#9519 NORM Not Tri: DSDT - possible loss of LID events

[Zarro Boogs per Child]

#9519: DSDT - possible loss of LID events
---------------------------------+------------------------------------------
 Reporter:  wmb at firmworks.com    |                 Owner:  pgf          
     Type:  defect               |                Status:  new          
 Priority:  normal               |             Milestone:  Not Triaged  
Component:  ofw - open firmware  |               Version:  not specified
 Keywords:                       |           Next_action:  never set    
 Verified:  0                    |   Deployment_affected:               
Blockedby:                       |              Blocking:               
---------------------------------+------------------------------------------
 The _LID method of the LID device in the XO-1.5 DSDT should return a value
 from the GPI7 register (port 0x448[7]) instead of from the toggled state
 variable that controls the edge detection polarity.

 The possible failure mode is as follows:

 If you open the lid then close it quickly, you would get the open event,
 but the close had already happened before the ACPI code got around to
 handling the open and flipping the edge polarity,
 so the edge detector got flipped, and then the next open doesn't get
 detected, because the detector was looking for a new close

 The _LID method should also compare the state variable with the GPI
 register, fixing the edge polarity if necessary.

-- 
Ticket URL: <http://dev.laptop.org/ticket/9519>
One Laptop Per Child <http://laptop.org/>
OLPC bug tracking system