From: Eric Dumazet <edumazet@google.com> mainline inclusion from mainline-v7.0-rc3 commit 2ef2b20cf4e04ac8a6ba68493f8780776ff84300 category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/13979 CVE: CVE-2026-23302 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i... -------------------------------- skmsg (and probably other layers) are changing these pointers while other cpus might read them concurrently. Add corresponding READ_ONCE()/WRITE_ONCE() annotations for UDP, TCP and AF_UNIX. Fixes: 604326b41a6f ("bpf, sockmap: convert to generic sk_msg interface") Reported-by: syzbot+87f770387a9e5dc6b79b@syzkaller.appspotmail.com Closes: https://lore.kernel.org/netdev/699ee9fc.050a0220.1cd54b.0009.GAE@google.com/ Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: John Fastabend <john.fastabend@gmail.com> Cc: Jakub Sitnicki <jakub@cloudflare.com> Cc: Willem de Bruijn <willemdebruijn.kernel@gmail.com> Reviewed-by: Kuniyuki Iwashima <kuniyu@google.com> Link: https://patch.msgid.link/20260225131547.1085509-1-edumazet@google.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Conflicts: net/ipv4/udp.c net/ipv4/tcp.c net/ipv4/tcp_input.c net/ipv4/tcp_minisocks.c net/unix/af_unix.c [conflicts due to not merge bd61848900bf ("net: devmem: Implement TX path") and 95b9a87c6a6b ("tcp: record last received ipv6 flowlabel") and f0db2bca0cf9 ("tcp: rework {__,}tcp_ecn_check_ce() -> tcp_data_ecn_check()") and 2bd99aef1b19 ("tcp: accept bare FIN packets under memory pressure") and b98256959305 ("tcp: make the dropreason really work when calling tcp_rcv_state_process()") and 85cb0757d7e1 ("net: Convert proto_ops connect() callbacks to use sockaddr_unsized") and f4e1fb04c123 ("af_unix: Use cached value for SOCK_STREAM in unix_inq_len().") and b650bf0977d3 ("udp: remove busylock and add per NUMA queues").] Signed-off-by: Li Xiasong <lixiasong1@huawei.com> --- net/core/skmsg.c | 14 +++++++------- net/ipv4/tcp.c | 4 ++-- net/ipv4/tcp_bpf.c | 2 +- net/ipv4/tcp_input.c | 14 ++++++++------ net/ipv4/tcp_minisocks.c | 2 +- net/ipv4/udp.c | 3 ++- net/ipv4/udp_bpf.c | 2 +- net/unix/af_unix.c | 8 ++++---- 8 files changed, 26 insertions(+), 23 deletions(-) diff --git a/net/core/skmsg.c b/net/core/skmsg.c index 3429c6e2ce21..e3696d510eec 100644 --- a/net/core/skmsg.c +++ b/net/core/skmsg.c @@ -1174,8 +1174,8 @@ void sk_psock_start_strp(struct sock *sk, struct sk_psock *psock) return; psock->saved_data_ready = sk->sk_data_ready; - sk->sk_data_ready = sk_psock_strp_data_ready; - sk->sk_write_space = sk_psock_write_space; + WRITE_ONCE(sk->sk_data_ready, sk_psock_strp_data_ready); + WRITE_ONCE(sk->sk_write_space, sk_psock_write_space); } void sk_psock_stop_strp(struct sock *sk, struct sk_psock *psock) @@ -1185,8 +1185,8 @@ void sk_psock_stop_strp(struct sock *sk, struct sk_psock *psock) if (!psock->saved_data_ready) return; - sk->sk_data_ready = psock->saved_data_ready; - psock->saved_data_ready = NULL; + WRITE_ONCE(sk->sk_data_ready, psock->saved_data_ready); + WRITE_ONCE(psock->saved_data_ready, NULL); strp_stop(&psock->strp); } @@ -1265,8 +1265,8 @@ void sk_psock_start_verdict(struct sock *sk, struct sk_psock *psock) return; psock->saved_data_ready = sk->sk_data_ready; - sk->sk_data_ready = sk_psock_verdict_data_ready; - sk->sk_write_space = sk_psock_write_space; + WRITE_ONCE(sk->sk_data_ready, sk_psock_verdict_data_ready); + WRITE_ONCE(sk->sk_write_space, sk_psock_write_space); } void sk_psock_stop_verdict(struct sock *sk, struct sk_psock *psock) @@ -1277,6 +1277,6 @@ void sk_psock_stop_verdict(struct sock *sk, struct sk_psock *psock) if (!psock->saved_data_ready) return; - sk->sk_data_ready = psock->saved_data_ready; + WRITE_ONCE(sk->sk_data_ready, psock->saved_data_ready); psock->saved_data_ready = NULL; } diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index bf23aa827031..1fe89e742f75 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -1342,7 +1342,7 @@ int tcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t size) err = sk_stream_error(sk, flags, err); /* make sure we wake any epoll edge trigger waiter */ if (unlikely(tcp_rtx_and_write_queues_empty(sk) && err == -EAGAIN)) { - sk->sk_write_space(sk); + READ_ONCE(sk->sk_write_space)(sk); tcp_chrono_stop(sk, TCP_CHRONO_SNDBUF_LIMITED); } return err; @@ -3705,7 +3705,7 @@ int do_tcp_setsockopt(struct sock *sk, int level, int optname, break; case TCP_NOTSENT_LOWAT: WRITE_ONCE(tp->notsent_lowat, val); - sk->sk_write_space(sk); + READ_ONCE(sk->sk_write_space)(sk); break; case TCP_INQ: if (val > 1 || val < 0) diff --git a/net/ipv4/tcp_bpf.c b/net/ipv4/tcp_bpf.c index 0e56f3baf98a..9ceb6e0c8b00 100644 --- a/net/ipv4/tcp_bpf.c +++ b/net/ipv4/tcp_bpf.c @@ -668,7 +668,7 @@ int tcp_bpf_update_proto(struct sock *sk, struct sk_psock *psock, bool restore) WRITE_ONCE(sk->sk_prot->unhash, psock->saved_unhash); tcp_update_ulp(sk, psock->sk_proto, psock->saved_write_space); } else { - sk->sk_write_space = psock->saved_write_space; + WRITE_ONCE(sk->sk_write_space, psock->saved_write_space); /* Pairs with lockless read in sk_clone_lock() */ sock_replace_proto(sk, psock->sk_proto); } diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 8d6757ec06aa..a10e7da1a612 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -4985,7 +4985,7 @@ static void tcp_data_queue_ofo(struct sock *sk, struct sk_buff *skb) if (unlikely(tcp_try_rmem_schedule(sk, skb, skb->truesize))) { NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPOFODROP); - sk->sk_data_ready(sk); + READ_ONCE(sk->sk_data_ready)(sk); tcp_drop_reason(sk, skb, SKB_DROP_REASON_PROTO_MEM); return; } @@ -5192,7 +5192,7 @@ int tcp_send_rcvq(struct sock *sk, struct msghdr *msg, size_t size) void tcp_data_ready(struct sock *sk) { if (tcp_epollin_ready(sk, sk->sk_rcvlowat) || sock_flag(sk, SOCK_DONE)) - sk->sk_data_ready(sk); + READ_ONCE(sk->sk_data_ready)(sk); } static void tcp_data_queue(struct sock *sk, struct sk_buff *skb) @@ -5238,7 +5238,7 @@ static void tcp_data_queue(struct sock *sk, struct sk_buff *skb) inet_csk(sk)->icsk_ack.pending |= (ICSK_ACK_NOMEM | ICSK_ACK_NOW); inet_csk_schedule_ack(sk); - sk->sk_data_ready(sk); + READ_ONCE(sk->sk_data_ready)(sk); if (skb_queue_len(&sk->sk_receive_queue)) { reason = SKB_DROP_REASON_PROTO_MEM; @@ -5675,7 +5675,9 @@ static void tcp_new_space(struct sock *sk) tp->snd_cwnd_stamp = tcp_jiffies32; } - INDIRECT_CALL_1(sk->sk_write_space, sk_stream_write_space, sk); + INDIRECT_CALL_1(READ_ONCE(sk->sk_write_space), + sk_stream_write_space, + sk); } /* Caller made space either from: @@ -5873,7 +5875,7 @@ static void