On 25 May 2017 at 18:53, Darrell Ball <[email protected]> wrote: > > > On 5/22/17, 4:44 PM, "[email protected] on behalf of Joe > Stringer" <[email protected] on behalf of [email protected]> wrote: > > On 20 May 2017 at 11:09, Darrell Ball <[email protected]> wrote: > > The checks to populate ct_orig_tuple in miniflow_extract > > includes recirc_id being non-zero. This is changed here > > to populate the ct_orig_tuple fields based only on ct_state > > per the logic of the design. > > > > Signed-off-by: Darrell Ball <[email protected]> > > I'm not exactly sure what "per the logic of the design" is supposed to > mean, and I suspect that someone reading through the git logs would be > just as lost. As far as I'm aware, any patch to OVS should be "per the > logic of the design", right? > > In general, a change can be consistent with an > existing design or a change to the existing design itself. > However, “per the logic of the design” can be assumed by default > and hence is usually considered superfluous, which I think is what you > want to convey – sure. > > I’ll include the text I used on the related thread, as I agree, the > motivation is not necessarily obvious, just by linking ct_orig_tuple to > ct_state, as I did in this commit message.
OK, thanks! _______________________________________________ dev mailing list [email protected] https://mail.openvswitch.org/mailman/listinfo/ovs-dev
