[net 0/3] tipc: add some patches

[net 0/3] tipc: add some patches

[Tuong Lien]

This series adds patches to fix some issues in TIPC streaming & service
subscription.

Tuong Lien (3):
  tipc: fix large latency in smart Nagle streaming
  tipc: fix memory leak in service subscripting
  tipc: fix failed service subscription deletion

 net/tipc/socket.c | 42 +++++++++++++++++++++++++++++++-----------
 net/tipc/subscr.h | 10 ++++++++++
 net/tipc/topsrv.c | 13 ++++++++-----
 3 files changed, 49 insertions(+), 16 deletions(-)

-- 
2.13.7

[net 1/3] tipc: fix large latency in smart Nagle streaming

[Tuong Lien]

Currently when a connection is in Nagle mode, we set the 'ack_required'
bit in the last sending buffer and wait for the corresponding ACK prior
to pushing more data. However, on the receiving side, the ACK is issued
only when application really  reads the whole data. Even if part of the
last buffer is received, we will not do the ACK as required. This might
cause an unnecessary delay since the receiver does not always fetch the
message as fast as the sender, resulting in a large latency in the user
message sending, which is: [one RTT + the receiver processing time].

The commit makes Nagle ACK as soon as possible i.e. when a message with
the 'ack_required' arrives in the receiving side's stack even before it
is processed or put in the socket receive queue...
This way, we can limit the streaming latency to one RTT as committed in
Nagle mode.

Acked-by: Ying Xue <ying.xue@windriver.com>
Acked-by: Jon Maloy <jmaloy@redhat.com>
Signed-off-by: Tuong Lien <tuong.t.lien@dektech.com.au>
---
 net/tipc/socket.c | 42 +++++++++++++++++++++++++++++++-----------
 1 file changed, 31 insertions(+), 11 deletions(-)

diff --git a/net/tipc/socket.c b/net/tipc/socket.c
index 87466607097f..e370ad0edd76 100644
--- a/net/tipc/socket.c
+++ b/net/tipc/socket.c
@@ -1739,22 +1739,21 @@ static int tipc_sk_anc_data_recv(struct msghdr *m, struct sk_buff *skb,
 	return 0;
 }
 
-static void tipc_sk_send_ack(struct tipc_sock *tsk)
+static struct sk_buff *tipc_sk_build_ack(struct tipc_sock *tsk)
 {
 	struct sock *sk = &tsk->sk;
-	struct net *net = sock_net(sk);
 	struct sk_buff *skb = NULL;
 	struct tipc_msg *msg;
 	u32 peer_port = tsk_peer_port(tsk);
 	u32 dnode = tsk_peer_node(tsk);
 
 	if (!tipc_sk_connected(sk))
-		return;
+		return NULL;
 	skb = tipc_msg_create(CONN_MANAGER, CONN_ACK, INT_H_SIZE, 0,
 			      dnode, tsk_own_node(tsk), peer_port,
 			      tsk->portid, TIPC_OK);
 	if (!skb)
-		return;
+		return NULL;
 	msg = buf_msg(skb);
 	msg_set_conn_ack(msg, tsk->rcv_unacked);
 	tsk->rcv_unacked = 0;
@@ -1764,7 +1763,19 @@ static void tipc_sk_send_ack(struct tipc_sock *tsk)
 		tsk->rcv_win = tsk_adv_blocks(tsk->sk.sk_rcvbuf);
 		msg_set_adv_win(msg, tsk->rcv_win);
 	}
-	tipc_node_xmit_skb(net, skb, dnode, msg_link_selector(msg));
+	return skb;
+}
+
+static void tipc_sk_send_ack(struct tipc_sock *tsk)
+{
+	struct sk_buff *skb;
+
+	skb = tipc_sk_build_ack(tsk);
+	if (!skb)
+		return;
+
+	tipc_node_xmit_skb(sock_net(&tsk->sk), skb, tsk_peer_node(tsk),
+			   msg_link_selector(buf_msg(skb)));
 }
 
 static int tipc_wait_for_rcvmsg(struct socket *sock, long *timeop)
@@ -1938,7 +1949,6 @@ static int tipc_recvstream(struct socket *sock, struct msghdr *m,
 	bool peek = flags & MSG_PEEK;
 	int offset, required, copy, copied = 0;
 	int hlen, dlen, err, rc;
-	bool ack = false;
 	long timeout;
 
 	/* Catch invalid receive attempts */
@@ -1983,7 +1993,6 @@ static int tipc_recvstream(struct socket *sock, struct msghdr *m,
 
 		/* Copy data if msg ok, otherwise return error/partial data */
 		if (likely(!err)) {
-			ack = msg_ack_required(hdr);
 			offset = skb_cb->bytes_read;
 			copy = min_t(int, dlen - offset, buflen - copied);
 			rc = skb_copy_datagram_msg(skb, hlen + offset, m, copy);
@@ -2011,7 +2020,7 @@ static int tipc_recvstream(struct socket *sock, struct msghdr *m,
 
