[PATCH] tipc: fix deadlock during socket release

[PATCH] tipc: fix deadlock during socket release

[Ying Xue]

A deadlock might occur if name table is withdrawn in socket release
routine, and while packets are still being received from bearer.

       CPU0                       CPU1
T0:   recv_msg()               release()
T1:   tipc_recv_msg()          tipc_withdraw()
T2:   [grab node lock]         [grab port lock]
T3:   tipc_link_wakeup_ports() tipc_nametbl_withdraw()
T4:   [grab port lock]*        named_cluster_distribute()
T5:   wakeupdispatch()         tipc_link_send()
T6:                            [grab node lock]*

The opposite order of holding port lock and node lock on above two
different paths may result in a deadlock. If socket lock instead of
port lock is used to protect port instance in tipc_withdraw(), the
reverse order of holding port lock and node lock will be eliminated,
as a result, the deadlock is killed as well.

Reported-by: Lars Everbrand <lars.everbrand@ericsson.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Signed-off-by: Ying Xue <ying.xue@windriver.com>
---
 net/tipc/port.c   |   45 +++++++++++++++------------------------------
 net/tipc/port.h   |    6 +++---
 net/tipc/socket.c |   46 +++++++++++++++++++++++++++++++---------------
 3 files changed, 49 insertions(+), 48 deletions(-)

diff --git a/net/tipc/port.c b/net/tipc/port.c
index c081a76..d43f318 100644
--- a/net/tipc/port.c
+++ b/net/tipc/port.c
@@ -251,18 +251,15 @@ struct tipc_port *tipc_createport(struct sock *sk,
 	return p_ptr;
 }
 
-int tipc_deleteport(u32 ref)
+int tipc_deleteport(struct tipc_port *p_ptr)
 {
-	struct tipc_port *p_ptr;
 	struct sk_buff *buf = NULL;
 
-	tipc_withdraw(ref, 0, NULL);
-	p_ptr = tipc_port_lock(ref);
-	if (!p_ptr)
-		return -EINVAL;
+	tipc_withdraw(p_ptr, 0, NULL);
 
-	tipc_ref_discard(ref);
-	tipc_port_unlock(p_ptr);
+	spin_lock_bh(p_ptr->lock);
+	tipc_ref_discard(p_ptr->ref);
+	spin_unlock_bh(p_ptr->lock);
 
 	k_cancel_timer(&p_ptr->timer);
 	if (p_ptr->connected) {
@@ -704,47 +701,36 @@ int tipc_set_portimportance(u32 ref, unsigned int imp)
 }
 
 
-int tipc_publish(u32 ref, unsigned int scope, struct tipc_name_seq const *seq)
+int tipc_publish(struct tipc_port *p_ptr, unsigned int scope,
+		 struct tipc_name_seq const *seq)
 {
-	struct tipc_port *p_ptr;
 	struct publication *publ;
 	u32 key;
-	int res = -EINVAL;
 
-	p_ptr = tipc_port_lock(ref);
-	if (!p_ptr)
+	if (p_ptr->connected)
 		return -EINVAL;
+	key = p_ptr->ref + p_ptr->pub_count + 1;
+	if (key == p_ptr->ref)
+		return -EADDRINUSE;
 
-	if (p_ptr->connected)
-		goto exit;
-	key = ref + p_ptr->pub_count + 1;
-	if (key == ref) {
-		res = -EADDRINUSE;
-		goto exit;
-	}
 	publ = tipc_nametbl_publish(seq->type, seq->lower, seq->upper,
 				    scope, p_ptr->ref, key);
 	if (publ) {
 		list_add(&publ->pport_list, &p_ptr->publications);
 		p_ptr->pub_count++;
 		p_ptr->published = 1;
-		res = 0;
+		return 0;
 	}
-exit:
-	tipc_port_unlock(p_ptr);
-	return res;
+	return -EINVAL;
 }
 
-int tipc_withdraw(u32 ref, unsigned int scope, struct tipc_name_seq const *seq)
+int tipc_withdraw(struct tipc_port *p_ptr, unsigned int scope,
+		  struct tipc_name_seq const *seq)
 {
-	struct tipc_port *p_ptr;
 	struct publication *publ;
 	struct publication *tpubl;
 	int res = -EINVAL;
 
-	p_ptr = tipc_port_lock(ref);
-	if (!p_ptr)
-		return -EINVAL;
 	if (!seq) {
 		list_for_each_entry_safe(publ, tpubl,
 					 &p_ptr->publications, pport_list) {
@@ -771,7 +757,6 @@ int tipc_withdraw(u32 ref, unsigned int scope, struct tipc_name_seq const *seq)
 	}
 	if (list_empty(&p_ptr->publications))
 		p_ptr->published = 0;
-	tipc_port_unlock(p_ptr);
 	return res;
 }
 
diff --git a/net/tipc/port.h b/net/tipc/port.h
index 9122535..34f12bd 100644
--- a/net/tipc/port.h
+++ b/net/tipc/port.h
@@ -116,7 +116,7 @@ int tipc_reject_msg(struct sk_buff *buf, u32 err);
 
 void tipc_acknowledge(u32 port_ref, u32 ack);
 
