On Tue, Apr 17, 2012 at 05:08:52PM -0700, Ethan Jackson wrote:
> > + facet = facet_lookup_valid(ofproto, flow, miss->hmap_node.hash);
>
> Pulling the hash out of hmap_node directly makes me a bit nervous.
> Does it make sense to store the hash directly in struct flow_miss with
> a comment about how it's computed? At minimum I think we should
> document that handle_flow_miss() expects 'miss' to be in a hash table
> which uses flow_hash() with a basis of zero as it's hash function.
OK, I added to handle_flow_miss():
uint32_t hash;
/* The caller must ensure that miss->hmap_node.hash contains
* flow_hash(miss->flow, 0). */
hash = miss->hmap_node.hash;
and used 'hash' in place of miss->hmap_node.hash.
Is that better?
_______________________________________________
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
