Content-Type: multipart/mixed; boundary="===============0421657642345281121==" MIME-Version: 1.0 From: Christoph Paasch To: mptcp at lists.01.org Subject: [MPTCP] [PATCH 4/5] mptcp: Always call sk_data_ready Date: Thu, 29 Nov 2018 17:05:13 -0800 Message-ID: <20181130010514.18251-5-cpaasch@apple.com> In-Reply-To: 20181130010514.18251-1-cpaasch@apple.com X-Status: X-Keywords: X-UID: 849 --===============0421657642345281121== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Since commit 03f45c883c6f ("tcp: avoid extra wakeups for SO_RCVLOWAT users"= ), tcp_data_queue checks for available bytes in the subflow's receive-queue before waking up the "app". However, with MPTCP we might have zero-length data-fins in the queue. Thus, we still need to wakeup the MPTCP-layer in case of a subflow. Fixes: dd64d1c40b48 ("Merge tag 'v4.18' into mptcp_trunk") Signed-off-by: Christoph Paasch Signed-off-by: Matthieu Baerts (cherry picked from commit 8562bd5e62968e4bbdbf897decf933e90210a139) Signed-off-by: Christoph Paasch --- net/ipv4/tcp_input.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index c0efc3de8683..0f1ac709f15d 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -4741,7 +4741,7 @@ void tcp_data_ready(struct sock *sk) const struct tcp_sock *tp =3D tcp_sk(sk); int avail =3D tp->rcv_nxt - tp->copied_seq; = - if (avail < sk->sk_rcvlowat && !sock_flag(sk, SOCK_DONE)) + if (avail < sk->sk_rcvlowat && !sock_flag(sk, SOCK_DONE) && !mptcp(tp)) return; = sk->sk_data_ready(sk); -- = 2.16.2 --===============0421657642345281121==--