> On Dec 8, 2015, at 5:08 PM, Ben Pfaff <[email protected]> wrote:
>
> +struct lport {
> + struct hmap_node name_node; /* Index by name. */
> + struct hmap_node key_node; /* Index by (dp->tunnel_key, tunnel_key). */
I could go either way, but do you think it might be clearler to change these
arguments to "dp_key, port_key"?
> + const struct sbrec_port_binding *sb;
What do you think about changing this member name to "pb"? I think it may be
clearer. It also matches the use in lflow.c.
> +struct mcgroup {
> + struct hmap_node dp_name_node; /* Index by (logical datapath, name). */
> + const struct sbrec_multicast_group *sb;
Similar to above, I wonder if it would be clearer to call this "mg" or
something.
Not a biggie, but the init/destroy functions and lookup functions are defined
in opposite order between lport and mcgroup.
> +/* Multicast group index
> + * =====================
> + *
> + * This is separate from the logical port index because of namespace issues:
> + * logical port names are globally unique, but multicast group names are only
> + * unique within the scope of a logical datapath. */
> +
> +/* A multicast group.
> + *
> + * Multicast groups could be indexed by number also, but so far the clients
> do
> + * not need this index. */
This seems redundant. Is that intentional?
Acked-by: Justin Pettit <[email protected]>
--Justin
_______________________________________________
dev mailing list
[email protected]
http://openvswitch.org/mailman/listinfo/dev
