#12320 NORM 4-softw: xo-4 thermal sensor anomalies

[Zarro Boogs per Child]

#12320: xo-4 thermal sensor anomalies
--------------------------+-------------------------------------------------
 Reporter:  pgf           |                 Owner:  pgf                              
     Type:  task          |                Status:  new                              
 Priority:  normal        |             Milestone:  4-software                       
Component:  not assigned  |               Version:  Development build as of this date
 Keywords:                |           Next_action:  reproduce                        
 Verified:  0             |   Deployment_affected:                                   
Blockedby:                |              Blocking:                                   
--------------------------+-------------------------------------------------
 i think the behavior of the thermal watchdog generator has changed,
 perhaps from B0 to B0B silicon for mmp3.  we used to (on A2 boards) be
 able to fairly accurately set the watchdog, based on reading the some or
 all of the sensors while heating the board.  i believe the sensor
 configured as watchdog was readable along with the other two.

 now, it seems that a sensor configured as watchdog can't be read (always
 shows 0), and while a thermal watchdog eventually does occur, it seems to
 be at a much higher temperature than configured.  just now on a B1 i set
 the watchdog threshold to 89.5C ("echo 2 > enable_watchdog"), but
 according to the other two sensors, the watchdog fired at 108C.

 this ticket is a reminder ticket -- i want to verify my recollections of
 A2 behavior, and look at the behavior of the three sensors on B0B (XO-4
 B1) silicon w/o the watchdog configured, to be sure we're not just using
 the coolest of the sensors.

-- 
Ticket URL: <http://dev.laptop.org/ticket/12320>
One Laptop Per Child <http://laptop.org/>
OLPC bug tracking system