From: Eric Dumazet Subject: [PATCH net-next 11/17] tcp: remove tcp_ecn_make_synack() socket argument Date: Fri, 25 Sep 2015 07:39:18 -0700 Message-ID: <1443191964-923-12-git-send-email-edumazet@google.com> References: <1443191964-923-1-git-send-email-edumazet@google.com> Cc: netdev , Eric Dumazet , Eric Dumazet To: "David S . Miller" Return-path: Received: from mail-pa0-f52.google.com ([209.85.220.52]:34286 "EHLO mail-pa0-f52.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932405AbbIYOkT (ORCPT ); Fri, 25 Sep 2015 10:40:19 -0400 Received: by padhy16 with SMTP id hy16so107771568pad.1 for ; Fri, 25 Sep 2015 07:40:19 -0700 (PDT) In-Reply-To: <1443191964-923-1-git-send-email-edumazet@google.com> Sender: netdev-owner@vger.kernel.org List-ID: SYNACK packets might be sent without holding socket lock. For DCTCP/ECN sake, we should call INET_ECN_xmit() while socket lock is owned, and only when we init/change congestion control. This also fixies a bug if congestion module is changed from dctcp to another one on a listener : we now clear ECN bits properly. Signed-off-by: Eric Dumazet --- net/ipv4/tcp_cong.c | 12 ++++++++++-- net/ipv4/tcp_output.c | 10 +++------- 2 files changed, 13 insertions(+), 9 deletions(-) diff --git a/net/ipv4/tcp_cong.c b/net/ipv4/tcp_cong.c index 93c4dc3ab23f..882caa4e72bc 100644 --- a/net/ipv4/tcp_cong.c +++ b/net/ipv4/tcp_cong.c @@ -173,6 +173,10 @@ out: */ if (ca->get_info) memset(icsk->icsk_ca_priv, 0, sizeof(icsk->icsk_ca_priv)); + if (ca->flags & TCP_CONG_NEEDS_ECN) + INET_ECN_xmit(sk); + else + INET_ECN_dontxmit(sk); } void tcp_init_congestion_control(struct sock *sk) @@ -181,6 +185,10 @@ void tcp_init_congestion_control(struct sock *sk) if (icsk->icsk_ca_ops->init) icsk->icsk_ca_ops->init(sk); + if (tcp_ca_needs_ecn(sk)) + INET_ECN_xmit(sk); + else + INET_ECN_dontxmit(sk); } static void tcp_reinit_congestion_control(struct sock *sk, @@ -192,8 +200,8 @@ static void tcp_reinit_congestion_control(struct sock *sk, icsk->icsk_ca_ops = ca; icsk->icsk_ca_setsockopt = 1; - if (sk->sk_state != TCP_CLOSE && icsk->icsk_ca_ops->init) - icsk->icsk_ca_ops->init(sk); + if (sk->sk_state != TCP_CLOSE) + tcp_init_congestion_control(sk); } /* Manage refcounts on socket close. */ diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index ab03018fb621..dcb5c144f8fa 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -357,14 +357,10 @@ static void tcp_ecn_clear_syn(struct sock *sk, struct sk_buff *skb) } static void -tcp_ecn_make_synack(const struct request_sock *req, struct tcphdr *th, - struct sock *sk) +tcp_ecn_make_synack(const struct request_sock *req, struct tcphdr *th) { - if (inet_rsk(req)->ecn_ok) { + if (inet_rsk(req)->ecn_ok) th->ece = 1; - if (tcp_ca_needs_ecn(sk)) - INET_ECN_xmit(sk); - } } /* Set up ECN state for a packet on a ESTABLISHED socket that is about to @@ -2999,7 +2995,7 @@ struct sk_buff *tcp_make_synack(struct sock *sk, struct dst_entry *dst, memset(th, 0, sizeof(struct tcphdr)); th->syn = 1; th->ack = 1; - tcp_ecn_make_synack(req, th, sk); + tcp_ecn_make_synack(req, th); th->source = htons(ireq->ir_num); th->dest = ireq->ir_rmt_port; /* Setting of flags are superfluous here for callers (and ECE is -- 2.6.0.rc2.230.g3dd15c0