#5024 HIGH Never A: Libertas BUG() with update1-635

[Zarro Boogs per Child]

#5024: Libertas BUG() with update1-635
---------------------+------------------------------------------------------
  Reporter:  cjb     |       Owner:  dwmw2         
      Type:  defect  |      Status:  assigned      
  Priority:  high    |   Milestone:  Never Assigned
 Component:  kernel  |     Version:                
Resolution:          |    Keywords:                
  Verified:  0       |  
---------------------+------------------------------------------------------

Comment(by dwmw2):

 Kim just showed me a C2 machine where the same thing happened -- the
 {{{usb 1-1: can't restore configuration #1 (error=-32)}}} message on
 resume.

 I can deal with the mess when the device goes away and comes back; that's
 fairly much the same mess I've been looking at for rmmod/insmod which I've
 been poking at already. But that just means we'll be able to reinitialise
 the device and recover (at least, the kernel will recover; NetworkManager
 will still SEGV as described in ticket #4960).

 The underlying problem -- that the device isn't accepting the
 reconfiguration after resume -- needs to be found. Firstly, I'd like to
 see logs with {{{libertas_debug=0x126000}}} to confirm that we really are
 correctly sending the 'HOST_SLEEP' commands before we suspend. And then I
 think we want to catch it on the USB analyser.

-- 
Ticket URL: <http://dev.laptop.org/ticket/5024#comment:4>
One Laptop Per Child <http://dev.laptop.org>
OLPC bug tracking system