Merge: vsock: backport latest commits for RHEL-9-2

MR: https://gitlab.com/redhat/centos-stream/src/kernel/centos-stream-9/-/merge_requests/1869

## Summary of Changes

Backport of vsock fixes, features (support for SO_RCVLOWAT) and tests (SO_RCVLOWAT)
included in Linux v6.2-rc3 for RHEL 9.2.

> git backport-diff -u upstream-local/master -r main..HEAD
Key:
[----] : patches are identical
[####] : number of functional differences between upstream/downstream patch
[down] : patch is downstream-only
The flags [FC] indicate (F)unctional and (C)ontextual differences, respectively

001/17:[----] [--] 'vsock: Fix memory leak in vsock_connect()'
002/17:[----] [--] 'vsock: Set socket state back to SS_UNCONNECTED in vsock_connect_timeout()'
003/17:[----] [--] 'vsock: SO_RCVLOWAT transport set callback'
004/17:[----] [--] 'hv_sock: disable SO_RCVLOWAT support'
005/17:[----] [--] 'virtio/vsock: use 'target' in notify_poll_in callback'
006/17:[----] [--] 'vmci/vsock: use 'target' in notify_poll_in callback'
007/17:[----] [--] 'vsock: pass sock_rcvlowat to notify_poll_in as target'
008/17:[----] [--] 'vsock: add API call for data ready'
009/17:[----] [--] 'virtio/vsock: check SO_RCVLOWAT before wake up reader'
010/17:[----] [--] 'vmci/vsock: check SO_RCVLOWAT before wake up reader'
011/17:[----] [--] 'vsock_test: POLLIN + SO_RCVLOWAT test'
012/17:[----] [--] 'vsock/vmci: fix repeated words in comments'
013/17:[----] [--] 'vhost/vsock: Use kvmalloc/kvfree for larger packets.'
014/17:[----] [--] 'vsock: remove the unused 'wait' in vsock_connectible_recvmsg()'
015/17:[----] [--] 'vsock: fix possible infinite sleep in vsock_connectible_wait_data()'
016/17:[----] [--] 'net: vmw_vsock: vmci: Check memcpy_from_msg()'
017/17:[----] [--] 'vhost/vsock: Fix error handling in vhost_vsock_init()'

## Approved Bugzilla Ticket

Bugzilla: https://bugzilla.redhat.com/2160028

Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>

Approved-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Approved-by: Mohamed Gamal Morsy <mgamal@redhat.com>

Signed-off-by: Herton R. Krzesinski <herton@redhat.com>

Author: Herton R. Krzesinski

drivers/vhost/vsock.c

@@ -393,7 +393,7 @@ vhost_vsock_alloc_pkt(struct vhost_virtqueue *vq,
 		return NULL;
 	}
 
-	pkt->buf = kmalloc(pkt->len, GFP_KERNEL);
+	pkt->buf = kvmalloc(pkt->len, GFP_KERNEL);
 	if (!pkt->buf) {
 		kfree(pkt);
 		return NULL;
@@ -964,7 +964,14 @@ static int __init vhost_vsock_init(void)
 				  VSOCK_TRANSPORT_F_H2G);
 	if (ret < 0)
 		return ret;
-	return misc_register(&vhost_vsock_misc);
+
+	ret = misc_register(&vhost_vsock_misc);
+	if (ret) {
+		vsock_core_unregister(&vhost_transport.transport);
+		return ret;
+	}
+
+	return 0;
 };
 
 static void __exit vhost_vsock_exit(void)

include/net/af_vsock.h

@@ -77,6 +77,7 @@ struct vsock_sock {
 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);
+void vsock_data_ready(struct sock *sk);
 
 /**** TRANSPORT ****/
 
@@ -134,6 +135,7 @@ struct vsock_transport {
 	u64 (*stream_rcvhiwat)(struct vsock_sock *);
 	bool (*stream_is_active)(struct vsock_sock *);
 	bool (*stream_allow)(u32 cid, u32 port);
+	int (*set_rcvlowat)(struct vsock_sock *vsk, int val);
 
 	/* SEQ_PACKET. */
 	ssize_t (*seqpacket_dequeue)(struct vsock_sock *vsk, struct msghdr *msg,

net/vmw_vsock/af_vsock.c

@@ -882,6 +882,16 @@ s64 vsock_stream_has_space(struct vsock_sock *vsk)
 }
 EXPORT_SYMBOL_GPL(vsock_stream_has_space);
 
