Commit 6c09fa09 authored Mar 05, 2015 by Eric Dumazet Committed by David S. Miller Mar 05, 2015

tcp: align tcp_xmit_size_goal() on tcp_tso_autosize()



With some mss values, it is possible tcp_xmit_size_goal() puts
one segment more in TSO packet than tcp_tso_autosize().

We send then one TSO packet followed by one single MSS.

It is not a serious bug, but we can do slightly better, especially
for drivers using netif_set_gso_max_size() to lower gso_max_size.

Using same formula avoids these corner cases and makes
tcp_xmit_size_goal() a bit faster.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Fixes: 605ad7f1 ("tcp: refine TSO autosizing")
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

parent e9647d1e

net/ipv4/tcp.c

+3 −7

Original line number	Diff line number	Diff line
		@@ -835,17 +835,13 @@ static unsigned int tcp_xmit_size_goal(struct sock *sk, u32 mss_now,
		int large_allowed)
		{
		struct tcp_sock *tp = tcp_sk(sk);
		u32 new_size_goal, size_goal, hlen;
		u32 new_size_goal, size_goal;

		if (!large_allowed \|\| !sk_can_gso(sk))
		return mss_now;

		/* Maybe we should/could use sk->sk_prot->max_header here ? */
		hlen = inet_csk(sk)->icsk_af_ops->net_header_len +
		inet_csk(sk)->icsk_ext_hdr_len +
		tp->tcp_header_len;

		new_size_goal = sk->sk_gso_max_size - 1 - hlen;
		/* Note : tcp_tso_autosize() will eventually split this later */
		new_size_goal = sk->sk_gso_max_size - 1 - MAX_TCP_HEADER;
		new_size_goal = tcp_bound_to_half_wnd(tp, new_size_goal);

		/* We try hard to avoid divides here */

Admin message