DHCPv6 working, some questions remain
Peter Mattern
Sat Jun 07 11:32:56 2014
Hello.
So once more a delayed answer; this time I ran into
https://github.com/archlinuxarm/PKGBUILDs/issues/861 which turned into a
problem as my card reader had decided to fail some days ago and
systemd's rescue and emergency target seem to rely on cgroups. Anyway it
took some time to get the device running again...
The good news is that it works now. Extract from the serial console
version 6.3.2 starting
eth0: executing `/usr/lib/dhcpcd/dhcpcd-run-hooks' PREINIT
eth0: executing `/usr/lib/dhcpcd/dhcpcd-run-hooks' CARRIER
DUID <DUID>
eth0: IAID <IAID>
eth0: soliciting an IPv6 router
eth0: sending Router Solicitation
eth0: reading lease `/var/lib/dhcpcd/dhcpcd-eth0.lease'
eth0: rebinding lease of <local IPv4 address>
eth0: sending REQUEST (xid 0x8d2d5ad6), next in 3.5 seconds
eth0: Router Advertisement from <link local unicast address
upstream router>
eth0: adding address <global unicast address>/64
eth0: vltime 7200 seconds, pltime 1501 seconds
eth0: waiting for Router Advertisement DAD to complete
0
1
2
3
setsockopt: SO_REUSEPORT: Protocol not available
4
5
eth0: reading lease `/var/lib/dhcpcd/dhcpcd-eth0.lease6'
eth0: confirming prior DHCPv6 lease
eth0: delaying CONFIRM6 (xid 0x82eebc), next in 0.9 seconds
eth0: acknowledged <local IPv4 address> from <IPv4 address
upstream router>
eth0: checking for <local IPv4 address>
eth0: sending ARP probe (1 of 3), next in 1.7 seconds
eth0: broadcasting CONFIRM6 (xid 0x82eebc), next in 1.0 seconds
eth0: REPLY6 received from <link local unicast address
upstream router>
eth0: adding address <global unicast address>/128
eth0: vltime 7200 seconds, pltime 3600 seconds
eth0: renew in 1800 seconds, rebind in 2880 seconds
eth0: writing lease `/var/lib/dhcpcd/dhcpcd-eth0.lease6'
eth0: waiting for DHCPv6 DAD to complete
eth0: sending ARP probe (2 of 3), next in 1.2 seconds
eth0: Router Advertisement DAD completed
eth0: executing `/usr/lib/dhcpcd/dhcpcd-run-hooks' ROUTERADVERT
eth0: DHCPv6 DAD completed
eth0: executing `/usr/lib/dhcpcd/dhcpcd-run-hooks' REBOOT6
forking to background
forked to background, child pid <PID>
Yet there remain some things strange at least to me:
First and trivial, those "marker numbers" introduced by patching dhcp6.c
can now only be seen on the serial console but no longer in systemd's
journal. I did a fresh install due to the problem stated above. But I
thought that everything was configured and compiled as before and thus
have no clue about that different logging behaviour. 'debug' is set in
dhcpcd.conf. If you should have any idea...
Second, you mentioned that SO_REUSEPORT was introduced in 3.9. But Arch
Linux i686 / x86_64, where the problem never appeared, is running 3.14.5
and Arch Linux ARM is using a fork from
https://github.com/linux-sunxi/linux-sunxi, right now 3.4.79.
If this should be of interest, I'm attaching an archive that contains
the files Arch Linux's build system is using to build the x86_64 Kernel.
Upstream Kernel and patches can be seen from the source variable in
script PKGBUILD. As for Arch Linux ARM, the relevant PKGBUILD and config
files can be found here:
https://github.com/archlinuxarm/PKGBUILDs/tree/master/core/linux-sun7i.
Thanks for your assistance and best regards
P. Mattern
Attachment:
Arch Linux x86_64 Kernel build.tar.xz
Description: application/xz
Archive administrator: postmaster@marples.name