Resguardo inicial

linux-7.2.3/net/vmw_vsock/Kconfig

@@ -0,0 +1,83 @@
+# SPDX-License-Identifier: GPL-2.0-only
+#
+# Vsock protocol
+#
+
+config VSOCKETS
+	tristate "Virtual Socket protocol"
+	help
+	  Virtual Socket Protocol is a socket protocol similar to TCP/IP
+	  allowing communication between Virtual Machines and hypervisor
+	  or host.
+
+	  You should also select one or more hypervisor-specific transports
+	  below.
+
+	  To compile this driver as a module, choose M here: the module
+	  will be called vsock. If unsure, say N.
+
+config VSOCKETS_DIAG
+	tristate "Virtual Sockets monitoring interface"
+	depends on VSOCKETS
+	default y
+	help
+	  Support for PF_VSOCK sockets monitoring interface used by the ss tool.
+	  If unsure, say Y.
+
+	  Enable this module so userspace applications can query open sockets.
+
+config VSOCKETS_LOOPBACK
+	tristate "Virtual Sockets loopback transport"
+	depends on VSOCKETS
+	default y
+	select VIRTIO_VSOCKETS_COMMON
+	help
+	  This module implements a loopback transport for Virtual Sockets,
+	  using vmw_vsock_virtio_transport_common.
+
+	  To compile this driver as a module, choose M here: the module
+	  will be called vsock_loopback. If unsure, say N.
+
+config VMWARE_VMCI_VSOCKETS
+	tristate "VMware VMCI transport for Virtual Sockets"
+	depends on VSOCKETS && VMWARE_VMCI
+	help
+	  This module implements a VMCI transport for Virtual Sockets.
+
+	  Enable this transport if your Virtual Machine runs on a VMware
+	  hypervisor.
+
+	  To compile this driver as a module, choose M here: the module
+	  will be called vmw_vsock_vmci_transport. If unsure, say N.
+
+config VIRTIO_VSOCKETS
+	tristate "virtio transport for Virtual Sockets"
+	depends on VSOCKETS && VIRTIO
+	select VIRTIO_VSOCKETS_COMMON
+	help
+	  This module implements a virtio transport for Virtual Sockets.
+
+	  Enable this transport if your Virtual Machine host supports Virtual
+	  Sockets over virtio.
+
+	  To compile this driver as a module, choose M here: the module will be
+	  called vmw_vsock_virtio_transport. If unsure, say N.
+
+config VIRTIO_VSOCKETS_COMMON
+	tristate
+	help
+	  This option is selected by any driver which needs to access
+	  the virtio_vsock.  The module will be called
+	  vmw_vsock_virtio_transport_common.
+
+config HYPERV_VSOCKETS
+	tristate "Hyper-V transport for Virtual Sockets"
+	depends on VSOCKETS && HYPERV_VMBUS
+	help
+	  This module implements a Hyper-V transport for Virtual Sockets.
+
+	  Enable this transport if your Virtual Machine host supports Virtual
+	  Sockets over Hyper-V VMBus.
+
+	  To compile this driver as a module, choose M here: the module will be
+	  called hv_sock. If unsure, say N.

linux-7.2.3/net/vmw_vsock/Makefile

@@ -0,0 +1,22 @@
+# SPDX-License-Identifier: GPL-2.0
+obj-$(CONFIG_VSOCKETS) += vsock.o
+obj-$(CONFIG_VSOCKETS_DIAG) += vsock_diag.o
+obj-$(CONFIG_VMWARE_VMCI_VSOCKETS) += vmw_vsock_vmci_transport.o
+obj-$(CONFIG_VIRTIO_VSOCKETS) += vmw_vsock_virtio_transport.o
+obj-$(CONFIG_VIRTIO_VSOCKETS_COMMON) += vmw_vsock_virtio_transport_common.o
+obj-$(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
+
+vmw_vsock_vmci_transport-y += vmci_transport.o vmci_transport_notify.o \
+	vmci_transport_notify_qstate.o
+
+vmw_vsock_virtio_transport-y += virtio_transport.o
+
+vmw_vsock_virtio_transport_common-y += virtio_transport_common.o
+
+hv_sock-y += hyperv_transport.o

linux-7.2.3/net/vmw_vsock/af_vsock_tap.c

@@ -0,0 +1,110 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Tap functions for AF_VSOCK sockets.
+ *
+ * Code based on net/netlink/af_netlink.c tap functions.
+ */
+
+#include <linux/module.h>
+#include <net/sock.h>
+#include <net/af_vsock.h>
+#include <linux/if_arp.h>
+
+static DEFINE_SPINLOCK(vsock_tap_lock);
+static struct list_head vsock_tap_all __read_mostly =
+				LIST_HEAD_INIT(vsock_tap_all);
+
+int vsock_add_tap(struct vsock_tap *vt)
+{
+	if (unlikely(vt->dev->type != ARPHRD_VSOCKMON))
+		return -EINVAL;
+
+	__module_get(vt->module);
+
+	spin_lock(&vsock_tap_lock);
+	list_add_rcu(&vt->list, &vsock_tap_all);
+	spin_unlock(&vsock_tap_lock);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(vsock_add_tap);
+
+int vsock_remove_tap(struct vsock_tap *vt)
+{
+	struct vsock_tap *tmp;
+	bool found = false;
+
+	spin_lock(&vsock_tap_lock);
+
+	list_for_each_entry(tmp, &vsock_tap_all, list) {
+		if (vt == tmp) {
+			list_del_rcu(&vt->list);
+			found = true;
+			goto out;
+		}
+	}
+
+	pr_warn("vsock_remove_tap: %p not found\n", vt);
+out:
+	spin_unlock(&vsock_tap_lock);
+
+	synchronize_net();
+
+	if (found)
+		module_put(vt->module);
+
+	return found ? 0 : -ENODEV;
+}
+EXPORT_SYMBOL_GPL(vsock_remove_tap);
+
+static int __vsock_deliver_tap_skb(struct sk_buff *skb,
+				   struct net_device *dev)
+{
+	int ret = 0;
+	struct sk_buff *nskb = skb_clone(skb, GFP_ATOMIC);
+
+	if (nskb) {
+		dev_hold(dev);
+
+		nskb->dev = dev;
+		ret = dev_queue_xmit(nskb);
+		if (unlikely(ret > 0))
+			ret = net_xmit_errno(ret);
+
+		dev_put(dev);
+	}
+
+	return ret;
+}
+
+static void __vsock_deliver_tap(struct sk_buff *skb)
+{
+	int ret;
+	struct vsock_tap *tmp;
+
+	list_for_each_entry_rcu(tmp, &vsock_tap_all, list) {
+		ret = __vsock_deliver_tap_skb(skb, tmp->dev);
+		if (unlikely(ret))
+			break;
+	}
+}
+
+void vsock_deliver_tap(struct sk_buff *build_skb(void *opaque), void *opaque)
+{
+	struct sk_buff *skb;
+
+	rcu_read_lock();
+
+	if (likely(list_empty(&vsock_tap_all)))
+		goto out;
+
+	skb = build_skb(opaque);
+	if (skb) {
+		__vsock_deliver_tap(skb);
+		consume_skb(skb);
+	}
+
+out:
+	rcu_read_unlock();
+}
+EXPORT_SYMBOL_GPL(vsock_deliver_tap);

linux-7.2.3/net/vmw_vsock/diag.c

@@ -0,0 +1,180 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * vsock sock_diag(7) module
+ *
+ * Copyright (C) 2017 Red Hat, Inc.
+ * Author: Stefan Hajnoczi <stefanha@redhat.com>
+ */
+
+#include <linux/module.h>
+#include <linux/sock_diag.h>
+#include <linux/vm_sockets_diag.h>
+#include <net/af_vsock.h>
+
+static int sk_diag_fill(struct sock *sk, struct sk_buff *skb,
+			u32 portid, u32 seq, u32 flags)
+{
+	struct vsock_sock *vsk = vsock_sk(sk);
+	struct vsock_diag_msg *rep;
+	struct nlmsghdr *nlh;
+
+	nlh = nlmsg_put(skb, portid, seq, SOCK_DIAG_BY_FAMILY, sizeof(*rep),
+			flags);
+	if (!nlh)
+		return -EMSGSIZE;
+
+	rep = nlmsg_data(nlh);
+	rep->vdiag_family = AF_VSOCK;
+
+	/* Lock order dictates that sk_lock is acquired before
+	 * vsock_table_lock, so we cannot lock here.  Simply don't take
+	 * sk_lock; sk is guaranteed to stay alive since vsock_table_lock is
+	 * held.
+	 */
+	rep->vdiag_type = sk->sk_type;
+	rep->vdiag_state = sk->sk_state;
+	rep->vdiag_shutdown = sk->sk_shutdown;
+	rep->vdiag_src_cid = vsk->local_addr.svm_cid;
+	rep->vdiag_src_port = vsk->local_addr.svm_port;
+	rep->vdiag_dst_cid = vsk->remote_addr.svm_cid;
+	rep->vdiag_dst_port = vsk->remote_addr.svm_port;
+	rep->vdiag_ino = sock_i_ino(sk);
+
+	sock_diag_save_cookie(sk, rep->vdiag_cookie);
+
+	return 0;
+}
+
+static int vsock_diag_dump(struct sk_buff *skb, struct netlink_callback *cb)
+{
+	struct vsock_diag_req *req;
+	struct vsock_sock *vsk;
+	unsigned int bucket;
+	unsigned int last_i;
+	unsigned int table;
+	struct net *net;
+	unsigned int i;
+
+	req = nlmsg_data(cb->nlh);
+	net = sock_net(skb->sk);
+
+	/* State saved between calls: */
+	table = cb->args[0];
+	bucket = cb->args[1];
+	i = last_i = cb->args[2];
+
+	/* TODO VMCI pending sockets? */
+
+	spin_lock_bh(&vsock_table_lock);
+
+	/* Bind table (locally created sockets) */
+	if (table == 0) {
+		while (bucket < ARRAY_SIZE(vsock_bind_table)) {
+			struct list_head *head = &vsock_bind_table[bucket];
+
+			i = 0;
+			list_for_each_entry(vsk, head, bound_table) {
+				struct sock *sk = sk_vsock(vsk);
+
+				if (!net_eq(sock_net(sk), net))
+					continue;
+				if (i < last_i)
+					goto next_bind;
+				if (!(req->vdiag_states & (1 << sk->sk_state)))
+					goto next_bind;
+				if (sk_diag_fill(sk, skb,
+						 NETLINK_CB(cb->skb).portid,
+						 cb->nlh->nlmsg_seq,
+						 NLM_F_MULTI) < 0)
+					goto done;
+next_bind:
+				i++;
+			}
+			last_i = 0;
+			bucket++;
+		}
+
+		table++;
+		bucket = 0;
+	}
+
+	/* Connected table (accepted connections) */
+	while (bucket < ARRAY_SIZE(vsock_connected_table)) {
+		struct list_head *head = &vsock_connected_table[bucket];
+
+		i = 0;
+		list_for_each_entry(vsk, head, connected_table) {
+			struct sock *sk = sk_vsock(vsk);
+
+			/* Skip sockets we've already seen above */
+			if (__vsock_in_bound_table(vsk))
+				continue;
+
+			if (!net_eq(sock_net(sk), net))
+				continue;
+			if (i < last_i)
+				goto next_connected;
+			if (!(req->vdiag_states & (1 << sk->sk_state)))
+				goto next_connected;
+			if (sk_diag_fill(sk, skb,
+					 NETLINK_CB(cb->skb).portid,
+					 cb->nlh->nlmsg_seq,
+					 NLM_F_MULTI) < 0)
+				goto done;
+next_connected:
+			i++;
+		}
+		last_i = 0;
+		bucket++;
+	}
+
+done:
+	spin_unlock_bh(&vsock_table_lock);
+
+	cb->args[0] = table;
+	cb->args[1] = bucket;
+	cb->args[2] = i;
+
+	return skb->len;
+}
+
+static int vsock_diag_handler_dump(struct sk_buff *skb, struct nlmsghdr *h)
+{
+	int hdrlen = sizeof(struct vsock_diag_req);
+	struct net *net = sock_net(skb->sk);
+
+	if (nlmsg_len(h) < hdrlen)
+		return -EINVAL;
+
+	if (h->nlmsg_flags & NLM_F_DUMP) {
+		struct netlink_dump_control c = {
+			.dump = vsock_diag_dump,
+		};
+		return netlink_dump_start(net->diag_nlsk, skb, h, &c);
+	}
+
+	return -EOPNOTSUPP;
+}
+
+static const struct sock_diag_handler vsock_diag_handler = {
+	.owner = THIS_MODULE,
+	.family = AF_VSOCK,
+	.dump = vsock_diag_handler_dump,
+};
+
+static int __init vsock_diag_init(void)
+{
+	return sock_diag_register(&vsock_diag_handler);
+}
+
+static void __exit vsock_diag_exit(void)
+{
+	sock_diag_unregister(&vsock_diag_handler);
+}
+
+module_init(vsock_diag_init);
+module_exit(vsock_diag_exit);
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("VMware Virtual Sockets monitoring via SOCK_DIAG");
+MODULE_ALIAS_NET_PF_PROTO_TYPE(PF_NETLINK, NETLINK_SOCK_DIAG,
+			       40 /* AF_VSOCK */);

