Merge branch 'vsock-sock_linger-rework'

Michal Luczaj says:

====================
vsock: SOCK_LINGER rework

Change vsock's lingerning to wait on close() until all data is sent, i.e.
until workers picked all the packets for processing.

v5: https://lore.kernel.org/r/20250521-vsock-linger-v5-0-94827860d1d6@rbox.co
v4: https://lore.kernel.org/r/20250501-vsock-linger-v4-0-beabbd8a0847@rbox.co
v3: https://lore.kernel.org/r/20250430-vsock-linger-v3-0-ddbe73b53457@rbox.co
v2: https://lore.kernel.org/r/20250421-vsock-linger-v2-0-fe9febd64668@rbox.co
v1: https://lore.kernel.org/r/20250407-vsock-linger-v1-0-1458038e3492@rbox.co

Signed-off-by: Michal Luczaj <mhal@rbox.co>
====================

Link: https://patch.msgid.link/20250522-vsock-linger-v6-0-2ad00b0e447e@rbox.co
Signed-off-by: Paolo Abeni <pabeni@redhat.com>

6 files changed, 134 insertions, 44 deletions

diff --git a/include/net/af_vsock.h b/include/net/af_vsock.h
index 9e85424c8343..d56e6e135158 100644
--- a/include/net/af_vsock.h
+++ b/include/net/af_vsock.h
@@ -221,6 +221,7 @@ void vsock_for_each_connected_socket(struct vsock_transport *transport,
 				     void (*fn)(struct sock *sk));
 int vsock_assign_transport(struct vsock_sock *vsk, struct vsock_sock *psk);
 bool vsock_find_cid(unsigned int cid);
+void vsock_linger(struct sock *sk);
 
 /**** TAP ****/
 
diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c
index fc6afbc8d680..2e7a3034e965 100644
--- a/net/vmw_vsock/af_vsock.c
+++ b/net/vmw_vsock/af_vsock.c
@@ -1013,6 +1013,39 @@ out:
 	return err;
 }
 
+void vsock_linger(struct sock *sk)
+{
+	DEFINE_WAIT_FUNC(wait, woken_wake_function);
+	ssize_t (*unsent)(struct vsock_sock *vsk);
+	struct vsock_sock *vsk = vsock_sk(sk);
+	long timeout;
+
+	if (!sock_flag(sk, SOCK_LINGER))
+		return;
+
+	timeout = sk->sk_lingertime;
+	if (!timeout)
+		return;
+
+	/* Transports must implement `unsent_bytes` if they want to support
+	 * SOCK_LINGER through `vsock_linger()` since we use it to check when
+	 * the socket can be closed.
+	 */
+	unsent = vsk->transport->unsent_bytes;
+	if (!unsent)
+		return;
+
+	add_wait_queue(sk_sleep(sk), &wait);
+
+	do {
+		if (sk_wait_event(sk, &timeout, unsent(vsk) == 0, &wait))
+			break;
+	} while (!signal_pending(current) && timeout);
+
+	remove_wait_queue(sk_sleep(sk), &wait);
+}
+EXPORT_SYMBOL_GPL(vsock_linger);
+
 static int vsock_shutdown(struct socket *sock, int mode)
 {
 	int err;
diff --git a/net/vmw_vsock/virtio_transport_common.c b/net/vmw_vsock/virtio_transport_common.c
index 6e7b727c781c..7897fd970dd8 100644
--- a/net/vmw_vsock/virtio_transport_common.c
+++ b/net/vmw_vsock/virtio_transport_common.c
@@ -1191,23 +1191,6 @@ static void virtio_transport_remove_sock(struct vsock_sock *vsk)
 	vsock_remove_sock(vsk);
 }
 
-static void virtio_transport_wait_close(struct sock *sk, long timeout)
-{
-	if (timeout) {
-		DEFINE_WAIT_FUNC(wait, woken_wake_function);
-
-		add_wait_queue(sk_sleep(sk), &wait);
-
-		do {
-			if (sk_wait_event(sk, &timeout,
-					  sock_flag(sk, SOCK_DONE), &wait))
-				break;
-		} while (!signal_pending(current) && timeout);
-
-		remove_wait_queue(sk_sleep(sk), &wait);
-	}
-}
-
 static void virtio_transport_cancel_close_work(struct vsock_sock *vsk,
 					       bool cancel_timeout)
 {
@@ -1277,8 +1260,8 @@ static bool virtio_transport_close(struct vsock_sock *vsk)
 	if ((sk->sk_shutdown & SHUTDOWN_MASK) != SHUTDOWN_MASK)
 		(void)virtio_transport_shutdown(vsk, SHUTDOWN_MASK);
 
