#9423 NORM 1.5-fir: Stabilize OFW USB handling on XO-1.5
Zarro Boogs per Child
bugtracker at laptop.org
Thu Nov 19 17:39:20 EST 2009
#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-firmware
Component: ofw - open firmware | Version: 1.5-B2
Resolution: fixed | Keywords:
Next_action: test in release | Verified: 0
Deployment_affected: | Blockedby:
Blocking: |
-------------------------------------------+--------------------------------
Changes (by wmb at firmworks.com):
* milestone: 1.5-software => 1.5-firmware
Comment:
I am convinced now that the problem with the Kingston DataTraveler stick
is related to descriptor polling.
I changed the timing of the descriptor polling routine. It has an initial
delay before it starts polling, then an interval between polls. Normally
the initial delay is 50 microseconds and the interval is 1 millisecond.
{{{
Initial Interval Failure Rate
50 uS 1 mS one failure per 8 tries, average
60 mS 1 mS never fails
20 mS 1 mS one in 8, average
20 mS 10 uS fails every time
}}}
The Kingston DataTraveler takes 44 mS (observed on analyzer) to respond to
the transaction (time from first NAK'ed IN to successful ACK'ed IN), so
the 60 mS
"never fails" case means that the CPU is not polling the descriptor area
until
after the transaction is finished.
In the other (failing) cases, the CPU is polling during the transaction.
By increasing the polling rate (10 uS interval), the failure can be made
to happen
every time.
--
Ticket URL: <http://dev.laptop.org/ticket/9423#comment:12>
One Laptop Per Child <http://laptop.org/>
OLPC bug tracking system
More information about the Bugs
mailing list