#9423 NORM 1.5-F11: Stabilize OFW USB handling on XO-1.5

[Zarro Boogs per Child]

#9423: Stabilize OFW USB handling on XO-1.5
-------------------------------------------+--------------------------------
           Reporter:  wmb at firmworks.com    |       Owner:  wmb at firmworks.com
               Type:  defect               |      Status:  closed           
           Priority:  normal               |   Milestone:  1.5-F11          
          Component:  ofw - open firmware  |     Version:  1.5-B2           
         Resolution:  fixed                |    Keywords:                   
        Next_action:  test in release      |    Verified:  0                
Deployment_affected:                       |   Blockedby:                   
           Blocking:                       |  
-------------------------------------------+--------------------------------

Comment(by wmb at firmworks.com):

 Via responded to the inquiry about this problem with the observation that
 their DOS stress test works.

 One thing I discovered in my testing is that you can make the problem - or
 a similar problem - happen more often if you set the EHCI sleep timer
 field in D16F4 Rx4b to 11 (register value = 0x60) and then poll the
 overlay area in the queue head immediately after linking the QH onto the
 async list.  The problem then is so severe that configuration descriptor
 reads will fail on most devices, so they won't show up in the probe list.
 If you delay 50 uS before polling, the config descriptor read goes away.

 I'm beginning to suspect a cache coherency problem.

-- 
Ticket URL: <http://dev.laptop.org/ticket/9423#comment:10>
One Laptop Per Child <http://laptop.org/>
OLPC bug tracking system