+void vsock_data_ready(struct sock *sk)
+{
+	struct vsock_sock *vsk = vsock_sk(sk);
+
+	if (vsock_stream_has_data(vsk) >= sk->sk_rcvlowat ||
+	    sock_flag(sk, SOCK_DONE))
+		sk->sk_data_ready(sk);
+}
+EXPORT_SYMBOL_GPL(vsock_data_ready);
+
 static int vsock_release(struct socket *sock)
 {
 	__vsock_release(sock->sk, 0);
@@ -1066,8 +1076,9 @@ static __poll_t vsock_poll(struct file *file, struct socket *sock,
 		if (transport && transport->stream_is_active(vsk) &&
 		    !(sk->sk_shutdown & RCV_SHUTDOWN)) {
 			bool data_ready_now = false;
+			int target = sock_rcvlowat(sk, 0, INT_MAX);
 			int ret = transport->notify_poll_in(
-					vsk, 1, &data_ready_now);
+					vsk, target, &data_ready_now);
 			if (ret < 0) {
 				mask |= EPOLLERR;
 			} else {
@@ -1286,6 +1297,7 @@ static void vsock_connect_timeout(struct work_struct *work)
 	if (sk->sk_state == TCP_SYN_SENT &&
 	    (sk->sk_shutdown != SHUTDOWN_MASK)) {
 		sk->sk_state = TCP_CLOSE;
+		sk->sk_socket->state = SS_UNCONNECTED;
 		sk->sk_err = ETIMEDOUT;
 		sk_error_report(sk);
 		vsock_transport_cancel_pkt(vsk);
@@ -1391,7 +1403,14 @@ static int vsock_connect(struct socket *sock, struct sockaddr *addr,
 			 * timeout fires.
 			 */
 			sock_hold(sk);
-			schedule_delayed_work(&vsk->connect_work, timeout);
+
+			/* If the timeout function is already scheduled,
+			 * reschedule it, then ungrab the socket refcount to
+			 * keep it balanced.
+			 */
+			if (mod_delayed_work(system_wq, &vsk->connect_work,
+					     timeout))
+				sock_put(sk);
 
 			/* Skip ahead to preserve error code set above. */
 			goto out_wait;
@@ -1886,8 +1905,11 @@ static int vsock_connectible_wait_data(struct sock *sk,
 	err = 0;
 	transport = vsk->transport;
 
-	while ((data = vsock_connectible_has_data(vsk)) == 0) {
+	while (1) {
 		prepare_to_wait(sk_sleep(sk), wait, TASK_INTERRUPTIBLE);
+		data = vsock_connectible_has_data(vsk);
+		if (data != 0)
+			break;
 
 		if (sk->sk_err != 0 ||
 		    (sk->sk_shutdown & RCV_SHUTDOWN) ||
@@ -2073,8 +2095,6 @@ vsock_connectible_recvmsg(struct socket *sock, struct msghdr *msg, size_t len,
 	const struct vsock_transport *transport;
 	int err;
 
-	DEFINE_WAIT(wait);
-
 	sk = sock->sk;
 	vsk = vsock_sk(sk);
 	err = 0;
@@ -2129,6 +2149,25 @@ vsock_connectible_recvmsg(struct socket *sock, struct msghdr *msg, size_t len,
 	return err;
 }
 
+static int vsock_set_rcvlowat(struct sock *sk, int val)
+{
+	const struct vsock_transport *transport;
+	struct vsock_sock *vsk;
+
+	vsk = vsock_sk(sk);
+
+	if (val > vsk->buffer_size)
+		return -EINVAL;
+
+	transport = vsk->transport;
+
+	if (transport && transport->set_rcvlowat)
+		return transport->set_rcvlowat(vsk, val);
+
+	WRITE_ONCE(sk->sk_rcvlowat, val ? : 1);
+	return 0;
+}
+
 static const struct proto_ops vsock_stream_ops = {
 	.family = PF_VSOCK,
 	.owner = THIS_MODULE,
@@ -2148,6 +2187,7 @@ static const struct proto_ops vsock_stream_ops = {
 	.recvmsg = vsock_connectible_recvmsg,
 	.mmap = sock_no_mmap,
 	.sendpage = sock_no_sendpage,
+	.set_rcvlowat = vsock_set_rcvlowat,
 };
 
 static const struct proto_ops vsock_seqpacket_ops = {

net/vmw_vsock/hyperv_transport.c

@@ -809,6 +809,12 @@ int hvs_notify_send_post_enqueue(struct vsock_sock *vsk, ssize_t written,
 	return 0;
 }
 
+static
+int hvs_set_rcvlowat(struct vsock_sock *vsk, int val)
+{
+	return -EOPNOTSUPP;
+}
+
 static struct vsock_transport hvs_transport = {
 	.module                   = THIS_MODULE,
 
@@ -844,6 +850,7 @@ static struct vsock_transport hvs_transport = {
 	.notify_send_pre_enqueue  = hvs_notify_send_pre_enqueue,
 	.notify_send_post_enqueue = hvs_notify_send_post_enqueue,
 
