Re: [PATCH 2/2] iio: dac: mcp47a1: add support for new device

On Sat, 25 Jul 2026 00:14:17 +0100
Jonathan Cameron <[email protected]> wrote:
> Given the really small number of registers, regmap does seem like
> it might be overkill

My thought was that regmap is preferred over i2c_smbus_* functions,
but that is a simple enough move.

...

> > +	/*
> > +	 * This is a volatile DAC which doesn't have an ID register, instead
> > +	 * the value register is set to 0x20 every Power-on-Reset (table 4-1).
> > +	 * Any other read value at startup could indicate that the device is
> > +	 * damaged etc.  
> 
> or some earlier software wrote it and the all the regulators are stubs
> provided as they can't be turned off by software.
> 
> I'd drop this test - it is a bit too flakey as a way to tell the chip
> is there and working. 

Yeah, fair enough, there's a reason why it just reports it as dev_info().

-- 
Kind regards,
Joshua Crofts