tcp: fix SO_RCVLOWAT and RCVBUF autotuning

Applications might use SO_RCVLOWAT on TCP socket hoping to receive
one [E]POLLIN event only when a given amount of bytes are ready in socket
receive queue.

Problem is that receive autotuning is not aware of this constraint,
meaning sk_rcvbuf might be too small to allow all bytes to be stored.

Add a new (struct proto_ops)->set_rcvlowat method so that a protocol
can override the default setsockopt(SO_RCVLOWAT) behavior.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

6 files changed, 29 insertions, 1 deletions

diff --git a/include/linux/net.h b/include/linux/net.h
index 2248a052061d..6554d3ba4396 100644
--- a/include/linux/net.h
+++ b/include/linux/net.h
@@ -197,6 +197,7 @@ struct proto_ops {
 					   int offset, size_t size, int flags);
 	int		(*sendmsg_locked)(struct sock *sk, struct msghdr *msg,
 					  size_t size);
+	int		(*set_rcvlowat)(struct sock *sk, int val);
 };
 
 #define DECLARE_SOCKADDR(type, dst, src)	\
diff --git a/include/net/tcp.h b/include/net/tcp.h
index 9c9b3768b350..b2318242cad8 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -402,6 +402,7 @@ void tcp_set_keepalive(struct sock *sk, int val);
 void tcp_syn_ack_timeout(const struct request_sock *req);
 int tcp_recvmsg(struct sock *sk, struct msghdr *msg, size_t len, int nonblock,
 		int flags, int *addr_len);
+int tcp_set_rcvlowat(struct sock *sk, int val);
 void tcp_parse_options(const struct net *net, const struct sk_buff *skb,
 		       struct tcp_options_received *opt_rx,
 		       int estab, struct tcp_fastopen_cookie *foc);
diff --git a/net/core/sock.c b/net/core/sock.c
index 6444525f610c..b2c3db169ca1 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -905,7 +905,10 @@ set_rcvbuf:
 	case SO_RCVLOWAT:
 		if (val < 0)
 			val = INT_MAX;
-		sk->sk_rcvlowat = val ? : 1;
+		if (sock->ops->set_rcvlowat)
+			ret = sock->ops->set_rcvlowat(sk, val);
+		else
+			sk->sk_rcvlowat = val ? : 1;
 		break;
 
 	case SO_RCVTIMEO:
diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c
index eaed0367e669..f5c562aaef35 100644
--- a/net/ipv4/af_inet.c
+++ b/net/ipv4/af_inet.c
@@ -1006,6 +1006,7 @@ const struct proto_ops inet_stream_ops = {
 	.compat_getsockopt = compat_sock_common_getsockopt,
 	.compat_ioctl	   = inet_compat_ioctl,
 #endif
+	.set_rcvlowat	   = tcp_set_rcvlowat,
 };
 EXPORT_SYMBOL(inet_stream_ops);
 
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index bccc4c270087..0abd8d1d3d1d 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -1701,6 +1701,27 @@ int tcp_peek_len(struct socket *sock)
 }
 EXPORT_SYMBOL(tcp_peek_len);
 
+/* Make sure sk_rcvbuf is big enough to satisfy SO_RCVLOWAT hint */
+int tcp_set_rcvlowat(struct sock *sk, int val)
+{
+	sk->sk_rcvlowat = val ? : 1;
+	if (sk->sk_userlocks & SOCK_RCVBUF_LOCK)
+		return 0;
+
+	/* val comes from user space and might be close to INT_MAX */
+	val <<= 1;
+	if (val < 0)
+		val = INT_MAX;
+
+	val = min(val, sock_net(sk)->ipv4.sysctl_tcp_rmem[2]);
+	if (val > sk->sk_rcvbuf) {
+		sk->sk_rcvbuf = val;
+		tcp_sk(sk)->window_clamp = tcp_win_from_space(sk, val);
+	}
+	return 0;
+}
+EXPORT_SYMBOL(tcp_set_rcvlowat);
+
 static void tcp_update_recv_tstamps(struct sk_buff *skb,
 				    struct scm_timestamping *tss)
 {
diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c
index 8da0b513f188..e70d59fb26e1 100644
--- a/net/ipv6/af_inet6.c
+++ b/net/ipv6/af_inet6.c
@@ -590,6 +590,7 @@ const struct proto_ops inet6_stream_ops = {
 	.compat_setsockopt = compat_sock_common_setsockopt,
 	.compat_getsockopt = compat_sock_common_getsockopt,
 #endif
+	.set_rcvlowat	   = tcp_set_rcvlowat,
 };
 
 const struct proto_ops inet6_dgram_ops = {