tcp: fix tcp_xmit_retransmit_queue() after rbtree introduction
authorEric Dumazet <edumazet@google.com>
Tue, 17 Oct 2017 02:38:35 +0000 (19:38 -0700)
committerDavid S. Miller <davem@davemloft.net>
Wed, 18 Oct 2017 13:19:26 +0000 (14:19 +0100)
commitb9f1f1ce866c28e3d9b86202441b220244754a69
treea4d2e4e627de150fbb6b72a754f511aaf7bd0537
parentb082af73bf527aa5180c8979d4620e450494c07b
tcp: fix tcp_xmit_retransmit_queue() after rbtree introduction

I tried to hard avoiding a call to rb_first() (via tcp_rtx_queue_head)
in tcp_xmit_retransmit_queue(). But this was probably too bold.

Quoting Yuchung :

We might miss re-arming the RTO if tp->retransmit_skb_hint is not NULL.
This can happen when RACK marks the first packet lost again and resets
tp->retransmit_skb_hint for example (tcp_rack_mark_skb_lost())

Fixes: 75c119afe14f ("tcp: implement rb-tree based retransmit queue")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/ipv4/tcp_output.c