RFC 6751, "Native IPv6 behind IPv4-to-IPv4 NAT Customer Premises Equipment (6a44)", October 2012Source of RFC: INDEPENDENT
Errata ID: 3385
Publication Format(s) : TEXT
Reported By: Andreas Cudok
Date Reported: 2012-10-18
Rejected by: Nevil Brownlee
Date Rejected: 2014-01-20
Section 6.6.2 says:
RR4-5 INVALID IPv6/UDP/IPv4 PACKET For ANY other case, the 6a44 relay MUST discard the packet.
It should say:
RR4-5 INVALID IPv6/UDP/IPv4 PACKET: INCONSISTENT SOURCE ADDRESSES If ALL the following conditions are satisfied, the 6a44 relay MUST discard the packet and return to the source an error-signaling bubble (i.e., with the "Bubble ID" field set to 0) which conveys the up-to-date IPv6 prefix of the client: (1) the IPv4 packet contains a complete UDP datagram (protocol = 17, offset = 0, more-fragment bit = 0); (2) the UDP payload is an IPv6 packet (length of at least 40 octets, version = 6); (3) the IPv6 source address starts with the 6a44-network IPv6 prefix followed by a value (in the field that is composed of bits 48-79) that is different from the UDP/IPv4 source address of the received packet; (4) the IPv6 destination address is not a Teredo address whose embedded IPv4 address is the 6a44-relay anycast address. RR4-6 INVALID IPv6/UDP/IPv4 PACKET: ANY OTHER INVALID PACKET For ANY other case, the 6a44 relay MUST silently discard the packet.
The conditions when to send an error-signaling bubble must be exactly specified. This is done by the modified RR4-5 rule. The orginal rule has moved to RR4-6 and changed from "discard" to "silently discard", i.e. without returning an error-signaling bubble to the source.
The reference "(i.e., not conforming to R44-2 condition (3) in Section 6.6.2)" in section 6.3 (which is wrong anyway because rule R44-2 doesn't exist) should be replaced by "(i.e., conformig to RR4-5 condition in Section 6.6.2)"
Replaced bby Erratum 3388