 		/* Send connection flow control advertisement when applicable */
 		tsk->rcv_unacked += tsk_inc(tsk, hlen + dlen);
-		if (ack || tsk->rcv_unacked >= tsk->rcv_win / TIPC_ACK_RATE)
+		if (tsk->rcv_unacked >= tsk->rcv_win / TIPC_ACK_RATE)
 			tipc_sk_send_ack(tsk);
 
 		/* Exit if all requested data or FIN/error received */
@@ -2105,9 +2114,11 @@ static void tipc_sk_proto_rcv(struct sock *sk,
  * tipc_sk_filter_connect - check incoming message for a connection-based socket
  * @tsk: TIPC socket
  * @skb: pointer to message buffer.
+ * @xmitq: for Nagle ACK if any
  * Returns true if message should be added to receive queue, false otherwise
  */
-static bool tipc_sk_filter_connect(struct tipc_sock *tsk, struct sk_buff *skb)
+static bool tipc_sk_filter_connect(struct tipc_sock *tsk, struct sk_buff *skb,
+				   struct sk_buff_head *xmitq)
 {
 	struct sock *sk = &tsk->sk;
 	struct net *net = sock_net(sk);
@@ -2171,8 +2182,17 @@ static bool tipc_sk_filter_connect(struct tipc_sock *tsk, struct sk_buff *skb)
 		if (!skb_queue_empty(&sk->sk_write_queue))
 			tipc_sk_push_backlog(tsk);
 		/* Accept only connection-based messages sent by peer */
-		if (likely(con_msg && !err && pport == oport && pnode == onode))
+		if (likely(con_msg && !err && pport == oport &&
+			   pnode == onode)) {
+			if (msg_ack_required(hdr)) {
+				struct sk_buff *skb;
+
+				skb = tipc_sk_build_ack(tsk);
+				if (skb)
+					__skb_queue_tail(xmitq, skb);
+			}
 			return true;
+		}
 		if (!tsk_peer_msg(tsk, hdr))
 			return false;
 		if (!err)
@@ -2267,7 +2287,7 @@ static void tipc_sk_filter_rcv(struct sock *sk, struct sk_buff *skb,
 	while ((skb = __skb_dequeue(&inputq))) {
 		hdr = buf_msg(skb);
 		limit = rcvbuf_limit(sk, skb);
-		if ((sk_conn && !tipc_sk_filter_connect(tsk, skb)) ||
+		if ((sk_conn && !tipc_sk_filter_connect(tsk, skb, xmitq)) ||
 		    (!sk_conn && msg_connected(hdr)) ||
 		    (!grp && msg_in_group(hdr)))
 			err = TIPC_ERR_NO_PORT;
-- 
2.13.7

[net 2/3] tipc: fix memory leak in service subscripting

[Tuong Lien]

Upon receipt of a service subscription request from user via a topology
connection, one 'sub' object will be allocated in kernel, so it will be
able to send an event of the service if any to the user correspondingly
then. Also, in case of any failure, the connection will be shutdown and
all the pertaining 'sub' objects will be freed.

However, there is a race condition as follows resulting in memory leak:

       receive-work       connection        send-work
              |                |                |
        sub-1 |<------//-------|                |
        sub-2 |<------//-------|                |
              |                |<---------------| evt for sub-x
        sub-3 |<------//-------|                |
              :                :                :
              :                :                :
              |       /--------|                |
              |       |        * peer closed    |
              |       |        |                |
              |       |        |<-------X-------| evt for sub-y
              |       |        |<===============|
        sub-n |<------/        X    shutdown    |
    -> orphan |                                 |

That is, the 'receive-work' may get the last subscription request while
the 'send-work' is shutting down the connection due to peer close.

We had a 'lock' on the connection, so the two actions cannot be carried
out simultaneously. If the last subscription is allocated e.g. 'sub-n',
before the 'send-work' closes the connection, there will be no issue at
all, the 'sub' objects will be freed. In contrast the last subscription
will become orphan since the connection was closed, and we released all
references.

This commit fixes the issue by simply adding one test if the connection
remains in 'connected' state right after we obtain the connection lock,
then a subscription object can be created as usual, otherwise we ignore
it.

