This question is in regard to USB issues with Montavista Pro 4.01 kernel
and the Davinci 6446
On the device side I am using the musb_hdrc controller driver in USB
peripheral mode (DMA enabled). Also, I am using gadget zero as our
standardized test for the USB subsytem. This is unmodified code as
supplied from the Montavista. On the host side I am using the usbtest
driver (from http://www.linux-usb.org/usbtest), again unmodified. I was
at version Pro 4.01 patchlevel pro-0785 with no issues - all tests
successfully completing.
I am also developing our own gadget driver utilizing bulk data endpoint
transfers - again this was working fine up to this point. Seeing that
there were updates for musb_hdrc (perfomance related) I applied patches
up to and including pro-1017 (9th Jan 07) Now the standardized gadget
zero test is failing (hanging) on Test 1 which is a bulk data transfer
(host to device).
Our gadget driver also fails consistently on all bulk data transfers. I
believe we have isolated the problem sufficiently to be in the musb_hdrc
controller driver which does not seem to be calling back into the
endpoint completion routine of the gadget driver even though data is
seen to go across the wire with a protocol analyzer.
Are there any known issues in the musb_hdrc controller driver that
could be causing this problem ?
_______________________________________________
Davinci-linux-open-source mailing list
Davinci-linux-open-source@linux.davincidsp.com
http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
