#9765 LOW 1.5-sof: frame battery icon lags charging initiation

[Zarro Boogs per Child]

#9765: frame battery icon lags charging initiation
--------------------------------+-------------------------------------------
           Reporter:  mikus     |       Owner:  pgf                              
               Type:  defect    |      Status:  new                              
           Priority:  low       |   Milestone:  1.5-software-later               
          Component:  kernel    |     Version:  Development build as of this date
         Resolution:            |    Keywords:                                   
        Next_action:  diagnose  |    Verified:  0                                
Deployment_affected:            |   Blockedby:                                   
           Blocking:            |  
--------------------------------+-------------------------------------------

Comment(by rsmith):

 Replying to [comment:18 pgf]:
 > cc'ing richard -- i think this might be a result of the way battery
 events are "batched" by the timer handling in CheckForBatterySCIEvents().
 i said above that it might be because the kernel or userlevel isn't
 watching for the event, but i think it's more likely that the EC is
 missing a condition for which it should do
 OEM_Put_SCI_Queue(SCI_BATTERY_STATUS_CHANGE).

 This appears to either be a subtle race condition in sugar or perhaps EC
 is not setting the bat status flags correctly or both.  Both q3a35 and my
 current dev tree send up a SCI 0x40 and then a 0x02 which is AC change and
 batt status change respectively.  With q3a35 I get the above behavior (ie
 charged) but with my dev tree it says charging.  Dealing such a case in
 the XO-1 is the reason that there is a delay between the AC SCI and the
 battery status SCI in the first place.

 If this continues to be a problem after I release new EC code ( > 1.9.24 )
 I'll dig deeper.

-- 
Ticket URL: <http://dev.laptop.org/ticket/9765#comment:20>
One Laptop Per Child <http://laptop.org/>
OLPC bug tracking system