Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next

Pull networking updates from Davic Miller:

 1) Support busy polling generically, for all NAPI drivers.  From Eric
    Dumazet.

 2) Add byte/packet counter support to nft_ct, from Floriani Westphal.

 3) Add RSS/XPS support to mvneta driver, from Gregory Clement.

 4) Implement IPV6_HDRINCL socket option for raw sockets, from Hannes
    Frederic Sowa.

 5) Add support for T6 adapter to cxgb4 driver, from Hariprasad Shenai.

 6) Add support for VLAN device bridging to mlxsw switch driver, from
    Ido Schimmel.

 7) Add driver for Netronome NFP4000/NFP6000, from Jakub Kicinski.

 8) Provide hwmon interface to mlxsw switch driver, from Jiri Pirko.

 9) Reorganize wireless drivers into per-vendor directories just like we
    do for ethernet drivers.  From Kalle Valo.

10) Provide a way for administrators "destroy" connected sockets via the
    SOCK_DESTROY socket netlink diag operation.  From Lorenzo Colitti.

11) Add support to add/remove multicast routes via netlink, from Nikolay
    Aleksandrov.

12) Make TCP keepalive settings per-namespace, from Nikolay Borisov.

13) Add forwarding and packet duplication facilities to nf_tables, from
    Pablo Neira Ayuso.

14) Dead route support in MPLS, from Roopa Prabhu.

15) TSO support for thunderx chips, from Sunil Goutham.

16) Add driver for IBM's System i/p VNIC protocol, from Thomas Falcon.

17) Rationalize, consolidate, and more completely document the checksum
    offloading facilities in the networking stack.  From Tom Herbert.

18) Support aborting an ongoing scan in mac80211/cfg80211, from
    Vidyullatha Kanchanapally.

19) Use per-bucket spinlock for bpf hash facility, from Tom Leiming.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next: (1375 commits)
  net: bnxt: always return values from _bnxt_get_max_rings
  net: bpf: reject invalid shifts
  phonet: properly unshare skbs in phonet_rcv()
  dwc_eth_qos: Fix dma address for multi-fragment skbs
  phy: remove an unneeded condition
  mdio: remove an unneed condition
  mdio_bus: NULL dereference on allocation error
  net: Fix typo in netdev_intersect_features
  net: freescale: mac-fec: Fix build error from phy_device API change
  net: freescale: ucc_geth: Fix build error from phy_device API change
  bonding: Prevent IPv6 link local address on enslaved devices
  IB/mlx5: Add flow steering support
  net/mlx5_core: Export flow steering API
  net/mlx5_core: Make ipv4/ipv6 location more clear
  net/mlx5_core: Enable flow steering support for the IB driver
  net/mlx5_core: Initialize namespaces only when supported by device
  net/mlx5_core: Set priority attributes
  net/mlx5_core: Connect flow tables
  net/mlx5_core: Introduce modify flow table command
  net/mlx5_core: Managing root flow table
  ...

2 files changed, 53 insertions, 27 deletions

diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c
index ef05cd9403d4..c5bf5ef2bf89 100644
--- a/net/unix/af_unix.c
+++ b/net/unix/af_unix.c
@@ -451,7 +451,7 @@ static void unix_write_space(struct sock *sk)
 	rcu_read_lock();
 	if (unix_writable(sk)) {
 		wq = rcu_dereference(sk->sk_wq);
-		if (wq_has_sleeper(wq))
+		if (skwq_has_sleeper(wq))
 			wake_up_interruptible_sync_poll(&wq->wait,
 				POLLOUT | POLLWRNORM | POLLWRBAND);
 		sk_wake_async(sk, SOCK_WAKE_SPACE, POLL_OUT);
@@ -1513,6 +1513,21 @@ static void unix_destruct_scm(struct sk_buff *skb)
 	sock_wfree(skb);
 }
 
+/*
+ * The "user->unix_inflight" variable is protected by the garbage
+ * collection lock, and we just read it locklessly here. If you go
+ * over the limit, there might be a tiny race in actually noticing
+ * it across threads. Tough.
+ */
+static inline bool too_many_unix_fds(struct task_struct *p)
+{
+	struct user_struct *user = current_user();
+
+	if (unlikely(user->unix_inflight > task_rlimit(p, RLIMIT_NOFILE)))
+		return !capable(CAP_SYS_RESOURCE) && !capable(CAP_SYS_ADMIN);
+	return false;
+}
+
 #define MAX_RECURSION_LEVEL 4
 
 static int unix_attach_fds(struct scm_cookie *scm, struct sk_buff *skb)
@@ -1521,6 +1536,9 @@ static int unix_attach_fds(struct scm_cookie *scm, struct sk_buff *skb)
 	unsigned char max_level = 0;
 	int unix_sock_count = 0;
 
+	if (too_many_unix_fds(current))
+		return -ETOOMANYREFS;
+
 	for (i = scm->fp->count - 1; i >= 0; i--) {
 		struct sock *sk = unix_get_socket(scm->fp->fp[i]);
 
@@ -1542,10 +1560,8 @@ static int unix_attach_fds(struct scm_cookie *scm, struct sk_buff *skb)
 	if (!UNIXCB(skb).fp)
 		return -ENOMEM;
 
