On Thu, May 28, 2015 at 12:07 PM, Peter Griffin peter.griffin@linaro.org wrote:
Hi Kevin,
<snip>
Do you have anything plugged into the USB OTG port?
Yes, I still have the cable used to fastboot plugged in.
I just managed to replicate Timeouts if I have both the USB ASIX adapter connected to a 'USB Host' port, and also a cable plugged into the 'USB OTG' port simultaneously.
Doh! I should've realized that because I now remember Tyler mentioning this limitation when he got it working in LAVA. So you basically have to choose between fastboot and USB host support. Nice tradeoff. ;)
If I unplug the OTG cable, then "usb start" detects the ASIX dongle. Cool.
Brilliant :-) that is excellent news
Next problem... it doesn't seem to DHCP. Does DHCP work for you?
I think I tried it initially when I got the port going, it didn't work and I set a static address to progress and so far never went back to debug why.
Heh, that's my plan as well. I've already spent way to much time getting annoyed at the firmware of this platform.
Using tcpdump on the LAN, I see it's doing the bootp request, but using a MAC addr of all zeros, even though u-boot tells me it's using the MAC from the adapter, and "usb start" has set usbethaddr accordingly.
Setting static IP addresses, I'm able to TFTP, but checking ARP tables on the TFTP server, it reports an incomplete MAC address also.
Ah interesting, you've debugged it more than me I'm afraid. It would certainly be good to get this working properly.
The good news is Iterating with U-Boot versions is now *much* quicker when you can tftp into RAM and don't have to faff with jumpers and fastboot not enumerating as a USB device ;-)
Agreed. Also, for my automation farm, u-boot with netboot is *so* easy to deal with. Now, as soon as hikey it's mainline, it will be included in the kernelci.org results.
Thanks again for all your help, and let me know if you want me to test mainline u-boot versions as they're coming out and fixing issues.
Kevin