-	if (sock_flag(sk, SOCK_LINGER) && !(current->flags & PF_EXITING))
-		virtio_transport_wait_close(sk, sk->sk_lingertime);
+	if (!(current->flags & PF_EXITING))
+		vsock_linger(sk);
 
 	if (sock_flag(sk, SOCK_DONE)) {
 		return true;
diff --git a/tools/testing/vsock/util.c b/tools/testing/vsock/util.c
index de25892f865f..0c7e9cbcbc85 100644
--- a/tools/testing/vsock/util.c
+++ b/tools/testing/vsock/util.c
@@ -17,6 +17,7 @@
 #include <assert.h>
 #include <sys/epoll.h>
 #include <sys/mman.h>
+#include <linux/sockios.h>
 
 #include "timeout.h"
 #include "control.h"
@@ -96,6 +97,30 @@ void vsock_wait_remote_close(int fd)
 	close(epollfd);
 }
 
+/* Wait until transport reports no data left to be sent.
+ * Return false if transport does not implement the unsent_bytes() callback.
+ */
+bool vsock_wait_sent(int fd)
+{
+	int ret, sock_bytes_unsent;
+
+	timeout_begin(TIMEOUT);
+	do {
+		ret = ioctl(fd, SIOCOUTQ, &sock_bytes_unsent);
+		if (ret < 0) {
+			if (errno == EOPNOTSUPP)
+				break;
+
+			perror("ioctl(SIOCOUTQ)");
+			exit(EXIT_FAILURE);
+		}
+		timeout_check("SIOCOUTQ");
+	} while (sock_bytes_unsent != 0);
+	timeout_end();
+
+	return !ret;
+}
+
 /* Create socket <type>, bind to <cid, port> and return the file descriptor. */
 int vsock_bind(unsigned int cid, unsigned int port, int type)
 {
@@ -798,3 +823,16 @@ void enable_so_zerocopy_check(int fd)
 	setsockopt_int_check(fd, SOL_SOCKET, SO_ZEROCOPY, 1,
 			     "setsockopt SO_ZEROCOPY");
 }
+
+void enable_so_linger(int fd, int timeout)
+{
+	struct linger optval = {
+		.l_onoff = 1,
+		.l_linger = timeout
+	};
+
+	if (setsockopt(fd, SOL_SOCKET, SO_LINGER, &optval, sizeof(optval))) {
+		perror("setsockopt(SO_LINGER)");
+		exit(EXIT_FAILURE);
+	}
+}
diff --git a/tools/testing/vsock/util.h b/tools/testing/vsock/util.h
index d1f765ce3eee..5e2db67072d5 100644
--- a/tools/testing/vsock/util.h
+++ b/tools/testing/vsock/util.h
@@ -54,6 +54,7 @@ int vsock_stream_listen(unsigned int cid, unsigned int port);
 int vsock_seqpacket_accept(unsigned int cid, unsigned int port,
 			   struct sockaddr_vm *clientaddrp);
 void vsock_wait_remote_close(int fd);
+bool vsock_wait_sent(int fd);
 void send_buf(int fd, const void *buf, size_t len, int flags,
 	      ssize_t expected_ret);
 void recv_buf(int fd, void *buf, size_t len, int flags, ssize_t expected_ret);
@@ -79,4 +80,5 @@ void setsockopt_int_check(int fd, int level, int optname, int val,
 void setsockopt_timeval_check(int fd, int level, int optname,
 			      struct timeval val, char const *errmsg);
 void enable_so_zerocopy_check(int fd);
+void enable_so_linger(int fd, int timeout);
 #endif /* UTIL_H */
diff --git a/tools/testing/vsock/vsock_test.c b/tools/testing/vsock/vsock_test.c
index 9ea33b78b9fc..f669baaa0dca 100644
--- a/tools/testing/vsock/vsock_test.c
+++ b/tools/testing/vsock/vsock_test.c
@@ -21,7 +21,6 @@
 #include <poll.h>
 #include <signal.h>
 #include <sys/ioctl.h>
-#include <linux/sockios.h>
 #include <linux/time64.h>
 