linux-7.2.3/net/vmw_vsock/hyperv_transport.c

@@ -0,0 +1,992 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Hyper-V transport for vsock
+ *
+ * Hyper-V Sockets supplies a byte-stream based communication mechanism
+ * between the host and the VM. This driver implements the necessary
+ * support in the VM by introducing the new vsock transport.
+ *
+ * Copyright (c) 2017, Microsoft Corporation.
+ */
+#include <linux/module.h>
+#include <linux/vmalloc.h>
+#include <linux/hyperv.h>
+#include <net/sock.h>
+#include <net/af_vsock.h>
+#include <hyperv/hvhdk.h>
+
+/* Older (VMBUS version 'VERSION_WIN10' or before) Windows hosts have some
+ * stricter requirements on the hv_sock ring buffer size of six 4K pages.
+ * HV_HYP_PAGE_SIZE is defined as 4K. Newer hosts don't have this limitation;
+ * but, keep the defaults the same for compat.
+ */
+#define RINGBUFFER_HVS_RCV_SIZE (HV_HYP_PAGE_SIZE * 6)
+#define RINGBUFFER_HVS_SND_SIZE (HV_HYP_PAGE_SIZE * 6)
+#define RINGBUFFER_HVS_MAX_SIZE (HV_HYP_PAGE_SIZE * 64)
+
+/* The MTU is 16KB per the host side's design */
+#define HVS_MTU_SIZE		(1024 * 16)
+
+/* How long to wait for graceful shutdown of a connection */
+#define HVS_CLOSE_TIMEOUT (8 * HZ)
+
+struct vmpipe_proto_header {
+	u32 pkt_type;
+	u32 data_size;
+};
+
+/* For recv, we use the VMBus in-place packet iterator APIs to directly copy
+ * data from the ringbuffer into the userspace buffer.
+ */
+struct hvs_recv_buf {
+	/* The header before the payload data */
+	struct vmpipe_proto_header hdr;
+
+	/* The payload */
+	u8 data[HVS_MTU_SIZE];
+};
+
+/* We can send up to HVS_MTU_SIZE bytes of payload to the host, but let's use
+ * a smaller size, i.e. HVS_SEND_BUF_SIZE, to maximize concurrency between the
+ * guest and the host processing as one VMBUS packet is the smallest processing
+ * unit.
+ *
+ * Note: the buffer can be eliminated in the future when we add new VMBus
+ * ringbuffer APIs that allow us to directly copy data from userspace buffer
+ * to VMBus ringbuffer.
+ */
+#define HVS_SEND_BUF_SIZE \
+		(HV_HYP_PAGE_SIZE - sizeof(struct vmpipe_proto_header))
+
+struct hvs_send_buf {
+	/* The header before the payload data */
+	struct vmpipe_proto_header hdr;
+
+	/* The payload */
+	u8 data[HVS_SEND_BUF_SIZE];
+};
+
+#define HVS_HEADER_LEN	(sizeof(struct vmpacket_descriptor) + \
+			 sizeof(struct vmpipe_proto_header))
+
+/* See 'prev_indices' in hv_ringbuffer_read(), hv_ringbuffer_write(), and
+ * __hv_pkt_iter_next().
+ */
+#define VMBUS_PKT_TRAILER_SIZE	(sizeof(u64))
+
+#define HVS_PKT_LEN(payload_len)	(HVS_HEADER_LEN + \
+					 ALIGN((payload_len), 8) + \
+					 VMBUS_PKT_TRAILER_SIZE)
+
+/* Upper bound on the size of a VMbus packet for hv_sock */
+#define HVS_MAX_PKT_SIZE	HVS_PKT_LEN(HVS_MTU_SIZE)
+
+union hvs_service_id {
+	guid_t	srv_id;
+
+	struct {
+		unsigned int svm_port;
+		unsigned char b[sizeof(guid_t) - sizeof(unsigned int)];
+	};
+};
+
+/* Per-socket state (accessed via vsk->trans) */
+struct hvsock {
+	struct vsock_sock *vsk;
+
+	guid_t vm_srv_id;
+	guid_t host_srv_id;
+
+	struct vmbus_channel *chan;
+	struct vmpacket_descriptor *recv_desc;
+
+	/* The length of the payload not delivered to userland yet */
+	u32 recv_data_len;
+	/* The offset of the payload */
+	u32 recv_data_off;
+
+	/* Have we sent the zero-length packet (FIN)? */
+	bool fin_sent;
+};
+
+/* In the VM, we support Hyper-V Sockets with AF_VSOCK, and the endpoint is
+ * <cid, port> (see struct sockaddr_vm). Note: cid is not really used here:
+ * when we write apps to connect to the host, we can only use VMADDR_CID_ANY
+ * or VMADDR_CID_HOST (both are equivalent) as the remote cid, and when we
+ * write apps to bind() & listen() in the VM, we can only use VMADDR_CID_ANY
+ * as the local cid.
+ *
+ * On the host, Hyper-V Sockets are supported by Winsock AF_HYPERV:
+ * https://docs.microsoft.com/en-us/virtualization/hyper-v-on-windows/user-
+ * guide/make-integration-service, and the endpoint is <VmID, ServiceId> with
+ * the below sockaddr:
+ *
+ * struct SOCKADDR_HV
+ * {
+ *    ADDRESS_FAMILY Family;
+ *    USHORT Reserved;
+ *    GUID VmId;
+ *    GUID ServiceId;
+ * };
+ * Note: VmID is not used by Linux VM and actually it isn't transmitted via
+ * VMBus, because here it's obvious the host and the VM can easily identify
+ * each other. Though the VmID is useful on the host, especially in the case
+ * of Windows container, Linux VM doesn't need it at all.
+ *
+ * To make use of the AF_VSOCK infrastructure in Linux VM, we have to limit
+ * the available GUID space of SOCKADDR_HV so that we can create a mapping
+ * between AF_VSOCK port and SOCKADDR_HV Service GUID. The rule of writing
+ * Hyper-V Sockets apps on the host and in Linux VM is:
+ *
+ ****************************************************************************
+ * The only valid Service GUIDs, from the perspectives of both the host and *
+ * Linux VM, that can be connected by the other end, must conform to this   *
+ * format: <port>-facb-11e6-bd58-64006a7986d3.                              *
+ ****************************************************************************
+ *
+ * When we write apps on the host to connect(), the GUID ServiceID is used.
+ * When we write apps in Linux VM to connect(), we only need to specify the
+ * port and the driver will form the GUID and use that to request the host.
+ *
+ */
+
+/* 00000000-facb-11e6-bd58-64006a7986d3 */
+static const guid_t srv_id_template =
+	GUID_INIT(0x00000000, 0xfacb, 0x11e6, 0xbd, 0x58,
+		  0x64, 0x00, 0x6a, 0x79, 0x86, 0xd3);
+
+static bool hvs_check_transport(struct vsock_sock *vsk);
+
+static bool is_valid_srv_id(const guid_t *id)
+{
+	return !memcmp(&id->b[4], &srv_id_template.b[4], sizeof(guid_t) - 4);
+}
+
+static unsigned int get_port_by_srv_id(const guid_t *svr_id)
+{
+	return *((unsigned int *)svr_id);
+}
+
+static void hvs_addr_init(struct sockaddr_vm *addr, const guid_t *svr_id)
+{
+	unsigned int port = get_port_by_srv_id(svr_id);
+
+	vsock_addr_init(addr, VMADDR_CID_ANY, port);
+}
+
+static void hvs_set_channel_pending_send_size(struct vmbus_channel *chan)
+{
+	set_channel_pending_send_size(chan,
+				      HVS_PKT_LEN(HVS_SEND_BUF_SIZE));
+
+	virt_mb();
+}
+
+static bool hvs_channel_readable(struct vmbus_channel *chan)
+{
+	u32 readable = hv_get_bytes_to_read(&chan->inbound);
+
+	/* 0-size payload means FIN */
+	return readable >= HVS_PKT_LEN(0);
+}
+
+static int hvs_channel_readable_payload(struct vmbus_channel *chan)
+{
+	u32 readable = hv_get_bytes_to_read(&chan->inbound);
+
+	if (readable > HVS_PKT_LEN(0)) {
+		/* At least we have 1 byte to read. We don't need to return
+		 * the exact readable bytes: see vsock_connectible_recvmsg() ->
+		 * vsock_stream_has_data().
+		 */
+		return 1;
+	}
+
+	if (readable == HVS_PKT_LEN(0)) {
+		/* 0-size payload means FIN */
+		return 0;
+	}
+
+	/* No payload or FIN */
+	return -1;
+}
+
+static size_t hvs_channel_writable_bytes(struct vmbus_channel *chan)
+{
+	u32 writeable = hv_get_bytes_to_write(&chan->outbound);
+	size_t ret;
+
+	/* The ringbuffer mustn't be 100% full, and we should reserve a
+	 * zero-length-payload packet for the FIN: see hv_ringbuffer_write()
+	 * and hvs_shutdown().
+	 */
+	if (writeable <= HVS_PKT_LEN(1) + HVS_PKT_LEN(0))
+		return 0;
+
+	ret = writeable - HVS_PKT_LEN(1) - HVS_PKT_LEN(0);
+
+	return round_down(ret, 8);
+}
+
+static int __hvs_send_data(struct vmbus_channel *chan,
+			   struct vmpipe_proto_header *hdr,
+			   size_t to_write)
+{
+	hdr->pkt_type = 1;
+	hdr->data_size = to_write;
+	return vmbus_sendpacket(chan, hdr, sizeof(*hdr) + to_write,
+				0, VM_PKT_DATA_INBAND, 0);
+}
+
+static int hvs_send_data(struct vmbus_channel *chan,
+			 struct hvs_send_buf *send_buf, size_t to_write)
+{
+	return __hvs_send_data(chan, &send_buf->hdr, to_write);
+}
+
+static void hvs_channel_cb(void *ctx)
+{
+	struct sock *sk = (struct sock *)ctx;
+	struct vsock_sock *vsk = vsock_sk(sk);
+	struct hvsock *hvs = vsk->trans;
+	struct vmbus_channel *chan = hvs->chan;
+
+	if (hvs_channel_readable(chan))
+		sk->sk_data_ready(sk);
+
+	if (hv_get_bytes_to_write(&chan->outbound) > 0)
+		sk->sk_write_space(sk);
+}
+
+static void hvs_do_close_lock_held(struct vsock_sock *vsk,
+				   bool cancel_timeout)
+{
+	struct sock *sk = sk_vsock(vsk);
+
+	sock_set_flag(sk, SOCK_DONE);
+	WRITE_ONCE(vsk->peer_shutdown, SHUTDOWN_MASK);
+	if (vsock_stream_has_data(vsk) <= 0)
+		sk->sk_state = TCP_CLOSING;
+	sk->sk_state_change(sk);
+	if (vsk->close_work_scheduled &&
+	    (!cancel_timeout || cancel_delayed_work(&vsk->close_work))) {
+		vsk->close_work_scheduled = false;
+		vsock_remove_sock(vsk);
+
+		/* Release the reference taken while scheduling the timeout */
+		sock_put(sk);
+	}
+}
+
+static void hvs_close_connection(struct vmbus_channel *chan)
+{
+	struct sock *sk = get_per_channel_state(chan);
+
+	lock_sock(sk);
+	hvs_do_close_lock_held(vsock_sk(sk), true);
+	release_sock(sk);
+
