dpif-linux: Use poll() internally in dpif_linux_recv().

Using poll() internally in dpif_linux_recv(), instead of relying
on the results of the main loop poll() call, brings netperf CRR
performance back within 1% of par versus the code base before the
poll_fd_woke() optimizations were introduced.  It also increases
the ovs-benchmark results by about 5% versus that baseline, too.

My theory is that this is because the main loop takes long enough
that a significant number of packets can arrive during the main
loop itself, so this reduces the time before OVS gets to those
packets.

1 files changed, 1 insertions, 0 deletions

diff --git a/lib/netlink-socket.h b/lib/netlink-socket.h
index b6356b9e..f2a5d3fa 100644
--- a/lib/netlink-socket.h
+++ b/lib/netlink-socket.h
@@ -60,6 +60,7 @@ int nl_sock_transact(struct nl_sock *, const struct ofpbuf *request,
 int nl_sock_drain(struct nl_sock *);
 
 void nl_sock_wait(const struct nl_sock *, short int events);
+int nl_sock_fd(const struct nl_sock *);
 
 uint32_t nl_sock_pid(const struct nl_sock *);