I have a tcpdump set in the background on OpenBSD 5.5-current from:

mercury$ sysctl kern.version
kern.version=OpenBSD 5.5-current (MERCURY.MP) #2: Sat Jun 21 08:24:41
CEST 2014
    r...@mercury.centroid.eu:/usr/src/sys/arch/amd64/compile/MERCURY.MP

late June (waiting for 5.6).

Now my problem is a certain WRT router, I'm supposed to be able to reach
it with an IPv6 address made of its MAC address, but ndp doesn't
recognize it, and it should.  While this was on my laptop which is off
(it is 5.5) right now I can duplicate what i saw in a tcpdump on my
workstation and it goes a little something like this:

mercury# ping6 fe80::1234:56ff:fe78:9abc%em0
PING6(56=40+8+8 bytes) fe80::beee:7bff:fedd:2e5a%em0 -->
fe80::1234:56ff:fe78:9abc%em0
18:43:39.537023 bc:ee:7b:dd:2e:5a 33:33:ff:78:9a:bc 86dd 86:
fe80::beee:7bff:fedd:2e5a > ff02::1:ff78:9abc: icmp6: neighbor sol: who
has fe80::1234:56ff:fe78:9abc(src lladdr: bc:ee:7b:dd:2e:5a) [icmp6
cksum ok] (len 32, hlim 255)
^C
--- fe80::1234:56ff:fe78:9abc%em0 ping6 statistics ---
2 packets transmitted, 0 packets received, 100.0% packet loss

mercury# 18:43:40.532913 bc:ee:7b:dd:2e:5a 33:33:ff:78:9a:bc 86dd 86:
fe80::beee:7bff:fedd:2e5a > ff02::1:ff78:9abc: icmp6: neighbor sol: who
has fe80::1234:56ff:fe78:9abc(src lladdr: bc:ee:7b:dd:2e:5a) [icmp6
cksum ok] (len 32, hlim 255)
18:43:41.532915 bc:ee:7b:dd:2e:5a 33:33:ff:78:9a:bc 86dd 86:
fe80::beee:7bff:fedd:2e5a > ff02::1:ff78:9abc: icmp6: neighbor sol: who
has fe80::1234:56ff:fe78:9abc(src lladdr: bc:ee:7b:dd:2e:5a) [icmp6
cksum ok] (len 32, hlim 255)

notice the 33:33:ff for the beginning mac address, I can not find the
code for that grepping through /sys/netinet6/*  where would that be
defined?  Or is this a brokenness that we encountered with this article
concerning vlan(4):

http://permalink.gmane.org/gmane.os.openbsd.misc/212691

Is it really a cause with 5.5, and fixed in 5.6, or is it normal
behaviour of ndp to add 33:33:ff, a multicast address after all but I'm
unsure if it's proper due to lack of knowledge of IPv6 and ND6.

Regards,

-peter

Reply via email to