+	/* Release the refcnt for the channel that's opened in
+	 * hvs_open_connection().
+	 */
+	sock_put(sk);
+}
+
+static void hvs_open_connection(struct vmbus_channel *chan)
+{
+	guid_t *if_instance, *if_type;
+	unsigned char conn_from_host;
+
+	struct sockaddr_vm addr;
+	struct sock *sk, *new = NULL;
+	struct vsock_sock *vnew = NULL;
+	struct hvsock *hvs = NULL;
+	struct hvsock *hvs_new = NULL;
+	int rcvbuf;
+	int ret;
+	int sndbuf;
+
+	if_type = &chan->offermsg.offer.if_type;
+	if_instance = &chan->offermsg.offer.if_instance;
+	conn_from_host = chan->offermsg.offer.u.pipe.user_def[0];
+	if (!is_valid_srv_id(if_type))
+		return;
+
+	hvs_addr_init(&addr, conn_from_host ? if_type : if_instance);
+	sk = vsock_find_bound_socket(&addr);
+	if (!sk)
+		return;
+
+	lock_sock(sk);
+	if ((conn_from_host && sk->sk_state != TCP_LISTEN) ||
+	    (!conn_from_host && sk->sk_state != TCP_SYN_SENT))
+		goto out;
+
+	if (conn_from_host) {
+		if (sk_acceptq_is_full(sk))
+			goto out;
+
+		new = vsock_create_connected(sk);
+		if (!new)
+			goto out;
+
+		new->sk_state = TCP_SYN_SENT;
+		vnew = vsock_sk(new);
+
+		hvs_addr_init(&vnew->local_addr, if_type);
+
+		/* Remote peer is always the host */
+		vsock_addr_init(&vnew->remote_addr,
+				VMADDR_CID_HOST, VMADDR_PORT_ANY);
+		vnew->remote_addr.svm_port = get_port_by_srv_id(if_instance);
+		ret = vsock_assign_transport(vnew, vsock_sk(sk));
+		/* Transport assigned (looking at remote_addr) must be the
+		 * same where we received the request.
+		 */
+		if (ret || !hvs_check_transport(vnew)) {
+			sock_put(new);
+			goto out;
+		}
+		hvs_new = vnew->trans;
+		hvs_new->chan = chan;
+	} else {
+		hvs = vsock_sk(sk)->trans;
+		hvs->chan = chan;
+	}
+
+	set_channel_read_mode(chan, HV_CALL_DIRECT);
+
+	/* Use the socket buffer sizes as hints for the VMBUS ring size. For
+	 * server side sockets, 'sk' is the parent socket and thus, this will
+	 * allow the child sockets to inherit the size from the parent. Keep
+	 * the mins to the default value and align to page size as per VMBUS
+	 * requirements.
+	 * For the max, the socket core library will limit the socket buffer
+	 * size that can be set by the user, but, since currently, the hv_sock
+	 * VMBUS ring buffer is physically contiguous allocation, restrict it
+	 * further.
+	 * Older versions of hv_sock host side code cannot handle bigger VMBUS
+	 * ring buffer size. Use the version number to limit the change to newer
+	 * versions.
+	 */
+	if (vmbus_proto_version < VERSION_WIN10_V5) {
+		sndbuf = RINGBUFFER_HVS_SND_SIZE;
+		rcvbuf = RINGBUFFER_HVS_RCV_SIZE;
+	} else {
+		sndbuf = max_t(int, sk->sk_sndbuf, RINGBUFFER_HVS_SND_SIZE);
+		sndbuf = min_t(int, sndbuf, RINGBUFFER_HVS_MAX_SIZE);
+		sndbuf = VMBUS_RING_SIZE(sndbuf);
+		rcvbuf = max_t(int, sk->sk_rcvbuf, RINGBUFFER_HVS_RCV_SIZE);
+		rcvbuf = min_t(int, rcvbuf, RINGBUFFER_HVS_MAX_SIZE);
+		rcvbuf = VMBUS_RING_SIZE(rcvbuf);
+	}
+
+	chan->max_pkt_size = HVS_MAX_PKT_SIZE;
+
+	ret = vmbus_open(chan, sndbuf, rcvbuf, NULL, 0, hvs_channel_cb,
+			 conn_from_host ? new : sk);
+	if (ret != 0) {
+		if (conn_from_host) {
+			hvs_new->chan = NULL;
+			sock_put(new);
+		} else {
+			hvs->chan = NULL;
+		}
+		goto out;
+	}
+
+	set_per_channel_state(chan, conn_from_host ? new : sk);
+
+	/* This reference will be dropped by hvs_close_connection(). */
+	sock_hold(conn_from_host ? new : sk);
+	vmbus_set_chn_rescind_callback(chan, hvs_close_connection);
+
+	/* Set the pending send size to max packet size to always get
+	 * notifications from the host when there is enough writable space.
+	 * The host is optimized to send notifications only when the pending
+	 * size boundary is crossed, and not always.
+	 */
+	hvs_set_channel_pending_send_size(chan);
+
+	if (conn_from_host) {
+		new->sk_state = TCP_ESTABLISHED;
+
+		hvs_new->vm_srv_id = *if_type;
+		hvs_new->host_srv_id = *if_instance;
+
+		vsock_insert_connected(vnew);
+
+		vsock_enqueue_accept(sk, new);
+	} else {
+		sk->sk_state = TCP_ESTABLISHED;
+		sk->sk_socket->state = SS_CONNECTED;
+
+		vsock_insert_connected(vsock_sk(sk));
+	}
+
+	sk->sk_state_change(sk);
+
+out:
+	/* Release refcnt obtained when we called vsock_find_bound_socket() */
+	sock_put(sk);
+
+	release_sock(sk);
+}
+
+static u32 hvs_get_local_cid(void)
+{
+	return VMADDR_CID_ANY;
+}
+
+static int hvs_sock_init(struct vsock_sock *vsk, struct vsock_sock *psk)
+{
+	struct hvsock *hvs;
+	struct sock *sk = sk_vsock(vsk);
+
+	hvs = kzalloc_obj(*hvs);
+	if (!hvs)
+		return -ENOMEM;
+
+	vsk->trans = hvs;
+	hvs->vsk = vsk;
+	sk->sk_sndbuf = RINGBUFFER_HVS_SND_SIZE;
+	sk->sk_rcvbuf = RINGBUFFER_HVS_RCV_SIZE;
+	return 0;
+}
+
+static int hvs_connect(struct vsock_sock *vsk)
+{
+	union hvs_service_id vm, host;
+	struct hvsock *h = vsk->trans;
+
+	vm.srv_id = srv_id_template;
+	vm.svm_port = vsk->local_addr.svm_port;
+	h->vm_srv_id = vm.srv_id;
+
+	host.srv_id = srv_id_template;
+	host.svm_port = vsk->remote_addr.svm_port;
+	h->host_srv_id = host.srv_id;
+
+	return vmbus_send_tl_connect_request(&h->vm_srv_id, &h->host_srv_id);
+}
+
+static void hvs_shutdown_lock_held(struct hvsock *hvs, int mode)
+{
+	struct vmpipe_proto_header hdr;
+
+	if (hvs->fin_sent || !hvs->chan)
+		return;
+
+	/* It can't fail: see hvs_channel_writable_bytes(). */
+	(void)__hvs_send_data(hvs->chan, &hdr, 0);
+	hvs->fin_sent = true;
+}
+
+static int hvs_shutdown(struct vsock_sock *vsk, int mode)
+{
+	if (!(mode & SEND_SHUTDOWN))
+		return 0;
+
+	hvs_shutdown_lock_held(vsk->trans, mode);
+	return 0;
+}
+
+static void hvs_close_timeout(struct work_struct *work)
+{
+	struct vsock_sock *vsk =
+		container_of(work, struct vsock_sock, close_work.work);
+	struct sock *sk = sk_vsock(vsk);
+
+	sock_hold(sk);
+	lock_sock(sk);
+	if (!sock_flag(sk, SOCK_DONE))
+		hvs_do_close_lock_held(vsk, false);
+
+	vsk->close_work_scheduled = false;
+	release_sock(sk);
+	sock_put(sk);
+}
+
+/* Returns true, if it is safe to remove socket; false otherwise */
+static bool hvs_close_lock_held(struct vsock_sock *vsk)
+{
+	struct sock *sk = sk_vsock(vsk);
+
+	if (!(sk->sk_state == TCP_ESTABLISHED ||
+	      sk->sk_state == TCP_CLOSING))
+		return true;
+
+	if ((sk->sk_shutdown & SHUTDOWN_MASK) != SHUTDOWN_MASK)
+		hvs_shutdown_lock_held(vsk->trans, SHUTDOWN_MASK);
+
+	if (sock_flag(sk, SOCK_DONE))
+		return true;
+
+	/* This reference will be dropped by the delayed close routine */
+	sock_hold(sk);
+	INIT_DELAYED_WORK(&vsk->close_work, hvs_close_timeout);
+	vsk->close_work_scheduled = true;
+	schedule_delayed_work(&vsk->close_work, HVS_CLOSE_TIMEOUT);
+	return false;
+}
+
+static void hvs_release(struct vsock_sock *vsk)
+{
+	bool remove_sock;
+
+	remove_sock = hvs_close_lock_held(vsk);
+	if (remove_sock)
+		vsock_remove_sock(vsk);
+}
+
+static void hvs_destruct(struct vsock_sock *vsk)
+{
+	struct hvsock *hvs = vsk->trans;
+	struct vmbus_channel *chan = hvs->chan;
+
+	if (chan)
+		vmbus_hvsock_device_unregister(chan);
+
+	kfree(hvs);
+	vsk->trans = NULL;
+}
+
+static int hvs_dgram_bind(struct vsock_sock *vsk, struct sockaddr_vm *addr)
+{
+	return -EOPNOTSUPP;
+}
+
+static int hvs_dgram_dequeue(struct vsock_sock *vsk, struct msghdr *msg,
+			     size_t len, int flags)
+{
+	return -EOPNOTSUPP;
+}
+
+static int hvs_dgram_enqueue(struct vsock_sock *vsk,
+			     struct sockaddr_vm *remote, struct msghdr *msg,
+			     size_t dgram_len)
+{
+	return -EOPNOTSUPP;
+}
+
+static bool hvs_dgram_allow(struct vsock_sock *vsk, u32 cid, u32 port)
+{
+	return false;
+}
+
+static int hvs_update_recv_data(struct hvsock *hvs)
+{
+	struct hvs_recv_buf *recv_buf;
+	u32 pkt_len, payload_len;
+
+	pkt_len = hv_pkt_len(hvs->recv_desc);
+
+	if (pkt_len < HVS_HEADER_LEN)
+		return -EIO;
+
+	recv_buf = (struct hvs_recv_buf *)(hvs->recv_desc + 1);
+	payload_len = recv_buf->hdr.data_size;
+
+	if (payload_len > pkt_len - HVS_HEADER_LEN ||
+	    payload_len > HVS_MTU_SIZE)
+		return -EIO;
+
+	if (payload_len == 0)
+		WRITE_ONCE(hvs->vsk->peer_shutdown,
+			   READ_ONCE(hvs->vsk->peer_shutdown) |
+			   SEND_SHUTDOWN);
+
+	hvs->recv_data_len = payload_len;
+	hvs->recv_data_off = 0;
+
+	return 0;
+}
+
+static ssize_t hvs_stream_dequeue(struct vsock_sock *vsk, struct msghdr *msg,
+				  size_t len, int flags)
+{
+	struct hvsock *hvs = vsk->trans;
+	bool need_refill = !hvs->recv_desc;
+	struct hvs_recv_buf *recv_buf;
+	u32 to_read;
+	int ret;
+
+	if (flags & MSG_PEEK)
+		return -EOPNOTSUPP;
+
+	if (need_refill) {
+		hvs->recv_desc = hv_pkt_iter_first(hvs->chan);
+		if (!hvs->recv_desc)
+			return -ENOBUFS;
+		ret = hvs_update_recv_data(hvs);
+		if (ret)
+			return ret;
+	}
+
+	recv_buf = (struct hvs_recv_buf *)(hvs->recv_desc + 1);
+	to_read = min_t(u32, len, hvs->recv_data_len);
+	ret = memcpy_to_msg(msg, recv_buf->data + hvs->recv_data_off, to_read);
+	if (ret != 0)
+		return ret;
+
+	hvs->recv_data_len -= to_read;
+	if (hvs->recv_data_len == 0) {
+		hvs->recv_desc = hv_pkt_iter_next(hvs->chan, hvs->recv_desc);
+		if (hvs->recv_desc) {
