700
> > Guenter Roeck <linux@roeck-us.net> wrote:
> >   
> > > On 8/30/24 06:14, Conor Dooley wrote:  
> > > > Hey Tobias, Guenter, Jonathan,
> > > >
> > > > On Fri, Aug 30, 2024 at 11:49:53AM +0000, Sperling, Tobias wrote:  
> > > >>  From b2e04ce5500faf274654be5284be9db4f3abefce Mon Sep 17 00:00:00  
> > 2001  
> > > >> From: Tobias Sperling <tobias.sperling@softing.com>
> > > >> Date: Fri, 23 Aug 2024 12:08:33 +0200
> > > >> Subject: [PATCH 1/2] dt-bindings: hwmon: Introduce ADS71x8
> > > >>
> > > >> Add documentation for the driver of ADS7128 and ADS7138 12-bit, 8-channel
> > > >> analog-to-digital converters. These ADCs have a wide operating range and
> > > >> a wide feature set. Communication is based on an I2C interface.
> > > >> The driver provides the functionality of manually reading single channels
> > > >> or sequentially reading all channels automatically.
> > > >>
> > > >> Signed-off-by: Tobias Sperling <tobias.sperling@softing.com>
> > > >> ---
> > > >>   .../devicetree/bindings/hwmon/ti,ads71x8.yaml |  85 +++++++++++  
> > > >
> > > > If this is a "generic" adc, why is it going into hwmon?
> > > > I would have expected this to be in iio/adc, and use more typical adc
> > > > bindings, even if the driver is in hwmon.
> > > >
> > > > Guenter/Jonathan wdyt?
> > > >  
> > >
> > > Same thought here. While the chip supports limits, making it suitable for
> > > hardware monitoring, its primary use seems to be as ADC, not as hardware
> > > monitoring device. The hardware monitoring API isn't well suited for the
> > > fast sample rate supported by this chip.  
> > 
> > Agreed, looks like a typical IIO ADC.
> > 
> > If the particular board needs it for hardware monitoring we have
> > the bridge that should work for that (iio-hwmon).  
> 
> Just some addition. In theory the chip also provides the possibility to use some
> channels as GPIO making it not only work as ADC.
> But yes, driver mainly implements reading of the ADC. Will try to make it an
> IIO ADC device then.
That's fairly common.  If you want to support it then provide the gpio_chip
etc and +CC the relevant maintainers and mailing lists.

Jonathan

> 
> > Jonathan
> >   
> > >
> > > Guenter
> > >  
> 
> Regards
> Tobias