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

Pull networking fixes from David Miller:

 1) Fix several scatter gather list issues in kTLS code, from Jakub
    Kicinski.

 2) macb driver device remove has to kill the hresp_err_tasklet. From
    Chuhong Yuan.

 3) Several memory leak and reference count bug fixes in tipc, from Tung
    Nguyen.

 4) Fix mlx5 build error w/o ipv6, from Yue Haibing.

 5) Fix jumbo frame and other regressions in r8169, from Heiner
    Kallweit.

 6) Undo some BUG_ON()'s and replace them with WARN_ON_ONCE and proper
    error propagation/handling. From Paolo Abeni.

* git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (24 commits)
  openvswitch: remove another BUG_ON()
  openvswitch: drop unneeded BUG_ON() in ovs_flow_cmd_build_info()
  net: phy: realtek: fix using paged operations with RTL8105e / RTL8208
  r8169: fix resume on cable plug-in
  r8169: fix jumbo configuration for RTL8168evl
  net: emulex: benet: indent a Kconfig depends continuation line
  selftests: forwarding: fix race between packet receive and tc check
  net: sched: fix `tc -s class show` no bstats on class with nolock subqueues
  net: ethernet: ti: ale: ensure vlan/mdb deleted when no members
  net/mlx5e: Fix build error without IPV6
  selftests: pmtu: use -oneline for ip route list cache
  tipc: fix duplicate SYN messages under link congestion
  tipc: fix wrong timeout input for tipc_wait_for_cond()
  tipc: fix wrong socket reference counter after tipc_sk_timeout() returns
  tipc: fix potential memory leak in __tipc_sendmsg()
  net: macb: add missed tasklet_kill
  selftests: bpf: correct perror strings
  selftests: bpf: test_sockmap: handle file creation failures gracefully
  net/tls: use sg_next() to walk sg entries
  net/tls: remove the dead inplace_crypto code
  ...

net/core/filter.c

@@ -2299,7 +2299,7 @@ BPF_CALL_4(bpf_msg_pull_data, struct sk_msg *, msg, u32, start,
 	WARN_ON_ONCE(last_sge == first_sge);
 	shift = last_sge > first_sge ?
 		last_sge - first_sge - 1 :
-		MAX_SKB_FRAGS - first_sge + last_sge - 1;
+		NR_MSG_FRAG_IDS - first_sge + last_sge - 1;
 	if (!shift)
 		goto out;
 
@@ -2308,8 +2308,8 @@ BPF_CALL_4(bpf_msg_pull_data, struct sk_msg *, msg, u32, start,
 	do {
 		u32 move_from;
 
-		if (i + shift >= MAX_MSG_FRAGS)
-			move_from = i + shift - MAX_MSG_FRAGS;
+		if (i + shift >= NR_MSG_FRAG_IDS)
+			move_from = i + shift - NR_MSG_FRAG_IDS;
 		else
 			move_from = i + shift;
 		if (move_from == msg->sg.end)
@@ -2323,7 +2323,7 @@ BPF_CALL_4(bpf_msg_pull_data, struct sk_msg *, msg, u32, start,
 	} while (1);
 
 	msg->sg.end = msg->sg.end - shift > msg->sg.end ?
-		      msg->sg.end - shift + MAX_MSG_FRAGS :
+		      msg->sg.end - shift + NR_MSG_FRAG_IDS :
 		      msg->sg.end - shift;
 out:
 	msg->data = sg_virt(&msg->sg.data[first_sge]) + start - offset;

net/core/skmsg.c

@@ -421,7 +421,7 @@ static int sk_psock_skb_ingress(struct sk_psock *psock, struct sk_buff *skb)
 	copied = skb->len;
 	msg->sg.start = 0;
 	msg->sg.size = copied;
-	msg->sg.end = num_sge == MAX_MSG_FRAGS ? 0 : num_sge;
+	msg->sg.end = num_sge;
 	msg->skb = skb;
 
 	sk_psock_queue_msg(psock, msg);