+			ret = hvs_update_recv_data(hvs);
+			if (ret)
+				return ret;
+		}
+	} else {
+		hvs->recv_data_off += to_read;
+	}
+
+	return to_read;
+}
+
+static ssize_t hvs_stream_enqueue(struct vsock_sock *vsk, struct msghdr *msg,
+				  size_t len)
+{
+	struct hvsock *hvs = vsk->trans;
+	struct vmbus_channel *chan = hvs->chan;
+	struct hvs_send_buf *send_buf;
+	ssize_t to_write, max_writable;
+	ssize_t ret = 0;
+	ssize_t bytes_written = 0;
+
+	BUILD_BUG_ON(sizeof(*send_buf) != HV_HYP_PAGE_SIZE);
+
+	send_buf = kmalloc_obj(*send_buf);
+	if (!send_buf)
+		return -ENOMEM;
+
+	/* Reader(s) could be draining data from the channel as we write.
+	 * Maximize bandwidth, by iterating until the channel is found to be
+	 * full.
+	 */
+	while (len) {
+		max_writable = hvs_channel_writable_bytes(chan);
+		if (!max_writable)
+			break;
+		to_write = min_t(ssize_t, len, max_writable);
+		to_write = min_t(ssize_t, to_write, HVS_SEND_BUF_SIZE);
+		/* memcpy_from_msg is safe for loop as it advances the offsets
+		 * within the message iterator.
+		 */
+		ret = memcpy_from_msg(send_buf->data, msg, to_write);
+		if (ret < 0)
+			goto out;
+
+		ret = hvs_send_data(hvs->chan, send_buf, to_write);
+		if (ret < 0)
+			goto out;
+
+		bytes_written += to_write;
+		len -= to_write;
+	}
+out:
+	/* If any data has been sent, return that */
+	if (bytes_written)
+		ret = bytes_written;
+	kfree(send_buf);
+	return ret;
+}
+
+static s64 hvs_stream_has_data(struct vsock_sock *vsk)
+{
+	struct hvsock *hvs = vsk->trans;
+	s64 ret;
+
+	if (hvs->recv_data_len > 0)
+		return hvs->recv_data_len;
+
+	switch (hvs_channel_readable_payload(hvs->chan)) {
+	case 1:
+		if (hvs->recv_desc) {
+			/* Here hvs->recv_data_len is 0, so hvs->recv_desc must
+			 * be NULL unless it points to the 0-byte-payload FIN
+			 * packet or a malformed/short packet: see
+			 * hvs_update_recv_data().
+			 *
+			 * If hvs->recv_desc points to the FIN packet, here all
+			 * the payload has been dequeued and the peer_shutdown
+			 * flag is set, but hvs_channel_readable_payload() still
+			 * returns 1, because the VMBus ringbuffer's read_index
+			 * is not updated for the FIN packet:
+			 * hvs_stream_dequeue() -> hv_pkt_iter_next() updates
+			 * the cached priv_read_index but has no opportunity to
+			 * update the read_index in hv_pkt_iter_close() as
+			 * hvs_stream_has_data() returns 0 for the FIN packet,
+			 * so it won't get dequeued.
+			 *
+			 * In case hvs->recv_desc points to a malformed/short
+			 * packet, return -EIO.
+			 */
+			if (!(vsk->peer_shutdown & SEND_SHUTDOWN))
+				return -EIO;
+
+			return 0;
+		}
+
+		hvs->recv_desc = hv_pkt_iter_first(hvs->chan);
+		if (!hvs->recv_desc)
+			return -ENOBUFS;
+
+		ret = hvs_update_recv_data(hvs);
+		if (ret)
+			return ret;
+		return hvs->recv_data_len;
+	case 0:
+		WRITE_ONCE(vsk->peer_shutdown,
+			   READ_ONCE(vsk->peer_shutdown) | SEND_SHUTDOWN);
+		ret = 0;
+		break;
+	default: /* -1 */
+		ret = 0;
+		break;
+	}
+
+	return ret;
+}
+
+static s64 hvs_stream_has_space(struct vsock_sock *vsk)
+{
+	struct hvsock *hvs = vsk->trans;
+
+	return hvs_channel_writable_bytes(hvs->chan);
+}
+
+static u64 hvs_stream_rcvhiwat(struct vsock_sock *vsk)
+{
+	return HVS_MTU_SIZE + 1;
+}
+
+static bool hvs_stream_is_active(struct vsock_sock *vsk)
+{
+	struct hvsock *hvs = vsk->trans;
+
+	return hvs->chan != NULL;
+}
+
+static bool hvs_stream_allow(struct vsock_sock *vsk, u32 cid, u32 port)
+{
+	if (!vsock_net_mode_global(vsk))
+		return false;
+
+	if (cid == VMADDR_CID_HOST)
+		return true;
+
+	return false;
+}
+
+static
+int hvs_notify_poll_in(struct vsock_sock *vsk, size_t target, bool *readable)
+{
+	struct hvsock *hvs = vsk->trans;
+
+	*readable = hvs_channel_readable(hvs->chan);
+	return 0;
+}
+
+static
+int hvs_notify_poll_out(struct vsock_sock *vsk, size_t target, bool *writable)
+{
+	*writable = hvs_stream_has_space(vsk) > 0;
+
+	return 0;
+}
+
+static
+int hvs_notify_recv_init(struct vsock_sock *vsk, size_t target,
+			 struct vsock_transport_recv_notify_data *d)
+{
+	return 0;
+}
+
+static
+int hvs_notify_recv_pre_block(struct vsock_sock *vsk, size_t target,
+			      struct vsock_transport_recv_notify_data *d)
+{
+	return 0;
+}
+
+static
+int hvs_notify_recv_pre_dequeue(struct vsock_sock *vsk, size_t target,
+				struct vsock_transport_recv_notify_data *d)
+{
+	return 0;
+}
+
+static
+int hvs_notify_recv_post_dequeue(struct vsock_sock *vsk, size_t target,
+				 ssize_t copied, bool data_read,
+				 struct vsock_transport_recv_notify_data *d)
+{
+	return 0;
+}
+
+static
+int hvs_notify_send_init(struct vsock_sock *vsk,
+			 struct vsock_transport_send_notify_data *d)
+{
+	return 0;
+}
+
+static
+int hvs_notify_send_pre_block(struct vsock_sock *vsk,
+			      struct vsock_transport_send_notify_data *d)
+{
+	return 0;
+}
+
+static
+int hvs_notify_send_pre_enqueue(struct vsock_sock *vsk,
+				struct vsock_transport_send_notify_data *d)
+{
+	return 0;
+}
+
+static
+int hvs_notify_send_post_enqueue(struct vsock_sock *vsk, ssize_t written,
+				 struct vsock_transport_send_notify_data *d)
+{
+	return 0;
+}
+
+static
+int hvs_notify_set_rcvlowat(struct vsock_sock *vsk, int val)
+{
+	return -EOPNOTSUPP;
+}
+
+static struct vsock_transport hvs_transport = {
+	.module                   = THIS_MODULE,
+
+	.get_local_cid            = hvs_get_local_cid,
+
+	.init                     = hvs_sock_init,
+	.destruct                 = hvs_destruct,
+	.release                  = hvs_release,
+	.connect                  = hvs_connect,
+	.shutdown                 = hvs_shutdown,
+
+	.dgram_bind               = hvs_dgram_bind,
+	.dgram_dequeue            = hvs_dgram_dequeue,
+	.dgram_enqueue            = hvs_dgram_enqueue,
+	.dgram_allow              = hvs_dgram_allow,
+
+	.stream_dequeue           = hvs_stream_dequeue,
+	.stream_enqueue           = hvs_stream_enqueue,
+	.stream_has_data          = hvs_stream_has_data,
+	.stream_has_space         = hvs_stream_has_space,
+	.stream_rcvhiwat          = hvs_stream_rcvhiwat,
+	.stream_is_active         = hvs_stream_is_active,
+	.stream_allow             = hvs_stream_allow,
+
+	.notify_poll_in           = hvs_notify_poll_in,
+	.notify_poll_out          = hvs_notify_poll_out,
+	.notify_recv_init         = hvs_notify_recv_init,
+	.notify_recv_pre_block    = hvs_notify_recv_pre_block,
+	.notify_recv_pre_dequeue  = hvs_notify_recv_pre_dequeue,
+	.notify_recv_post_dequeue = hvs_notify_recv_post_dequeue,
+	.notify_send_init         = hvs_notify_send_init,
+	.notify_send_pre_block    = hvs_notify_send_pre_block,
+	.notify_send_pre_enqueue  = hvs_notify_send_pre_enqueue,
+	.notify_send_post_enqueue = hvs_notify_send_post_enqueue,
+
+	.notify_set_rcvlowat      = hvs_notify_set_rcvlowat
+};
+
+static bool hvs_check_transport(struct vsock_sock *vsk)
+{
+	return vsk->transport == &hvs_transport;
+}
+
+static int hvs_probe(struct hv_device *hdev,
+		     const struct hv_vmbus_device_id *dev_id)
+{
+	struct vmbus_channel *chan = hdev->channel;
+
+	hvs_open_connection(chan);
+
+	/* Always return success to suppress the unnecessary error message
+	 * in vmbus_probe(): on error the host will rescind the device in
+	 * 30 seconds and we can do cleanup at that time in
+	 * vmbus_onoffer_rescind().
+	 */
+	return 0;
+}
+
+static void hvs_remove(struct hv_device *hdev)
+{
+	struct vmbus_channel *chan = hdev->channel;
+
+	vmbus_close(chan);
+}
+
+/* hv_sock connections can not persist across hibernation, and all the hv_sock
+ * channels are forced to be rescinded before hibernation: see
+ * vmbus_bus_suspend(). Here the dummy hvs_suspend() and hvs_resume()
+ * are only needed because hibernation requires that every vmbus device's
+ * driver should have a .suspend and .resume callback: see vmbus_suspend().
+ */
+static int hvs_suspend(struct hv_device *hv_dev)
+{
+	/* Dummy */
+	return 0;
+}
+
+static int hvs_resume(struct hv_device *dev)
+{
+	/* Dummy */
+	return 0;
+}
+
+/* This isn't really used. See vmbus_match() and vmbus_probe() */
+static const struct hv_vmbus_device_id id_table[] = {
+	{},
+};
+
+static struct hv_driver hvs_drv = {
+	.name		= "hv_sock",
+	.hvsock		= true,
+	.id_table	= id_table,
+	.probe		= hvs_probe,
+	.remove		= hvs_remove,
+	.suspend	= hvs_suspend,
+	.resume		= hvs_resume,
+};
+
+static int __init hvs_init(void)
+{
+	int ret;
+
+	if (vmbus_proto_version < VERSION_WIN10)
+		return -ENODEV;
+
+	ret = vmbus_driver_register(&hvs_drv);
+	if (ret != 0)
+		return ret;
+
+	ret = vsock_core_register(&hvs_transport, VSOCK_TRANSPORT_F_G2H);
+	if (ret) {
+		vmbus_driver_unregister(&hvs_drv);
+		return ret;
+	}
+
+	return 0;
+}
+
+static void __exit hvs_exit(void)
+{
+	vsock_core_unregister(&hvs_transport);
+	vmbus_driver_unregister(&hvs_drv);
+}
+
+module_init(hvs_init);
+module_exit(hvs_exit);
+
+MODULE_DESCRIPTION("Hyper-V Sockets");
+MODULE_VERSION("1.0.0");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS_NETPROTO(PF_VSOCK);

