Received-SPF: pass (google.com: domain of linux-kernel+bounces-173118-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45d1:ec00::1 as permitted sender) client-ip=2604:1380:45d1:ec00::1;
From: Breno Leitao <leitao@debian.org>
To: kuniyu@amazon.com,
	"David S. Miller" <davem@davemloft.net>,
	Eric Dumazet <edumazet@google.com>,
	Jakub Kicinski <kuba@kernel.org>,
	Paolo Abeni <pabeni@redhat.com>
Cc: netdev@vger.kernel.org,
	horms@kernel.org,
	paulmck@kernel.org,
	David Howells <dhowells@redhat.com>,
	Alexander Mikhalitsyn <alexander@mihalicyn.com>,
	John Fastabend <john.fastabend@gmail.com>,
	Daan De Meyer <daan.j.demeyer@gmail.com>,
	linux-kernel@vger.kernel.org (open list)
Subject: [PATCH net-next] af_unix: Fix data races in unix_release_sock/unix_stream_sendmsg
Date: Wed,  8 May 2024 04:17:45 -0700
Message-ID: <20240508111749.2386649-1-leitao@debian.org>
Precedence: bulk
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit

A data-race condition has been identified in af_unix. In one data path,
the write function unix_release_sock() atomically writes to
sk->sk_shutdown using WRITE_ONCE. However, on the reader side,
unix_stream_sendmsg() does not read it atomically. Consequently, this
issue is causing the following KCSAN splat to occur:

	BUG: KCSAN: data-race in unix_release_sock / unix_stream_sendmsg

	write (marked) to 0xffff88867256ddbb of 1 bytes by task 7270 on cpu 28:
	unix_release_sock (net/unix/af_unix.c:640)
	unix_release (net/unix/af_unix.c:1050)
	sock_close (net/socket.c:659 net/socket.c:1421)
	__fput (fs/file_table.c:422)
	__fput_sync (fs/file_table.c:508)
	__se_sys_close (fs/open.c:1559 fs/open.c:1541)
	__x64_sys_close (fs/open.c:1541)
	x64_sys_call (arch/x86/entry/syscall_64.c:33)
	do_syscall_64 (arch/x86/entry/common.c:?)
	entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:130)

	read to 0xffff88867256ddbb of 1 bytes by task 989 on cpu 14:
	unix_stream_sendmsg (net/unix/af_unix.c:2273)
	__sock_sendmsg (net/socket.c:730 net/socket.c:745)
	____sys_sendmsg (net/socket.c:2584)
	__sys_sendmmsg (net/socket.c:2638 net/socket.c:2724)
	__x64_sys_sendmmsg (net/socket.c:2753 net/socket.c:2750 net/socket.c:2750)
	x64_sys_call (arch/x86/entry/syscall_64.c:33)
	do_syscall_64 (arch/x86/entry/common.c:?)
	entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:130)

	value changed: 0x01 -> 0x03

The line numbers are related to commit dd5a440a31fa ("Linux 6.9-rc7").

Commit e1d09c2c2f57 ("af_unix: Fix data races around sk->sk_shutdown.")
addressed a comparable issue in the past regarding sk->sk_shutdown.
However, it overlooked resolving this particular data path.

To prevent potential race conditions in the future, all read accesses to
sk->sk_shutdown in af_unix need be marked with READ_ONCE(). Although
there are additional reads in other->sk_shutdown without atomic reads,
I'm excluding them as I'm uncertain about their potential parallel
execution.

Signed-off-by: Breno Leitao <leitao@debian.org>
---
 net/unix/af_unix.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c
index 9a6ad5974dff..74795e6d13c6 100644
--- a/net/unix/af_unix.c
+++ b/net/unix/af_unix.c
@@ -2270,7 +2270,7 @@ static int unix_stream_sendmsg(struct socket *sock, struct msghdr *msg,
 			goto out_err;
 	}
 
-	if (sk->sk_shutdown & SEND_SHUTDOWN)
+	if (READ_ONCE(sk->sk_shutdown) & SEND_SHUTDOWN)
 		goto pipe_err;
 
 	while (sent < len) {
@@ -2446,7 +2446,7 @@ int __unix_dgram_recvmsg(struct sock *sk, struct msghdr *msg, size_t size,
 		unix_state_lock(sk);
 		/* Signal EOF on disconnected non-blocking SEQPACKET socket. */
 		if (sk->sk_type == SOCK_SEQPACKET && err == -EAGAIN &&
-		    (sk->sk_shutdown & RCV_SHUTDOWN))
+		    (READ_ONCE(sk->sk_shutdown) & RCV_SHUTDOWN))
 			err = 0;
 		unix_state_unlock(sk);
 		goto out;
@@ -2566,7 +2566,7 @@ static long unix_stream_data_wait(struct sock *sk, long timeo,
 		if (tail != last ||
 		    (tail && tail->len != last_len) ||
 		    sk->sk_err ||
-		    (sk->sk_shutdown & RCV_SHUTDOWN) ||
+		    (READ_ONCE(sk->sk_shutdown) & RCV_SHUTDOWN) ||
 		    signal_pending(current) ||
 		    !timeo)
 			break;
@@ -2764,7 +2764,7 @@ static int unix_stream_read_generic(struct unix_stream_read_state *state,
 			err = sock_error(sk);
 			if (err)
 				goto unlock;
-			if (sk->sk_shutdown & RCV_SHUTDOWN)
+			if (READ_ONCE(sk->sk_shutdown) & RCV_SHUTDOWN)
 				goto unlock;
 
 			unix_state_unlock(sk);
-- 
2.43.0