net/ipv4/tcp_bpf.c

@@ -301,7 +301,7 @@ EXPORT_SYMBOL_GPL(tcp_bpf_sendmsg_redir);
 static int tcp_bpf_send_verdict(struct sock *sk, struct sk_psock *psock,
 				struct sk_msg *msg, int *copied, int flags)
 {
-	bool cork = false, enospc = msg->sg.start == msg->sg.end;
+	bool cork = false, enospc = sk_msg_full(msg);
 	struct sock *sk_redir;
 	u32 tosend, delta = 0;
 	int ret;

net/openvswitch/datapath.c

@@ -905,7 +905,10 @@ static struct sk_buff *ovs_flow_cmd_build_info(const struct sw_flow *flow,
 	retval = ovs_flow_cmd_fill_info(flow, dp_ifindex, skb,
 					info->snd_portid, info->snd_seq, 0,
 					cmd, ufid_flags);
-	BUG_ON(retval < 0);
+	if (WARN_ON_ONCE(retval < 0)) {
+		kfree_skb(skb);
+		skb = ERR_PTR(retval);
+	}
 	return skb;
 }
 
@@ -1369,7 +1372,10 @@ static int ovs_flow_cmd_del(struct sk_buff *skb, struct genl_info *info)
 						     OVS_FLOW_CMD_DEL,
 						     ufid_flags);
 			rcu_read_unlock();
-			BUG_ON(err < 0);
+			if (WARN_ON_ONCE(err < 0)) {
+				kfree_skb(reply);
+				goto out_free;
+			}
 
 			ovs_notify(&dp_flow_genl_family, reply, info);
 		} else {
@@ -1377,6 +1383,7 @@ static int ovs_flow_cmd_del(struct sk_buff *skb, struct genl_info *info)
 		}
 	}
 
+out_free:
 	ovs_flow_free(flow, true);
 	return 0;
 unlock:

net/sched/sch_mq.c

@@ -245,7 +245,8 @@ static int mq_dump_class_stats(struct Qdisc *sch, unsigned long cl,
 	struct netdev_queue *dev_queue = mq_queue_get(sch, cl);
 
 	sch = dev_queue->qdisc_sleeping;
-	if (gnet_stats_copy_basic(&sch->running, d, NULL, &sch->bstats) < 0 ||
+	if (gnet_stats_copy_basic(&sch->running, d, sch->cpu_bstats,
+				  &sch->bstats) < 0 ||
 	    qdisc_qstats_copy(d, sch) < 0)
 		return -1;
 	return 0;

net/sched/sch_mqprio.c

@@ -557,8 +557,8 @@ static int mqprio_dump_class_stats(struct Qdisc *sch, unsigned long cl,
 		struct netdev_queue *dev_queue = mqprio_queue_get(sch, cl);
 
 		sch = dev_queue->qdisc_sleeping;
-		if (gnet_stats_copy_basic(qdisc_root_sleeping_running(sch),
-					  d, NULL, &sch->bstats) < 0 ||
+		if (gnet_stats_copy_basic(qdisc_root_sleeping_running(sch), d,
+					  sch->cpu_bstats, &sch->bstats) < 0 ||
 		    qdisc_qstats_copy(d, sch) < 0)
 			return -1;
 	}

net/sched/sch_multiq.c

@@ -339,7 +339,7 @@ static int multiq_dump_class_stats(struct Qdisc *sch, unsigned long cl,
 
 	cl_q = q->queues[cl - 1];
 	if (gnet_stats_copy_basic(qdisc_root_sleeping_running(sch),
-				  d, NULL, &cl_q->bstats) < 0 ||
+				  d, cl_q->cpu_bstats, &cl_q->bstats) < 0 ||
 	    qdisc_qstats_copy(d, cl_q) < 0)
 		return -1;
 

net/sched/sch_prio.c

@@ -356,7 +356,7 @@ static int prio_dump_class_stats(struct Qdisc *sch, unsigned long cl,
 
 	cl_q = q->queues[cl - 1];
 	if (gnet_stats_copy_basic(qdisc_root_sleeping_running(sch),
-				  d, NULL, &cl_q->bstats) < 0 ||
+				  d, cl_q->cpu_bstats, &cl_q->bstats) < 0 ||
 	    qdisc_qstats_copy(d, cl_q) < 0)
 		return -1;
 

