#8901 HIGH Not Tri: Kernel/EC loose keyboard data when running in Synaptics protocol mode

[Zarro Boogs per Child]

#8901: Kernel/EC loose keyboard data when running in Synaptics protocol mode
-------------------------------------------+--------------------------------
           Reporter:  dilinger             |       Owner:  rsmith       
               Type:  defect               |      Status:  new          
           Priority:  high                 |   Milestone:  Not Triaged  
          Component:  embedded controller  |     Version:  not specified
         Resolution:                       |    Keywords:               
        Next_action:  diagnose             |    Verified:  0            
Deployment_affected:                       |   Blockedby:               
           Blocking:  8894                 |  
-------------------------------------------+--------------------------------

Comment(by dsaxena):

 Replying to [comment:19 pgf]:
 >
 > re: the dropped key down events -- i verified yesterday that we're not
 seeing them at the input side of the EC -- everything that's coming in is
 going out.  it's not clear to me, however, that this is a synaptics issue
 -- we're not losing their bytes, right?  we're losing keyboard bytes.

 Good point. It only shows up in Synaptics mode, which is why I asked. The
 only guess I had was too much data for the EC to process but if we're
 loosing it at input side I'm a bit baffled. Do you have ability @ 1CC to
 oscope the signals and see what's happening at HW level? Is there anything
 else buffering the PS/2 bytes before they hit the EC?

-- 
Ticket URL: <http://dev.laptop.org/ticket/8901#comment:20>
One Laptop Per Child <http://laptop.org/>
OLPC bug tracking system