[Nut-upsdev] Newhidups gets unbound after a while

Peter Selinger selinger at mathstat.dal.ca
Wed May 3 15:49:55 UTC 2006


Patrick Agrain wrote:
> 
> Hello Peter,
> 
> This morning, I told about a problem when CONFIG_USB_HID is not set. The 
> problem exist, but not as I described it.
> [...]
> 
> But I think you're right. The reconnection problem is due to the "unknown" 
> state of Vendor and Product and not to the fact that Driver=(none).

Yes, the kernel driver has nothing to do with that. I usually use a
USB mouse, so I have to have USB support in the kernel. The kernel
will usually attach itself to any device that it thinks might be a
mouse or a keyboard, even if it turns out not to be. This is when you
see "Driver=hid". NUT causes the kernel to detach its internal driver
from the device; this is why you see:

failed to claim USB device, trying 2 more time(s)...
detaching kernel driver from USB device...
trying again to claim USB device...
HID descriptor retrieved (Reportlen = 769)

After this, NUT attaches its own driver to the device, which is when
you see "Driver=usbdevfs". The details of this could vary between
operating systems and versions. 

Anyway, I agree that none of this is the source of your problems. The
problem seems to be an inability to get the vendor and product strings
after a reconnect. This is a bit unusual. Do you have the latest
libusb installed? Check /usr/lib/libusb*. Should be at least
libusb-0.1.so.4.4.2 or so.

-- Peter



More information about the Nut-upsdev mailing list