linux-7.2.3/net/vmw_vsock/virtio_transport.c

@@ -0,0 +1,989 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * virtio transport for vsock
+ *
+ * Copyright (C) 2013-2015 Red Hat, Inc.
+ * Author: Asias He <asias@redhat.com>
+ *         Stefan Hajnoczi <stefanha@redhat.com>
+ *
+ * Some of the code is take from Gerd Hoffmann <kraxel@redhat.com>'s
+ * early virtio-vsock proof-of-concept bits.
+ */
+#include <linux/spinlock.h>
+#include <linux/module.h>
+#include <linux/list.h>
+#include <linux/atomic.h>
+#include <linux/virtio.h>
+#include <linux/virtio_ids.h>
+#include <linux/virtio_config.h>
+#include <linux/virtio_vsock.h>
+#include <linux/dma-mapping.h>
+#include <net/sock.h>
+#include <linux/mutex.h>
+#include <net/af_vsock.h>
+
+static struct workqueue_struct *virtio_vsock_workqueue;
+static struct virtio_vsock __rcu *the_virtio_vsock;
+static DEFINE_MUTEX(the_virtio_vsock_mutex); /* protects the_virtio_vsock */
+static struct virtio_transport virtio_transport; /* forward declaration */
+
+struct virtio_vsock {
+	struct virtio_device *vdev;
+	struct virtqueue *vqs[VSOCK_VQ_MAX];
+
+	/* Virtqueue processing is deferred to a workqueue */
+	struct work_struct tx_work;
+	struct work_struct rx_work;
+	struct work_struct event_work;
+
+	/* The following fields are protected by tx_lock.  vqs[VSOCK_VQ_TX]
+	 * must be accessed with tx_lock held.
+	 */
+	struct mutex tx_lock;
+	bool tx_run;
+
+	struct work_struct send_pkt_work;
+	struct sk_buff_head send_pkt_queue;
+
+	atomic_t queued_replies;
+
+	/* The following fields are protected by rx_lock.  vqs[VSOCK_VQ_RX]
+	 * must be accessed with rx_lock held.
+	 */
+	struct mutex rx_lock;
+	bool rx_run;
+	int rx_buf_nr;
+	int rx_buf_max_nr;
+
+	u32 guest_cid;
+	bool seqpacket_allow;
+
+	/* These fields are used only in tx path in function
+	 * 'virtio_transport_send_pkt_work()', so to save
+	 * stack space in it, place both of them here. Each
+	 * pointer from 'out_sgs' points to the corresponding
+	 * element in 'out_bufs' - this is initialized in
+	 * 'virtio_vsock_probe()'. Both fields are protected
+	 * by 'tx_lock'. +1 is needed for packet header.
+	 */
+	struct scatterlist *out_sgs[MAX_SKB_FRAGS + 1];
+	struct scatterlist out_bufs[MAX_SKB_FRAGS + 1];
+
+	/* The following fields are protected by event_lock.
+	 * vqs[VSOCK_VQ_EVENT] must be accessed with event_lock held.
+	 */
+	struct mutex event_lock;
+	bool event_run;
+	__dma_from_device_group_begin();
+	struct virtio_vsock_event event_list[8];
+	__dma_from_device_group_end();
+};
+
+static u32 virtio_transport_get_local_cid(void)
+{
+	struct virtio_vsock *vsock;
+	u32 ret;
+
+	rcu_read_lock();
+	vsock = rcu_dereference(the_virtio_vsock);
+	if (!vsock) {
+		ret = VMADDR_CID_ANY;
+		goto out_rcu;
+	}
+
+	ret = vsock->guest_cid;
+out_rcu:
+	rcu_read_unlock();
+	return ret;
+}
+
+/* Caller need to hold vsock->tx_lock on vq */
+static int virtio_transport_send_skb(struct sk_buff *skb, struct virtqueue *vq,
+				     struct virtio_vsock *vsock, gfp_t gfp)
+{
+	int ret, in_sg = 0, out_sg = 0;
+	struct scatterlist **sgs;
+
+	sgs = vsock->out_sgs;
+	sg_init_one(sgs[out_sg], virtio_vsock_hdr(skb),
+		    sizeof(*virtio_vsock_hdr(skb)));
+	out_sg++;
+
+	if (!skb_is_nonlinear(skb)) {
+		if (skb->len > 0) {
+			sg_init_one(sgs[out_sg], skb->data, skb->len);
+			out_sg++;
+		}
+	} else {
+		struct skb_shared_info *si;
+		int i;
+
+		/* If skb is nonlinear, then its buffer must contain
+		 * only header and nothing more. Data is stored in
+		 * the fragged part.
+		 */
+		WARN_ON_ONCE(skb_headroom(skb) != sizeof(*virtio_vsock_hdr(skb)));
+
+		si = skb_shinfo(skb);
+
+		for (i = 0; i < si->nr_frags; i++) {
+			skb_frag_t *skb_frag = &si->frags[i];
+			void *va;
+
+			/* We will use 'page_to_virt()' for the userspace page
+			 * here, because virtio or dma-mapping layers will call
+			 * 'virt_to_phys()' later to fill the buffer descriptor.
+			 * We don't touch memory at "virtual" address of this page.
+			 */
+			va = page_to_virt(skb_frag_page(skb_frag));
+			sg_init_one(sgs[out_sg],
+				    va + skb_frag_off(skb_frag),
+				    skb_frag_size(skb_frag));
+			out_sg++;
+		}
+	}
+
+	ret = virtqueue_add_sgs(vq, sgs, out_sg, in_sg, skb, gfp);
+	/* Usually this means that there is no more space available in
+	 * the vq
+	 */
+	if (ret < 0)
+		return ret;
+
+	virtio_transport_deliver_tap_pkt(skb);
+	return 0;
+}
+
+static void
+virtio_transport_send_pkt_work(struct work_struct *work)
+{
+	struct virtio_vsock *vsock =
+		container_of(work, struct virtio_vsock, send_pkt_work);
+	struct virtqueue *vq;
+	bool added = false;
+	bool restart_rx = false;
+
+	mutex_lock(&vsock->tx_lock);
+
+	if (!vsock->tx_run)
+		goto out;
+
+	vq = vsock->vqs[VSOCK_VQ_TX];
+
+	for (;;) {
+		struct sk_buff *skb;
+		bool reply;
+		int ret;
+
+		skb = virtio_vsock_skb_dequeue(&vsock->send_pkt_queue);
+		if (!skb)
+			break;
+
+		reply = virtio_vsock_skb_reply(skb);
+
+		ret = virtio_transport_send_skb(skb, vq, vsock, GFP_KERNEL);
+		if (ret < 0) {
+			virtio_vsock_skb_queue_head(&vsock->send_pkt_queue, skb);
+			break;
+		}
+
+		if (reply) {
+			struct virtqueue *rx_vq = vsock->vqs[VSOCK_VQ_RX];
+			int val;
+
+			val = atomic_dec_return(&vsock->queued_replies);
+
+			/* Do we now have resources to resume rx processing? */
+			if (val + 1 == virtqueue_get_vring_size(rx_vq))
+				restart_rx = true;
+		}
+
+		added = true;
+	}
+
+	if (added)
+		virtqueue_kick(vq);
+
+out:
+	mutex_unlock(&vsock->tx_lock);
+
+	if (restart_rx)
+		queue_work(virtio_vsock_workqueue, &vsock->rx_work);
+}
+
+/* Caller need to hold RCU for vsock.
+ * Returns 0 if the packet is successfully put on the vq.
+ */
+static int virtio_transport_send_skb_fast_path(struct virtio_vsock *vsock, struct sk_buff *skb)
+{
+	struct virtqueue *vq = vsock->vqs[VSOCK_VQ_TX];
+	int ret;
+
+	/* Inside RCU, can't sleep! */
+	ret = mutex_trylock(&vsock->tx_lock);
+	if (unlikely(ret == 0))
+		return -EBUSY;
+
+	ret = virtio_transport_send_skb(skb, vq, vsock, GFP_ATOMIC);
+	if (ret == 0)
+		virtqueue_kick(vq);
+
+	mutex_unlock(&vsock->tx_lock);
+
+	return ret;
+}
+
+static int
+virtio_transport_send_pkt(struct sk_buff *skb, struct net *net)
+{
+	struct virtio_vsock_hdr *hdr;
+	struct virtio_vsock *vsock;
+	int len = skb->len;
+
+	hdr = virtio_vsock_hdr(skb);
+
+	rcu_read_lock();
+	vsock = rcu_dereference(the_virtio_vsock);
+	if (!vsock) {
+		kfree_skb(skb);
+		len = -ENODEV;
+		goto out_rcu;
+	}
+
+	if (le64_to_cpu(hdr->dst_cid) == vsock->guest_cid) {
+		kfree_skb(skb);
+		len = -ENODEV;
+		goto out_rcu;
+	}
+
+	/* If send_pkt_queue is empty, we can safely bypass this queue
+	 * because packet order is maintained and (try) to put the packet
+	 * on the virtqueue using virtio_transport_send_skb_fast_path.
+	 * If this fails we simply put the packet on the intermediate
+	 * queue and schedule the worker.
+	 */
+	if (!skb_queue_empty_lockless(&vsock->send_pkt_queue) ||
+	    virtio_transport_send_skb_fast_path(vsock, skb)) {
+		if (virtio_vsock_skb_reply(skb))
+			atomic_inc(&vsock->queued_replies);
+
+		virtio_vsock_skb_queue_tail(&vsock->send_pkt_queue, skb);
+		queue_work(virtio_vsock_workqueue, &vsock->send_pkt_work);
+	}
+
+out_rcu:
+	rcu_read_unlock();
+	return len;
+}
+
+static int
+virtio_transport_cancel_pkt(struct vsock_sock *vsk)
+{
+	struct virtio_vsock *vsock;
+	int cnt = 0, ret;
+
+	rcu_read_lock();
+	vsock = rcu_dereference(the_virtio_vsock);
+	if (!vsock) {
+		ret = -ENODEV;
+		goto out_rcu;
+	}
+
+	cnt = virtio_transport_purge_skbs(vsk, &vsock->send_pkt_queue);
+
+	if (cnt) {
+		struct virtqueue *rx_vq = vsock->vqs[VSOCK_VQ_RX];
+		int new_cnt;
+
+		new_cnt = atomic_sub_return(cnt, &vsock->queued_replies);
+		if (new_cnt + cnt >= virtqueue_get_vring_size(rx_vq) &&
+		    new_cnt < virtqueue_get_vring_size(rx_vq))
+			queue_work(virtio_vsock_workqueue, &vsock->rx_work);
+	}
+
+	ret = 0;
+
+out_rcu:
+	rcu_read_unlock();
+	return ret;
+}
+
+static void virtio_vsock_rx_fill(struct virtio_vsock *vsock)
+{
+	int total_len = VIRTIO_VSOCK_DEFAULT_RX_BUF_SIZE;
+	struct scatterlist pkt, *p;
+	struct virtqueue *vq;
+	struct sk_buff *skb;
+	int ret;
+
+	vq = vsock->vqs[VSOCK_VQ_RX];
+
+	do {
+		skb = virtio_vsock_alloc_linear_skb(total_len, GFP_KERNEL);
+		if (!skb)
+			break;
+
+		memset(skb->head, 0, VIRTIO_VSOCK_SKB_HEADROOM);
+		sg_init_one(&pkt, virtio_vsock_hdr(skb), total_len);
+		p = &pkt;
+		ret = virtqueue_add_sgs(vq, &p, 0, 1, skb, GFP_KERNEL);
+		if (ret < 0) {
+			kfree_skb(skb);
+			break;
+		}
+
+		vsock->rx_buf_nr++;
+	} while (vq->num_free);
+	if (vsock->rx_buf_nr > vsock->rx_buf_max_nr)
+		vsock->rx_buf_max_nr = vsock->rx_buf_nr;
