tcp: tso: remove tp->tso_deferred (5f852eb5) · Commits · 戴 / test

TSO relies on ability to defer sending a small amount of packets. Heuristic is to wait for future ACKS in hope to send more packets at once. Current algorithm uses a per socket tso_deferred field as a pseudo timer. This pseudo timer relies on future ACK, but there is no guarantee we receive them in time. Fix would be to use a real timer, but cost of such timer is probably too expensive for typical cases. This patch changes the logic to test the time of last transmit, because we should not add bursts of more than 1ms for any given flow. We've used this patch for about two years at Google, before FQ/pacing as it would reduce a fair amount of bursts. Signed-off-by:

Eric Dumazet <edumazet@google.com> Signed-off-by:

Yuchung Cheng <ycheng@google.com> Signed-off-by:

Neal Cardwell <ncardwell@google.com> Signed-off-by:

David S. Miller <davem@davemloft.net>

include/linux/tcp.h

+0 −1

Original line number	Diff line number	Diff line
		@@ -236,7 +236,6 @@ struct tcp_sock {
		u32 lost_out; /* Lost packets */
		u32 sacked_out; /* SACK'd packets */
		u32 fackets_out; /* FACK'd packets */
		u32 tso_deferred;

		/* from STCP, retrans queue hinting */
		struct sk_buff* lost_skb_hint;

net/ipv4/tcp_output.c

+5 −9

Original line number	Diff line number	Diff line
		@@ -1763,9 +1763,10 @@ static bool tcp_tso_should_defer(struct sock sk, struct sk_buff skb,
		if (icsk->icsk_ca_state != TCP_CA_Open)
		goto send_now;

		/* Defer for less than two clock ticks. */
		if (tp->tso_deferred &&
		(((u32)jiffies << 1) >> 1) - (tp->tso_deferred >> 1) > 1)
		/* Avoid bursty behavior by allowing defer
		* only if the last write was recent.
		*/
		if ((s32)(tcp_time_stamp - tp->lsndtime) > 0)
		goto send_now;

		in_flight = tcp_packets_in_flight(tp);
		@@ -1807,11 +1808,7 @@ static bool tcp_tso_should_defer(struct sock sk, struct sk_buff skb,
		goto send_now;
		}

		/* Ok, it looks like it is advisable to defer.
		* Do not rearm the timer if already set to not break TCP ACK clocking.
		*/
		if (!tp->tso_deferred)
		tp->tso_deferred = 1 \| (jiffies << 1);
		/* Ok, it looks like it is advisable to defer. */

		if (cong_win < send_win && cong_win < skb->len)
		*is_cwnd_limited = true;
		@@ -1819,7 +1816,6 @@ static bool tcp_tso_should_defer(struct sock sk, struct sk_buff skb,
		return true;

		send_now:
		tp->tso_deferred = 0;
		return false;
		}

Admin message