IANA has allocated the registered numbers - DHCPv6 registry lists the following option numbers:
89 OPTION_S46_RULE [draft-ietf-softwire-map-dhcp] 90 OPTION_S46_BR [draft-ietf-softwire-map-dhcp] 91 OPTION_S46_DMR [draft-ietf-softwire-map-dhcp] 92 OPTION_S46_V4V6BIND [draft-ietf-softwire-map-dhcp] 93 OPTION_S46_PORTPARAMS [draft-ietf-softwire-map-dhcp] 94 OPTION_S46_CONT_MAPE [draft-ietf-softwire-map-dhcp] 95 OPTION_S46_CONT_MAPT [draft-ietf-softwire-map-dhcp] 96 OPTION_S46_CONT_LW [draft-ietf-softwire-map-dhcp]
Before I even noticed that, our friends on OpenWRT development team have already reacted. With a side effect of MAP-E/LW4o6/MAP-T provisioning being compiled in and activated by default in the latest images!
Actually running MAP-T will need the installation of "nat46" kernel module, but the good news is that you no longer need to pretend to be a hacker to try the code ! :) The latest nightly images of Chaos Calmer will work!
Also, the configuration is much simpler as well - the odhcp6c sends the OPTION_S46_CONT_MAPE, OPTION_S46_CONT_MAPT, OPTION_S46_CONT_LW in ORO (Option Request Option) by default as soon as the IPv6 interface is marked as "DHCPv6" - and as soon as the DHCPv6 server provisions the correct option, the MAP setup comes up!
Magic!
So, this is probably the last series of "let's build the images" exercise, and it is aimed at completely virtual setups - hence the CPE is done as a KVM guest. I've tested the KVM guest in a MAP-T setup and it works fine.
NB: The move from the high number to low number uncovered a bug in the provisioning script, sorry for that, I fixed it. Shift-reload to get the latest version of the page. Also - now that the options have been finalized, I'll start the work to integrate this script into the MAP Simulation Tool to avoid copypaste in the future.
Stay tuned & enjoy your MAPping!