net/tipc/socket.c

@@ -532,7 +532,7 @@ static void __tipc_shutdown(struct socket *sock, int error)
 	struct sock *sk = sock->sk;
 	struct tipc_sock *tsk = tipc_sk(sk);
 	struct net *net = sock_net(sk);
-	long timeout = CONN_TIMEOUT_DEFAULT;
+	long timeout = msecs_to_jiffies(CONN_TIMEOUT_DEFAULT);
 	u32 dnode = tsk_peer_node(tsk);
 	struct sk_buff *skb;
 
@@ -540,12 +540,10 @@ static void __tipc_shutdown(struct socket *sock, int error)
 	tipc_wait_for_cond(sock, &timeout, (!tsk->cong_link_cnt &&
 					    !tsk_conn_cong(tsk)));
 
-	/* Push out unsent messages or remove if pending SYN */
-	skb = skb_peek(&sk->sk_write_queue);
-	if (skb && !msg_is_syn(buf_msg(skb)))
-		tipc_sk_push_backlog(tsk);
-	else
-		__skb_queue_purge(&sk->sk_write_queue);
+	/* Push out delayed messages if in Nagle mode */
+	tipc_sk_push_backlog(tsk);
+	/* Remove pending SYN */
+	__skb_queue_purge(&sk->sk_write_queue);
 
 	/* Reject all unreceived messages, except on an active connection
 	 * (which disconnects locally & sends a 'FIN+' to peer).
@@ -1248,9 +1246,14 @@ static void tipc_sk_push_backlog(struct tipc_sock *tsk)
 	struct sk_buff_head *txq = &tsk->sk.sk_write_queue;
 	struct net *net = sock_net(&tsk->sk);
 	u32 dnode = tsk_peer_node(tsk);
+	struct sk_buff *skb = skb_peek(txq);
 	int rc;
 
-	if (skb_queue_empty(txq) || tsk->cong_link_cnt)
+	if (!skb || tsk->cong_link_cnt)
+		return;
+
+	/* Do not send SYN again after congestion */
+	if (msg_is_syn(buf_msg(skb)))
 		return;
 
 	tsk->snt_unacked += tsk->snd_backlog;
@@ -1447,8 +1450,10 @@ static int __tipc_sendmsg(struct socket *sock, struct msghdr *m, size_t dlen)
 	rc = tipc_msg_build(hdr, m, 0, dlen, mtu, &pkts);
 	if (unlikely(rc != dlen))
 		return rc;
-	if (unlikely(syn && !tipc_msg_skb_clone(&pkts, &sk->sk_write_queue)))
+	if (unlikely(syn && !tipc_msg_skb_clone(&pkts, &sk->sk_write_queue))) {
+		__skb_queue_purge(&pkts);
 		return -ENOMEM;
+	}
 
 	trace_tipc_sk_sendmsg(sk, skb_peek(&pkts), TIPC_DUMP_SK_SNDQ, " ");
 	rc = tipc_node_xmit(net, &pkts, dnode, tsk->portid);
@@ -2757,6 +2762,7 @@ static void tipc_sk_timeout(struct timer_list *t)
 	if (sock_owned_by_user(sk)) {
 		sk_reset_timer(sk, &sk->sk_timer, jiffies + HZ / 20);
 		bh_unlock_sock(sk);
+		sock_put(sk);
 		return;
 	}
 

net/tls/tls_main.c

@@ -209,24 +209,15 @@ int tls_push_partial_record(struct sock *sk, struct tls_context *ctx,
 	return tls_push_sg(sk, ctx, sg, offset, flags);
 }
 
-bool tls_free_partial_record(struct sock *sk, struct tls_context *ctx)
+void tls_free_partial_record(struct sock *sk, struct tls_context *ctx)
 {
 	struct scatterlist *sg;
 
