On Thu, Dec 01, 2011 at 04:24:07PM -0800, Jesse Gross wrote: > We currently have a version of ipv6_skip_exthdr() which is > identical to the main one with the addition of fragment reporting. > We can propose our version for upstream and then use it directly > without duplication. > > Signed-off-by: Jesse Gross <je...@nicira.com>
The case of a IPv6 fragment header with frag_off of 0 is really odd. I guess that this code handles it acceptably (treating it the same as an unfragmented packet) but this new behavior differs, I believe, from userspace treatment of such a packet, and so I believe that the patch should also update lib/flow.c I believe that "ntohs(*fp) & ~0x7" can be written slightly more efficiently, or maybe just in a way that is more obviously optimizable, as "*fp & htons(~0x7)". This is a pretty big function to mark inline. _______________________________________________ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
