tcp: allow again tcp_disconnect() when threads are waiting

As reported by Tom, .NET and applications build on top of it rely
on connect(AF_UNSPEC) to async cancel pending I/O operations on TCP
socket.

The blamed commit below caused a regression, as such cancellation
can now fail.

As suggested by Eric, this change addresses the problem explicitly
causing blocking I/O operation to terminate immediately (with an error)
when a concurrent disconnect() is executed.

Instead of tracking the number of threads blocked on a given socket,
track the number of disconnect() issued on such socket. If such counter
changes after a blocking operation releasing and re-acquiring the socket
lock, error out the current operation.

Fixes: 4faeee0cf8a5 ("tcp: deny tcp_disconnect() when threads are waiting")
Reported-by: Tom Deseyn <tdeseyn@redhat.com>
Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1886305
Suggested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/f3b95e47e3dbed840960548aebaa8d954372db41.1697008693.git.pabeni@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

2 files changed, 20 insertions, 9 deletions

diff --git a/net/tls/tls_main.c b/net/tls/tls_main.c
index 02f583ff9239..002483e60c19 100644
--- a/net/tls/tls_main.c
+++ b/net/tls/tls_main.c
@@ -139,8 +139,8 @@ void update_sk_prot(struct sock *sk, struct tls_context *ctx)
 
 int wait_on_pending_writer(struct sock *sk, long *timeo)
 {
-	int rc = 0;
 	DEFINE_WAIT_FUNC(wait, woken_wake_function);
+	int ret, rc = 0;
 
 	add_wait_queue(sk_sleep(sk), &wait);
 	while (1) {
@@ -154,9 +154,13 @@ int wait_on_pending_writer(struct sock *sk, long *timeo)
 			break;
 		}
 
-		if (sk_wait_event(sk, timeo,
-				  !READ_ONCE(sk->sk_write_pending), &wait))
+		ret = sk_wait_event(sk, timeo,
+				    !READ_ONCE(sk->sk_write_pending), &wait);
+		if (ret) {
+			if (ret < 0)
+				rc = ret;
 			break;
+		}
 	}
 	remove_wait_queue(sk_sleep(sk), &wait);
 	return rc;
diff --git a/net/tls/tls_sw.c b/net/tls/tls_sw.c
index d1fc295b83b5..e9d1e83a859d 100644
--- a/net/tls/tls_sw.c
+++ b/net/tls/tls_sw.c
@@ -1291,6 +1291,7 @@ tls_rx_rec_wait(struct sock *sk, struct sk_psock *psock, bool nonblock,
 	struct tls_context *tls_ctx = tls_get_ctx(sk);
 	struct tls_sw_context_rx *ctx = tls_sw_ctx_rx(tls_ctx);
 	DEFINE_WAIT_FUNC(wait, woken_wake_function);
+	int ret = 0;
 	long timeo;
 
 	timeo = sock_rcvtimeo(sk, nonblock);
@@ -1302,6 +1303,9 @@ tls_rx_rec_wait(struct sock *sk, struct sk_psock *psock, bool nonblock,
 		if (sk->sk_err)
 			return sock_error(sk);
 
+		if (ret < 0)
+			return ret;
+
 		if (!skb_queue_empty(&sk->sk_receive_queue)) {
 			tls_strp_check_rcv(&ctx->strp);
 			if (tls_strp_msg_ready(ctx))
@@ -1320,10 +1324,10 @@ tls_rx_rec_wait(struct sock *sk, struct sk_psock *psock, bool nonblock,
 		released = true;
 		add_wait_queue(sk_sleep(sk), &wait);
 		sk_set_bit(SOCKWQ_ASYNC_WAITDATA, sk);
-		sk_wait_event(sk, &timeo,
-			      tls_strp_msg_ready(ctx) ||
-			      !sk_psock_queue_empty(psock),
-			      &wait);
+		ret = sk_wait_event(sk, &timeo,
+				    tls_strp_msg_ready(ctx) ||
+				    !sk_psock_queue_empty(psock),
+				    &wait);
 		sk_clear_bit(SOCKWQ_ASYNC_WAITDATA, sk);
 		remove_wait_queue(sk_sleep(sk), &wait);
 
@@ -1852,6 +1856,7 @@ static int tls_rx_reader_acquire(struct sock *sk, struct tls_sw_context_rx *ctx,
 				 bool nonblock)
 {
 	long timeo;
+	int ret;
 
 	timeo = sock_rcvtimeo(sk, nonblock);
 
@@ -1861,14 +1866,16 @@ static int tls_rx_reader_acquire(struct sock *sk, struct tls_sw_context_rx *ctx,
 		ctx->reader_contended = 1;
 
 		add_wait_queue(&ctx->wq, &wait);
-		sk_wait_event(sk, &timeo,
-			      !READ_ONCE(ctx->reader_present), &wait);
+		ret = sk_wait_event(sk, &timeo,
+				    !READ_ONCE(ctx->reader_present), &wait);
 		remove_wait_queue(&ctx->wq, &wait);
 
 		if (timeo <= 0)
 			return -EAGAIN;
 		if (signal_pending(current))
 			return sock_intr_errno(timeo);
+		if (ret < 0)
+			return ret;
 	}
 
 	WRITE_ONCE(ctx->reader_present, 1);