[Ipsec] draft-ietf-ipsec-esp-v3-10.txt

Hi Steve,

(CC'ing Pekka for IPv6 inputs) 
Sorry for getting this late. I noticed in section 3.3.4 it is stated
that: -

   NOTE: For IPv6 -- For bump-in-the-stack and bump-in-the-wire
   implementations, it will be necessary to examine all the extension
   headers to determine if there is a fragmentation header and hence
   that the packet needs reassembling prior to IPsec processing.

However I do not think having a fragment header itself means the packet
is a fragment. Only if the offset field or the more fragment field is
set to non-zero, only then is a packet a fragment. 

The fragment header is also used to signal to an IPv4-to-IPv6 translator
that the packet can be fragmented, DF bit setting in the outer IPv4
header (SIIT - RFC2765). I think the same text is there for AH draft
draft-ietf-ipsec-rfc2402bis-11.txt too.

We could change the text to 

   NOTE: For IPv6 -- For bump-in-the-stack and bump-in-the-wire
   implementations, it will be necessary to examine all the extension
   headers to determine if there is a fragmentation header, and either
   the More flag or the Fragment Offset is non-zero. If so that packet
   needs reassembling prior to IPsec processing.

We can bounce the text off the IPv6 list too if we desire.

Thanks,
Vishwas

_______________________________________________
Ipsec mailing list
Ipsec@ietf.org
https://www1.ietf.org/mailman/listinfo/ipsec