Re: [PATCH] hwmon: mp2891: add MP2891 driver

[Guenter Roeck <linux@roeck-us.net>]

On Fri, May 26, 2023 at 09:23:12AM -0700, Guenter Roeck wrote:
[ ... ]
> 
> The above functions really do not make any sense whatsoever, meaning I do
> not trust that any of the remaining functions are needed either. Please
> only provide necessary private functions. For each function which is
> actually needed please explain why the core function can not be used.
> 
> Note that access to the datasheet would help me to determine which of
> those functions are really needed. Requiring me to create an account
> just to read a non-descriptive one-page "datasheet" doesn't help at all
> (and, FWIW, is really pointless).
> 

Adding to that: An internal source with access to the datasheet stated
that almost all registers/command would be standard, and that only VID
handling would be special. Even for that I am not sure if that is really
correct, and that the VID stepping is so far out of standard VID coding
that it can't be modeled as standard VID stepping. As such, I am not
inclined to accept a driver which re-implements pretty much all telemetry
registers without some very specific and confirmable information that
and why this is really needed. A vague reference to the datasheet won't
help, sorry.

Guenter