Subject: Re: Ehci/Umass bug ?
To: Charles M. Hannum <abuse@spamalicious.com>
From: Vincent <10.50@free.fr>
List: tech-kern
Date: 03/22/2005 17:35:22
This message got lost somewhere in the multiverse? ;)

I have given up with that problem, but I am sure it is still there. What 
is the current status of the EHCI driver?

Thanks
Vincent

Charles M. Hannum a écrit :
> On Thursday 25 November 2004 08:44, Lennart Augustsson wrote:
> 
>>Vincent wrote:
>>
>>>Hi,
>>>
>>>
>>>>BABBLE certainly means the device is talking when it shouldn't.
>>>>It could be a consequence of the earlier XACTERR, maybe.
>>>
>>>What does that XACTERR mean ?
>>
>> From the EHCI spec:
>>"Set to a 1 by the Host Controller during status update
>>in the case where the host did not recieve a valid response
>>from the device (Timeout, CRC, Bad PID, etc).  This bit will
>>only be set for IN transactions."
> 
> 
> Unless you have a bad cable, it's usually caused by a "bad PID" -- which 
> typically means the data toggle is in the wrong state.
> 
> There are some weird cases that can cause this to happen under normal use 
> (IIRC, aue can cause it by doing a control transfer while there's an 
> outstanding bulk transfer queued).  I've thought about how to fix this all 
> cleanly, but it's a pain.
> 
> As for what's wrong with this particular device, I haven't looked at it in 
> detail.  :-)
> 
>