> > Similar to v4, we should ensure that at least a minimum fragmentation > length can be achieved. > > Signed-off-by: Aaron Conole <acon...@redhat.com> > --- > lib/librte_ip_frag/rte_ipv6_fragmentation.c | 11 +++++++++++ > 1 file changed, 11 insertions(+) > > diff --git a/lib/librte_ip_frag/rte_ipv6_fragmentation.c > b/lib/librte_ip_frag/rte_ipv6_fragmentation.c > index 43449970e5..820a5dc725 100644 > --- a/lib/librte_ip_frag/rte_ipv6_fragmentation.c > +++ b/lib/librte_ip_frag/rte_ipv6_fragmentation.c > @@ -79,6 +79,17 @@ rte_ipv6_fragment_packet(struct rte_mbuf *pkt_in, > uint16_t fragment_offset, frag_size; > uint64_t frag_bytes_remaining; > > + /* > + * The rules on ipv6 fragmentation means we need at least to be > + * greater than the smallest fragment possible - ipv6hdr + frag header > + * + min octets. We rely on the ipv6_extension_fragment header to > + * include 'data' octets. > + */ > + if (unlikely(mtu_size < > + (sizeof(struct rte_ipv6_hdr) + > + sizeof(struct ipv6_extension_fragment)))) > + return -EINVAL; > +
Don't know where it comes from. Reading RFC 8200, it clearly states that min MTU for ipv6 is 1280B. In fact, I don't see much point in these extra checking. > /* > * Ensure the IP payload length of all fragments (except the > * the last fragment) are a multiple of 8 bytes per RFC2460. > -- > 2.25.1
