#2258 HIGH 9.1.0: Power savings in keyboard/touchpad in suspend/resume in *new version for MP*

[Zarro Boogs per Child]

#2258: Power savings in keyboard/touchpad in suspend/resume in *new version for
MP*
------------------------+---------------------------------------------------
   Reporter:  jg        |       Owner:  dilinger
       Type:  defect    |      Status:  reopened
   Priority:  high      |   Milestone:  9.1.0   
  Component:  kernel    |     Version:          
 Resolution:            |    Keywords:  power   
Next_action:  diagnose  |    Verified:  0       
  Blockedby:            |    Blocking:          
------------------------+---------------------------------------------------
Changes (by gnu):

 * cc: wad, bernie (removed)
 * cc: cjb (added)
  * status:  closed => reopened
  * next_action:  => diagnose
  * resolution:  fixed =>
  * milestone:  Trial-3 => 9.1.0


Comment:

 Reopened because we never implemented powering down the keyboard/touchpad
 in production software.

 See also #2770 about *when* to power it off and on.  The current bug is
 more detailed on *how* to power it off and on.

 (One question, does merely pulling the data line low *briefly* (e.g. for
 one bit-time or eight bit-times) power the keyboard/touchpad back up?  Or
 do you have to pull it low for a whole second?  It seemed to me in reading
 the spec that a brief low would do: "The method of operation again is a
 thing to make the signal line of MS-DAT Low. This operation is the same as
 doing operation that sends the command to HGPK (device)."  If it only
 takes sending a zero byte, and then waiting a second, it's much easier to
 power it back up, and this power-saving improvement can get done faster.)

-- 
Ticket URL: <http://dev.laptop.org/ticket/2258#comment:7>
One Laptop Per Child <http://laptop.org/>
OLPC bug tracking system