tcp: add tcp_sock_set_nodelay

Add a helper to directly set the TCP_NODELAY sockopt from kernel space
without going through a fake uaccess.  Cleanup the callers to avoid
pointless wrappers now that this is a simple function call.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Sagi Grimberg <sagi@grimberg.me>
Acked-by: Jason Gunthorpe <jgg@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

5 files changed, 29 insertions, 35 deletions

diff --git a/net/ceph/messenger.c b/net/ceph/messenger.c
index f8ca5edc5f2c..27d6ab11f9ee 100644
--- a/net/ceph/messenger.c
+++ b/net/ceph/messenger.c
@@ -490,15 +490,8 @@ static int ceph_tcp_connect(struct ceph_connection *con)
 		return ret;
 	}
 
-	if (ceph_test_opt(from_msgr(con->msgr), TCP_NODELAY)) {
-		int optval = 1;
-
-		ret = kernel_setsockopt(sock, SOL_TCP, TCP_NODELAY,
-					(char *)&optval, sizeof(optval));
-		if (ret)
-			pr_err("kernel_setsockopt(TCP_NODELAY) failed: %d",
-			       ret);
-	}
+	if (ceph_test_opt(from_msgr(con->msgr), TCP_NODELAY))
+		tcp_sock_set_nodelay(sock->sk);
 
 	con->sock = sock;
 	return 0;
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index e6cf702e16d6..a65f293a19fa 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -2832,6 +2832,30 @@ void tcp_sock_set_cork(struct sock *sk, bool on)
 }
 EXPORT_SYMBOL(tcp_sock_set_cork);
 
+/* TCP_NODELAY is weaker than TCP_CORK, so that this option on corked socket is
+ * remembered, but it is not activated until cork is cleared.
+ *
+ * However, when TCP_NODELAY is set we make an explicit push, which overrides
+ * even TCP_CORK for currently queued segments.
+ */
+static void __tcp_sock_set_nodelay(struct sock *sk, bool on)
+{
+	if (on) {
+		tcp_sk(sk)->nonagle |= TCP_NAGLE_OFF|TCP_NAGLE_PUSH;
+		tcp_push_pending_frames(sk);
+	} else {
+		tcp_sk(sk)->nonagle &= ~TCP_NAGLE_OFF;
+	}
+}
+
+void tcp_sock_set_nodelay(struct sock *sk)
+{
+	lock_sock(sk);
+	__tcp_sock_set_nodelay(sk, true);
+	release_sock(sk);
+}
+EXPORT_SYMBOL(tcp_sock_set_nodelay);
+
 /*
  *	Socket option code for TCP.
  */
@@ -2929,20 +2953,7 @@ static int do_tcp_setsockopt(struct sock *sk, int level,
 		break;
 
 	case TCP_NODELAY:
-		if (val) {
-			/* TCP_NODELAY is weaker than TCP_CORK, so that
-			 * this option on corked socket is remembered, but
-			 * it is not activated until cork is cleared.
-			 *
-			 * However, when TCP_NODELAY is set we make
-			 * an explicit push, which overrides even TCP_CORK
-			 * for currently queued segments.
-			 */
-			tp->nonagle |= TCP_NAGLE_OFF|TCP_NAGLE_PUSH;
-			tcp_push_pending_frames(sk);
-		} else {
-			tp->nonagle &= ~TCP_NAGLE_OFF;
-		}
+		__tcp_sock_set_nodelay(sk, val);
 		break;
 
 	case TCP_THIN_LINEAR_TIMEOUTS:
diff --git a/net/rds/tcp.c b/net/rds/tcp.c
index 46782fac4c16..43db0eca911f 100644
--- a/net/rds/tcp.c
+++ b/net/rds/tcp.c
@@ -89,15 +89,6 @@ static struct ctl_table rds_tcp_sysctl_table[] = {
 	{ }
 };
 
-/* doing it this way avoids calling tcp_sk() */
-void rds_tcp_nonagle(struct socket *sock)
-{
-	int val = 1;
-
-	kernel_setsockopt(sock, SOL_TCP, TCP_NODELAY, (void *)&val,
-			      sizeof(val));
-}
-
 u32 rds_tcp_write_seq(struct rds_tcp_connection *tc)
 {
 	/* seq# of the last byte of data in tcp send buffer */
@@ -502,7 +493,7 @@ void rds_tcp_tune(struct socket *sock)
 	struct net *net = sock_net(sk);
 	struct rds_tcp_net *rtn = net_generic(net, rds_tcp_netid);
 
-	rds_tcp_nonagle(sock);
+	tcp_sock_set_nodelay(sock->sk);
 	lock_sock(sk);
 	if (rtn->sndbuf_size > 0) {
 		sk->sk_sndbuf = rtn->sndbuf_size;
diff --git a/net/rds/tcp.h b/net/rds/tcp.h
index d640e210b97b..f6d75d8cb167 100644
--- a/net/rds/tcp.h
+++ b/net/rds/tcp.h
@@ -50,7 +50,6 @@ struct rds_tcp_statistics {
 
 /* tcp.c */
 void rds_tcp_tune(struct socket *sock);
-void rds_tcp_nonagle(struct socket *sock);
 void rds_tcp_set_callbacks(struct socket *sock, struct rds_conn_path *cp);
 void rds_tcp_reset_callbacks(struct socket *sock, struct rds_conn_path *cp);
 void rds_tcp_restore_callbacks(struct socket *sock,
diff --git a/net/rds/tcp_listen.c b/net/rds/tcp_listen.c
index d8bd13276959..6f90ea077adc 100644
--- a/net/rds/tcp_listen.c
+++ b/net/rds/tcp_listen.c
@@ -288,7 +288,7 @@ struct socket *rds_tcp_listen_init(struct net *net, bool isv6)
 	}
 
 	sock->sk->sk_reuse = SK_CAN_REUSE;
-	rds_tcp_nonagle(sock);
+	tcp_sock_set_nodelay(sock->sk);
 
 	write_lock_bh(&sock->sk->sk_callback_lock);
 	sock->sk->sk_user_data = sock->sk->sk_data_ready;