> On Mon, 27 Jun 2005, Dag Nygren wrote:
> 
> > All the problems seems to occur after a report back from the
> > USB controller with the cc register set at 6. This means that
> > the controller detected a PID check error. If I read the specs right
> > this transfer should then be retried (by the USB layer I presume)
> > and not reported back as failed to upper layers. Am I right?
> 
> I'm not very familiar with the OHCI spec, but if you're seeing a PID check 
> error then probably the transfer _was_ retried already (by the hardware).  
> It has to be reported as failed to the upper layers if the PID check 
> occurs every time the transfer is retried.

Well, I just checked a bit more and it seems like the OHCI device has already
tried 3 times before it reports a failure. What is interesting though is that 
the OHCI
specs states that a field called ErrorCount in the should be initialized to 0 
by the
host driver when the General TD is queued  and I haven't been able to locate 
this
in the code so far. This ErrorCount is what the hardware use as the counter of 
how
many retries we had so far.

> > Additionally there really seems to be (as you pointed out) problems
> > in the whole SCSI (and possibly USB) layer in the way it handles
> > error conditions.
> 
> There definitely are some bugs in the SCSI layer.  They are being worked 
> on.

Just tell me if I can help.

Dag




-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
_______________________________________________
[email protected]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel

Reply via email to