Merge branch 'vsock-sockmap-support'

Bobby Eshleman says:

====================
Add support for sockmap to vsock.

We're testing usage of vsock as a way to redirect guest-local UDS
requests to the host and this patch series greatly improves the
performance of such a setup.

Compared to copying packets via userspace, this improves throughput by
121% in basic testing.

Tested as follows.

Setup: guest unix dgram sender -> guest vsock redirector -> host vsock
       server
Threads: 1
Payload: 64k
No sockmap:
- 76.3 MB/s
- The guest vsock redirector was
  "socat VSOCK-CONNECT:2:1234 UNIX-RECV:/path/to/sock"
Using sockmap (this patch):
- 168.8 MB/s (+121%)
- The guest redirector was a simple sockmap echo server,
  redirecting unix ingress to vsock 2:1234 egress.
- Same sender and server programs

*Note: these numbers are from RFC v1

Only the virtio transport has been tested. The loopback transport was
used in writing bpf/selftests, but not thoroughly tested otherwise.

This series requires the skb patch.

Changes in v4:
- af_vsock: fix parameter alignment in vsock_dgram_recvmsg()
- af_vsock: add TCP_ESTABLISHED comment in vsock_dgram_connect()
- vsock/bpf: change ret type to bool

Changes in v3:
- vsock/bpf: Refactor wait logic in vsock_bpf_recvmsg() to avoid
  backwards goto
- vsock/bpf: Check psock before acquiring slock
- vsock/bpf: Return bool instead of int of 0 or 1
- vsock/bpf: Wrap macro args __sk/__psock in parens
- vsock/bpf: Place comment trailer */ on separate line

Changes in v2:
- vsock/bpf: rename vsock_dgram_* -> vsock_*
- vsock/bpf: change sk_psock_{get,put} and {lock,release}_sock() order
  to minimize slock hold time
- vsock/bpf: use "new style" wait
- vsock/bpf: fix bug in wait log
- vsock/bpf: add check that recvmsg sk_type is one dgram, seqpacket, or
  stream.  Return error if not one of the three.
- virtio/vsock: comment __skb_recv_datagram() usage
- virtio/vsock: do not init copied in read_skb()
- vsock/bpf: add ifdef guard around struct proto in dgram_recvmsg()
- selftests/bpf: add vsock loopback config for aarch64
- selftests/bpf: add vsock loopback config for s390x
- selftests/bpf: remove vsock device from vmtest.sh qemu machine
- selftests/bpf: remove CONFIG_VIRTIO_VSOCKETS=y from config.x86_64
- vsock/bpf: move transport-related (e.g., if (!vsk->transport)) checks
  out of fast path
====================
Signed-off-by: Bobby Eshleman <bobby.eshleman@bytedance.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/vhost/vsock.c

@@ -439,6 +439,7 @@ static struct virtio_transport vhost_transport = {
 		.notify_send_post_enqueue = virtio_transport_notify_send_post_enqueue,
 		.notify_buffer_size       = virtio_transport_notify_buffer_size,
 
+		.read_skb = virtio_transport_read_skb,
 	},
 
 	.send_pkt = vhost_transport_send_pkt,

include/linux/virtio_vsock.h

@@ -245,4 +245,5 @@ u32 virtio_transport_get_credit(struct virtio_vsock_sock *vvs, u32 wanted);
 void virtio_transport_put_credit(struct virtio_vsock_sock *vvs, u32 credit);
 void virtio_transport_deliver_tap_pkt(struct sk_buff *skb);
 int virtio_transport_purge_skbs(void *vsk, struct sk_buff_head *list);
+int virtio_transport_read_skb(struct vsock_sock *vsk, skb_read_actor_t read_actor);
 #endif /* _LINUX_VIRTIO_VSOCK_H */

include/net/af_vsock.h

@@ -75,6 +75,7 @@ struct vsock_sock {
 	void *trans;
 };
 
