Re: [PATCH 0/5] scsi: ufs: ufs device as a temperature sensor

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

On Mon, Feb 03, 2020 at 09:29:57PM +0000, Avri Altman wrote:
> > >> Can you add an explanation why this can't be added to the just-
> > introduced
> > >> 'drivetemp' driver in the hwmon subsystem, and why it make sense to
> > have
> > >> proprietary attributes for temperature and temperature limits ?
> 
> 
> Guenter hi,
> Yeah - I see your point. But here is the thing - 
> UFS devices support only a subset of scsi commands.
> It does not support ATA_16 nor SMART attributes.
> Moreover, you can't read UFS attributes using any other scsi/ATA/SATA
> Commands, nor it obey the ATA temperature sensing conventions.
> So unless you want to totally break the newly born drivetemp - 
> Better to leave ufs devices out of it.
> 

drivetemp is written with extensibility in mind. For example, Martin has a
prototype enhancement which supports SCSI drive temperature sensors. 
As long as a device can be identified as ufs device, and as long as there
is a means to pass-through commands, adding a new type would be easy.

> Another option is to put a ufs module under hwmon.
> Do you see why would that be more advantageous to using the thermal core?
> Provided that you are not going to deprecate it (Intel's wifi card is still using it)...
> 

Deprecate what, and what does this discussion have to do with Intel's wifi
card ?

Either case, the hardware monitoring subsystem provides standard attributes,
and it provides a bridge to the thermal subsystem. The question should be
why _not_ to use the hwmon subsystem, and this question should be answered
as part of this patch series.

Guenter