IPv6 Local Peer Discovery is not working about kadnode HOT 11 CLOSED

hm, the error string is not in the current code anymore. Could you try to compile Kadnode from sources and give it a try?

from kadnode.

Hi
On the router and on my Mac is the latest Master Branch running f14b8af
I can see https://github.com/mwarning/KadNode/blob/master/src/ext-lpd.c#L119 for mac os x and https://github.com/mwarning/KadNode/blob/master/src/ext-lpd.c#L75 for Linux.
The function mcast_set_group() is calling for every ip6 addr on the Interface. In my case 6 times on Mac and 3 times on Linux Router.
If you want you can have access via ipv6 to a test Router. I Plan to debug this on Saturday.

from kadnode.

ah, ok. I see now why I could not find that error string.
Anyway, seeing the error message I would think that SO_REUSEADDR was not set. But it looks like the code does that. No idea right now what went wrong. Maybe something specific for MacOS.

from kadnode.

Same Problem on Linux!

from kadnode.

Hi
I think i have Fix for that Issue. :-)
RFC: 80ffb46

from kadnode.

@stargieg thanks for looking into the issue. :-)

from kadnode.

I will give the multicast code a full review in a few days. Let's see if there are also other issues.

from kadnode.

Hi @mwarning
This patch works only for one Interface. I am working on it.
But the bigger Problem is the fe80 or not fe80 question.
DHT receive any but not fe80 https://github.com/mwarning/KadNode/blob/master/src/dht.c#L376
LPD send only on fe80 https://github.com/mwarning/KadNode/blob/master/src/ext-lpd.c#L143
I think the DHT Code is the right way, because of fe80 dup on openwrt

root@wdr43-007:~# ip -6 addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 
    inet6 ::1/128 scope host 
       valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qlen 1000
    inet6 fe80::6666:xxxx:xxxx:55fc/64 scope link 
       valid_lft forever preferred_lft forever
5: br-lan: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 
    inet6 2a00:xxxx:xxxx:8405::1/64 scope global noprefixroute 
       valid_lft forever preferred_lft forever
    inet6 fd0d:8a84:7a92:1::1/64 scope global noprefixroute 
       valid_lft forever preferred_lft forever
    inet6 fe80::6666:xxxx:xxxx:55fc/64 scope link 
       valid_lft forever preferred_lft forever
7: eth0.2@eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 
    inet6 fe80::6666:xxxx:xxxx:55fc/64 scope link 
       valid_lft forever preferred_lft forever
16: br-fflandhcp: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 
    inet6 fd0d:8a84:7a92::1/64 scope global noprefixroute 
       valid_lft forever preferred_lft forever
    inet6 2a00:xxxx:xxxx:8404::1/64 scope global noprefixroute 
       valid_lft forever preferred_lft forever
    inet6 fe80::6666:xxxx:xxxx:55fe/64 scope link 
       valid_lft forever preferred_lft forever
17: wlan1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qlen 1000
    inet6 fe80::6666:xxxx:xxxx:55fe/64 scope link 
       valid_lft forever preferred_lft forever
18: wlan0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qlen 1000
    inet6 fe80::6666:xxxx:xxxx:55fd/64 scope link 
       valid_lft forever preferred_lft forever

The wlan0 and wlan1 iface are br-fflandhcp members, but have a fe80 addr where kadnode ist listening
and sending. :-(
The same with eth0 and br-lan.

from kadnode.

Please review this patch also.
stargieg@7191c40

from kadnode.

@stargieg fyi, I did a partial rewrite of the multicast code to simplify the mechanism and included some fixes of yours. It still needs to be tested. I will upload it when it proves to work.

from kadnode.

I simplified the discovery code a lot. This will probably help IPv6 support.

from kadnode.