+	virtqueue_kick(vq);
+}
+
+static void virtio_transport_tx_work(struct work_struct *work)
+{
+	struct virtio_vsock *vsock =
+		container_of(work, struct virtio_vsock, tx_work);
+	struct virtqueue *vq;
+	bool added = false;
+
+	mutex_lock(&vsock->tx_lock);
+
+	if (!vsock->tx_run)
+		goto out;
+
+	vq = vsock->vqs[VSOCK_VQ_TX];
+
+	do {
+		struct sk_buff *skb;
+		unsigned int len;
+
+		virtqueue_disable_cb(vq);
+		while ((skb = virtqueue_get_buf(vq, &len)) != NULL) {
+			virtio_transport_consume_skb_sent(skb, true);
+			added = true;
+		}
+	} while (!virtqueue_enable_cb(vq));
+
+out:
+	mutex_unlock(&vsock->tx_lock);
+
+	if (added)
+		queue_work(virtio_vsock_workqueue, &vsock->send_pkt_work);
+}
+
+/* Is there space left for replies to rx packets? */
+static bool virtio_transport_more_replies(struct virtio_vsock *vsock)
+{
+	struct virtqueue *vq = vsock->vqs[VSOCK_VQ_RX];
+	int val;
+
+	smp_rmb(); /* paired with atomic_inc() and atomic_dec_return() */
+	val = atomic_read(&vsock->queued_replies);
+
+	return val < virtqueue_get_vring_size(vq);
+}
+
+/* event_lock must be held */
+static int virtio_vsock_event_fill_one(struct virtio_vsock *vsock,
+				       struct virtio_vsock_event *event)
+{
+	struct scatterlist sg;
+	struct virtqueue *vq;
+
+	vq = vsock->vqs[VSOCK_VQ_EVENT];
+
+	sg_init_one(&sg, event, sizeof(*event));
+
+	return virtqueue_add_inbuf_cache_clean(vq, &sg, 1, event, GFP_KERNEL);
+}
+
+/* event_lock must be held */
+static void virtio_vsock_event_fill(struct virtio_vsock *vsock)
+{
+	size_t i;
+
+	for (i = 0; i < ARRAY_SIZE(vsock->event_list); i++) {
+		struct virtio_vsock_event *event = &vsock->event_list[i];
+
+		virtio_vsock_event_fill_one(vsock, event);
+	}
+
+	virtqueue_kick(vsock->vqs[VSOCK_VQ_EVENT]);
+}
+
+static void virtio_vsock_reset_sock(struct sock *sk)
+{
+	/* vmci_transport.c doesn't take sk_lock here either.  At least we're
+	 * under vsock_table_lock so the sock cannot disappear while we're
+	 * executing.
+	 */
+
+	sk->sk_state = TCP_CLOSE;
+	sk->sk_err = ECONNRESET;
+	sk_error_report(sk);
+}
+
+static void virtio_vsock_update_guest_cid(struct virtio_vsock *vsock)
+{
+	struct virtio_device *vdev = vsock->vdev;
+	__le64 guest_cid;
+
+	vdev->config->get(vdev, offsetof(struct virtio_vsock_config, guest_cid),
+			  &guest_cid, sizeof(guest_cid));
+	vsock->guest_cid = le64_to_cpu(guest_cid);
+}
+
+/* event_lock must be held */
+static void virtio_vsock_event_handle(struct virtio_vsock *vsock,
+				      struct virtio_vsock_event *event)
+{
+	switch (le32_to_cpu(event->id)) {
+	case VIRTIO_VSOCK_EVENT_TRANSPORT_RESET:
+		virtio_vsock_update_guest_cid(vsock);
+		vsock_for_each_connected_socket(&virtio_transport.transport,
+						virtio_vsock_reset_sock);
+		break;
+	}
+}
+
+static void virtio_transport_event_work(struct work_struct *work)
+{
+	struct virtio_vsock *vsock =
+		container_of(work, struct virtio_vsock, event_work);
+	struct virtqueue *vq;
+
+	mutex_lock(&vsock->event_lock);
+
+	if (!vsock->event_run)
+		goto out;
+
+	vq = vsock->vqs[VSOCK_VQ_EVENT];
+
+	do {
+		struct virtio_vsock_event *event;
+		unsigned int len;
+
+		virtqueue_disable_cb(vq);
+		while ((event = virtqueue_get_buf(vq, &len)) != NULL) {
+			if (len == sizeof(*event))
+				virtio_vsock_event_handle(vsock, event);
+
+			virtio_vsock_event_fill_one(vsock, event);
+		}
+	} while (!virtqueue_enable_cb(vq));
+
+	virtqueue_kick(vsock->vqs[VSOCK_VQ_EVENT]);
+out:
+	mutex_unlock(&vsock->event_lock);
+}
+
+static void virtio_vsock_event_done(struct virtqueue *vq)
+{
+	struct virtio_vsock *vsock = vq->vdev->priv;
+
+	if (!vsock)
+		return;
+	queue_work(virtio_vsock_workqueue, &vsock->event_work);
+}
+
+static void virtio_vsock_tx_done(struct virtqueue *vq)
+{
+	struct virtio_vsock *vsock = vq->vdev->priv;
+
+	if (!vsock)
+		return;
+	queue_work(virtio_vsock_workqueue, &vsock->tx_work);
+}
+
+static void virtio_vsock_rx_done(struct virtqueue *vq)
+{
+	struct virtio_vsock *vsock = vq->vdev->priv;
+
+	if (!vsock)
+		return;
+	queue_work(virtio_vsock_workqueue, &vsock->rx_work);
+}
+
+static bool virtio_transport_can_msgzerocopy(int bufs_num)
+{
+	struct virtio_vsock *vsock;
+	bool res = false;
+
+	rcu_read_lock();
+
+	vsock = rcu_dereference(the_virtio_vsock);
+	if (vsock) {
+		struct virtqueue *vq = vsock->vqs[VSOCK_VQ_TX];
+
+		/* Check that tx queue is large enough to keep whole
+		 * data to send. This is needed, because when there is
+		 * not enough free space in the queue, current skb to
+		 * send will be reinserted to the head of tx list of
+		 * the socket to retry transmission later, so if skb
+		 * is bigger than whole queue, it will be reinserted
+		 * again and again, thus blocking other skbs to be sent.
+		 * Each page of the user provided buffer will be added
+		 * as a single buffer to the tx virtqueue, so compare
+		 * number of pages against maximum capacity of the queue.
+		 */
+		if (bufs_num <= vq->num_max)
+			res = true;
+	}
+
+	rcu_read_unlock();
+
+	return res;
+}
+
+static bool virtio_transport_msgzerocopy_allow(void)
+{
+	return true;
+}
+
+bool virtio_transport_stream_allow(struct vsock_sock *vsk, u32 cid, u32 port)
+{
+	return vsock_net_mode_global(vsk);
+}
+
+static bool virtio_transport_seqpacket_allow(struct vsock_sock *vsk,
+					     u32 remote_cid);
+
+static bool virtio_transport_has_remote_cid(struct vsock_sock *vsk, u32 cid)
+{
+	/* The CID could be implemented by the host. Always assume it is. */
+	return true;
+}
+
+static struct virtio_transport virtio_transport = {
+	.transport = {
+		.module                   = THIS_MODULE,
+
+		.get_local_cid            = virtio_transport_get_local_cid,
+		.has_remote_cid           = virtio_transport_has_remote_cid,
+
+		.init                     = virtio_transport_do_socket_init,
+		.destruct                 = virtio_transport_destruct,
+		.release                  = virtio_transport_release,
+		.connect                  = virtio_transport_connect,
+		.shutdown                 = virtio_transport_shutdown,
+		.cancel_pkt               = virtio_transport_cancel_pkt,
+
+		.dgram_bind               = virtio_transport_dgram_bind,
+		.dgram_dequeue            = virtio_transport_dgram_dequeue,
+		.dgram_enqueue            = virtio_transport_dgram_enqueue,
+		.dgram_allow              = virtio_transport_dgram_allow,
+
+		.stream_dequeue           = virtio_transport_stream_dequeue,
+		.stream_enqueue           = virtio_transport_stream_enqueue,
+		.stream_has_data          = virtio_transport_stream_has_data,
+		.stream_has_space         = virtio_transport_stream_has_space,
+		.stream_rcvhiwat          = virtio_transport_stream_rcvhiwat,
+		.stream_is_active         = virtio_transport_stream_is_active,
+		.stream_allow             = virtio_transport_stream_allow,
+
+		.seqpacket_dequeue        = virtio_transport_seqpacket_dequeue,
+		.seqpacket_enqueue        = virtio_transport_seqpacket_enqueue,
+		.seqpacket_allow          = virtio_transport_seqpacket_allow,
+		.seqpacket_has_data       = virtio_transport_seqpacket_has_data,
+
+		.msgzerocopy_allow        = virtio_transport_msgzerocopy_allow,
+
+		.notify_poll_in           = virtio_transport_notify_poll_in,
+		.notify_poll_out          = virtio_transport_notify_poll_out,
+		.notify_recv_init         = virtio_transport_notify_recv_init,
+		.notify_recv_pre_block    = virtio_transport_notify_recv_pre_block,
+		.notify_recv_pre_dequeue  = virtio_transport_notify_recv_pre_dequeue,
+		.notify_recv_post_dequeue = virtio_transport_notify_recv_post_dequeue,
+		.notify_send_init         = virtio_transport_notify_send_init,
+		.notify_send_pre_block    = virtio_transport_notify_send_pre_block,
+		.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,
+		.notify_set_rcvlowat      = virtio_transport_notify_set_rcvlowat,
+
+		.unsent_bytes             = virtio_transport_unsent_bytes,
+
+		.read_skb = virtio_transport_read_skb,
+	},
+
+	.send_pkt = virtio_transport_send_pkt,
+	.can_msgzerocopy = virtio_transport_can_msgzerocopy,
+};
+
+static bool
+virtio_transport_seqpacket_allow(struct vsock_sock *vsk, u32 remote_cid)
+{
+	struct virtio_vsock *vsock;
+	bool seqpacket_allow;
+
+	if (!vsock_net_mode_global(vsk))
+		return false;
+
+	seqpacket_allow = false;
+	rcu_read_lock();
+	vsock = rcu_dereference(the_virtio_vsock);
+	if (vsock)
+		seqpacket_allow = vsock->seqpacket_allow;
+	rcu_read_unlock();
+
+	return seqpacket_allow;
+}
+
+static void virtio_transport_rx_work(struct work_struct *work)
+{
+	struct virtio_vsock *vsock =
+		container_of(work, struct virtio_vsock, rx_work);
+	struct virtqueue *vq;
+
+	mutex_lock(&vsock->rx_lock);
+
+	if (!vsock->rx_run)
+		goto out_nofill;
