Timo Juhani Lindfors <timo.lindf...@iki.fi> writes:
Hi,
> Arnaud Patard (Rtp) <arnaud.pat...@rtp-net.org> writes:
>> update other fields in the structure accordingly) like in the *not*
>> *tested* appended patch. I didn't even build tested it :)
>> If it works for you, I'll try to submit it for next merge window.
>
> Whoa, thanks a lot! Catting /dev/usbmon1 on PC side now shows
> interrupt transfers. However, I have not managed to get my usb
> keyboard gadget to work reliably so I can not say much about
> reliability yet.
There are others gadget driver using interrupt endpoints so with them
you should be able to tell if the problem comes from your gadget driver
or from some bugs in the udc driver (I've never tested interrupt
endpoint before so you may be just unlucky and have hitten some bugs).
> It works with libusb but the kernel usbhid.ko driver
> seems to just flood logs with
>
> usbhid: input irq status -75 received
>From Documentation/usb/error-codes.txt:
-EOVERFLOW (*) The amount of data returned by the endpoint was
greater than either the max packet size of the
endpoint or the remaining buffer size.
"Babble".
Maybe you can have a look at the request you're allocating and sending
to the udc driver and check their size (should be <= at the endpoint
maxpacket size). You can also look at the usb endpoint descriptor you're
sending to the host and check that the value is fine. [ fwiw, ep[1-4]
hardware maxpacket size is 64 or 128 and the udc is configured to use
it but one can configure it through the ops->ep_enable stuff ]
Regards,
Arnaud
_______________________________________________
devel mailing list
devel@lists.openmoko.org
https://lists.openmoko.org/mailman/listinfo/devel
