Re: [Fwd: [Fwd: Some precisions about CRC32 description]]

[Gorry Fairhurst <gorry@erg.abdn.ac.uk>]

So, I like this contribution, and agree that we should include some guidance
in an appendix to the next revision of the ULE-ID. I think we should await a
summary of the experience of the Salzburg Interop tests (next week) between
the two implementation that were described at the last IETF.

I've also made a note that the CRC description needs to be updated: One word
on the polynomial used for the CRC - the current ULE text is correct - but
the problem is that the specifying the polynomial does not provide
sufficient guidance on how to construct the CRC bytes.  I had thought
orginally we could refer to the use in SCTP, however since this placed the
CRC *WITHIN* the header, the procedure differs significantly. I propose we
explicitly talk about the polynomial - and THEN explicitly about the
computation. Text to be rewritten (thanks for the pointer).


Gorry

On 21/1/04 9:08 am, "Bernhard Collini-Nocker" <bnocker@cosy.sbg.ac.at>
wrote:

> Hi Gorry,
> 
> sorry to answer late, I am sick suffering from a cold.
> I am also convinced that this type of information could well fit into the
> draft, maybe only with a different example or an additional IPv4 example.
> 
> Bernhard
> 
>> -----Original Message-----
>> From: Gorry Fairhurst [mailto:gorry@erg.abdn.ac.uk]
>> Sent: Donnerstag, 15. Janner 2004 20:26
>> To: Bernhard Collini-Nocker
>> Subject: [Fwd: [Fwd: Some precisions about CRC32 description]]
>> 
>> 
>> 
>> This looks like the right sort of input to the ULE draft, but
>> does this make sense to you?
>> >> 
>> Gorry
>> 
>> -------- Original Message --------
>> Subject:     Some precisions about CRC32 description
>> Date:     Wed, 14 Jan 2004 08:21:51 +0100
>> From:     Alain RITOUX <alain.ritoux@6wind.com>
>> Reply-To:     ip-dvb@erg.abdn.ac.uk
>> To:     IPDVB <ip-dvb@erg.abdn.ac.uk>
>> 
>> 
>> 
>> Hi all,
>> 
>> The text in draft is :
>> 
>> In the CRC32 description, it should be added that the checksum to be
>> computed is the forward CRC-32 (as opposed to the reverse way, as I
>> learnt recently). So I would propose a sloght modification to the CRC
>> descritption in 4.6 :
>> 
>> "Each SNDU MUST carry a 32-bit CRC field in the last four bytes of
>> the SNDU. This position eases CRC computation by hardware.  The forward
>>                                                            ^^^^^^^^^^^^
>> CRC-32 polynomial is to be used. This is a 32 bit value calculated
>> according to the generator polynomial represented 0x04C11DB7 in
>> hexadecimal:
>> x^32+x^26+x^23+x^22+x^16+x^12+x^11+x^10+x^8+x^7+x^5+x^4+x^2+x^1+x^0. "
>> 
>> 
>> More over, there are reference to Ethernet, MPE, and AAL5,
>> "Examples where this polynomial is also employed include Ethernet,
>> DSM-CC section syntax [ISO-DSMCC} and AAL5 [ITU3563]. The use
>> resembles, but is different to that in SCTP [RFC3309]."
>> 
>> can anyone confirm, those are exactly the same computation
>> (IV=0xffffffff, forward computation, no final XOR) ? if not, the
>> reference should be removed for it might lead to errors.
>> 
>> An exemple wouldn't hurt, so here is what I computed, for a small ping6
>> from 2001:660:3008:1789::5 to 2001:660:3008:1789::6, with the associated
>> DVB MAC addr being 01:02:03:04:05:06. It gives the following SNDU :
>> 
>> 0000:  00 3f 86 dd 01 02 03 04 05 06 60 00 00 00 00 0d  .?........`.....
>> 0010:  3a 40 20 01 06 60 30 08 17 89 00 00 00 00 00 00  :@ ..`0.........
>> 0020:  00 05 20 01 06 60 30 08 17 89 00 00 00 00 00 00  .. ..`0.........
>> 0030:  00 06 80 00 9c 58 07 70 00 00 00 00 00 00 00 02  .....X.p........
>> 0040:  72 c9 c2                                         r..
>> 
>> 
>> Regards.
>> Alain.
>> --
>> Alain RITOUX
>> Tel +33-1-39-30-92-32
>> Fax +33-1-39-30-92-11
>> visit our web http://www.6wind.com
>> 
>> 
>> 
>> 
>> 
>> 
>