+
+	vq = vsock->vqs[VSOCK_VQ_RX];
+
+	do {
+		virtqueue_disable_cb(vq);
+		for (;;) {
+			unsigned int len, payload_len;
+			struct virtio_vsock_hdr *hdr;
+			struct sk_buff *skb;
+
+			if (!virtio_transport_more_replies(vsock)) {
+				/* Stop rx until the device processes already
+				 * pending replies.  Leave rx virtqueue
+				 * callbacks disabled.
+				 */
+				goto out;
+			}
+
+			skb = virtqueue_get_buf(vq, &len);
+			if (!skb)
+				break;
+
+			vsock->rx_buf_nr--;
+
+			/* Drop short/long packets */
+			if (unlikely(len < sizeof(*hdr) ||
+				     len > virtio_vsock_skb_len(skb))) {
+				kfree_skb(skb);
+				continue;
+			}
+
+			hdr = virtio_vsock_hdr(skb);
+			payload_len = le32_to_cpu(hdr->len);
+			if (unlikely(payload_len > len - sizeof(*hdr))) {
+				kfree_skb(skb);
+				continue;
+			}
+
+			if (payload_len)
+				virtio_vsock_skb_put(skb, payload_len);
+
+			virtio_transport_deliver_tap_pkt(skb);
+
+			/* Force virtio-transport into global mode since it
+			 * does not yet support local-mode namespacing.
+			 */
+			virtio_transport_recv_pkt(&virtio_transport, skb, NULL);
+		}
+	} while (!virtqueue_enable_cb(vq));
+
+out:
+	if (vsock->rx_buf_nr < vsock->rx_buf_max_nr / 2)
+		virtio_vsock_rx_fill(vsock);
+out_nofill:
+	mutex_unlock(&vsock->rx_lock);
+}
+
+static int virtio_vsock_vqs_init(struct virtio_vsock *vsock)
+{
+	struct virtio_device *vdev = vsock->vdev;
+	struct virtqueue_info vqs_info[] = {
+		{ "rx", virtio_vsock_rx_done },
+		{ "tx", virtio_vsock_tx_done },
+		{ "event", virtio_vsock_event_done },
+	};
+	int ret;
+
+	mutex_lock(&vsock->rx_lock);
+	vsock->rx_buf_nr = 0;
+	vsock->rx_buf_max_nr = 0;
+	mutex_unlock(&vsock->rx_lock);
+
+	atomic_set(&vsock->queued_replies, 0);
+
+	ret = virtio_find_vqs(vdev, VSOCK_VQ_MAX, vsock->vqs, vqs_info, NULL);
+	if (ret < 0)
+		return ret;
+
+	virtio_vsock_update_guest_cid(vsock);
+
+	virtio_device_ready(vdev);
+
+	return 0;
+}
+
+static void virtio_vsock_vqs_start(struct virtio_vsock *vsock)
+{
+	mutex_lock(&vsock->tx_lock);
+	vsock->tx_run = true;
+	mutex_unlock(&vsock->tx_lock);
+
+	mutex_lock(&vsock->rx_lock);
+	virtio_vsock_rx_fill(vsock);
+	vsock->rx_run = true;
+	mutex_unlock(&vsock->rx_lock);
+
+	mutex_lock(&vsock->event_lock);
+	virtio_vsock_event_fill(vsock);
+	vsock->event_run = true;
+	mutex_unlock(&vsock->event_lock);
+
+	/* virtio_transport_send_pkt() can queue packets once
+	 * the_virtio_vsock is set, but they won't be processed until
+	 * vsock->tx_run is set to true. We queue vsock->send_pkt_work
+	 * when initialization finishes to send those packets queued
+	 * earlier.
+	 * We don't need to queue the other workers (rx, event) because
+	 * as long as we don't fill the queues with empty buffers, the
+	 * host can't send us any notification.
+	 */
+	queue_work(virtio_vsock_workqueue, &vsock->send_pkt_work);
+}
+
+static void virtio_vsock_vqs_del(struct virtio_vsock *vsock)
+{
+	struct virtio_device *vdev = vsock->vdev;
+	struct sk_buff *skb;
+
+	/* Reset all connected sockets when the VQs disappear */
+	vsock_for_each_connected_socket(&virtio_transport.transport,
+					virtio_vsock_reset_sock);
+
+	/* Stop all work handlers to make sure no one is accessing the device,
+	 * so we can safely call virtio_reset_device().
+	 */
+	mutex_lock(&vsock->rx_lock);
+	vsock->rx_run = false;
+	mutex_unlock(&vsock->rx_lock);
+
+	mutex_lock(&vsock->tx_lock);
+	vsock->tx_run = false;
+	mutex_unlock(&vsock->tx_lock);
+
+	mutex_lock(&vsock->event_lock);
+	vsock->event_run = false;
+	mutex_unlock(&vsock->event_lock);
+
+	/* Flush all device writes and interrupts, device will not use any
+	 * more buffers.
+	 */
+	virtio_reset_device(vdev);
+
+	mutex_lock(&vsock->rx_lock);
+	while ((skb = virtqueue_detach_unused_buf(vsock->vqs[VSOCK_VQ_RX])))
+		kfree_skb(skb);
+	mutex_unlock(&vsock->rx_lock);
+
+	mutex_lock(&vsock->tx_lock);
+	while ((skb = virtqueue_detach_unused_buf(vsock->vqs[VSOCK_VQ_TX])))
+		kfree_skb(skb);
+	mutex_unlock(&vsock->tx_lock);
+
+	virtio_vsock_skb_queue_purge(&vsock->send_pkt_queue);
+
+	/* Delete virtqueues and flush outstanding callbacks if any */
+	vdev->config->del_vqs(vdev);
+}
+
+static int virtio_vsock_probe(struct virtio_device *vdev)
+{
+	struct virtio_vsock *vsock = NULL;
+	int ret;
+	int i;
+
+	ret = mutex_lock_interruptible(&the_virtio_vsock_mutex);
+	if (ret)
+		return ret;
+
+	/* Only one virtio-vsock device per guest is supported */
+	if (rcu_dereference_protected(the_virtio_vsock,
+				lockdep_is_held(&the_virtio_vsock_mutex))) {
+		ret = -EBUSY;
+		goto out;
+	}
+
+	vsock = kzalloc_obj(*vsock);
+	if (!vsock) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	vsock->vdev = vdev;
+
+
+	mutex_init(&vsock->tx_lock);
+	mutex_init(&vsock->rx_lock);
+	mutex_init(&vsock->event_lock);
+	skb_queue_head_init(&vsock->send_pkt_queue);
+	INIT_WORK(&vsock->rx_work, virtio_transport_rx_work);
+	INIT_WORK(&vsock->tx_work, virtio_transport_tx_work);
+	INIT_WORK(&vsock->event_work, virtio_transport_event_work);
+	INIT_WORK(&vsock->send_pkt_work, virtio_transport_send_pkt_work);
+
+	if (virtio_has_feature(vdev, VIRTIO_VSOCK_F_SEQPACKET))
+		vsock->seqpacket_allow = true;
+
+	vdev->priv = vsock;
+
+	ret = virtio_vsock_vqs_init(vsock);
+	if (ret < 0)
+		goto out;
+
+	for (i = 0; i < ARRAY_SIZE(vsock->out_sgs); i++)
+		vsock->out_sgs[i] = &vsock->out_bufs[i];
+
+	rcu_assign_pointer(the_virtio_vsock, vsock);
+	virtio_vsock_vqs_start(vsock);
+
+	mutex_unlock(&the_virtio_vsock_mutex);
+
+	return 0;
+
+out:
+	kfree(vsock);
+	mutex_unlock(&the_virtio_vsock_mutex);
+	return ret;
+}
+
+static void virtio_vsock_remove(struct virtio_device *vdev)
+{
+	struct virtio_vsock *vsock = vdev->priv;
+
+	mutex_lock(&the_virtio_vsock_mutex);
+
+	vdev->priv = NULL;
+	rcu_assign_pointer(the_virtio_vsock, NULL);
+	synchronize_rcu();
+
+	virtio_vsock_vqs_del(vsock);
+
+	/* Other works can be queued before 'config->del_vqs()', so we flush
+	 * all works before to free the vsock object to avoid use after free.
+	 */
+	flush_work(&vsock->rx_work);
+	flush_work(&vsock->tx_work);
+	flush_work(&vsock->event_work);
+	flush_work(&vsock->send_pkt_work);
+
+	mutex_unlock(&the_virtio_vsock_mutex);
+
+	kfree(vsock);
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int virtio_vsock_freeze(struct virtio_device *vdev)
+{
+	struct virtio_vsock *vsock = vdev->priv;
+
+	mutex_lock(&the_virtio_vsock_mutex);
+
+	rcu_assign_pointer(the_virtio_vsock, NULL);
+	synchronize_rcu();
+
+	virtio_vsock_vqs_del(vsock);
+
+	mutex_unlock(&the_virtio_vsock_mutex);
+
+	return 0;
+}
+
+static int virtio_vsock_restore(struct virtio_device *vdev)
+{
+	struct virtio_vsock *vsock = vdev->priv;
+	int ret;
+
+	mutex_lock(&the_virtio_vsock_mutex);
+
+	/* Only one virtio-vsock device per guest is supported */
+	if (rcu_dereference_protected(the_virtio_vsock,
+				lockdep_is_held(&the_virtio_vsock_mutex))) {
+		ret = -EBUSY;
+		goto out;
+	}
+
+	ret = virtio_vsock_vqs_init(vsock);
+	if (ret < 0)
+		goto out;
+
+	rcu_assign_pointer(the_virtio_vsock, vsock);
+	virtio_vsock_vqs_start(vsock);
+
+out:
+	mutex_unlock(&the_virtio_vsock_mutex);
+	return ret;
+}
+#endif /* CONFIG_PM_SLEEP */
+
+static struct virtio_device_id id_table[] = {
+	{ VIRTIO_ID_VSOCK, VIRTIO_DEV_ANY_ID },
+	{ 0 },
+};
+
+static unsigned int features[] = {
+	VIRTIO_VSOCK_F_SEQPACKET
+};
+
+static struct virtio_driver virtio_vsock_driver = {
+	.feature_table = features,
+	.feature_table_size = ARRAY_SIZE(features),
+	.driver.name = KBUILD_MODNAME,
+	.id_table = id_table,
+	.probe = virtio_vsock_probe,
+	.remove = virtio_vsock_remove,
+#ifdef CONFIG_PM_SLEEP
+	.freeze = virtio_vsock_freeze,
+	.restore = virtio_vsock_restore,
+#endif
+};
+
+static int __init virtio_vsock_init(void)
+{
+	int ret;
+
+	virtio_vsock_workqueue = alloc_workqueue("virtio_vsock", WQ_PERCPU, 0);
+	if (!virtio_vsock_workqueue)
+		return -ENOMEM;
+
+	ret = vsock_core_register(&virtio_transport.transport,
+				  VSOCK_TRANSPORT_F_G2H);
+	if (ret)
+		goto out_wq;
+
+	ret = register_virtio_driver(&virtio_vsock_driver);
+	if (ret)
+		goto out_vci;
+
+	return 0;
+
+out_vci:
+	vsock_core_unregister(&virtio_transport.transport);
+out_wq:
+	destroy_workqueue(virtio_vsock_workqueue);
+	return ret;
+}
+
+static void __exit virtio_vsock_exit(void)
+{
+	unregister_virtio_driver(&virtio_vsock_driver);
+	vsock_core_unregister(&virtio_transport.transport);
+	destroy_workqueue(virtio_vsock_workqueue);
+}
+
+module_init(virtio_vsock_init);
+module_exit(virtio_vsock_exit);
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR("Asias He");
+MODULE_DESCRIPTION("virtio transport for vsock");
+MODULE_DEVICE_TABLE(virtio, id_table);