+s64 vsock_connectible_has_data(struct vsock_sock *vsk);
 s64 vsock_stream_has_data(struct vsock_sock *vsk);
 s64 vsock_stream_has_space(struct vsock_sock *vsk);
 struct sock *vsock_create_connected(struct sock *parent);
@@ -173,6 +174,9 @@ struct vsock_transport {
 
 	/* Addressing. */
 	u32 (*get_local_cid)(void);
+
+	/* Read a single skb */
+	int (*read_skb)(struct vsock_sock *, skb_read_actor_t);
 };
 
 /**** CORE ****/
@@ -225,5 +229,18 @@ int vsock_init_tap(void);
 int vsock_add_tap(struct vsock_tap *vt);
 int vsock_remove_tap(struct vsock_tap *vt);
 void vsock_deliver_tap(struct sk_buff *build_skb(void *opaque), void *opaque);
+int vsock_connectible_recvmsg(struct socket *sock, struct msghdr *msg, size_t len,
+			      int flags);
+int vsock_dgram_recvmsg(struct socket *sock, struct msghdr *msg,
+			size_t len, int flags);
+
+#ifdef CONFIG_BPF_SYSCALL
+extern struct proto vsock_proto;
+int vsock_bpf_update_proto(struct sock *sk, struct sk_psock *psock, bool restore);
+void __init vsock_bpf_build_proto(void);
+#else
+static inline void __init vsock_bpf_build_proto(void)
+{}
+#endif
 
 #endif /* __AF_VSOCK_H__ */

net/vmw_vsock/Makefile

@@ -8,6 +8,7 @@ obj-$(CONFIG_HYPERV_VSOCKETS) += hv_sock.o
 obj-$(CONFIG_VSOCKETS_LOOPBACK) += vsock_loopback.o
 
 vsock-y += af_vsock.o af_vsock_tap.o vsock_addr.o
+vsock-$(CONFIG_BPF_SYSCALL) += vsock_bpf.o
 
 vsock_diag-y += diag.o
 

net/vmw_vsock/af_vsock.c

@@ -116,10 +116,13 @@ static void vsock_sk_destruct(struct sock *sk);
 static int vsock_queue_rcv_skb(struct sock *sk, struct sk_buff *skb);
 
 /* Protocol family. */
-static struct proto vsock_proto = {
+struct proto vsock_proto = {
 	.name = "AF_VSOCK",
 	.owner = THIS_MODULE,
 	.obj_size = sizeof(struct vsock_sock),
+#ifdef CONFIG_BPF_SYSCALL
+	.psock_update_sk_prot = vsock_bpf_update_proto,
+#endif
 };
 
 /* The default peer timeout indicates how long we will wait for a peer response
@@ -865,7 +868,7 @@ s64 vsock_stream_has_data(struct vsock_sock *vsk)
 }
 EXPORT_SYMBOL_GPL(vsock_stream_has_data);
 
-static s64 vsock_connectible_has_data(struct vsock_sock *vsk)
+s64 vsock_connectible_has_data(struct vsock_sock *vsk)
 {
 	struct sock *sk = sk_vsock(vsk);
 
@@ -874,6 +877,7 @@ static s64 vsock_connectible_has_data(struct vsock_sock *vsk)
 	else
 		return vsock_stream_has_data(vsk);
 }
+EXPORT_SYMBOL_GPL(vsock_connectible_has_data);
 
 s64 vsock_stream_has_space(struct vsock_sock *vsk)
 {
@@ -1131,6 +1135,13 @@ static __poll_t vsock_poll(struct file *file, struct socket *sock,
 	return mask;
 }
 
+static int vsock_read_skb(struct sock *sk, skb_read_actor_t read_actor)
+{
+	struct vsock_sock *vsk = vsock_sk(sk);
+
+	return vsk->transport->read_skb(vsk, read_actor);
+}
+
 static int vsock_dgram_sendmsg(struct socket *sock, struct msghdr *msg,
 			       size_t len)
 {
@@ -1242,18 +1253,42 @@ static int vsock_dgram_connect(struct socket *sock,
 	memcpy(&vsk->remote_addr, remote_addr, sizeof(vsk->remote_addr));
 	sock->state = SS_CONNECTED;
 
