#4732 BLOC Never A: The touchpad events that wake us up are not seen by X.

[Zarro Boogs per Child]

#4732: The touchpad events that wake us up are not seen by X.
----------------------+-----------------------------------------------------
  Reporter:  cjb      |       Owner:  dilinger      
      Type:  defect   |      Status:  new           
  Priority:  blocker  |   Milestone:  Never Assigned
 Component:  kernel   |     Version:                
Resolution:           |    Keywords:                
  Verified:  0        |  
----------------------+-----------------------------------------------------

Comment(by dilinger):

 What would that solve, if it's still taking us 1+ second to awaken?

 User experience:  dcon is frozen, machine is suspended, they drag their
 finger along the touchpad.
 The machine awakens, the kernel starts resuming.  Even if we somehow are
 able to record the touchpad packets and then send them to the input layer,
 it still takes 1.1s before we hit "Restarting tasks".  Thus, the machine
 will appear to not do anything for 1.1s, and then the cursor will jump
 across the screen.  We *really* need to get resume time reduced
 significantly..

 It looks like the worst resume time offenders are lxfb (50ms), usb (50ms
 for *hci_hcd together, 150ms for usb1-0 hub stuff, 220ms for usb2-0 hub
 stuff, 300ms for usb1-1 hub stuff), psmouse (50ms),  and some sort of
 150ms pause between when we start with ac97, and when we resume tasks.

 Of course, those times are different across suspend/resume cycles, but the
 point is still that resume is taking way too long..

-- 
Ticket URL: <http://dev.laptop.org/ticket/4732#comment:3>
One Laptop Per Child <http://dev.laptop.org>
OLPC bug tracking system