[lm-sensors] Suggested additions to hwmon sysfs ABI

Jonathan Cameron jic23 at cam.ac.uk
Mon Jul 11 10:53:09 CEST 2011


On 07/09/11 18:07, Guenter Roeck wrote:
> Hi all,
> 
> while adding support for more PMBus devices, I noticed that many of the chips
> support attributes which are not currently available in the sysfs ABI.
> 
> Here is a (not necessarily complete) list.
> 
> ---
> 
> Historic minimum attributes:
> 		vout_min	vin_min		temp_min
> MAX34440	x
> MAX34441	x
> LTC2978		x		x		x
> 
> Peak (historic maximum) attributes:
> 
> 		pin_peak	vin_peak	vout_peak	iout_peak	temp_peak
> MAX34440					x		x		x
> MAX34441					x		x		x
> MAX8688						x		x		x
> MAX16064					x				x
> ADM1275				x		x		x
> ADM1276		x		x		x		x
> LTC2978				x		x				x
> LM25066		x (1)
> LM5064		x (1)
> LM5066		x (1)
> ZLxxxx (2)							x (3)
> 
> Average attributes:
> 
> 		vin_avg		vout_avg	iin_avg		pin_avg
> LM25066		x		x		x		x (1)
> LM5064		x		x		x		x (1)
> LM5066		x		x		x		x (1)
> 
> Notes:
> 1) In addition to instantaneous power measurements
> 2) All Intersil / Zilker Labs power management chips
> 3) Only available through SNAPSHOT command, which reads a total of 32 bytes.
>    Also see Intersil / Zilker Labs AN2033.
> 
> ---
> 
> Based on this list, I suggest adding the following attributes to the ABI.
> 
> in[0-*]_input_lowest
> in[0-*]_input_highest
> in[0-*]_average
> in[0-*]_reset_history
> 
> curr[1-*]_input_lowest
> curr[1-*]_input_highest
> curr[1-*]_average
> curr[1-*]_reset_history
> 
> temp[1-*]_input_lowest
> temp[1-*]_input_highest
> temp[1-*]_average
> temp[1-*]_reset_history
> 
> It might also be necessary to modify the sensors command to support both instantaneous
> and average/peak power readings at the same time.
How are you going to do peak power?

Otherwise, Looks sensible.  Ran into this in IIO as well, we went with peak, but highest / lowest
works just as well. (not seen the minimum version yet)
> 
> Thoughts/comments ?
> 
> Thanks,
> Guenter
> 
> _______________________________________________
> lm-sensors mailing list
> lm-sensors at lm-sensors.org
> http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
> 





More information about the lm-sensors mailing list