+	/* sock map disallows redirection of non-TCP sockets with sk_state !=
+	 * TCP_ESTABLISHED (see sock_map_redirect_allowed()), so we set
+	 * TCP_ESTABLISHED here to allow redirection of connected vsock dgrams.
+	 *
+	 * This doesn't seem to be abnormal state for datagram sockets, as the
+	 * same approach can be see in other datagram socket types as well
+	 * (such as unix sockets).
+	 */
+	sk->sk_state = TCP_ESTABLISHED;
+
 out:
 	release_sock(sk);
 	return err;
 }
 
-static int vsock_dgram_recvmsg(struct socket *sock, struct msghdr *msg,
-			       size_t len, int flags)
+int vsock_dgram_recvmsg(struct socket *sock, struct msghdr *msg,
+			size_t len, int flags)
 {
-	struct vsock_sock *vsk = vsock_sk(sock->sk);
+#ifdef CONFIG_BPF_SYSCALL
+	const struct proto *prot;
+#endif
+	struct vsock_sock *vsk;
+	struct sock *sk;
+
+	sk = sock->sk;
+	vsk = vsock_sk(sk);
+
+#ifdef CONFIG_BPF_SYSCALL
+	prot = READ_ONCE(sk->sk_prot);
+	if (prot != &vsock_proto)
+		return prot->recvmsg(sk, msg, len, flags, NULL);
+#endif
 
 	return vsk->transport->dgram_dequeue(vsk, msg, len, flags);
 }
+EXPORT_SYMBOL_GPL(vsock_dgram_recvmsg);
 
 static const struct proto_ops vsock_dgram_ops = {
 	.family = PF_VSOCK,
@@ -1272,6 +1307,7 @@ static const struct proto_ops vsock_dgram_ops = {
 	.recvmsg = vsock_dgram_recvmsg,
 	.mmap = sock_no_mmap,
 	.sendpage = sock_no_sendpage,
+	.read_skb = vsock_read_skb,
 };
 
 static int vsock_transport_cancel_pkt(struct vsock_sock *vsk)
@@ -2086,13 +2122,16 @@ static int __vsock_seqpacket_recvmsg(struct sock *sk, struct msghdr *msg,
 	return err;
 }
 
-static int
+int
 vsock_connectible_recvmsg(struct socket *sock, struct msghdr *msg, size_t len,
 			  int flags)
 {
 	struct sock *sk;
 	struct vsock_sock *vsk;
 	const struct vsock_transport *transport;
+#ifdef CONFIG_BPF_SYSCALL
+	const struct proto *prot;
+#endif
 	int err;
 
 	sk = sock->sk;
@@ -2139,6 +2178,14 @@ vsock_connectible_recvmsg(struct socket *sock, struct msghdr *msg, size_t len,
 		goto out;
 	}
 
+#ifdef CONFIG_BPF_SYSCALL
+	prot = READ_ONCE(sk->sk_prot);
+	if (prot != &vsock_proto) {
+		release_sock(sk);
+		return prot->recvmsg(sk, msg, len, flags, NULL);
+	}
+#endif
+
 	if (sk->sk_type == SOCK_STREAM)
 		err = __vsock_stream_recvmsg(sk, msg, len, flags);
 	else
@@ -2148,6 +2195,7 @@ vsock_connectible_recvmsg(struct socket *sock, struct msghdr *msg, size_t len,
 	release_sock(sk);
 	return err;
 }
+EXPORT_SYMBOL_GPL(vsock_connectible_recvmsg);
 
