Hi, Akisada-san.
I can see from the test specification that I was right about the test
purpose. And so my problem still stays the same. Although RFC 4861
expects that the prefix will be on-link, RFC 4862 permits invalidating
this prefix because of invalid prefix length.
Please correct me if I'm wrong.
Thank you for your time.
Best regards,
Marvell Software Solutions Israel Ltd.
Atidim Technological Park Bldg #4
Tel Aviv, 61581, Israel
Email: vkuk@marvell.com
Office: +972-3-645-8955
Web site: http://www.marvell.com
-----Original Message-----
From: Yukiyo Akisada [mailto:akisada@tahi.org]
Sent: Wednesday, June 18, 2008 3:33 AM
To: Vladimir Kuk
Cc: users@tahi.org
Subject: Re: [users:00777] Self_Test v4.0.1 NDP test v6LC.2.2.19
Hi, Vladimir.
The official test specification can be downloaded from
<http://www.ipv6ready.org/about_phase2_test.html>.
<http://www.ipv6ready.org/pdf/IPv6_Ready_Test_Specification_Core_Protoco
lsv4_0_0.pdf> is the newest.
The documents in the test script have differences from it
because of my loose.
Please refer to the official specification in this moment.
If you still have questions after checking the official document,
please let me know again.
Thanks,
On Tue, 17 Jun 2008 12:00:04 +0300
"Vladimir Kuk" <vkuk@marvell.com> wrote:
> Hello.
>
> I'm running Self Tests for phase 2 host part and ran into some problem
> with NDP test 2.2.19.
>
> In it a little problematic to understand the test since the
description
> in it, is for another test (2.2.12.A).
>
> As far as I can it tests for discarding the invalid prefix for
purposes
> of autoconfiguration but expects this prefix to be accepted for onlink
> determination.
>
> If I understood the test correctly, then there is a problem since RFC
> 4892 states:
>
> 5.5.3. Router Advertisement Processing p.19:
>
> If the sum of the prefix length and interface identifier length
> does not equal 128 bits, the Prefix Information option MUST be
> ignored. An implementation MAY wish to log a system management
> error in this case. The length of the interface identifier is
> defined in a separate link-type specific document, which should
> also be consistent with the address architecture [RFC4291] (see
> Section 2).
>
> Also, in same section:
> It should be noted, however, that this does not mean
> the advertised prefix length is meaningless. In fact, the
> advertised length has non-trivial meaning for on-link
> determination in [RFC4861] where the sum of the prefix length and
> the interface identifier length may not be equal to 128. Thus, it
> should be safe to validate the advertised prefix length here, in
> order to detect and avoid a configuration error specifying an
> invalid prefix length in the context of address autoconfiguration.
>
>
> Can anyone shed a light on the matter?
>
>
> Thank you.
>
> Best regards,
>
> Vladimir Kuk
>
> Marvell Software Solutions Israel Ltd.
> Atidim Technological Park Bldg #4
> Tel Aviv, 61581, Israel
> Email: vkuk@marvell.com
> Office: +972-3-645-8955
> Web site: http://www.marvell.com
>
>
>
>
>
>
--
Yukiyo Akisada <akisada@tahi.org>