net: implement lockless SO_PRIORITY

This is a followup of 8bf43be799d4 ("net: annotate data-races
around sk->sk_priority").

sk->sk_priority can be read and written without holding the socket lock.

Signed-off-by: Eric Dumazet <[e-mail]>
Reviewed-by: Wenjia Zhang <[e-mail]>
Signed-off-by: David S. Miller <[e-mail]>

drivers/net/ppp/pppoe.c

···
 
 	skb->dev = dev;
 
-	skb->priority = sk->sk_priority;
+	skb->priority = READ_ONCE(sk->sk_priority);
 	skb->protocol = cpu_to_be16(ETH_P_PPP_SES);
 
 	ph = skb_put(skb, total_len + sizeof(struct pppoe_hdr));

include/net/bluetooth/bluetooth.h

···
 		return ERR_PTR(-EFAULT);
 	}
 
-	skb->priority = sk->sk_priority;
+	skb->priority = READ_ONCE(sk->sk_priority);
 
 	return skb;
 }

net/appletalk/aarp.c

···
 
 sendit:
 	if (skb->sk)
-		skb->priority = skb->sk->sk_priority;
+		skb->priority = READ_ONCE(skb->sk->sk_priority);
 	if (dev_queue_xmit(skb))
 		goto drop;
 sent:

net/ax25/af_ax25.c

···
 	sock_init_data(NULL, sk);
 
 	sk->sk_type     = osk->sk_type;
-	sk->sk_priority = osk->sk_priority;
+	sk->sk_priority = READ_ONCE(osk->sk_priority);
 	sk->sk_protocol = osk->sk_protocol;
 	sk->sk_rcvbuf   = osk->sk_rcvbuf;
 	sk->sk_sndbuf   = osk->sk_sndbuf;

net/bluetooth/l2cap_sock.c

···
 		return ERR_PTR(-ENOTCONN);
 	}
 
-	skb->priority = sk->sk_priority;
+	skb->priority = READ_ONCE(sk->sk_priority);
 
 	bt_cb(skb)->l2cap.chan = chan;
 

net/can/j1939/socket.c

···
 	skcb = j1939_skb_to_cb(skb);
 	memset(skcb, 0, sizeof(*skcb));
 	skcb->addr = jsk->addr;
-	skcb->priority = j1939_prio(sk->sk_priority);
+	skcb->priority = j1939_prio(READ_ONCE(sk->sk_priority));
 
 	if (msg->msg_name) {
 		struct sockaddr_can *addr = msg->msg_name;

net/can/raw.c

···
 	}
 
 	skb->dev = dev;
-	skb->priority = sk->sk_priority;
+	skb->priority = READ_ONCE(sk->sk_priority);
 	skb->mark = READ_ONCE(sk->sk_mark);
 	skb->tstamp = sockc.transmit_time;
 

net/core/sock.c

···
 
 void sock_set_priority(struct sock *sk, u32 priority)
 {
-	lock_sock(sk);
 	WRITE_ONCE(sk->sk_priority, priority);
-	release_sock(sk);
 }
 EXPORT_SYMBOL(sock_set_priority);
 
···
 
 	valbool = val ? 1 : 0;
 
+	/* handle options which do not require locking the socket. */
+	switch (optname) {
+	case SO_PRIORITY:
+		if ((val >= 0 && val <= 6) ||
+		    sockopt_ns_capable(sock_net(sk)->user_ns, CAP_NET_RAW) ||
+		    sockopt_ns_capable(sock_net(sk)->user_ns, CAP_NET_ADMIN)) {
+			sock_set_priority(sk, val);
+			return 0;
+		}
+		return -EPERM;
+	}
+
 	sockopt_lock_sock(sk);
 
