Re: (resend) [rfc2462bis] prefix length check for existing addresses

[Pekka Savola <pekkas@xxxxxxxxxx>]

On Tue, 8 Jun 2004, JINMEI Tatuya / [ISO-2022-JP] $B?@L@C#:H(B wrote:
> Assuming I'm correct for both the questions, I'd like to propose to
> revise the logic in Section 5.5.3 as follows:
> 
> a) Check for the Autonomous flag
> b) Ignore the link-local prefix
> c) Check for the case of preferred lifetime > valid lifetime
> d) Check if the prefix length is valid for the receiving interface.
>    If it is not valid, ignore the prefix.

How do you check that the length is valid for the *receiving 
interface*?  I guess you make the assumption, based on IPv6 over Foo 
documents, that for each interface, it's clear which kind of prefix 
lengths for SLAAC are used?

> e) If the prefix advertised does not match the prefix of an address
>    already in the list, make a new address from the prefix
> f) If the prefix advertised does match the prefix of an address
>    already in the list, update the lifetimes of the address

Note that you can avoid the wording nit above if you just checked the 
prefix and length explicitly both in e) and f) -- that is, don't 
configure a new address if it doesn't have the right prefix length 
(this already happens today), and don't update the prefix lifetimes 
unless the prefix *and* the length matches an existing prefix 
completely.

So, would a more straightforward change be a modification in f) above?
(The code might be a bit cleaner with a common rule before e or f 
though.)

-- 
Pekka Savola                 "You each name yourselves king, yet the
Netcore Oy                    kingdom bleeds."
Systems. Networks. Security. -- George R.R. Martin: A Clash of Kings



--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------