 static int vsock_set_rcvlowat(struct sock *sk, int val)
 {
@@ -2188,6 +2236,7 @@ static const struct proto_ops vsock_stream_ops = {
 	.mmap = sock_no_mmap,
 	.sendpage = sock_no_sendpage,
 	.set_rcvlowat = vsock_set_rcvlowat,
+	.read_skb = vsock_read_skb,
 };
 
 static const struct proto_ops vsock_seqpacket_ops = {
@@ -2209,6 +2258,7 @@ static const struct proto_ops vsock_seqpacket_ops = {
 	.recvmsg = vsock_connectible_recvmsg,
 	.mmap = sock_no_mmap,
 	.sendpage = sock_no_sendpage,
+	.read_skb = vsock_read_skb,
 };
 
 static int vsock_create(struct net *net, struct socket *sock,
@@ -2348,6 +2398,8 @@ static int __init vsock_init(void)
 		goto err_unregister_proto;
 	}
 
+	vsock_bpf_build_proto();
+
 	return 0;
 
 err_unregister_proto:

net/vmw_vsock/virtio_transport.c

@@ -457,6 +457,8 @@ static struct virtio_transport virtio_transport = {
 		.notify_send_pre_enqueue  = virtio_transport_notify_send_pre_enqueue,
 		.notify_send_post_enqueue = virtio_transport_notify_send_post_enqueue,
 		.notify_buffer_size       = virtio_transport_notify_buffer_size,
+
+		.read_skb = virtio_transport_read_skb,
 	},
 
 	.send_pkt = virtio_transport_send_pkt,

net/vmw_vsock/virtio_transport_common.c

@@ -1418,6 +1418,31 @@ int virtio_transport_purge_skbs(void *vsk, struct sk_buff_head *queue)
 }
 EXPORT_SYMBOL_GPL(virtio_transport_purge_skbs);
 
+int virtio_transport_read_skb(struct vsock_sock *vsk, skb_read_actor_t recv_actor)
+{
+	struct virtio_vsock_sock *vvs = vsk->trans;
+	struct sock *sk = sk_vsock(vsk);
+	struct sk_buff *skb;
+	int off = 0;
+	int copied;
+	int err;
+
+	spin_lock_bh(&vvs->rx_lock);
+	/* Use __skb_recv_datagram() for race-free handling of the receive. It
+	 * works for types other than dgrams.
+	 */
+	skb = __skb_recv_datagram(sk, &vvs->rx_queue, MSG_DONTWAIT, &off, &err);
+	spin_unlock_bh(&vvs->rx_lock);
+
+	if (!skb)
+		return err;
+
+	copied = recv_actor(sk, skb);
+	kfree_skb(skb);
+	return copied;
+}
+EXPORT_SYMBOL_GPL(virtio_transport_read_skb);
+
 MODULE_LICENSE("GPL v2");
 MODULE_AUTHOR("Asias He");
 MODULE_DESCRIPTION("common code for virtio vsock");

