summaryrefslogtreecommitdiffstats
path: root/net/ipv4/tcp_ulp.c
diff options
context:
space:
mode:
authorJakub Sitnicki <jakub@cloudflare.com>2020-02-18 18:10:13 +0100
committerDaniel Borkmann <daniel@iogearbox.net>2020-02-21 22:29:45 +0100
commitb8e202d1d1d0f182f01062804efb523ea9a9008c (patch)
tree57a1c1d075e2d470fd9c1487744286c2f3bed747 /net/ipv4/tcp_ulp.c
parentdocs/bpf: Update bpf development Q/A file (diff)
downloadlinux-b8e202d1d1d0f182f01062804efb523ea9a9008c.tar.xz
linux-b8e202d1d1d0f182f01062804efb523ea9a9008c.zip
net, sk_msg: Annotate lockless access to sk_prot on clone
sk_msg and ULP frameworks override protocol callbacks pointer in sk->sk_prot, while tcp accesses it locklessly when cloning the listening socket, that is with neither sk_lock nor sk_callback_lock held. Once we enable use of listening sockets with sockmap (and hence sk_msg), there will be shared access to sk->sk_prot if socket is getting cloned while being inserted/deleted to/from the sockmap from another CPU: Read side: tcp_v4_rcv sk = __inet_lookup_skb(...) tcp_check_req(sk) inet_csk(sk)->icsk_af_ops->syn_recv_sock tcp_v4_syn_recv_sock tcp_create_openreq_child inet_csk_clone_lock sk_clone_lock READ_ONCE(sk->sk_prot) Write side: sock_map_ops->map_update_elem sock_map_update_elem sock_map_update_common sock_map_link_no_progs tcp_bpf_init tcp_bpf_update_sk_prot sk_psock_update_proto WRITE_ONCE(sk->sk_prot, ops) sock_map_ops->map_delete_elem sock_map_delete_elem __sock_map_delete sock_map_unref sk_psock_put sk_psock_drop sk_psock_restore_proto tcp_update_ulp WRITE_ONCE(sk->sk_prot, proto) Mark the shared access with READ_ONCE/WRITE_ONCE annotations. Signed-off-by: Jakub Sitnicki <jakub@cloudflare.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/20200218171023.844439-2-jakub@cloudflare.com
Diffstat (limited to 'net/ipv4/tcp_ulp.c')
-rw-r--r--net/ipv4/tcp_ulp.c3
1 files changed, 2 insertions, 1 deletions
diff --git a/net/ipv4/tcp_ulp.c b/net/ipv4/tcp_ulp.c
index 38d3ad141161..6c43fa189195 100644
--- a/net/ipv4/tcp_ulp.c
+++ b/net/ipv4/tcp_ulp.c
@@ -106,7 +106,8 @@ void tcp_update_ulp(struct sock *sk, struct proto *proto,
if (!icsk->icsk_ulp_ops) {
sk->sk_write_space = write_space;
- sk->sk_prot = proto;
+ /* Pairs with lockless read in sk_clone_lock() */
+ WRITE_ONCE(sk->sk_prot, proto);
return;
}