From 4e4f1fc226816905c937f9b29dabe351075dfe0f Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 6 Sep 2013 10:35:58 -0700 Subject: tcp: properly increase rcv_ssthresh for ofo packets TCP receive window handling is multi staged. A socket has a memory budget, static or dynamic, in sk_rcvbuf. Because we do not really know how this memory budget translates to a TCP window (payload), TCP announces a small initial window (about 20 MSS). When a packet is received, we increase TCP rcv_win depending on the payload/truesize ratio of this packet. Good citizen packets give a hint that it's reasonable to have rcv_win = sk_rcvbuf/2 This heuristic takes place in tcp_grow_window() Problem is : We currently call tcp_grow_window() only for in-order packets. This means that reorders or packet losses stop proper grow of rcv_win, and senders are unable to benefit from fast recovery, or proper reordering level detection. Really, a packet being stored in OFO queue is not a bad citizen. It should be part of the game as in-order packets. In our traces, we very often see sender is limited by linux small receive windows, even if linux hosts use autotuning (DRS) and should allow rcv_win to grow to ~3MB. Signed-off-by: Eric Dumazet Acked-by: Neal Cardwell Signed-off-by: David S. Miller --- net/ipv4/tcp_input.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'net/ipv4/tcp_input.c') diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 894bc174f472..25a89eaa669d 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -4139,6 +4139,7 @@ static void tcp_data_queue_ofo(struct sock *sk, struct sk_buff *skb) if (!tcp_try_coalesce(sk, skb1, skb, &fragstolen)) { __skb_queue_after(&tp->out_of_order_queue, skb1, skb); } else { + tcp_grow_window(sk, skb); kfree_skb_partial(skb, fragstolen); skb = NULL; } @@ -4214,8 +4215,10 @@ add_sack: if (tcp_is_sack(tp)) tcp_sack_new_ofo_skb(sk, seq, end_seq); end: - if (skb) + if (skb) { + tcp_grow_window(sk, skb); skb_set_owner_r(skb, sk); + } } static int __must_check tcp_queue_rcv(struct sock *sk, struct sk_buff *skb, int hdrlen, -- cgit v1.2.3-55-g7522