+	.set_rcvlowat             = hvs_set_rcvlowat
 };
 
 static bool hvs_check_transport(struct vsock_sock *vsk)

net/vmw_vsock/virtio_transport_common.c

@@ -634,10 +634,7 @@ virtio_transport_notify_poll_in(struct vsock_sock *vsk,
 				size_t target,
 				bool *data_ready_now)
 {
-	if (vsock_stream_has_data(vsk))
-		*data_ready_now = true;
-	else
-		*data_ready_now = false;
+	*data_ready_now = vsock_stream_has_data(vsk) >= target;
 
 	return 0;
 }
@@ -1084,7 +1081,7 @@ virtio_transport_recv_connected(struct sock *sk,
 	switch (le16_to_cpu(pkt->hdr.op)) {
 	case VIRTIO_VSOCK_OP_RW:
 		virtio_transport_recv_enqueue(vsk, pkt);
-		sk->sk_data_ready(sk);
+		vsock_data_ready(sk);
 		return err;
 	case VIRTIO_VSOCK_OP_CREDIT_REQUEST:
 		virtio_transport_send_credit_update(vsk);
@@ -1342,7 +1339,7 @@ EXPORT_SYMBOL_GPL(virtio_transport_recv_pkt);
 
 void virtio_transport_free_pkt(struct virtio_vsock_pkt *pkt)
 {
-	kfree(pkt->buf);
+	kvfree(pkt->buf);
 	kfree(pkt);
 }
 EXPORT_SYMBOL_GPL(virtio_transport_free_pkt);

net/vmw_vsock/vmci_transport.c

@@ -951,7 +951,7 @@ static int vmci_transport_recv_listen(struct sock *sk,
 	 * for ourself or any previous connection requests that we received.
 	 * If it's the latter, we try to find a socket in our list of pending
 	 * connections and, if we do, call the appropriate handler for the
-	 * state that that socket is in.  Otherwise we try to service the
+	 * state that socket is in.  Otherwise we try to service the
 	 * connection request.
 	 */
 	pending = vmci_transport_get_pending(sk, pkt);
@@ -1711,7 +1711,11 @@ static int vmci_transport_dgram_enqueue(
 	if (!dg)
 		return -ENOMEM;
 
-	memcpy_from_msg(VMCI_DG_PAYLOAD(dg), msg, len);
+	err = memcpy_from_msg(VMCI_DG_PAYLOAD(dg), msg, len);
+	if (err) {
+		kfree(dg);
+		return err;
+	}
 
 	dg->dst = vmci_make_handle(remote_addr->svm_cid,
 				   remote_addr->svm_port);

net/vmw_vsock/vmci_transport_notify.c

@@ -307,7 +307,7 @@ vmci_transport_handle_wrote(struct sock *sk,
 	struct vsock_sock *vsk = vsock_sk(sk);
 	PKT_FIELD(vsk, sent_waiting_read) = false;
 #endif
-	sk->sk_data_ready(sk);
+	vsock_data_ready(sk);
 }
 
 static void vmci_transport_notify_pkt_socket_init(struct sock *sk)
@@ -340,12 +340,12 @@ vmci_transport_notify_pkt_poll_in(struct sock *sk,
 {
 	struct vsock_sock *vsk = vsock_sk(sk);
 
-	if (vsock_stream_has_data(vsk)) {
+	if (vsock_stream_has_data(vsk) >= target) {
 		*data_ready_now = true;
 	} else {
-		/* We can't read right now because there is nothing in the
-		 * queue. Ask for notifications when there is something to
-		 * read.
+		/* We can't read right now because there is not enough data
+		 * in the queue. Ask for notifications when there is something
+		 * to read.
 		 */
 		if (sk->sk_state == TCP_ESTABLISHED) {
 			if (!send_waiting_read(sk, 1))

net/vmw_vsock/vmci_transport_notify_qstate.c

@@ -84,7 +84,7 @@ vmci_transport_handle_wrote(struct sock *sk,
 			    bool bottom_half,
 			    struct sockaddr_vm *dst, struct sockaddr_vm *src)
 {
-	sk->sk_data_ready(sk);
+	vsock_data_ready(sk);
 }
 
 static void vsock_block_update_write_window(struct sock *sk)
@@ -161,12 +161,12 @@ vmci_transport_notify_pkt_poll_in(struct sock *sk,
 {
 	struct vsock_sock *vsk = vsock_sk(sk);
 
-	if (vsock_stream_has_data(vsk)) {
+	if (vsock_stream_has_data(vsk) >= target) {
 		*data_ready_now = true;
 	} else {
-		/* We can't read right now because there is nothing in the
-		 * queue. Ask for notifications when there is something to
-		 * read.
+		/* We can't read right now because there is not enough data
+		 * in the queue. Ask for notifications when there is something
+		 * to read.
 		 */
 		if (sk->sk_state == TCP_ESTABLISHED)
 			vsock_block_update_write_window(sk);
@@ -282,7 +282,7 @@ vmci_transport_notify_pkt_recv_post_dequeue(
 		/* See the comment in
 		 * vmci_transport_notify_pkt_send_post_enqueue().
 		 */
-		sk->sk_data_ready(sk);
+		vsock_data_ready(sk);
 	}
 
 	return err;