Re: RFC4862 and 64-bit IID (Re: draft-bourbaki-6man-classless-ipv6-00)

[Brian E Carpenter <[e-mail]>]

On 20/06/2017 05:25, JINMEI Tatuya / 神明達哉 wrote:
> At Mon, 19 Jun 2017 18:45:12 +0200,
> Job Snijders <[e-mail]> wrote:
> 
>>>> I'd also note once again that this discussion has nothing to do with
>>>> the fact that we can perform 'ifconfig en0 2001:db8::1/120'.  This
>>>> operation configures a 128-bit IPv6 address with 120-bit on-link
>>>> prefix.  On-link prefixes have always been variable, and they have
>>>> nothing to do with IID length or SLAAC.  We don't have to update
>>>> addr-arch or RFC4862 because of this.  (draft-bourbaki-6man-
>>>> classless-ipv6-00 seems to be confused on this point, and I suspect
>>>> it increases the confusion and controversy in this whole thread).
>>>
>>> Agreed.
>>
>> Is there a document that states that IPv6 routers and hosts must support
>> on-link prefixes of all sizes? 
> 
> In my own draft draft-jinmei-6man-prefix-clarify-00 I tried to explain
> that's at least the case for configuring on-link prefixes on a host
> advertised via Router Advertisements.  RFC4861, or any other RFCs that
> I know of, doesn't have an exact phrase like "a host MUST support
> on-link prefixes of all lengths", but I believe I provided sufficient
> evidences that that's the reasonable interpretation of the current
> specification and it matches what's currently implemented.
> 
> For manual configuration on hosts and routers, I don't think there is
> an RFC that has this exact phrase or equally strong evidence as the
> host autoconfiguration case.  But at least for host implementations I
> believe it's quite natural to extend that interpretation to manual
> configuration.  And same for routers - after all, it's a router to
> advertise an on-link prefix of arbitrary lengths via RA PIO.  So, it
> shouldn't be unreasonable to assume it should be able to use that
> prefix for its own on-link determination.
> 
> And, perhaps more important in the context of
> draft-bourbaki-6man-classless-ipv6-00, (as far as I know) no one,
> even strong 64bit-IID advocates, disagrees on this interpretation.

I would say it is a completely separate issue that the draft simply
doesn't care about.

    Brian