Agree with Carsten here.  The referenced RFC 2464 complements the U/L bit when 
an EUI-64 is converted into an interface identifier. So the meaning of the U/L 
bit is also reversed in the interface identifier so it differs from the values 
as written in the IEEE "Guidelines for EUI-64" document.

Just noting that the naming in RFC 4944 may confuse some readers: both bits are 
now named "U/L bit".  The complemented bit could have been named "L/U bit" 
instead; or "the bit in the U/L bit position". Also the RFC 2464 reference URL 
to the "Guidelines for EUI-64" document is not valid anymore. There's a 2017 
version located at: 
https://standards.ieee.org/wp-content/uploads/import/documents/tutorials/eui.pdf
 

Esko

-----Original Message-----
From: 6lo <[email protected]> On Behalf Of Carsten Bormann
Sent: Tuesday, April 23, 2024 21:22
To: Eric Vyncke (evyncke) <[email protected]>
Cc: [email protected]
Subject: Re: [6lo] Processing old erratum 6194 on RFC 4944, "Transmission of 
IPv6 Packets over IEEE 802.15.4 Networks"

RFC 4944 does what (AFAIK) all 802-derived IP-over-foos do when creating IP 
addresses from MAC addresses/EUIs:

   The Interface Identifier is then formed from the EUI-64 by
   complementing the "Universal/Local" (U/L) bit, 

(Section 4 of RFC 2464).

The text cited from IEEE talks about EUIs.
EUIs need to have the bit complemented to build the Interface Identifier.
So the U/L bit in the IID is inverted from that sense, i.e., 0 for local and 1 
for universal.
This semantics also applies to IIDs built from 16-bit addresses, which are 
never globally unique, and therefore always have the bit set to 0 (local).

The text in RFC 4944 is as intended.

Grüße, Carsten


> On 23. Apr 2024, at 15:40, Eric Vyncke (evyncke) 
> <[email protected]> wrote:
> 
> As the new shepherding AD for 6LO, I just found a reported (i.e., neither 
> verified nor closed) erratum on RFC 4944, "Transmission of IPv6 Packets over 
> IEEE 802.15.4 Networks". 
> https://www.rfc-editor.org/errata_search.php?eid=6194This erratum was never 
> posted to the 6LO list probably due to a technical issue.
>  What is the WG view on this erratum ?
>  Thank you in advance for the discussion.
>  Regards
>  -éric
>  The reported errata is:
>  Reported By: Tommaso Pecorella
> Date Reported: 2020-05-30
> Section 6 says:
> However, in the
>    resultant interface identifier, the "Universal/Local" (U/L) bit SHALL
>    be set to zero in keeping with the fact that this is not a globally
>    unique value.
>  It should say:
> However, in the
>    resultant interface identifier, the "Universal/Local" (U/L) bit SHALL
>    be set to one in keeping with the fact that this is not a globally
>    unique value.
>  Notes:
> IEEE (see 
> https://standards.ieee.org/content/dam/ieee-standards/standards/web/documents/tutorials/eui.pdf)
>  states that:
> "the second least significant bit of Octet 0 (the U/L bit) indicates 
> universal (U/L=0) or local (U/L=1) administration of the address."
> Thus, the U/L bit in the "pseudo 48-bit address" shall have its U/L bit set 
> to one, not to zero.
>  _______________________________________________
> 6lo mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/6lo


_______________________________________________
6lo mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6lo
_______________________________________________
6lo mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6lo

Reply via email to