On Thu, 17 Jul 2003, Charles Lepple wrote:

> The firmware that I am loading is actually the ep2_inout firmware. (I am
> trying to establish that the USB physical layer is sound, and I assume
> that the EZ-USB chip is functional.)
> 
> I just finished testing with several other chips, and I can confirm that
> test 10 (-g > 2) acts consistently on the following devices:
> 
> - the original homebrew EZ-USB FX board
> - a Keyspan USA-19 (AN2131Q EZ-USB)
> - a Keyspan USA-19Qi (AN2131Q EZ-USB)
> - another commercial EZ-USB FX board
> 
> None of the chips in these devices seem to have outstanding errata which
> might be affecting usbtest.

When you say it acts consistently, do you mean that it provokes the same
error response as in the log you posted?  That would certainly indicate
another error in the UHCI driver rather than one in your board.

> I am still more than willing to test patches; however, it's not going to
> hold me up if it would make sense for you to tackle more pressing USB
> issues first.

Thanks for the offer; I'm sure I'll take you up on it.  Considering the 
importance of the host controller drivers to the USB subsystem, a problem 
like this deserves a good deal of attention.

At the moment, I'm not quite sure what additional diagnostics would help.  
Perhaps printing the urb values in addition to the urbp's, and also 
printing an indication of when the URBs are queued and dequeued would be 
good.  Maybe also printing the transfer length at the time of submission.

But I can't think how to determine if somehow the driver is mistakenly
reporting a stall or NAK as a short read.  From the code and the debugging
output it certainly _looks_ correct.  The only surefire way to tell is to
use a bus analyzer, which I assume is impractical here.  I suppose it's
vaguely possible that the USB controller itself is broken -- talk about
grasping at straws!

Alan Stern




-------------------------------------------------------
This SF.net email is sponsored by: VM Ware
With VMware you can run multiple operating systems on a single machine.
WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the
same time. Free trial click here: http://www.vmware.com/wl/offer/345/0
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel

Reply via email to