 	switch (optname) {
···
 
 	case SO_NO_CHECK:
 		sk->sk_no_check_tx = valbool;
-		break;
-
-	case SO_PRIORITY:
-		if ((val >= 0 && val <= 6) ||
-		    sockopt_ns_capable(sock_net(sk)->user_ns, CAP_NET_RAW) ||
-		    sockopt_ns_capable(sock_net(sk)->user_ns, CAP_NET_ADMIN))
-			WRITE_ONCE(sk->sk_priority, val);
-		else
-			ret = -EPERM;
 		break;
 
 	case SO_LINGER:

net/dccp/ipv6.c

···
 		if (!opt)
 			opt = rcu_dereference(np->opt);
 		err = ip6_xmit(sk, skb, &fl6, READ_ONCE(sk->sk_mark), opt,
-			       np->tclass, sk->sk_priority);
+			       np->tclass, READ_ONCE(sk->sk_priority));
 		rcu_read_unlock();
 		err = net_xmit_eval(err);
 	}

net/ipv4/inet_diag.c

···
 		 * For cgroup2 classid is always zero.
 		 */
 		if (!classid)
-			classid = sk->sk_priority;
+			classid = READ_ONCE(sk->sk_priority);
 
 		if (nla_put_u32(skb, INET_DIAG_CLASS_ID, classid))
 			goto errout;

net/ipv4/ip_output.c

···
 		ip_options_build(skb, opt, cork->addr, rt);
 	}
 
-	skb->priority = (cork->tos != -1) ? cork->priority: sk->sk_priority;
+	skb->priority = (cork->tos != -1) ? cork->priority: READ_ONCE(sk->sk_priority);
 	skb->mark = cork->mark;
 	skb->tstamp = cork->transmit_time;
 	/*

net/ipv4/tcp_ipv4.c

···
 		ctl_sk->sk_mark = (sk->sk_state == TCP_TIME_WAIT) ?
 				   inet_twsk(sk)->tw_mark : sk->sk_mark;
 		ctl_sk->sk_priority = (sk->sk_state == TCP_TIME_WAIT) ?
-				   inet_twsk(sk)->tw_priority : sk->sk_priority;
+				   inet_twsk(sk)->tw_priority : READ_ONCE(sk->sk_priority);
 		transmit_time = tcp_transmit_time(sk);
 		xfrm_sk_clone_policy(ctl_sk, sk);
 		txhash = (sk->sk_state == TCP_TIME_WAIT) ?

net/ipv4/tcp_minisocks.c

···
 
 		tw->tw_transparent	= inet_test_bit(TRANSPARENT, sk);
 		tw->tw_mark		= sk->sk_mark;
-		tw->tw_priority		= sk->sk_priority;
+		tw->tw_priority		= READ_ONCE(sk->sk_priority);
 		tw->tw_rcv_wscale	= tp->rx_opt.rcv_wscale;
 		tcptw->tw_rcv_nxt	= tp->rcv_nxt;
 		tcptw->tw_snd_nxt	= tp->snd_nxt;

net/ipv6/inet6_connection_sock.c

···
 	fl6.daddr = sk->sk_v6_daddr;
 
 	res = ip6_xmit(sk, skb, &fl6, sk->sk_mark, rcu_dereference(np->opt),
-		       np->tclass,  sk->sk_priority);
+		       np->tclass, READ_ONCE(sk->sk_priority));
 	rcu_read_unlock();
 	return res;
 }

net/ipv6/ip6_output.c

···
 	hdr->saddr = fl6->saddr;
 	hdr->daddr = *final_dst;
 
