net/packet: remove locking from packet_rcv_has_room()

__packet_rcv_has_room() can now be run without lock being held.

po->pressure is only a non persistent hint, we can mark
all read/write accesses with READ_ONCE()/WRITE_ONCE()
to document the fact that the field could be written
without any synchronization.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c
index 860ca3e6abf5198214612e9acc095530b61dac40..d409e2fdaa7ee8ddf261354f91b682e403f40e9e 100644 (file)
--- a/net/packet/af_packet.c
+++ b/net/packet/af_packet.c
@@ -1260,15 +1260,13 @@ static int __packet_rcv_has_room(const struct packet_sock *po,
 
 static int packet_rcv_has_room(struct packet_sock *po, struct sk_buff *skb)
 {
-       int ret;
-       bool has_room;
+       int pressure, ret;
 
-       spin_lock_bh(&po->sk.sk_receive_queue.lock);
        ret = __packet_rcv_has_room(po, skb);
-       has_room = ret == ROOM_NORMAL;
-       if (po->pressure == has_room)
-               po->pressure = !has_room;
-       spin_unlock_bh(&po->sk.sk_receive_queue.lock);
+       pressure = ret != ROOM_NORMAL;
+
+       if (READ_ONCE(po->pressure) != pressure)
+               WRITE_ONCE(po->pressure, pressure);
 
        return ret;
 }
@@ -1353,7 +1351,7 @@ static unsigned int fanout_demux_rollover(struct packet_fanout *f,
        i = j = min_t(int, po->rollover->sock, num - 1);
        do {
                po_next = pkt_sk(f->arr[i]);
-               if (po_next != po_skip && !po_next->pressure &&
+               if (po_next != po_skip && !READ_ONCE(po_next->pressure) &&
                    packet_rcv_has_room(po_next, skb) == ROOM_NORMAL) {
                        if (i != j)
                                po->rollover->sock = i;
@@ -3310,7 +3308,7 @@ static int packet_recvmsg(struct socket *sock, struct msghdr *msg, size_t len,
        if (skb == NULL)
                goto out;
 
-       if (pkt_sk(sk)->pressure)
+       if (READ_ONCE(pkt_sk(sk)->pressure))
                packet_rcv_has_room(pkt_sk(sk), NULL);
 
        if (pkt_sk(sk)->has_vnet_hdr) {
@@ -4129,8 +4127,8 @@ static __poll_t packet_poll(struct file *file, struct socket *sock,
                        TP_STATUS_KERNEL))
                        mask |= EPOLLIN | EPOLLRDNORM;
        }
-       if (po->pressure && __packet_rcv_has_room(po, NULL) == ROOM_NORMAL)
-               po->pressure = 0;
+       if (READ_ONCE(po->pressure) && __packet_rcv_has_room(po, NULL) == ROOM_NORMAL)
+               WRITE_ONCE(po->pressure, 0);
        spin_unlock_bh(&sk->sk_receive_queue.lock);
        spin_lock_bh(&sk->sk_write_queue.lock);
        if (po->tx_ring.pg_vec) {