On Tue, 18 Dec 2007, Graeme Gill wrote:
> Alan Stern wrote:
>
> > That's right. The semantics of USBDEVFS_DISCONNECT has changed since
> > 2.6.9. It now requires that there be no drivers bound to any
> > interfaces; thus you have to call usb_detach_kernel_driver_np() first.
>
> Right, it does make sense. The confusing part is usb_detach_kernel_driver_np()
> taking the interface as an argument.
It seems logical to me. A device can have several interfaces, each
with its own driver. So to detach a driver from an interface, your
program should have to specify which one.
> > There's also an excellent chance you don't need to call
> > usb_set_configuration() at all. Your devices have got only one
> > configuration, right?
>
> Well, I don't know. All the current devices may only have one
> interface, but this is in common code that is intended to
> cope with any number of currently unknown future devices, so
> I am just following l ibusb recommendations, and doing a
> usb_set_configuration(), to make sure they work.
I suspect that it acts merely as a weak kind of reinitialization.
> > It sounds like you're not getting the same USB data sent on Linux and
> > on the other operating systems. Have you tried monitoring the USB
> > transfers and comparing the low-level data?
>
> I don't have (or have access) to a USB bus analizer. I can
> spy on the MSWindows stuff to a degree using snoopypro.
Yes, that will work.
> I'm not familiar enough with the Linux USB driver yet to know
> what facilities it has in this regard. Pointers ?
See the kernel source file Documentation/usb/usbmon.txt.
Alternatively, for debugging a userspace driver you can do:
echo 1 >/sys/module/usbcore/parameters/usbfs_snoop
and get the information from the system log. usbmon is somewhat more
reliable.
Alan Stern
-
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