-int tipc_deleteport(u32 portref);
+int tipc_deleteport(struct tipc_port *p_ptr);
 
 int tipc_portimportance(u32 portref, unsigned int *importance);
 int tipc_set_portimportance(u32 portref, unsigned int importance);
@@ -127,9 +127,9 @@ int tipc_set_portunreliable(u32 portref, unsigned int isunreliable);
 int tipc_portunreturnable(u32 portref, unsigned int *isunreturnable);
 int tipc_set_portunreturnable(u32 portref, unsigned int isunreturnable);
 
-int tipc_publish(u32 portref, unsigned int scope,
+int tipc_publish(struct tipc_port *p_ptr, unsigned int scope,
 		 struct tipc_name_seq const *name_seq);
-int tipc_withdraw(u32 portref, unsigned int scope,
+int tipc_withdraw(struct tipc_port *p_ptr, unsigned int scope,
 		  struct tipc_name_seq const *name_seq);
 
 int tipc_connect(u32 portref, struct tipc_portid const *port);
diff --git a/net/tipc/socket.c b/net/tipc/socket.c
index 3b61851..e741416 100644
--- a/net/tipc/socket.c
+++ b/net/tipc/socket.c
@@ -354,7 +354,7 @@ static int release(struct socket *sock)
 	 * Delete TIPC port; this ensures no more messages are queued
 	 * (also disconnects an active connection & sends a 'FIN-' to peer)
 	 */
-	res = tipc_deleteport(tport->ref);
+	res = tipc_deleteport(tport);
 
 	/* Discard any remaining (connection-based) messages in receive queue */
 	__skb_queue_purge(&sk->sk_receive_queue);
@@ -386,30 +386,46 @@ static int release(struct socket *sock)
  */
 static int bind(struct socket *sock, struct sockaddr *uaddr, int uaddr_len)
 {
+	struct sock *sk = sock->sk;
 	struct sockaddr_tipc *addr = (struct sockaddr_tipc *)uaddr;
-	u32 portref = tipc_sk_port(sock->sk)->ref;
+	struct tipc_port *tport = tipc_sk_port(sock->sk);
+	int res = -EINVAL;
 
-	if (unlikely(!uaddr_len))
-		return tipc_withdraw(portref, 0, NULL);
+	lock_sock(sk);
+	if (unlikely(!uaddr_len)) {
+		res = tipc_withdraw(tport, 0, NULL);
+		goto exit;
+	}
 
-	if (uaddr_len < sizeof(struct sockaddr_tipc))
-		return -EINVAL;
-	if (addr->family != AF_TIPC)
-		return -EAFNOSUPPORT;
+	if (uaddr_len < sizeof(struct sockaddr_tipc)) {
+		res = -EINVAL;
+		goto exit;
+	}
+	if (addr->family != AF_TIPC) {
+		res = -EAFNOSUPPORT;
+		goto exit;
+	}
 
 	if (addr->addrtype == TIPC_ADDR_NAME)
 		addr->addr.nameseq.upper = addr->addr.nameseq.lower;
-	else if (addr->addrtype != TIPC_ADDR_NAMESEQ)
-		return -EAFNOSUPPORT;
+	else if (addr->addrtype != TIPC_ADDR_NAMESEQ) {
+		res = -EAFNOSUPPORT;
+		goto exit;
+	}
 
 	if ((addr->addr.nameseq.type < TIPC_RESERVED_TYPES) &&
 	    (addr->addr.nameseq.type != TIPC_TOP_SRV) &&
-	    (addr->addr.nameseq.type != TIPC_CFG_SRV))
-		return -EACCES;
+	    (addr->addr.nameseq.type != TIPC_CFG_SRV)) {
+		res = -EACCES;
+		goto exit;
+	}
 
-	return (addr->scope > 0) ?
-		tipc_publish(portref, addr->scope, &addr->addr.nameseq) :
-		tipc_withdraw(portref, -addr->scope, &addr->addr.nameseq);
+	res = (addr->scope > 0) ?
+		tipc_publish(tport, addr->scope, &addr->addr.nameseq) :
+		tipc_withdraw(tport, -addr->scope, &addr->addr.nameseq);
+exit:
+	release_sock(sk);
+	return res;
 }
 
 /**
-- 
1.7.9.5

Re: [PATCH] tipc: fix deadlock during socket release

[David Miller]

From: Ying Xue <ying.xue@windriver.com>
Date: Fri, 27 Dec 2013 10:18:28 +0800

> A deadlock might occur if name table is withdrawn in socket release
> routine, and while packets are still being received from bearer.
> 
>        CPU0                       CPU1
> T0:   recv_msg()               release()
> T1:   tipc_recv_msg()          tipc_withdraw()
> T2:   [grab node lock]         [grab port lock]
> T3:   tipc_link_wakeup_ports() tipc_nametbl_withdraw()
> T4:   [grab port lock]*        named_cluster_distribute()
> T5:   wakeupdispatch()         tipc_link_send()
> T6:                            [grab node lock]*
> 
> The opposite order of holding port lock and node lock on above two
> different paths may result in a deadlock. If socket lock instead of
> port lock is used to protect port instance in tipc_withdraw(), the
> reverse order of holding port lock and node lock will be eliminated,
> as a result, the deadlock is killed as well.
> 
> Reported-by: Lars Everbrand <lars.everbrand@ericsson.com>
> Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
> Signed-off-by: Ying Xue <ying.xue@windriver.com>

Applied, thanks.