[dev] babeld: minimal updates for compliance with RFC 6126bis

Donald Sharp sharpd at cumulusnetworks.com
Wed Jun 19 10:35:46 EDT 2019


Laure -

This looks like the meat of the change:

commit 7426386b14d7800be769c419c119c9532d9e3402 (HEAD -> babel)
Author: Laure Merat <laure.merat at etu.univ-paris-diderot.fr>
Date:   Thu Jun 13 17:11:04 2019 +0200

    babeld: minimal updates for compliance with RFC 6126bis.

    The older Babel specification (RFC 6126) is in the process of
being supressed by RFC 6126. This implements the minimum amount of
changes to make FRR comply with 6126bis, but doesn't actually use any
of the new features.

    More precisely:

      - we ignore any Hello TLV that carries the Unicast bit in its
"Flags" field (formerly "Reserved"); it would be incorrect to treat
such a Hello as an ordinary Hello;

      - we don't reset any timers when we receive a Hello with the
Interval field set to 0;

      - we ignore any TLV that carries an unknown subTLV with the
Mandatory bit set (bit 0x80 of the Type field).


I would suggest signing off your commit as well as fixing the
tab/spacing issues that I see that are inconsistent with the rest of
babel.  Other than that from my perspective if your tests show proper
interopability with other implementations of babel, I'm not too
concerned about the contents because my assumption is that the code
mirrors what is in the non-FRR babel code.

Thanks!

donald

On Wed, Jun 19, 2019 at 10:10 AM laure merat <merat.laure at gmail.com> wrote:
>
> Hi,
>
> I'm a master's student doing an internship with Juliusz Chroboczek.
>
> I've updated Babel to the last specification (you can find my work at https://github.com/lmerat46/frr).
>
> I would be very grateful if you could review it and tell me if you have any comments before I do the merge request (should be next week, Juliusz still has to check some things)!
>
> Best regards,
>
> Laure Mérat
>
> _______________________________________________
> dev mailing list
> dev at lists.frrouting.org
> https://lists.frrouting.org/listinfo/dev



More information about the dev mailing list