Acked-by: Jarno Rajahalme <jrajaha...@nicira.com>
Jarno
On Jan 13, 2015, at 4:47 PM, Ben Pfaff <b...@nicira.com> wrote:
> Conjunction actions shouldn't normally get executed, so it's worth
> logging if it happens.
>
> Suggested-by: Jarno Rajahalme <jrajaha...@nicira.com>
> Signed-off-by: Ben Pfaff <b...@nicira.com>
> ---
> ofproto/ofproto-dpif-xlate.c | 9 ++++++++-
> 1 file changed, 8 insertions(+), 1 deletion(-)
>
> diff --git a/ofproto/ofproto-dpif-xlate.c b/ofproto/ofproto-dpif-xlate.c
> index 2efcbb9..0786513 100644
> --- a/ofproto/ofproto-dpif-xlate.c
> +++ b/ofproto/ofproto-dpif-xlate.c
> @@ -4056,8 +4056,15 @@ do_xlate_actions(const struct ofpact *ofpacts, size_t
> ofpacts_len,
> xlate_learn_action(ctx, ofpact_get_LEARN(a));
> break;
>
> - case OFPACT_CONJUNCTION:
> + case OFPACT_CONJUNCTION: {
> + /* A flow with a "conjunction" action represents part of a
> special
> + * kind of "set membership match". Such a flow should not
> actually
> + * get executed, but it could via, say, a "packet-out", even
> though
> + * that wouldn't be useful. Log it to help debugging. */
> + static struct vlog_rate_limit rl = VLOG_RATE_LIMIT_INIT(1, 1);
> + VLOG_INFO_RL(&rl, "executing no-op conjunction action");
> break;
> + }
>
> case OFPACT_EXIT:
> ctx->exit = true;
> --
> 1.7.10.4
>
_______________________________________________
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
