tcp: fix tcp_cong_avoid_ai() credit accumulation bug with decreases in w
authorNeal Cardwell <ncardwell@google.com>
Tue, 10 Mar 2015 21:17:03 +0000 (17:17 -0400)
committerDavid S. Miller <davem@davemloft.net>
Wed, 11 Mar 2015 20:51:51 +0000 (16:51 -0400)
commit9949afa42be0b76f5832db112ce51bb6b35b2abb
treedded55a25200f202f9904369f39ccad9769da00d
parent366c1bd191c49380d81b15b96cf7d4e3528a82a2
tcp: fix tcp_cong_avoid_ai() credit accumulation bug with decreases in w

The recent change to tcp_cong_avoid_ai() to handle stretch ACKs
introduced a bug where snd_cwnd_cnt could accumulate a very large
value while w was large, and then if w was reduced snd_cwnd could be
incremented by a large delta, leading to a large burst and high packet
loss. This was tickled when CUBIC's bictcp_update() sets "ca->cnt =
100 * cwnd".

This bug crept in while preparing the upstream version of
814d488c6126.

Testing: This patch has been tested in datacenter netperf transfers
and live youtube.com and google.com servers.

Fixes: 814d488c6126 ("tcp: fix the timid additive increase on stretch ACKs")
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/ipv4/tcp_cong.c