net/vmw_vsock/vsock_bpf.c

+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2022 Bobby Eshleman <bobby.eshleman@bytedance.com>
+ *
+ * Based off of net/unix/unix_bpf.c
+ */
+
+#include <linux/bpf.h>
+#include <linux/module.h>
+#include <linux/skmsg.h>
+#include <linux/socket.h>
+#include <linux/wait.h>
+#include <net/af_vsock.h>
+#include <net/sock.h>
+
+#define vsock_sk_has_data(__sk, __psock)				\
+		({	!skb_queue_empty(&(__sk)->sk_receive_queue) ||	\
+			!skb_queue_empty(&(__psock)->ingress_skb) ||	\
+			!list_empty(&(__psock)->ingress_msg);		\
+		})
+
+static struct proto *vsock_prot_saved __read_mostly;
+static DEFINE_SPINLOCK(vsock_prot_lock);
+static struct proto vsock_bpf_prot;
+
+static bool vsock_has_data(struct sock *sk, struct sk_psock *psock)
+{
+	struct vsock_sock *vsk = vsock_sk(sk);
+	s64 ret;
+
+	ret = vsock_connectible_has_data(vsk);
+	if (ret > 0)
+		return true;
+
+	return vsock_sk_has_data(sk, psock);
+}
+
+static bool vsock_msg_wait_data(struct sock *sk, struct sk_psock *psock, long timeo)
+{
+	bool ret;
+
+	DEFINE_WAIT_FUNC(wait, woken_wake_function);
+
+	if (sk->sk_shutdown & RCV_SHUTDOWN)
+		return true;
+
+	if (!timeo)
+		return false;
+
+	add_wait_queue(sk_sleep(sk), &wait);
+	sk_set_bit(SOCKWQ_ASYNC_WAITDATA, sk);
+	ret = vsock_has_data(sk, psock);
+	if (!ret) {
+		wait_woken(&wait, TASK_INTERRUPTIBLE, timeo);
+		ret = vsock_has_data(sk, psock);
+	}
+	sk_clear_bit(SOCKWQ_ASYNC_WAITDATA, sk);
+	remove_wait_queue(sk_sleep(sk), &wait);
+	return ret;
+}
+
+static int __vsock_recvmsg(struct sock *sk, struct msghdr *msg, size_t len, int flags)
+{
+	struct socket *sock = sk->sk_socket;
+	int err;
+
+	if (sk->sk_type == SOCK_STREAM || sk->sk_type == SOCK_SEQPACKET)
+		err = vsock_connectible_recvmsg(sock, msg, len, flags);
+	else if (sk->sk_type == SOCK_DGRAM)
+		err = vsock_dgram_recvmsg(sock, msg, len, flags);
+	else
+		err = -EPROTOTYPE;
+
+	return err;
+}
+
+static int vsock_bpf_recvmsg(struct sock *sk, struct msghdr *msg,
+			     size_t len, int flags, int *addr_len)
+{
+	struct sk_psock *psock;
+	int copied;
+
+	psock = sk_psock_get(sk);
+	if (unlikely(!psock))
+		return __vsock_recvmsg(sk, msg, len, flags);
+
+	lock_sock(sk);
+	if (vsock_has_data(sk, psock) && sk_psock_queue_empty(psock)) {
+		release_sock(sk);
+		sk_psock_put(sk, psock);
+		return __vsock_recvmsg(sk, msg, len, flags);
+	}
+
+	copied = sk_msg_recvmsg(sk, psock, msg, len, flags);
+	while (copied == 0) {
+		long timeo = sock_rcvtimeo(sk, flags & MSG_DONTWAIT);
+
+		if (!vsock_msg_wait_data(sk, psock, timeo)) {
+			copied = -EAGAIN;
+			break;
+		}
+
+		if (sk_psock_queue_empty(psock)) {
+			release_sock(sk);
+			sk_psock_put(sk, psock);
+			return __vsock_recvmsg(sk, msg, len, flags);
+		}
+
+		copied = sk_msg_recvmsg(sk, psock, msg, len, flags);
+	}
+
+	release_sock(sk);
+	sk_psock_put(sk, psock);
+
+	return copied;
+}
+
+/* Copy of original proto with updated sock_map methods */
+static struct proto vsock_bpf_prot = {
+	.close = sock_map_close,
+	.recvmsg = vsock_bpf_recvmsg,
+	.sock_is_readable = sk_msg_is_readable,
+	.unhash = sock_map_unhash,
+};
+
+static void vsock_bpf_rebuild_protos(struct proto *prot, const struct proto *base)
+{
+	*prot        = *base;
+	prot->close  = sock_map_close;
+	prot->recvmsg = vsock_bpf_recvmsg;
+	prot->sock_is_readable = sk_msg_is_readable;
+}
+
+static void vsock_bpf_check_needs_rebuild(struct proto *ops)
+{
+	/* Paired with the smp_store_release() below. */
+	if (unlikely(ops != smp_load_acquire(&vsock_prot_saved))) {
+		spin_lock_bh(&vsock_prot_lock);
+		if (likely(ops != vsock_prot_saved)) {
+			vsock_bpf_rebuild_protos(&vsock_bpf_prot, ops);
+			/* Make sure proto function pointers are updated before publishing the
+			 * pointer to the struct.
+			 */
+			smp_store_release(&vsock_prot_saved, ops);
+		}
+		spin_unlock_bh(&vsock_prot_lock);
+	}
+}
+
+int vsock_bpf_update_proto(struct sock *sk, struct sk_psock *psock, bool restore)
+{
+	struct vsock_sock *vsk;
+
+	if (restore) {
+		sk->sk_write_space = psock->saved_write_space;
+		sock_replace_proto(sk, psock->sk_proto);
+		return 0;
+	}
+
+	vsk = vsock_sk(sk);
+	if (!vsk->transport)
+		return -ENODEV;
+
+	if (!vsk->transport->read_skb)
+		return -EOPNOTSUPP;
+
+	vsock_bpf_check_needs_rebuild(psock->sk_proto);
+	sock_replace_proto(sk, &vsock_bpf_prot);
+	return 0;
+}
+
+void __init vsock_bpf_build_proto(void)
+{
+	vsock_bpf_rebuild_protos(&vsock_bpf_prot, &vsock_proto);
+}

