Juha Ylönen wrote:
As far as I can see only data from 00:22:fa:dc:9a:ce can be seen (send from us aka laptop?). Have you tested if the adhoc connection between this two devices (arm <-> laptop) is really working (just set up the adhoc network as usual and then give them an ip instead of adding to /proc/net/batman-adv/interfaces. A simple ping test should be sufficient to say that it works(tm)).
Yes, ad-hoc itself works pretty well. the device boots up to his default ad-hoc, and I've sent the module over with scp and did some configuring over ssh. I did run couple of tests with IP disabled in the interfaces (set up exactly as in the readme) but recently I've just kept them in the default setup with IP's, will it cause problems? The problem existed with non-IP interfaces too.
batman-adv works on a lower level - so it shouldn't be a really problem in that situation. Maybe you can "tcpdump -s 0 -w test.dump" on the arm too. I am currently not sure why the arm cannot or doesn't send packets at all. The logs in 20100118132935.91200@gmx.net say that it was send, but I currently don't understand why they don't reach their destination. Maybe a dump from the arm reveals the problem - or at least we know if data gets send at all or if the data is stopped somewhere else.
Best regards, Sven