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

Pull networking fixes from David Miller:

 1) Fix double-free in batman-adv, from Sven Eckelmann.

 2) Fix packet stats for fast-RX path, from Joannes Berg.

 3) Netfilter's ip_route_me_harder() doesn't handle request sockets
    properly, fix from Florian Westphal.

 4) Fix sendmsg deadlock in rxrpc, from David Howells.

 5) Add missing RCU locking to transport hashtable scan, from Xin Long.

 6) Fix potential packet loss in mlxsw driver, from Ido Schimmel.

 7) Fix race in NAPI handling between poll handlers and busy polling,
    from Eric Dumazet.

 8) TX path in vxlan and geneve need proper RCU locking, from Jakub
    Kicinski.

 9) SYN processing in DCCP and TCP need to disable BH, from Eric
    Dumazet.

10) Properly handle net_enable_timestamp() being invoked from IRQ
    context, also from Eric Dumazet.

11) Fix crash on device-tree systems in xgene driver, from Alban Bedel.

12) Do not call sk_free() on a locked socket, from Arnaldo Carvalho de
    Melo.

13) Fix use-after-free in netvsc driver, from Dexuan Cui.

14) Fix max MTU setting in bonding driver, from WANG Cong.

15) xen-netback hash table can be allocated from softirq context, so use
    GFP_ATOMIC. From Anoob Soman.

16) Fix MAC address change bug in bgmac driver, from Hari Vyas.

17) strparser needs to destroy strp_wq on module exit, from WANG Cong.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (69 commits)
  strparser: destroy workqueue on module exit
  sfc: fix IPID endianness in TSOv2
  sfc: avoid max() in array size
  rds: remove unnecessary returned value check
  rxrpc: Fix potential NULL-pointer exception
  nfp: correct DMA direction in XDP DMA sync
  nfp: don't tell FW about the reserved buffer space
  net: ethernet: bgmac: mac address change bug
  net: ethernet: bgmac: init sequence bug
  xen-netback: don't vfree() queues under spinlock
  xen-netback: keep a local pointer for vif in backend_disconnect()
  netfilter: nf_tables: don't call nfnetlink_set_err() if nfnetlink_send() fails
  netfilter: nft_set_rbtree: incorrect assumption on lower interval lookups
  netfilter: nf_conntrack_sip: fix wrong memory initialisation
  can: flexcan: fix typo in comment
  can: usb_8dev: Fix memory leak of priv->cmd_msg_buffer
  can: gs_usb: fix coding style
  can: gs_usb: Don't use stack memory for USB transfers
  ixgbe: Limit use of 2K buffers on architectures with 256B or larger cache lines
  ixgbe: update the rss key on h/w, when ethtool ask for it
  ...

1 files changed, 48 insertions, 0 deletions

diff --git a/net/rxrpc/call_accept.c b/net/rxrpc/call_accept.c
index 7c4c64ab8da2..0ed181f53f32 100644
--- a/net/rxrpc/call_accept.c
+++ b/net/rxrpc/call_accept.c
@@ -323,6 +323,8 @@ static struct rxrpc_call *rxrpc_alloc_incoming_call(struct rxrpc_sock *rx,
  *
  * If we want to report an error, we mark the skb with the packet type and
  * abort code and return NULL.
+ *
+ * The call is returned with the user access mutex held.
  */
 struct rxrpc_call *rxrpc_new_incoming_call(struct rxrpc_local *local,
 					   struct rxrpc_connection *conn,
@@ -371,6 +373,18 @@ found_service:
 	trace_rxrpc_receive(call, rxrpc_receive_incoming,
 			    sp->hdr.serial, sp->hdr.seq);
 
+	/* Lock the call to prevent rxrpc_kernel_send/recv_data() and
+	 * sendmsg()/recvmsg() inconveniently stealing the mutex once the
+	 * notification is generated.
+	 *
+	 * The BUG should never happen because the kernel should be well
+	 * behaved enough not to access the call before the first notification
+	 * event and userspace is prevented from doing so until the state is
+	 * appropriate.
+	 */
+	if (!mutex_trylock(&call->user_mutex))
+		BUG();
+
 	/* Make the call live. */
 	rxrpc_incoming_call(rx, call, skb);
 	conn = call->conn;
@@ -429,10 +443,12 @@ out:
 /*
  * handle acceptance of a call by userspace
  * - assign the user call ID to the call at the front of the queue
+ * - called with the socket locked.
  */
 struct rxrpc_call *rxrpc_accept_call(struct rxrpc_sock *rx,
 				     unsigned long user_call_ID,
 				     rxrpc_notify_rx_t notify_rx)
+	__releases(&rx->sk.sk_lock.slock)
 {
 	struct rxrpc_call *call;
 	struct rb_node *parent, **pp;
@@ -446,6 +462,7 @@ struct rxrpc_call *rxrpc_accept_call(struct rxrpc_sock *rx,
 
 	if (list_empty(&rx->to_be_accepted)) {
 		write_unlock(&rx->call_lock);
+		release_sock(&rx->sk);
 		kleave(" = -ENODATA [empty]");
 		return ERR_PTR(-ENODATA);
 	}
@@ -470,10 +487,39 @@ struct rxrpc_call *rxrpc_accept_call(struct rxrpc_sock *rx,
 	 */
 	call = list_entry(rx->to_be_accepted.next,
 			  struct rxrpc_call, accept_link);
+	write_unlock(&rx->call_lock);
+
+	/* We need to gain the mutex from the interrupt handler without
+	 * upsetting lockdep, so we have to release it there and take it here.
+	 * We are, however, still holding the socket lock, so other accepts
+	 * must wait for us and no one can add the user ID behind our backs.
+	 */
+	if (mutex_lock_interruptible(&call->user_mutex) < 0) {
+		release_sock(&rx->sk);
+		kleave(" = -ERESTARTSYS");
+		return ERR_PTR(-ERESTARTSYS);
+	}
+
+	write_lock(&rx->call_lock);
 	list_del_init(&call->accept_link);
 	sk_acceptq_removed(&rx->sk);
 	rxrpc_see_call(call);
 
+	/* Find the user ID insertion point. */
+	pp = &rx->calls.rb_node;
+	parent = NULL;
+	while (*pp) {
+		parent = *pp;
+		call = rb_entry(parent, struct rxrpc_call, sock_node);
+
+		if (user_call_ID < call->user_call_ID)
+			pp = &(*pp)->rb_left;
+		else if (user_call_ID > call->user_call_ID)
+			pp = &(*pp)->rb_right;
+		else
+			BUG();
+	}
+
 	write_lock_bh(&call->state_lock);
 	switch (call->state) {
 	case RXRPC_CALL_SERVER_ACCEPTING:
@@ -499,6 +545,7 @@ struct rxrpc_call *rxrpc_accept_call(struct rxrpc_sock *rx,
 	write_unlock(&rx->call_lock);
 	rxrpc_notify_socket(call);
 	rxrpc_service_prealloc(rx, GFP_KERNEL);
+	release_sock(&rx->sk);
 	_leave(" = %p{%d}", call, call->debug_id);
 	return call;
 
@@ -515,6 +562,7 @@ id_in_use:
 	write_unlock(&rx->call_lock);
 out:
 	rxrpc_service_prealloc(rx, GFP_KERNEL);
+	release_sock(&rx->sk);
 	_leave(" = %d", ret);
 	return ERR_PTR(ret);
 }