On Fri, 14 Feb 2014, Paul Zimmerman wrote: > > The problem may have something to do with the fact that the keyboard's > > internal hub runs at full speed. The other devices you tested, which > > worked okay, would have been low speed and high speed but not full > > speed. > > The original email said: > > > The status returned from the hcd is 0x4a (Stalled, XactErr, SplitXState) > > Could the SplitXState be menaingful? Maybe something to do with transaction- > translator support in u-boot? No doubt the problem is related to the TT support somehow. In itself, the SplitXState flag doesn't mean much -- just that an error occurred while the controller was trying to complete a split transaction. It probably would be more useful to stick the bus analyzer on the high-speed part of the transfer, i.e., between the root hub and the internal hub. I assume this isn't feasible, though. About the only path forward is to do what Peter suggested: dump the QH and qTD data structures to see if they are correct. For example, is the EPS field in the QH set to indicate Full Speed? Alan Stern -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html