net/vmw_vsock/vsock_loopback.c

@@ -94,6 +94,8 @@ static struct virtio_transport loopback_transport = {
 		.notify_send_pre_enqueue  = virtio_transport_notify_send_pre_enqueue,
 		.notify_send_post_enqueue = virtio_transport_notify_send_post_enqueue,
 		.notify_buffer_size       = virtio_transport_notify_buffer_size,
+
+		.read_skb = virtio_transport_read_skb,
 	},
 
 	.send_pkt = vsock_loopback_send_pkt,

tools/testing/selftests/bpf/config.aarch64

@@ -176,6 +176,8 @@ CONFIG_VIRTIO_MMIO_CMDLINE_DEVICES=y
 CONFIG_VIRTIO_MMIO=y
 CONFIG_VIRTIO_NET=y
 CONFIG_VIRTIO_PCI=y
+CONFIG_VIRTIO_VSOCKETS_COMMON=y
 CONFIG_VLAN_8021Q=y
 CONFIG_VSOCKETS=y
+CONFIG_VSOCKETS_LOOPBACK=y
 CONFIG_XFRM_USER=y

tools/testing/selftests/bpf/config.s390x

@@ -140,5 +140,8 @@ CONFIG_VIRTIO_BALLOON=y
 CONFIG_VIRTIO_BLK=y
 CONFIG_VIRTIO_NET=y
 CONFIG_VIRTIO_PCI=y
+CONFIG_VIRTIO_VSOCKETS_COMMON=y
 CONFIG_VLAN_8021Q=y
+CONFIG_VSOCKETS=y
+CONFIG_VSOCKETS_LOOPBACK=y
 CONFIG_XFRM_USER=y

tools/testing/selftests/bpf/config.x86_64

@@ -234,7 +234,10 @@ CONFIG_VIRTIO_BLK=y
 CONFIG_VIRTIO_CONSOLE=y
 CONFIG_VIRTIO_NET=y
 CONFIG_VIRTIO_PCI=y
+CONFIG_VIRTIO_VSOCKETS_COMMON=y
 CONFIG_VLAN_8021Q=y
+CONFIG_VSOCKETS=y
+CONFIG_VSOCKETS_LOOPBACK=y
 CONFIG_X86_ACPI_CPUFREQ=y
 CONFIG_X86_CPUID=y
 CONFIG_X86_MSR=y

tools/testing/selftests/bpf/prog_tests/sockmap_listen.c

@@ -18,6 +18,7 @@
 #include <string.h>
 #include <sys/select.h>
 #include <unistd.h>
+#include <linux/vm_sockets.h>
 
