On Thu, Jul 19, 2012 at 02:20:08PM -0400, Ed Maste wrote: > > It's a little sad that the pcap architecture means that we have to do > > an extra copy of the packet. > > Yes. With more significant changes to the overall netdev datapath we > could eventually eliminate these copies, but for now we're hoping for > a minimally invasive change set.
I'm happy with that. > > My docs for libpcap mention a "pcap_next_ex" function that can be used > > to distinguish errors from no available packets. That might be easier > > to use than pcap_dispatch(). > > Yes, this looks like a sensible change. Any objection to having that > come in as a subsequent commit? OK. > Thanks again for the review and feedback! You're welcome. I look forward to the next version. _______________________________________________ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
