#10108 NORM Not Tri: USB ports can't be powered during suspend

[Zarro Boogs per Child]

#10108: USB ports can't be powered during suspend
----------------------+-----------------------------------------------------
 Reporter:  pgf       |                 Owner:  wad          
     Type:  defect    |                Status:  new          
 Priority:  normal    |             Milestone:  Not Triaged  
Component:  hardware  |               Version:  not specified
 Keywords:            |           Next_action:  never set    
 Verified:  0         |   Deployment_affected:               
Blockedby:            |              Blocking:               
----------------------+-----------------------------------------------------
 although USB port power is controlled via the VX855 GPO10 output, it seems
 that that output may not be functional during suspend.

 ----

 <pgf> Mitch_Bradley: i'm looking at pdf page 63 of DS_VX855_VX875_112.pdf,
 and am wondering what Note 2, regarding GPO10, might mean: "For GPO10, GPO
 function does not work under system suspend state."

 <pgf> Mitch_Bradley: i'm afraid it may mean that there's no way to keep
 USB devices powered during suspend.  unless OFW is explicitly turning off
 USB power somewhere i'm not seeing it, i see no reason for the power to
 drop when i use the 's' command.

 ...

 <Mitch_Bradley> pgf: I'm pretty sure that OFW doesn't explicitly turn off
 the USB power

 <Mitch_Bradley> pgf: I just tested it by stepping all the way to the "ax-
 call" in s3 (USB power still on), then manually executing all the
 instructions that touch the hardware in the "suspend-entry" code in
 resume.bth

 <Mitch_Bradley> the power stayed on during that whole exercise, but went
 off as soon as I executed the ax-call

 <Mitch_Bradley> It also went off if I just did "2400 4040 pw!" instead of
 the ax-call, thus pounding the register that does the actual "go to sleep
 now" step

 <pgf> Mitch_Bradley: thanks for checking.  so either there's another bit
 somewhere (No! not in a VIA chip!), or we can't leave the power on...


 ----

 we need to confirm this (mis)behavior with Via.

-- 
Ticket URL: <http://dev.laptop.org/ticket/10108>
One Laptop Per Child <http://laptop.org/>
OLPC bug tracking system