Alan Stern wrote:
>
> That looks like you managed to trigger the VIA USB bug. You can tell for
> sure by searching for error lines in the log that report -75 or "babble".
>
> Once that bug is triggered the USB controller is completely useless.
> There's currently no way to rescusitate it short of rebooting.
I do not get any -75 or babble. If you mean the entire controller is
useless, that isn't the case either as my USB mouse still works properly.
David Brownell wrote:
> Brian S. Stephan wrote:
>
>> usb-storage: Status code -32; transferred 31/31
>>...
>> usb-storage: Status code -32; transferred 31/31
>
> Those messages seem rather bizarre. If maxpacket is any legal
> value (512 for high speed, else 64/32/16/8 at full speed) then
> the only way it should see a STALL (status -EPIPE) is to have
> gotten normal treatment for the 31 bytes -- ending in a short
> packet -- then have the HCD issue *another* packet, to which
> the device responds with a STALL packet.
>
> But there's no reason to issue that last packet, since all
> requested 31 bytes were already transferred (written).
>
> More typically I'd suspect a STALL in response to either
> the first write (so "transferred 0/31"), or else with a
> bizarre full speed device some other full-packet write
> ("8/31", "16/31", or "24/31").
>
> Something's clearly going wrong with your HC/HCD if it's
> giving you those status messages.
Certainly, and it's twice as bizarre because as far as I can tell it's
something in 2.6 that triggers the whole mess, although the mess is most
probably not the kernel's.
Another (one last?) thing noticed, when doing the math for the combination
of such large transfer messages:
usb-storage: Status code 0; transferred 65536/65536
the total amount that was supposedly transferred is much larger than the
source file, and that addition is well before the timeouts. I assume the
64kb transfers are the file data and not protocol, so it would seem
something is going wrong there, too.
Anyway, this is mainly just a last followup, is it time to declare this a
lost cause?
Brian
-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
