IPFW flaws with IPv6 fragments

[Matthew Luckie]

Hi

I'm just wondering if I can interest anyone in an IPFW PR with a tested 
patch, which I submitted a few weeks ago.

http://www.freebsd.org/cgi/query-pr.cgi?pr=145733

The flaws that the patch fixes:

  - Rejection of packets with an IPv6 Fragmentation header if the packet
    is not actually fragmented (offset and mf both zero).  This type of
    packet is allowed by RFC 2460.

  - Rejection of fragments with offset != 0 if they are small (because
    the code tries to pullup a transport layer header which isn't there)

  - No check of the transport layer fields with for the first fragment
    offset zero because the mf bit is masked into the offset field.

I'm happy to address any concerns with the patch if there are any.

Thanks,

Matthew