net: percpu_counter_inc() should not be called in BH-disabled section

Based upon a lockdep report by Alexey Dobriyan.

I checked all per_cpu_counter_xxx() usages in network tree, and I
think all call sites are BH enabled except one in
inet_csk_listen_stop().

commit dd24c00191d5e4a1ae896aafe33c6b8095ab4bd1
(net: Use a percpu_counter for orphan_count)
replaced atomic_t orphan_count to a percpu_counter.

atomic_inc()/atomic_dec() can be called from any context, while
percpu_counter_xxx() should be called from a consistent state.

For orphan_count, this context can be the BH-enabled one.

Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c
index 1ccdbba..fe32255 100644 (file)
--- a/net/ipv4/inet_connection_sock.c
+++ b/net/ipv4/inet_connection_sock.c
@@ -632,6 +632,8 @@ void inet_csk_listen_stop(struct sock *sk)
 
                acc_req = req->dl_next;
 
+               percpu_counter_inc(sk->sk_prot->orphan_count);
+
                local_bh_disable();
                bh_lock_sock(child);
                WARN_ON(sock_owned_by_user(child));
@@ -641,8 +643,6 @@ void inet_csk_listen_stop(struct sock *sk)
 
                sock_orphan(child);
 
-               percpu_counter_inc(sk->sk_prot->orphan_count);
-
                inet_csk_destroy_sock(child);
 
                bh_unlock_sock(child);