 #include "vsock_test_zerocopy.h"
@@ -1280,7 +1279,7 @@ static void test_unsent_bytes_server(const struct test_opts *opts, int type)
 static void test_unsent_bytes_client(const struct test_opts *opts, int type)
 {
 	unsigned char buf[MSG_BUF_IOCTL_LEN];
-	int ret, fd, sock_bytes_unsent;
+	int fd;
 
 	fd = vsock_connect(opts->peer_cid, opts->peer_port, type);
 	if (fd < 0) {
@@ -1297,22 +1296,12 @@ static void test_unsent_bytes_client(const struct test_opts *opts, int type)
 	/* SIOCOUTQ isn't guaranteed to instantly track sent data. Even though
 	 * the "RECEIVED" message means that the other side has received the
 	 * data, there can be a delay in our kernel before updating the "unsent
-	 * bytes" counter. Repeat SIOCOUTQ until it returns 0.
+	 * bytes" counter. vsock_wait_sent() will repeat SIOCOUTQ until it
+	 * returns 0.
 	 */
-	timeout_begin(TIMEOUT);
-	do {
-		ret = ioctl(fd, SIOCOUTQ, &sock_bytes_unsent);
-		if (ret < 0) {
-			if (errno == EOPNOTSUPP) {
-				fprintf(stderr, "Test skipped, SIOCOUTQ not supported.\n");
-				break;
-			}
-			perror("ioctl");
-			exit(EXIT_FAILURE);
-		}
-		timeout_check("SIOCOUTQ");
-	} while (sock_bytes_unsent != 0);
-	timeout_end();
+	if (!vsock_wait_sent(fd))
+		fprintf(stderr, "Test skipped, SIOCOUTQ not supported.\n");
+
 	close(fd);
 }
 
@@ -1824,10 +1813,6 @@ static void test_stream_connect_retry_server(const struct test_opts *opts)
 
 static void test_stream_linger_client(const struct test_opts *opts)
 {
-	struct linger optval = {
-		.l_onoff = 1,
-		.l_linger = 1
-	};
 	int fd;
 
 	fd = vsock_stream_connect(opts->peer_cid, opts->peer_port);
@@ -1836,15 +1821,58 @@ static void test_stream_linger_client(const struct test_opts *opts)
 		exit(EXIT_FAILURE);
 	}
 
-	if (setsockopt(fd, SOL_SOCKET, SO_LINGER, &optval, sizeof(optval))) {
-		perror("setsockopt(SO_LINGER)");
+	enable_so_linger(fd, 1);
+	close(fd);
+}
+
+static void test_stream_linger_server(const struct test_opts *opts)
+{
+	int fd;
+
+	fd = vsock_stream_accept(VMADDR_CID_ANY, opts->peer_port, NULL);
+	if (fd < 0) {
+		perror("accept");
 		exit(EXIT_FAILURE);
 	}
 
+	vsock_wait_remote_close(fd);
 	close(fd);
 }
 
-static void test_stream_linger_server(const struct test_opts *opts)
+/* Half of the default to not risk timing out the control channel */
+#define LINGER_TIMEOUT	(TIMEOUT / 2)
+
+static void test_stream_nolinger_client(const struct test_opts *opts)
+{
+	bool waited;
+	time_t ns;
+	int fd;
+
+	fd = vsock_stream_connect(opts->peer_cid, opts->peer_port);
+	if (fd < 0) {
+		perror("connect");
+		exit(EXIT_FAILURE);
+	}
+
+	enable_so_linger(fd, LINGER_TIMEOUT);
+	send_byte(fd, 1, 0); /* Left unread to expose incorrect behaviour. */
+	waited = vsock_wait_sent(fd);
+
+	ns = current_nsec();
+	close(fd);
+	ns = current_nsec() - ns;
+
+	if (!waited) {
+		fprintf(stderr, "Test skipped, SIOCOUTQ not supported.\n");
+	} else if (DIV_ROUND_UP(ns, NSEC_PER_SEC) >= LINGER_TIMEOUT) {
+		fprintf(stderr, "Unexpected lingering\n");
+		exit(EXIT_FAILURE);
+	}
+
+	control_writeln("DONE");
+}
+
+static void test_stream_nolinger_server(const struct test_opts *opts)
 {
 	int fd;
 
@@ -1854,7 +1882,7 @@ static void test_stream_linger_server(const struct test_opts *opts)
 		exit(EXIT_FAILURE);
 	}
 
-	vsock_wait_remote_close(fd);
+	control_expectln("DONE");
 	close(fd);
 }
 
@@ -2018,6 +2046,11 @@ static struct test_case test_cases[] = {
 		.run_client = test_stream_linger_client,
 		.run_server = test_stream_linger_server,
 	},
+	{
+		.name = "SOCK_STREAM SO_LINGER close() on unread",
+		.run_client = test_stream_nolinger_client,
+		.run_server = test_stream_nolinger_server,
+	},
 	{},
 };