#12095 NORM 4-softw: runin-cpu_temp fails on strangely high temperature readings

[Zarro Boogs per Child]

#12095: runin-cpu_temp fails on strangely high temperature readings
---------------------------------------+------------------------------------
           Reporter:  martin.langhoff  |       Owner:  pgf          
               Type:  defect           |      Status:  new          
           Priority:  normal           |   Milestone:  4-software   
          Component:  kernel           |     Version:  not specified
         Resolution:                   |    Keywords:               
        Next_action:  never set        |    Verified:  0            
Deployment_affected:                   |   Blockedby:               
           Blocking:                   |  
---------------------------------------+------------------------------------
Changes (by wad):

 * cc: pgf (added)


Comment:

 Thanks for opening this.

 Several of the units generating these problems have had a double thermal
 pad replaced with a single pad and a liberal dose of "Artic Silver"
 (supposedly high quality thermal grease hanging around the hardware lab).
 And while overall temperature readings are better, they still trigger
 occasional errors.

 When testing B1, we will be using thermocouples on the processor and can
 verify if the occasional failing temperature is real.

 I'm worried about a mass failure of B1 runin due to noise on the thermal
 diodes... I suggest that instead of triggering on the max of the
 temperatures read from the sensors, we instead trigger on a max of the
 average of the three temperatures read from any one sensor.

-- 
Ticket URL: <http://dev.laptop.org/ticket/12095#comment:2>
One Laptop Per Child <http://laptop.org/>
OLPC bug tracking system