Hi everyone,
I've been working on those patches for seperating the link quality measurements from the OGMs and putting that into a seperate protocol with its own packets a little more. This Neighbour Discovery Protocol will allow us to have faster intervals for NDP or different intervals on different interface types, as well as creating individual convergence speed optimizations for the link quality measurements (NDP) or the path quality propagation (OGM) later.
This is a first draft and basically what we've been discussing on the WBMv3 in Italy and which was documented a little in our wiki.
I haven't added my Signed-off-by yet as I'd like to get some feedback first, especially for the last patch. Here I'm not sure if that's sufficient and ok to do it like that (the limit check seemed redundant to me) or if I'm missing something. Especially I'm not really sure what tq_own was representing as the link quality seemed to be substracted with the real_packet_count.
After the overall way of doing it is acknowledged, I'd start cleaning up the OGM handling and removing variables and redundant LQ measurements.
Cheers, Linus