tcp_urg(struct sock *sk, struct sk_buff *skb, const struct tcphdr *t BUG(); WRITE_ONCE(tp->urg_data, TCP_URG_VALID | tmp); if (!sock_flag(sk, SOCK_DEAD)) - sk->sk_data_ready(sk); + READ_ONCE(sk->sk_data_ready)(sk); } } } @@ -7279,7 +7281,7 @@ int tcp_conn_request(struct request_sock_ops *rsk_ops, sock_put(fastopen_sk); goto drop_and_free; } - sk->sk_data_ready(sk); + READ_ONCE(sk->sk_data_ready)(sk); bh_unlock_sock(fastopen_sk); sock_put(fastopen_sk); } else { diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c index 770b1ffb2e22..8d72532143bf 100644 --- a/net/ipv4/tcp_minisocks.c +++ b/net/ipv4/tcp_minisocks.c @@ -900,7 +900,7 @@ int tcp_child_process(struct sock *parent, struct sock *child, ret = tcp_rcv_state_process(child, skb); /* Wakeup parent, send SIGIO */ if (state == TCP_SYN_RECV && child->sk_state != state) - parent->sk_data_ready(parent); + READ_ONCE(parent->sk_data_ready)(parent); } else { /* Alas, it is possible again, because we do lookup * in main socket hash table and lock on listening diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index 4c1f82f6777f..40100fed5c67 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -1618,7 +1618,8 @@ int __udp_enqueue_schedule_skb(struct sock *sk, struct sk_buff *skb) spin_unlock(&list->lock); if (!sock_flag(sk, SOCK_DEAD)) - INDIRECT_CALL_1(sk->sk_data_ready, sock_def_readable, sk); + INDIRECT_CALL_1(READ_ONCE(sk->sk_data_ready), + sock_def_readable, sk); busylock_release(busy); return 0; diff --git a/net/ipv4/udp_bpf.c b/net/ipv4/udp_bpf.c index 0735d820e413..44271ba1adec 100644 --- a/net/ipv4/udp_bpf.c +++ b/net/ipv4/udp_bpf.c @@ -143,7 +143,7 @@ int udp_bpf_update_proto(struct sock *sk, struct sk_psock *psock, bool restore) int family = sk->sk_family == AF_INET ? UDP_BPF_IPV4 : UDP_BPF_IPV6; if (restore) { - sk->sk_write_space = psock->saved_write_space; + WRITE_ONCE(sk->sk_write_space, psock->saved_write_space); sock_replace_proto(sk, psock->sk_proto); return 0; } diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c index abf9fe35e620..51f444ecebf6 100644 --- a/net/unix/af_unix.c +++ b/net/unix/af_unix.c @@ -1678,7 +1678,7 @@ static int unix_stream_connect(struct socket *sock, struct sockaddr *uaddr, __skb_queue_tail(&other->sk_receive_queue, skb); spin_unlock(&other->sk_receive_queue.lock); unix_state_unlock(other); - other->sk_data_ready(other); + READ_ONCE(other->sk_data_ready)(other); sock_put(other); return 0; @@ -2138,7 +2138,7 @@ static int unix_dgram_sendmsg(struct socket *sock, struct msghdr *msg, scm_stat_add(other, skb); skb_queue_tail(&other->sk_receive_queue, skb); unix_state_unlock(other); - other->sk_data_ready(other); + READ_ONCE(other->sk_data_ready)(other); sock_put(other); scm_destroy(&scm); return len; @@ -2206,7 +2206,7 @@ static int queue_oob(struct socket *sock, struct msghdr *msg, struct sock *other sk_send_sigurg(other); unix_state_unlock(other); - other->sk_data_ready(other); + READ_ONCE(other->sk_data_ready)(other); return err; } @@ -2317,7 +2317,7 @@ static int unix_stream_sendmsg(struct socket *sock, struct msghdr *msg, scm_stat_add(other, skb); skb_queue_tail(&other->sk_receive_queue, skb); unix_state_unlock(other); - other->sk_data_ready(other); + READ_ONCE(other->sk_data_ready)(other); sent += size; } -- 2.34.1