Hi All, I am going through the linux IPV6 routing code and have a query on the route cache being maintained for IPV6 packets. Why is there no route-cahe hash list for IPv6 similar to IPv4? For ipv4 routing, we have ip_mkroute_input which creates the route-cache entry and adds it to the "rt_hash_table" Hash Table. For all the subsequent packets, the route look up from the routing database being maintained in trie is avoided. The route cache from the hast list is used. But in IPv6, it appears that the route-cache entry, is being maintained in the routing tree data structure itself. This would require a route-look to happen over the routing database for every packet . What is the basic idea behind this change. What would be the disadvantages if a hash list similar to ipv4 is maintained for IPv6 route cache. Thanks. Regards, Sai Naidu _______________________________________________ Kernelnewbies mailing list Kernelnewbies@xxxxxxxxxxxxxxxxx http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