-	if (unix_sock_count) {
-		for (i = scm->fp->count - 1; i >= 0; i--)
-			unix_inflight(scm->fp->fp[i]);
-	}
+	for (i = scm->fp->count - 1; i >= 0; i--)
+		unix_inflight(scm->fp->fp[i]);
 	return max_level;
 }
 
@@ -2092,8 +2108,8 @@ static int unix_dgram_recvmsg(struct socket *sock, struct msghdr *msg,
 	struct scm_cookie scm;
 	struct sock *sk = sock->sk;
 	struct unix_sock *u = unix_sk(sk);
-	int noblock = flags & MSG_DONTWAIT;
-	struct sk_buff *skb;
+	struct sk_buff *skb, *last;
+	long timeo;
 	int err;
 	int peeked, skip;
 
@@ -2101,30 +2117,38 @@ static int unix_dgram_recvmsg(struct socket *sock, struct msghdr *msg,
 	if (flags&MSG_OOB)
 		goto out;
 
-	err = mutex_lock_interruptible(&u->readlock);
-	if (unlikely(err)) {
-		/* recvmsg() in non blocking mode is supposed to return -EAGAIN
-		 * sk_rcvtimeo is not honored by mutex_lock_interruptible()
-		 */
-		err = noblock ? -EAGAIN : -ERESTARTSYS;
-		goto out;
-	}
+	timeo = sock_rcvtimeo(sk, flags & MSG_DONTWAIT);
 
-	skip = sk_peek_offset(sk, flags);
+	do {
+		mutex_lock(&u->readlock);
 
-	skb = __skb_recv_datagram(sk, flags, &peeked, &skip, &err);
-	if (!skb) {
+		skip = sk_peek_offset(sk, flags);
+		skb = __skb_try_recv_datagram(sk, flags, &peeked, &skip, &err,
+					      &last);
+		if (skb)
+			break;
+
+		mutex_unlock(&u->readlock);
+
+		if (err != -EAGAIN)
+			break;
+	} while (timeo &&
+		 !__skb_wait_for_more_packets(sk, &err, &timeo, last));
+
+	if (!skb) { /* implies readlock unlocked */
 		unix_state_lock(sk);
 		/* Signal EOF on disconnected non-blocking SEQPACKET socket. */
 		if (sk->sk_type == SOCK_SEQPACKET && err == -EAGAIN &&
 		    (sk->sk_shutdown & RCV_SHUTDOWN))
 			err = 0;
 		unix_state_unlock(sk);
-		goto out_unlock;
+		goto out;
 	}
 
-	wake_up_interruptible_sync_poll(&u->peer_wait,
-					POLLOUT | POLLWRNORM | POLLWRBAND);
+	if (wq_has_sleeper(&u->peer_wait))
+		wake_up_interruptible_sync_poll(&u->peer_wait,
+						POLLOUT | POLLWRNORM |
+						POLLWRBAND);
 
 	if (msg->msg_name)
 		unix_copy_addr(msg, skb->sk);
@@ -2176,7 +2200,6 @@ static int unix_dgram_recvmsg(struct socket *sock, struct msghdr *msg,
 
 out_free:
 	skb_free_datagram(sk, skb);
-out_unlock:
 	mutex_unlock(&u->readlock);
 out:
 	return err;
diff --git a/net/unix/garbage.c b/net/unix/garbage.c
index a73a226f2d33..8fcdc2283af5 100644
--- a/net/unix/garbage.c
+++ b/net/unix/garbage.c
@@ -120,11 +120,11 @@ void unix_inflight(struct file *fp)
 {
 	struct sock *s = unix_get_socket(fp);
 
+	spin_lock(&unix_gc_lock);
+
 	if (s) {
 		struct unix_sock *u = unix_sk(s);
 
-		spin_lock(&unix_gc_lock);
-
 		if (atomic_long_inc_return(&u->inflight) == 1) {
 			BUG_ON(!list_empty(&u->link));
 			list_add_tail(&u->link, &gc_inflight_list);
@@ -132,25 +132,28 @@ void unix_inflight(struct file *fp)
 			BUG_ON(list_empty(&u->link));
 		}
 		unix_tot_inflight++;
-		spin_unlock(&unix_gc_lock);
 	}
+	fp->f_cred->user->unix_inflight++;
+	spin_unlock(&unix_gc_lock);
 }
 
 void unix_notinflight(struct file *fp)
 {
 	struct sock *s = unix_get_socket(fp);
 
+	spin_lock(&unix_gc_lock);
+
 	if (s) {
 		struct unix_sock *u = unix_sk(s);
 
-		spin_lock(&unix_gc_lock);
 		BUG_ON(list_empty(&u->link));
 
 		if (atomic_long_dec_and_test(&u->inflight))
 			list_del_init(&u->link);
 		unix_tot_inflight--;
-		spin_unlock(&unix_gc_lock);
 	}
+	fp->f_cred->user->unix_inflight--;
+	spin_unlock(&unix_gc_lock);
 }
 
 static void scan_inflight(struct sock *x, void (*func)(struct unix_sock *),