 #include <bpf/bpf.h>
 #include <bpf/libbpf.h>
@@ -251,6 +252,16 @@ static void init_addr_loopback6(struct sockaddr_storage *ss, socklen_t *len)
 	*len = sizeof(*addr6);
 }
 
+static void init_addr_loopback_vsock(struct sockaddr_storage *ss, socklen_t *len)
+{
+	struct sockaddr_vm *addr = memset(ss, 0, sizeof(*ss));
+
+	addr->svm_family = AF_VSOCK;
+	addr->svm_port = VMADDR_PORT_ANY;
+	addr->svm_cid = VMADDR_CID_LOCAL;
+	*len = sizeof(*addr);
+}
+
 static void init_addr_loopback(int family, struct sockaddr_storage *ss,
 			       socklen_t *len)
 {
@@ -261,6 +272,9 @@ static void init_addr_loopback(int family, struct sockaddr_storage *ss,
 	case AF_INET6:
 		init_addr_loopback6(ss, len);
 		return;
+	case AF_VSOCK:
+		init_addr_loopback_vsock(ss, len);
+		return;
 	default:
 		FAIL("unsupported address family %d", family);
 	}
@@ -1478,6 +1492,8 @@ static const char *family_str(sa_family_t family)
 		return "IPv6";
 	case AF_UNIX:
 		return "Unix";
+	case AF_VSOCK:
+		return "VSOCK";
 	default:
 		return "unknown";
 	}
@@ -1689,6 +1705,151 @@ static void test_unix_redir(struct test_sockmap_listen *skel, struct bpf_map *ma
 	unix_skb_redir_to_connected(skel, map, sotype);
 }
 
