FMIPv6: Mobility Header Link Layer Address Option]

Hello all,


I insist on this point, but does someone have an answer ?


Regards,
Nicolas

-------- Message original --------
Sujet: 	[Mipshop] Mobility Header Link Layer Address Option
Date: 	Thu, 08 Jun 2006 17:49:02 +0200
De: 	Nicolas DICHTEL <nicolas.dichtel6wind.com>
Répondre à: 	nicolas.dichtel6wind.com
Pour: 	mipshopietf.org



Hi,

sorry for the noise, but I have other questions concerning
the MH-LLA 
Option.

1/ About Length of the MH-LLA Option

"Length
      The size of this option in octets not including the
Type,
      Length and Option-Code fields."

In RFC 3775, Section 6.2, Length field is describe as
follow:
"Option Length

    8-bit unsigned integer, representing the length in
octets of the
    mobility option, not including the Option Type and
Option Length
    fields."

Why not following the same way for this option (i.e.,
include the size of
the Option-Code field in the Length field) ?

2/ About alignment

When the LLA size is 6 bytes, what is the required alignment
? Is it 4n+2,
8n+6 ?
And so, what's the purpose of the field Pad0 ? If the LLA
size is 12 bytes,
is this field already there ? Is it longer ?


Thanks in advance for your answer,
Nicolas


_______________________________________________
Mipshop mailing list
Mipshopietf.org
https:
//www1.ietf.org/mailman/listinfo/mipshop





_______________________________________________
Mipshop mailing list
Mipshopietf.org
https:
//www1.ietf.org/mailman/listinfo/mipshop

Nicolas DICHTEL wrote:

> 1/ About Length of the MH-LLA Option
> 
> "Length
>      The size of this option in octets not including
the Type,
>      Length and Option-Code fields."
> 
> In RFC 3775, Section 6.2, Length field is describe as
follow:
> "Option Length
> 
>    8-bit unsigned integer, representing the length in
octets of the
>    mobility option, not including the Option Type and
Option Length
>    fields."
> 
> Why not following the same way for this option (i.e.,
include the size of
> the Option-Code field in the Length field) ?

I agree.

> 2/ About alignment
> 
> When the LLA size is 6 bytes, what is the required
alignment ? Is it 4n+2,
> 8n+6 ?
> And so, what's the purpose of the field Pad0 ? If the
LLA size is 12 bytes,
> is this field already there ? Is it longer ?

the LLA option has no alignment requirement since the LLA
field is just a string. so the LLA option can start
immediately after the previous option. for padding after
the LLA option, the alignment requirement for the option
that follows the LLA option will decide how much padding
is needed.

Vijay

_______________________________________________
Mipshop mailing list
Mipshopietf.org
https:
//www1.ietf.org/mailman/listinfo/mipshop

Vijay Devarapalli a écrit :
> Nicolas DICHTEL wrote:
>> 2/ About alignment
>>
>> When the LLA size is 6 bytes, what is the required
alignment ? Is it 
>> 4n+2,
>> 8n+6 ?
>> And so, what's the purpose of the field Pad0 ? If
the LLA size is 12 
>> bytes,
>> is this field already there ? Is it longer ?
>
> the LLA option has no alignment requirement since the
LLA
> field is just a string. so the LLA option can start
> immediately after the previous option. for padding
after
> the LLA option, the alignment requirement for the
option
> that follows the LLA option will decide how much
padding
> is needed.
Hmm, I was talking about MH-LLA. Section 6.4.4 says:

                                                    
"When LLA size is
   different, the option MUST be aligned
appropriately."

And so, what about Pad0 field if LLA size is different from
6 ?


Regards,
Nicolas

_______________________________________________
Mipshop mailing list
Mipshopietf.org
https:
//www1.ietf.org/mailman/listinfo/mipshop

Nicolas DICHTEL wrote:
> Vijay Devarapalli a écrit :
>> Nicolas DICHTEL wrote:
>>> 2/ About alignment
>>>
>>> When the LLA size is 6 bytes, what is the
required alignment ? Is it 
>>> 4n+2,
>>> 8n+6 ?
>>> And so, what's the purpose of the field Pad0 ?
If the LLA size is 12 
>>> bytes,
>>> is this field already there ? Is it longer ?
>>
>> the LLA option has no alignment requirement since
the LLA
>> field is just a string. so the LLA option can start
>> immediately after the previous option. for padding
after
>> the LLA option, the alignment requirement for the
option
>> that follows the LLA option will decide how much
padding
>> is needed.
> Hmm, I was talking about MH-LLA. Section 6.4.4 says:
> 
>                                                    
"When LLA size is
>   different, the option MUST be aligned
appropriately."
> 
> And so, what about Pad0 field if LLA size is different
from 6 ?

the pad0 field needs to be removed. not sure why it is
there.

Vijay

_______________________________________________
Mipshop mailing list
Mipshopietf.org
https:
//www1.ietf.org/mailman/listinfo/mipshop

 

>> And so, what about Pad0 field if LLA size is
different from 6 ?
>
>the pad0 field needs to be removed. not sure why it is
there.

I think we added it to show how to use the option with LLA=6
bytes.
Agree we can remove it.

So, what's the consensus on LLA-Length field? Current
implementations
use it in Proxy Router messages, so they have to change. FNA
also uses
it,
so it too has to be updated. 

Perhaps better to define the MH-LLA size in octets?

