On Thu, 5 Jun 2003, Major A wrote:

> I tried various UHCI hosts, they all work fine under all kernel
> versions I tried (2.4.{20,21}, 2.5.70), but at a much lower speed.

That reinforces my suspicion that your original problem stemmed from
something in the EHCI driver or hardware.  David Brownell is really the
best person to try to solve this.  The strongest hint I've seen so far was
in the earlier log you sent, that showed the -19 error code from
usb_reset_device().

Maybe the best thing to try would be to repeat that test, but with more 
debugging output.  Do you think you can insert

        printk(KERN_INFO "%s %d\n", __FUNCTION__, __LINE__);

statements at various judicious places in the usb_physical_reset_device() 
function near the end of drivers/usb/core/hub.c?  That might help pinpoint 
just where the problem strikes.

> > Maybe another possibility is a flaw in the filesystem code (I'm guessing 
> > wildly here).  Is it possible to use a different sort of filesystem and 
> > repeat the tests?  And related to that, did your original command ever 
> > end up writing anything to the md5sums output file?
> 
> I'm afraid not -- the disk is in one partition, and I need the data
> that's on there. Currently I have no way of moving it elsewhere for a
> test.
> 
> But why should a filesystem or SCSI problem cause a timeout in the USB
> subsystem? There must be something wrong with USB, even if other parts
> of the kernel aren't spotless either.

It wouldn't have caused the original problem; that clearly had to do with
the USB system.  I was just speculating about why your latest test ended
up hanging in spite of no errors occurring.  Or apparently no errors -- is
there any possibility that some additional kernel messages failed to make
their way into your logfile?

Alan Stern



-------------------------------------------------------
This SF.net email is sponsored by:  Etnus, makers of TotalView, The best
thread debugger on the planet. Designed with thread debugging features
you've never dreamed of, try TotalView 6 free at www.etnus.com.
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel

Reply via email to