+/* Returns two connected loopback vsock sockets */
+static int vsock_socketpair_connectible(int sotype, int *v0, int *v1)
+{
+	struct sockaddr_storage addr;
+	socklen_t len = sizeof(addr);
+	int s, p, c;
+
+	s = socket_loopback(AF_VSOCK, sotype);
+	if (s < 0)
+		return -1;
+
+	c = xsocket(AF_VSOCK, sotype | SOCK_NONBLOCK, 0);
+	if (c == -1)
+		goto close_srv;
+
+	if (getsockname(s, sockaddr(&addr), &len) < 0)
+		goto close_cli;
+
+	if (connect(c, sockaddr(&addr), len) < 0 && errno != EINPROGRESS) {
+		FAIL_ERRNO("connect");
+		goto close_cli;
+	}
+
+	len = sizeof(addr);
+	p = accept_timeout(s, sockaddr(&addr), &len, IO_TIMEOUT_SEC);
+	if (p < 0)
+		goto close_cli;
+
+	*v0 = p;
+	*v1 = c;
+
+	return 0;
+
+close_cli:
+	close(c);
+close_srv:
+	close(s);
+
+	return -1;
+}
+
+static void vsock_unix_redir_connectible(int sock_mapfd, int verd_mapfd,
+					 enum redir_mode mode, int sotype)
+{
+	const char *log_prefix = redir_mode_str(mode);
+	char a = 'a', b = 'b';
+	int u0, u1, v0, v1;
+	int sfd[2];
+	unsigned int pass;
+	int err, n;
+	u32 key;
+
+	zero_verdict_count(verd_mapfd);
+
+	if (socketpair(AF_UNIX, SOCK_STREAM | SOCK_NONBLOCK, 0, sfd))
+		return;
+
+	u0 = sfd[0];
+	u1 = sfd[1];
+
+	err = vsock_socketpair_connectible(sotype, &v0, &v1);
+	if (err) {
+		FAIL("vsock_socketpair_connectible() failed");
+		goto close_uds;
+	}
+
+	err = add_to_sockmap(sock_mapfd, u0, v0);
+	if (err) {
+		FAIL("add_to_sockmap failed");
+		goto close_vsock;
+	}
+
+	n = write(v1, &a, sizeof(a));
+	if (n < 0)
+		FAIL_ERRNO("%s: write", log_prefix);
+	if (n == 0)
+		FAIL("%s: incomplete write", log_prefix);
+	if (n < 1)
+		goto out;
+
+	n = recv(mode == REDIR_INGRESS ? u0 : u1, &b, sizeof(b), MSG_DONTWAIT);
+	if (n < 0)
+		FAIL("%s: recv() err, errno=%d", log_prefix, errno);
+	if (n == 0)
+		FAIL("%s: incomplete recv", log_prefix);
+	if (b != a)
+		FAIL("%s: vsock socket map failed, %c != %c", log_prefix, a, b);
+
+	key = SK_PASS;
+	err = xbpf_map_lookup_elem(verd_mapfd, &key, &pass);
+	if (err)
+		goto out;
+	if (pass != 1)
+		FAIL("%s: want pass count 1, have %d", log_prefix, pass);
+out:
+	key = 0;
+	bpf_map_delete_elem(sock_mapfd, &key);
+	key = 1;
+	bpf_map_delete_elem(sock_mapfd, &key);
+
+close_vsock:
+	close(v0);
+	close(v1);
+
+close_uds:
+	close(u0);
+	close(u1);
+}
+
+static void vsock_unix_skb_redir_connectible(struct test_sockmap_listen *skel,
+					     struct bpf_map *inner_map,
+					     int sotype)
+{
+	int verdict = bpf_program__fd(skel->progs.prog_skb_verdict);
+	int verdict_map = bpf_map__fd(skel->maps.verdict_map);
+	int sock_map = bpf_map__fd(inner_map);
+	int err;
+
+	err = xbpf_prog_attach(verdict, sock_map, BPF_SK_SKB_VERDICT, 0);
+	if (err)
+		return;
+
+	skel->bss->test_ingress = false;
+	vsock_unix_redir_connectible(sock_map, verdict_map, REDIR_EGRESS, sotype);
+	skel->bss->test_ingress = true;
+	vsock_unix_redir_connectible(sock_map, verdict_map, REDIR_INGRESS, sotype);
+
+	xbpf_prog_detach2(verdict, sock_map, BPF_SK_SKB_VERDICT);
+}
+
+static void test_vsock_redir(struct test_sockmap_listen *skel, struct bpf_map *map)
+{
+	const char *family_name, *map_name;
+	char s[MAX_TEST_NAME];
+
+	family_name = family_str(AF_VSOCK);
+	map_name = map_type_str(map);
+	snprintf(s, sizeof(s), "%s %s %s", map_name, family_name, __func__);
+	if (!test__start_subtest(s))
+		return;
+
+	vsock_unix_skb_redir_connectible(skel, map, SOCK_STREAM);
+	vsock_unix_skb_redir_connectible(skel, map, SOCK_SEQPACKET);
+}
+
 static void test_reuseport(struct test_sockmap_listen *skel,
 			   struct bpf_map *map, int family, int sotype)
 {
@@ -2060,12 +2221,14 @@ void serial_test_sockmap_listen(void)
 	run_tests(skel, skel->maps.sock_map, AF_INET6);
 	test_unix_redir(skel, skel->maps.sock_map, SOCK_DGRAM);
 	test_unix_redir(skel, skel->maps.sock_map, SOCK_STREAM);
+	test_vsock_redir(skel, skel->maps.sock_map);
 
 	skel->bss->test_sockmap = false;
 	run_tests(skel, skel->maps.sock_hash, AF_INET);
 	run_tests(skel, skel->maps.sock_hash, AF_INET6);
 	test_unix_redir(skel, skel->maps.sock_hash, SOCK_DGRAM);
 	test_unix_redir(skel, skel->maps.sock_hash, SOCK_STREAM);
+	test_vsock_redir(skel, skel->maps.sock_hash);
 
 	test_sockmap_listen__destroy(skel);
 }