>> Plugging them into the same port (!) on my device one of them is
>> recognized as SuperSpeed, the other as high speed ???
>> (working on 4.14.129 LTS)
>> From dmesg, the "faulty" one:
>> [  530.585871] usb 1-2: new high-speed USB device number 4 using
>> xhci_hcd   <<<<<<<<< HUH ????
> XHCI is not like EHCI. It needs no companion controller, as it serves
> all speeds.
> This is on a lower layer than ax88179. This comes from xhci_hcd.
> Is this a regression?
I don't think its a regression.

Meanwhile some more strange things:
Depending on how fast i plug in the "faulty" one, sometimes
it is recognized as a SuperSpeed too.
But then, i have made the following observation:
- ping from the device to a host works (and vice versa)
- scp-ing a file from host to to device the adapter suddenly stalls,
  but i don't see any error message in dmesg
- device then is stalled, ping doesn't work any more
- unbind/bind driver recovers the adapter
  (and always returns SuperSpeed mode)
- ping works, scp breaks it again
- leaving the device in stall, scp tells "lost connection"
  on the host after ~10 minutes

Is there something i can do to force an error message to be seen
when the ETH2USB adapter stalls?

My current assumption is that the signal quality of the USB port is at a
corner case, and therefore some "better" Adapters work, some "bad ones"
don't. But as there is no error message seen in the dmesg, i am somehow lost.

Best regards

Reply via email to