-Rajeev


>
>Vijay
>
>_______________________________________________
>Mipshop mailing list
>Mipshopietf.org
>https:
//www1.ietf.org/mailman/listinfo/mipshop
>

_______________________________________________
Mipshop mailing list
Mipshopietf.org
https:
//www1.ietf.org/mailman/listinfo/mipshop

rajeev.koodlinokia.com a écrit :
>>> And so, what about Pad0 field if LLA size is
different from 6 ?
>> the pad0 field needs to be removed. not sure why it
is there.
>
> I think we added it to show how to use the option with
LLA=6 bytes.
But why padding before LLA field and not after it ?
> Agree we can remove it.
>
> So, what's the consensus on LLA-Length field? Current
implementations
> use it in Proxy Router messages, so they have to
change. FNA also uses
> it,
> so it too has to be updated. 
>
> Perhaps better to define the MH-LLA size in octets?
MH-LLA size is already in octects, Section 6.4:

            "The Length field is in units of 8 octets,
except for the
     Mobility Header Link-Layer Address option, whose Length
field
     is in units of octets in accordance with [3], Section
6.2."

And I think LLA option must follow RFC2461 Section 4.6 ...

Nicolas


_______________________________________________
Mipshop mailing list
Mipshopietf.org
https:
//www1.ietf.org/mailman/listinfo/mipshop

Hi Nicolas,
 
>> I think we added it to show how to use the option
with LLA=6 bytes.
>But why padding before LLA field and not after it ?

We should remove it. You can pad after the LLA. The size of
the LLA
is determined by the link type, as in 2461.

>> Perhaps better to define the MH-LLA size in octets?
>MH-LLA size is already in octects, Section 6.4:

Sorry about the confusion. Going back to it, now I am
confused
as to what is the issue.. Is it only the use of Pad0 in
MH-LLA?

>And I think LLA option must follow RFC2461 Section 4.6
...

It does, right? Are you referring to 

"Depending on the size of the individual LLA option,
appropriate
   padding MUST be used to ensure that the entire option
size is a
   multiple of 8 octets." ?

What do we need to do here? 

-Rajeev


>
>Nicolas
>
>

_______________________________________________
Mipshop mailing list
Mipshopietf.org
https:
//www1.ietf.org/mailman/listinfo/mipshop

Hi Rajeev,

I'm confused too. There was mutiple questions in the last
emails ...

1/ About pad0 field in MH-LLA option
-> We can remove it.

2/ About unit of length field in LLA option
-> we must follow RFC2461, the size of the LLA is
determined by the link 
type.

3/ About value of length field in MH-LLA option
-> it must follow RFC3775, so it must include size of
Option-code field.

4/ About alignment requirement of MH-LLA option
No alignment requirement ?


Hope that clarifies the issue 


Regards,
Nicolas


Le 09.08.2006 19:47, rajeev.koodlinokia.com a écrit :
> Hi Nicolas,
>  
>>> I think we added it to show how to use the
option with LLA=6 bytes.
>> But why padding before LLA field and not after it ?
>
> We should remove it. You can pad after the LLA. The
size of the LLA
> is determined by the link type, as in 2461.
>
>>> Perhaps better to define the MH-LLA size in
octets?
>> MH-LLA size is already in octects, Section 6.4:
>
> Sorry about the confusion. Going back to it, now I am
confused
> as to what is the issue.. Is it only the use of Pad0 in
MH-LLA?
>
>> And I think LLA option must follow RFC2461 Section
4.6 ...
>
> It does, right? Are you referring to 
>
> "Depending on the size of the individual LLA
option, appropriate
>    padding MUST be used to ensure that the entire
option size is a
>    multiple of 8 octets." ?
>
> What do we need to do here? 
>
> -Rajeev
>
>
>> Nicolas
>>
>>
>
>
>

_______________________________________________
Mipshop mailing list
Mipshopietf.org
https:
//www1.ietf.org/mailman/listinfo/mipshop

sounds good to me.

Vijay

Nicolas DICHTEL wrote:
> Hi Rajeev,
> 
> I'm confused too. There was mutiple questions in the
last emails ...
> 
> 1/ About pad0 field in MH-LLA option
> -> We can remove it.
> 
> 2/ About unit of length field in LLA option
> -> we must follow RFC2461, the size of the LLA is
determined by the link 
> type.
> 
> 3/ About value of length field in MH-LLA option
> -> it must follow RFC3775, so it must include size
of Option-code field.
> 
> 4/ About alignment requirement of MH-LLA option
> No alignment requirement ?
> 
> 
> Hope that clarifies the issue 
> 
> 
> Regards,
> Nicolas
> 
> 
> Le 09.08.2006 19:47, rajeev.koodlinokia.com
a écrit :
>> Hi Nicolas,
>>  
>>>> I think we added it to show how to use the
option with LLA=6 bytes.
>>> But why padding before LLA field and not after
it ?
>>
>> We should remove it. You can pad after the LLA. The
size of the LLA
>> is determined by the link type, as in 2461.
>>
>>>> Perhaps better to define the MH-LLA size in
octets?
>>> MH-LLA size is already in octects, Section 6.4:
>>
>> Sorry about the confusion. Going back to it, now I
am confused
>> as to what is the issue.. Is it only the use of
Pad0 in MH-LLA?
>>
>>> And I think LLA option must follow RFC2461
Section 4.6 ...
>>
>> It does, right? Are you referring to
>> "Depending on the size of the individual LLA
option, appropriate
>>    padding MUST be used to ensure that the entire
option size is a
>>    multiple of 8 octets." ?
>>
>> What do we need to do here?
>> -Rajeev
>>
>>
>>> Nicolas
>>>
>>>
>>
>>
>>


_______________________________________________
Mipshop mailing list
Mipshopietf.org
https:
//www1.ietf.org/mailman/listinfo/mipshop