On Mon, Feb 08, 2016 at 11:14:23AM -0800, Jarno Rajahalme wrote: > > > On Feb 5, 2016, at 7:27 PM, Ben Pfaff <b...@ovn.org> wrote: > > > > On Fri, Feb 05, 2016 at 04:41:29PM -0800, Jarno Rajahalme wrote: > >> > >>> On Feb 5, 2016, at 3:30 PM, Ben Pfaff <b...@ovn.org> wrote: > >>> > >>> Mirroring is supposed to happen at most once for any destination on a > >>> given > >>> packet, so the implementation keeps track of which mirrors have already > >>> been used. However, until this commit it did that incorrectly: it > >>> considered a mirror "used" even if it had been rejected on the basis of > >>> VLAN. This commit fixes the problem. > >> > >> So even if a mirror has been rejected on the basis of a VLAN, it > >> should still be considered for output (later)? Can you describe a > >> scenario where this makes a difference? E.g., is there a case where a > >> packet is not sent when it should have been sent, or did we get > >> duplicate mirroring due to this, as tested against by the new test > >> case? > > > > I think the best way to answer this is to rewrite the commit message. > > Here's a new version. It also adds more comments to the code. > > > > Ben, > > Thanks for taking the time to explain this in detail. I guess we must have > the documentary piece of the commit message in a document somewhere, but I > had missed that so was not aware of it in this level of detail. It makes all > sense now, > > Acked-by: Jarno Rajahalme <ja...@ovn.org>
Thanks for the reviews. I applied these patches to master and branch-2.5. _______________________________________________ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
