[sane-devel] usb_control/bulk_msg: timeout

Henning Meier-Geinitz henning@meier-geinitz.de
Thu, 2 Jan 2003 23:05:25 +0100


On Thu, Jan 02, 2003 at 01:53:06PM -0800, Bill Moseley wrote:
> It would be too bad if it's a motherboard.  Seems unlikely, though.
> If I were able to test the the hardware and find out it's ok, then where
> should I focus my debugging?  

I guess then it's somewhere in the kernel USb core. The
linux-usb-users mailing list may be more appropraite in this case

> Is the usb_control/bulk_msg: timeout message generated in the scanner
> module?

No, it's from the USB core. The "NAK received" messages are from the
scanner module. While a write to the scanner resulting in timeout will
just return an error immediately (NAK received), a read will try
multiple times (12 by default, I think) before returning an error
(Excessive NAKs received).

> And is that what is causing the front-end X apps to report the
> "I/O" error? 

Probably yes.

> I wonder if I could get the scanner module to tell more about what's
> timing out.

You can enable debugging in drivers/usb/scanner.h. And/or add your own
debug messages.

> I assume the scanner module talks to the usb module to get out to the
> device.

The scanner module uses the function "usb_bulk_msg" ffrom the USB core
to read/write to the device.

> It seems odd to me that one X frontend works yet the other fails with an
> I/O error. 

If you have only one installation of SANE, it may be just a timing

> I don't know the Sane API but I assumed that the all apps would see
> the same problem using the same API.

Usually that's the case.