Re: [PATCH v3 1/2] scsi: ufs: Probe for temperature notification support

[Guenter Roeck]

On 9/13/21 12:49 AM, Avri Altman wrote:
> > > > The "enable" attribute only makes sense if it can be used to actually
> > > > enable or disable a specific sensor, and is not tied to limit
> > > > attributes but to the actual sensor values.
> > > See explanation above.
> > >    Will make it writable as well.
> >
> > That only makes sense if the information is passed to the chip. What is going to
> > happen if the user writes 0 into the attribute ?
> Will turn off the temperature exception bits, so that Tcase is no longer valid,
> and the device will always return Tcase = 0.

Ok. Then attempts to read the temperature should return -ENODATA, not -EINVAL,
if Tcase == 0.

Guenter