Re: [PATCH 4/7] hwmon: (lm90) Add support for 2nd remote channel's offset register

[Slawomir Stepien <sst@poczta.fm>]

On cze 05, 2022 23:50, Guenter Roeck wrote:
> On 6/5/22 23:30, Slawomir Stepien wrote:
> > On cze 05, 2022 11:03, Guenter Roeck wrote:
> > > On Wed, May 25, 2022 at 09:36:54AM +0200, Slawomir Stepien wrote:
> > > > From: Slawomir Stepien <slawomir.stepien@nokia.com>
> > > > 
> > > > The ADT7461 supports offset register for both remote channels it has.
> > > 
> > > ADT7481
> > 
> > Oops. I will fix that in new version.
> > 
> > > > Both registers have the same bit width (resolution).
> > > > 
> > > > In the code, this device has LM90_HAVE_TEMP3 and LM90_HAVE_OFFSET flags,
> > > > but the support of second remote channel's offset is missing. Add that
> > > > implementation.
> > > > 
> > > > Signed-off-by: Slawomir Stepien <slawomir.stepien@nokia.com>
> > > > ---
> > > >   drivers/hwmon/lm90.c | 37 ++++++++++++++++++++++++++++++++-----
> > > >   1 file changed, 32 insertions(+), 5 deletions(-)
> > > > 
> > > > diff --git a/drivers/hwmon/lm90.c b/drivers/hwmon/lm90.c
> > > > index 02b211a4e571..d226f1dea2ba 100644
> > > > --- a/drivers/hwmon/lm90.c
> > > > +++ b/drivers/hwmon/lm90.c
> > > > @@ -153,6 +153,8 @@ enum chips { adm1023, adm1032, adt7461, adt7461a, adt7481,
> > > >   #define LM90_REG_REMOTE_TEMPL		0x10
> > > >   #define LM90_REG_REMOTE_OFFSH		0x11
> > > >   #define LM90_REG_REMOTE_OFFSL		0x12
> > > > +#define LM90_REG_REMOTE2_OFFSH		0x34
> > > > +#define LM90_REG_REMOTE2_OFFSL		0x35
> > > 
> > > I don't think those are needed.
> > 
> > In lm90_temp_write() (unlike in lm90_update_limits()) the remote channel is *not* set. I find
> 
> ... unless lm90_set_temp() is used to write the values. If I recall correctly
> I didn't do that because selecting the remote channel seemed unnecessary.
> 
> > setting it (the remote channel) in lm90_temp_write() a waste of xfers, if we can address the
> > registers directly. But if you prefer to have just one set of register and setting the remote
> > channel bit, then sure I can do it like that.
> > 
> It isn't as if setting the offset happens all the time, so I'd prefer
> to use lm90_set_temp() if that is possible.

Good point! So I will use just one set of the registers.

-- 
Slawomir Stepien