Skip to contentSkip to conversation

Improved handling of IPv6 atomic fragments (FO=0, MF=0)

DISCUSSION Fernando Gont 2 participants 2 messages Jan 30, 2012

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

Folks,

FYI. We have published an IETF I-D that proposes an improved handling of IPv6 atomic fragments (IPv6 fragments that have an offset of 0, and MF=0).

The I-D is available here:
<http://tools.ietf.org/id/draft-gont-6man-ipv6-atomic-fragments-00.txt>

The aforementioned behaviour eliminates fragmentation-based attacks against traffic that employs atomic fragments, and has already been implemented by OpenBSD.

Thanks,

Signature · 3 lines
Fernando Gont
e-mail: [email protected] || [email protected]
PGP Fingerprint: 7809 84F5 322E 45C7 F1C9 3945 96EE A9EF D076 FFF1
#1 lore raw
Quoted 5 lines · Fernando Gont

Folks,

FYI. We have published an IETF I-D that proposes an improved handling of IPv6 atomic fragments (IPv6 fragments that have an offset of 0, and MF=0).

Quoted 8 lines

The I-D is available here:
<http://tools.ietf.org/id/draft-gont-6man-ipv6-atomic-fragments-00.txt>

The aforementioned behaviour eliminates fragmentation-based attacks against traffic that employs atomic fragments, and has already been implemented by OpenBSD.

Thanks,

Unless I missed something, linux already does that.

ipv6_frag_rcv()

        if (!(fhdr->frag_off & htons(0xFFF9))) {
                /* It is not a fragmented frame */
                skb->transport_header += sizeof(struct frag_hdr);
                IP6_INC_STATS_BH(net,
                                 ip6_dst_idev(skb_dst(skb)), IPSTATS_MIB_REASMOKS);

                IP6CB(skb)->nhoff = (u8 *)fhdr - skb_network_header(skb);
                return 1;
        }