-	skb->priority = sk->sk_priority;
+	skb->priority = READ_ONCE(sk->sk_priority);
 	skb->mark = cork->base.mark;
 	skb->tstamp = cork->base.transmit_time;
 

net/ipv6/tcp_ipv6.c

···
 		if (!opt)
 			opt = rcu_dereference(np->opt);
 		err = ip6_xmit(sk, skb, fl6, skb->mark ? : READ_ONCE(sk->sk_mark),
-			       opt, tclass, sk->sk_priority);
+			       opt, tclass, READ_ONCE(sk->sk_priority));
 		rcu_read_unlock();
 		err = net_xmit_eval(err);
 	}
···
 			trace_tcp_send_reset(sk, skb);
 			if (inet6_test_bit(REPFLOW, sk))
 				label = ip6_flowlabel(ipv6h);
-			priority = sk->sk_priority;
+			priority = READ_ONCE(sk->sk_priority);
 			txhash = sk->sk_txhash;
 		}
 		if (sk->sk_state == TCP_TIME_WAIT) {

net/mptcp/sockopt.c

···
 			sock_valbool_flag(ssk, SOCK_KEEPOPEN, !!val);
 			break;
 		case SO_PRIORITY:
-			ssk->sk_priority = val;
+			WRITE_ONCE(ssk->sk_priority, val);
 			break;
 		case SO_SNDBUF:
 		case SO_SNDBUFFORCE:

net/netrom/af_netrom.c

···
 	sock_init_data(NULL, sk);
 
 	sk->sk_type     = osk->sk_type;
-	sk->sk_priority = osk->sk_priority;
+	sk->sk_priority = READ_ONCE(osk->sk_priority);
 	sk->sk_protocol = osk->sk_protocol;
 	sk->sk_rcvbuf   = osk->sk_rcvbuf;
 	sk->sk_sndbuf   = osk->sk_sndbuf;

net/rose/af_rose.c

···
 #endif
 
 	sk->sk_type     = osk->sk_type;
-	sk->sk_priority = osk->sk_priority;
+	sk->sk_priority = READ_ONCE(osk->sk_priority);
 	sk->sk_protocol = osk->sk_protocol;
 	sk->sk_rcvbuf   = osk->sk_rcvbuf;
 	sk->sk_sndbuf   = osk->sk_sndbuf;

net/sched/em_meta.c

···
 		*err = -1;
 		return;
 	}
-	dst->value = sk->sk_priority;
+	dst->value = READ_ONCE(sk->sk_priority);
 }
 
 META_COLLECTOR(int_sk_rcvlowat)

net/sctp/ipv6.c

···
 		rcu_read_lock();
 		res = ip6_xmit(sk, skb, fl6, sk->sk_mark,
 			       rcu_dereference(np->opt),
-			       tclass, sk->sk_priority);
+			       tclass, READ_ONCE(sk->sk_priority));
 		rcu_read_unlock();
 		return res;
 	}

net/smc/af_smc.c

···
 	nsk->sk_sndtimeo = osk->sk_sndtimeo;
 	nsk->sk_rcvtimeo = osk->sk_rcvtimeo;
 	nsk->sk_mark = READ_ONCE(osk->sk_mark);
-	nsk->sk_priority = osk->sk_priority;
+	nsk->sk_priority = READ_ONCE(osk->sk_priority);
 	nsk->sk_rcvlowat = osk->sk_rcvlowat;
 	nsk->sk_bound_dev_if = osk->sk_bound_dev_if;
 	nsk->sk_err = osk->sk_err;

net/x25/af_x25.c

···
 	x25 = x25_sk(sk);
 
 	sk->sk_type        = osk->sk_type;
-	sk->sk_priority    = osk->sk_priority;
+	sk->sk_priority    = READ_ONCE(osk->sk_priority);
 	sk->sk_protocol    = osk->sk_protocol;
 	sk->sk_rcvbuf      = osk->sk_rcvbuf;
 	sk->sk_sndbuf      = osk->sk_sndbuf;

net/xdp/xsk.c

···
 	}
 
 	skb->dev = dev;
-	skb->priority = xs->sk.sk_priority;
+	skb->priority = READ_ONCE(xs->sk.sk_priority);
 	skb->mark = READ_ONCE(xs->sk.sk_mark);
 	skb->destructor = xsk_destruct_skb;
 	xsk_set_destructor_arg(skb);