-	sg = ctx->partially_sent_record;
-	if (!sg)
-		return false;
-
-	while (1) {
+	for (sg = ctx->partially_sent_record; sg; sg = sg_next(sg)) {
 		put_page(sg_page(sg));
 		sk_mem_uncharge(sk, sg->length);
-
-		if (sg_is_last(sg))
-			break;
-		sg++;
 	}
 	ctx->partially_sent_record = NULL;
-	return true;
 }
 
 static void tls_write_space(struct sock *sk)

net/tls/tls_sw.c

@@ -710,8 +710,7 @@ static int tls_push_record(struct sock *sk, int flags,
 	}
 
 	i = msg_pl->sg.start;
-	sg_chain(rec->sg_aead_in, 2, rec->inplace_crypto ?
-		 &msg_en->sg.data[i] : &msg_pl->sg.data[i]);
+	sg_chain(rec->sg_aead_in, 2, &msg_pl->sg.data[i]);
 
 	i = msg_en->sg.end;
 	sk_msg_iter_var_prev(i);
@@ -771,8 +770,14 @@ static int bpf_exec_tx_verdict(struct sk_msg *msg, struct sock *sk,
 
 	policy = !(flags & MSG_SENDPAGE_NOPOLICY);
 	psock = sk_psock_get(sk);
-	if (!psock || !policy)
-		return tls_push_record(sk, flags, record_type);
+	if (!psock || !policy) {
+		err = tls_push_record(sk, flags, record_type);
+		if (err) {
+			*copied -= sk_msg_free(sk, msg);
+			tls_free_open_rec(sk);
+		}
+		return err;
+	}
 more_data:
 	enospc = sk_msg_full(msg);
 	if (psock->eval == __SK_NONE) {
@@ -970,8 +975,6 @@ alloc_encrypted:
 			if (ret)
 				goto fallback_to_reg_send;
 
-			rec->inplace_crypto = 0;
-
 			num_zc++;
 			copied += try_to_copy;
 
@@ -984,7 +987,7 @@ alloc_encrypted:
 					num_async++;
 				else if (ret == -ENOMEM)
 					goto wait_for_memory;
-				else if (ret == -ENOSPC)
+				else if (ctx->open_rec && ret == -ENOSPC)
 					goto rollback_iter;
 				else if (ret != -EAGAIN)
 					goto send_end;
@@ -1053,11 +1056,12 @@ wait_for_memory:
 		ret = sk_stream_wait_memory(sk, &timeo);
 		if (ret) {
 trim_sgl:
-			tls_trim_both_msgs(sk, orig_size);
+			if (ctx->open_rec)
+				tls_trim_both_msgs(sk, orig_size);
 			goto send_end;
 		}
 
-		if (msg_en->sg.size < required_size)
+		if (ctx->open_rec && msg_en->sg.size < required_size)
 			goto alloc_encrypted;
 	}
 
@@ -1169,7 +1173,6 @@ alloc_payload:
 
 		tls_ctx->pending_open_record_frags = true;
 		if (full_record || eor || sk_msg_full(msg_pl)) {
-			rec->inplace_crypto = 0;
 			ret = bpf_exec_tx_verdict(msg_pl, sk, full_record,
 						  record_type, &copied, flags);
 			if (ret) {
@@ -1190,11 +1193,13 @@ wait_for_sndbuf:
 wait_for_memory:
 		ret = sk_stream_wait_memory(sk, &timeo);
 		if (ret) {
-			tls_trim_both_msgs(sk, msg_pl->sg.size);
+			if (ctx->open_rec)
+				tls_trim_both_msgs(sk, msg_pl->sg.size);
 			goto sendpage_end;
 		}
 
-		goto alloc_payload;
+		if (ctx->open_rec)
+			goto alloc_payload;
 	}
 
 	if (num_async) {
@@ -2084,7 +2089,8 @@ void tls_sw_release_resources_tx(struct sock *sk)
 	/* Free up un-sent records in tx_list. First, free
 	 * the partially sent record if any at head of tx_list.
 	 */
-	if (tls_free_partial_record(sk, tls_ctx)) {
+	if (tls_ctx->partially_sent_record) {
+		tls_free_partial_record(sk, tls_ctx);
 		rec = list_first_entry(&ctx->tx_list,
 				       struct tls_rec, list);
 		list_del(&rec->list);