Michel Machado

This issue is going to be affected by issue #249, which will add one more field to the statistics.

`rte_strdup()` in `main/main.c` is a small, useful function that should go upstream as well. It's being used in a number of places in Gatekeeper.

Note that our `rte_strdup()` will return `NULL` when the source string is `NULL`, so we must check that is the expected behavior of `strdup(3)` before pushing the patch upstream. If...

One should search the code of Gatekeeper to identify all small functions that should be pushed upstream. For example: `rte_priv_to_mbuf()` and `rte_mbuf_prefetch_part2_non_temporal()`.

Currently, Gatekeeper's DPDK adds `rte_hash_lookup_bulk_with_hash()`, `rte_hash_prefetch_buckets()`, and `rte_hash_prefetch_buckets_non_temporal()`. If we implement a flow table independently of DPDK, we can avoid having a custom version of DPDK.

While we review the fields `data_len` and `pkt_len` of `struct rte_mbuf`, we should also adopt the macros `rte_pktmbuf_pkt_len()` and `rte_pktmbuf_data_len()` to use instead of the fields.

The first patch of pull request #538 largely increases the memory allocation capacity of Gatekeeper's version of DPDK. Thus, one will no longer need to work around this issue. Nevertheless,...

Pull request #538 addressed this issue for our patched version of DPDK.

A long-term solution would be for DPDK to have a [meson option](http://doc.dpdk.org/guides/prog_guide/build-sdk-meson.html) for the maximum amount of hugepages memory that it can manage like it has options for the maximum...

Do the missed IPv4 addresses belong to protected prefixes (i.e. Grantor FIB entries on Gatekeeper servers)? If not, could you explain what the missed addresses are in your environment?