From: Eric Dumazet Subject: [PATCH net-next 13/14] tcp: constify tcp_syn_flood_action() socket argument Date: Tue, 29 Sep 2015 07:42:51 -0700 Message-ID: <1443537772-15305-14-git-send-email-edumazet@google.com> References: <1443537772-15305-1-git-send-email-edumazet@google.com> Cc: netdev , Eric Dumazet , Eric Dumazet To: "David S . Miller" Return-path: Received: from mail-pa0-f45.google.com ([209.85.220.45]:33724 "EHLO mail-pa0-f45.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754494AbbI2Or5 (ORCPT ); Tue, 29 Sep 2015 10:47:57 -0400 Received: by pacex6 with SMTP id ex6so8504515pac.0 for ; Tue, 29 Sep 2015 07:47:56 -0700 (PDT) In-Reply-To: <1443537772-15305-1-git-send-email-edumazet@google.com> Sender: netdev-owner@vger.kernel.org List-ID: tcp_syn_flood_action() will soon be called with unlocked socket. In order to avoid SYN flood warning being emitted multiple times, use xchg(). Extend max_qlen_log and synflood_warned fields in struct listen_sock to u32 Signed-off-by: Eric Dumazet --- include/net/request_sock.h | 5 ++--- net/ipv4/tcp_input.c | 9 +++++---- 2 files changed, 7 insertions(+), 7 deletions(-) diff --git a/include/net/request_sock.h b/include/net/request_sock.h index 90247ec7955b..c146b5284786 100644 --- a/include/net/request_sock.h +++ b/include/net/request_sock.h @@ -129,9 +129,8 @@ struct listen_sock { atomic_t qlen_dec; /* qlen = qlen_inc - qlen_dec */ atomic_t young_dec; - u8 max_qlen_log ____cacheline_aligned_in_smp; - u8 synflood_warned; - /* 2 bytes hole, try to use */ + u32 max_qlen_log ____cacheline_aligned_in_smp; + u32 synflood_warned; u32 hash_rnd; u32 nr_table_entries; struct request_sock *syn_table[0]; diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 67b27aee8d28..e58cbcd2f07e 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -6064,7 +6064,7 @@ EXPORT_SYMBOL(inet_reqsk_alloc); /* * Return true if a syncookie should be sent */ -static bool tcp_syn_flood_action(struct sock *sk, +static bool tcp_syn_flood_action(const struct sock *sk, const struct sk_buff *skb, const char *proto) { @@ -6082,11 +6082,12 @@ static bool tcp_syn_flood_action(struct sock *sk, NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPREQQFULLDROP); lopt = inet_csk(sk)->icsk_accept_queue.listen_opt; - if (!lopt->synflood_warned && sysctl_tcp_syncookies != 2) { - lopt->synflood_warned = 1; + if (!lopt->synflood_warned && + sysctl_tcp_syncookies != 2 && + xchg(&lopt->synflood_warned, 1) == 0) pr_info("%s: Possible SYN flooding on port %d. %s. Check SNMP counters.\n", proto, ntohs(tcp_hdr(skb)->dest), msg); - } + return want_cookie; } -- 2.6.0.rc2.230.g3dd15c0