Acked-by: Ying Xue <ying.xue@windriver.com>
Acked-by: Jon Maloy <jmaloy@redhat.com>
Reported-by: Thang Ngo <thang.h.ngo@dektech.com.au>
Signed-off-by: Tuong Lien <tuong.t.lien@dektech.com.au>
---
 net/tipc/topsrv.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/net/tipc/topsrv.c b/net/tipc/topsrv.c
index 73dbed0c4b6b..931c426673c0 100644
--- a/net/tipc/topsrv.c
+++ b/net/tipc/topsrv.c
@@ -400,7 +400,9 @@ static int tipc_conn_rcv_from_sock(struct tipc_conn *con)
 		return -EWOULDBLOCK;
 	if (ret == sizeof(s)) {
 		read_lock_bh(&sk->sk_callback_lock);
-		ret = tipc_conn_rcv_sub(srv, con, &s);
+		/* RACE: the connection can be closed in the meantime */
+		if (likely(connected(con)))
+			ret = tipc_conn_rcv_sub(srv, con, &s);
 		read_unlock_bh(&sk->sk_callback_lock);
 		if (!ret)
 			return 0;
-- 
2.13.7

[net 3/3] tipc: fix failed service subscription deletion

[Tuong Lien]

When a service subscription is expired or canceled by user, it needs to
be deleted from the subscription list, so that new subscriptions can be
registered (max = 65535 per net). However, there are two issues in code
that can cause such an unused subscription to persist:

1) The 'tipc_conn_delete_sub()' has a loop on the subscription list but
it makes a break shortly when the 1st subscription differs from the one
specified, so the subscription will not be deleted.

2) In case a subscription is canceled, the code to remove the
'TIPC_SUB_CANCEL' flag from the subscription filter does not work if it
is a local subscription (i.e. the little endian isn't involved). So, it
will be no matches when looking for the subscription to delete later.

The subscription(s) will be removed eventually when the user terminates
its topology connection but that could be a long time later. Meanwhile,
the number of available subscriptions may be exhausted.

This commit fixes the two issues above, so as needed a subscription can
be deleted correctly.

Acked-by: Ying Xue <ying.xue@windriver.com>
Acked-by: Jon Maloy <jmaloy@redhat.com>
Signed-off-by: Tuong Lien <tuong.t.lien@dektech.com.au>
---
 net/tipc/subscr.h | 10 ++++++++++
 net/tipc/topsrv.c |  9 +++++----
 2 files changed, 15 insertions(+), 4 deletions(-)

diff --git a/net/tipc/subscr.h b/net/tipc/subscr.h
index aa015c233898..6ebbec1bedd1 100644
--- a/net/tipc/subscr.h
+++ b/net/tipc/subscr.h
@@ -96,6 +96,16 @@ void tipc_sub_get(struct tipc_subscription *subscription);
 		(swap_ ? swab32(val__) : val__);			\
 	})
 
+/* tipc_sub_write - write val_ to field_ of struct sub_ in user endian format
+ */
+#define tipc_sub_write(sub_, field_, val_)				\
+	({								\
+		struct tipc_subscr *sub__ = sub_;			\
+		u32 val__ = val_;					\
+		int swap_ = !((sub__)->filter & TIPC_FILTER_MASK);	\
+		(sub__)->field_ = swap_ ? swab32(val__) : val__;	\
+	})
+
 /* tipc_evt_write - write val_ to field_ of struct evt_ in user endian format
  */
 #define tipc_evt_write(evt_, field_, val_)				\
diff --git a/net/tipc/topsrv.c b/net/tipc/topsrv.c
index 931c426673c0..446af7bbd13e 100644
--- a/net/tipc/topsrv.c
+++ b/net/tipc/topsrv.c
@@ -237,8 +237,8 @@ static void tipc_conn_delete_sub(struct tipc_conn *con, struct tipc_subscr *s)
 		if (!s || !memcmp(s, &sub->evt.s, sizeof(*s))) {
 			tipc_sub_unsubscribe(sub);
 			atomic_dec(&tn->subscription_count);
-		} else if (s) {
-			break;
+			if (s)
+				break;
 		}
 	}
 	spin_unlock_bh(&con->sub_lock);
@@ -362,9 +362,10 @@ static int tipc_conn_rcv_sub(struct tipc_topsrv *srv,
 {
 	struct tipc_net *tn = tipc_net(srv->net);
 	struct tipc_subscription *sub;
+	u32 s_filter = tipc_sub_read(s, filter);
 
-	if (tipc_sub_read(s, filter) & TIPC_SUB_CANCEL) {
-		s->filter &= __constant_ntohl(~TIPC_SUB_CANCEL);
+	if (s_filter & TIPC_SUB_CANCEL) {
+		tipc_sub_write(s, filter, s_filter & ~TIPC_SUB_CANCEL);
 		tipc_conn_delete_sub(con, s);
 		return 0;
 	}
-- 
2.13.7

Re: [net 0/3] tipc: add some patches

[David Miller]

From: Tuong Lien <tuong.t.lien@dektech.com.au>
Date: Wed, 13 May 2020 19:33:15 +0700

> This series adds patches to fix some issues in TIPC streaming & service
> subscription.

Series applied.