On 03/01/2012 16:52, Rick Macklem wrote: > The basics are in RFC4291, but I think that inet_pton(3) knows how to > deal with it. (I think "::" can be used once to specify the longest # > of 16bit fields that are all zeros.)
RFC 4291 has a basic description of the textual representation of IPv6 addresses, but it is ambiguous: there are several different ways to present the same address according to the RFC 4291 rules. inet_pton(3) follows RFC 5952 which is a superset of the 4291 rules, only allowing a single, unambiguous representation for each IPv6 number. > After inet_pton() has translated it to a binary address, then the macros > in sys/netinet6/in6.h can be used to determine if the address is a loopback, > etc. While 5952 describes how to correctly present an IPv6 address, there's still lots of important other stuff in 4291. For instance bit 70 in an IPv6 address flags that the address is derived from a number hardwired into the interface -- typically the ethernet MAC address, as is commonly done for SLAAC (StateLess Address Autoconfiguration: RFC 4862, rtsold(8), rtadvd(8)). So an arbitrarily invented address should have that bit set to zero. Bit 71 is also special, indicating manycast vs unicast, and should also be zero for the vast majority of uses. See http://www.infracaninophile.co.uk/articles/hotchpotch.html#rand-aaaa.pl for some perl code that operates in this area. Also of interest: RFC 5156 which lists IPv6 address ranges dedicated to special purpose usages, and RFC 4193 which roughly is the IPv6 equivalent to RFC 1918, but somewhat more complicated. You might find https://www.sixxs.net/tools/grh/ula/ relevant too, although actually using that as a registry is pretty pointless. Cheers, Matthew -- Dr Matthew J Seaman MA, D.Phil. 7 Priory Courtyard Flat 3 PGP: http://www.infracaninophile.co.uk/pgpkey Ramsgate JID: matt...@infracaninophile.co.uk Kent, CT11 9PW
signature.asc
Description: OpenPGP digital signature