linux-7.2.3/net/vmw_vsock/vmci_transport.h

@@ -0,0 +1,130 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * VMware vSockets Driver
+ *
+ * Copyright (C) 2013 VMware, Inc. All rights reserved.
+ */
+
+#ifndef _VMCI_TRANSPORT_H_
+#define _VMCI_TRANSPORT_H_
+
+#include <linux/vmw_vmci_defs.h>
+#include <linux/vmw_vmci_api.h>
+
+#include <net/vsock_addr.h>
+#include <net/af_vsock.h>
+
+/* If the packet format changes in a release then this should change too. */
+#define VMCI_TRANSPORT_PACKET_VERSION 1
+
+/* The resource ID on which control packets are sent. */
+#define VMCI_TRANSPORT_PACKET_RID 1
+
+/* The resource ID on which control packets are sent to the hypervisor. */
+#define VMCI_TRANSPORT_HYPERVISOR_PACKET_RID 15
+
+#define VSOCK_PROTO_INVALID        0
+#define VSOCK_PROTO_PKT_ON_NOTIFY (1 << 0)
+#define VSOCK_PROTO_ALL_SUPPORTED (VSOCK_PROTO_PKT_ON_NOTIFY)
+
+#define vmci_trans(_vsk) ((struct vmci_transport *)((_vsk)->trans))
+
+enum vmci_transport_packet_type {
+	VMCI_TRANSPORT_PACKET_TYPE_INVALID = 0,
+	VMCI_TRANSPORT_PACKET_TYPE_REQUEST,
+	VMCI_TRANSPORT_PACKET_TYPE_NEGOTIATE,
+	VMCI_TRANSPORT_PACKET_TYPE_OFFER,
+	VMCI_TRANSPORT_PACKET_TYPE_ATTACH,
+	VMCI_TRANSPORT_PACKET_TYPE_WROTE,
+	VMCI_TRANSPORT_PACKET_TYPE_READ,
+	VMCI_TRANSPORT_PACKET_TYPE_RST,
+	VMCI_TRANSPORT_PACKET_TYPE_SHUTDOWN,
+	VMCI_TRANSPORT_PACKET_TYPE_WAITING_WRITE,
+	VMCI_TRANSPORT_PACKET_TYPE_WAITING_READ,
+	VMCI_TRANSPORT_PACKET_TYPE_REQUEST2,
+	VMCI_TRANSPORT_PACKET_TYPE_NEGOTIATE2,
+	VMCI_TRANSPORT_PACKET_TYPE_MAX
+};
+
+struct vmci_transport_waiting_info {
+	u64 generation;
+	u64 offset;
+};
+
+/* Control packet type for STREAM sockets.  DGRAMs have no control packets nor
+ * special packet header for data packets, they are just raw VMCI DGRAM
+ * messages.  For STREAMs, control packets are sent over the control channel
+ * while data is written and read directly from queue pairs with no packet
+ * format.
+ */
+struct vmci_transport_packet {
+	struct vmci_datagram dg;
+	u8 version;
+	u8 type;
+	u16 proto;
+	u32 src_port;
+	u32 dst_port;
+	u32 _reserved2;
+	union {
+		u64 size;
+		u64 mode;
+		struct vmci_handle handle;
+		struct vmci_transport_waiting_info wait;
+	} u;
+};
+
+struct vmci_transport_notify_pkt {
+	u64 write_notify_window;
+	u64 write_notify_min_window;
+	bool peer_waiting_read;
+	bool peer_waiting_write;
+	bool peer_waiting_write_detected;
+	bool sent_waiting_read;
+	bool sent_waiting_write;
+	struct vmci_transport_waiting_info peer_waiting_read_info;
+	struct vmci_transport_waiting_info peer_waiting_write_info;
+	u64 produce_q_generation;
+	u64 consume_q_generation;
+};
+
+struct vmci_transport_notify_pkt_q_state {
+	u64 write_notify_window;
+	u64 write_notify_min_window;
+	bool peer_waiting_write;
+	bool peer_waiting_write_detected;
+};
+
+union vmci_transport_notify {
+	struct vmci_transport_notify_pkt pkt;
+	struct vmci_transport_notify_pkt_q_state pkt_q_state;
+};
+
+/* Our transport-specific data. */
+struct vmci_transport {
+	/* For DGRAMs. */
+	struct vmci_handle dg_handle;
+	/* For STREAMs. */
+	struct vmci_handle qp_handle;
+	struct vmci_qp *qpair;
+	u64 produce_size;
+	u64 consume_size;
+	u32 detach_sub_id;
+	union vmci_transport_notify notify;
+	const struct vmci_transport_notify_ops *notify_ops;
+	struct list_head elem;
+	struct sock *sk;
+	spinlock_t lock; /* protects sk. */
+};
+
+int vmci_transport_send_wrote_bh(struct sockaddr_vm *dst,
+				 struct sockaddr_vm *src);
+int vmci_transport_send_read_bh(struct sockaddr_vm *dst,
+				struct sockaddr_vm *src);
+int vmci_transport_send_wrote(struct sock *sk);
+int vmci_transport_send_read(struct sock *sk);
+int vmci_transport_send_waiting_write(struct sock *sk,
+				      struct vmci_transport_waiting_info *wait);
+int vmci_transport_send_waiting_read(struct sock *sk,
+				     struct vmci_transport_waiting_info *wait);
+
+#endif

linux-7.2.3/net/vmw_vsock/vmci_transport_notify.c

@@ -0,0 +1,672 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * VMware vSockets Driver
+ *
+ * Copyright (C) 2009-2013 VMware, Inc. All rights reserved.
+ */
+
+#include <linux/types.h>
+#include <linux/socket.h>
+#include <linux/stddef.h>
+#include <net/sock.h>
+
+#include "vmci_transport_notify.h"
+
+#define PKT_FIELD(vsk, field_name) (vmci_trans(vsk)->notify.pkt.field_name)
+
+static bool vmci_transport_notify_waiting_write(struct vsock_sock *vsk)
+{
+#if defined(VSOCK_OPTIMIZATION_WAITING_NOTIFY)
+	bool retval;
+	u64 notify_limit;
+
+	if (!PKT_FIELD(vsk, peer_waiting_write))
+		return false;
+
+#ifdef VSOCK_OPTIMIZATION_FLOW_CONTROL
+	/* When the sender blocks, we take that as a sign that the sender is
+	 * faster than the receiver. To reduce the transmit rate of the sender,
+	 * we delay the sending of the read notification by decreasing the
+	 * write_notify_window. The notification is delayed until the number of
+	 * bytes used in the queue drops below the write_notify_window.
+	 */
+
+	if (!PKT_FIELD(vsk, peer_waiting_write_detected)) {
+		PKT_FIELD(vsk, peer_waiting_write_detected) = true;
+		if (PKT_FIELD(vsk, write_notify_window) < PAGE_SIZE) {
+			PKT_FIELD(vsk, write_notify_window) =
+			    PKT_FIELD(vsk, write_notify_min_window);
+		} else {
+			PKT_FIELD(vsk, write_notify_window) -= PAGE_SIZE;
+			if (PKT_FIELD(vsk, write_notify_window) <
+			    PKT_FIELD(vsk, write_notify_min_window))
+				PKT_FIELD(vsk, write_notify_window) =
+				    PKT_FIELD(vsk, write_notify_min_window);
+
+		}
+	}
+	notify_limit = vmci_trans(vsk)->consume_size -
+		PKT_FIELD(vsk, write_notify_window);
+#else
+	notify_limit = 0;
+#endif
+
+	/* For now we ignore the wait information and just see if the free
+	 * space exceeds the notify limit.  Note that improving this function
+	 * to be more intelligent will not require a protocol change and will
+	 * retain compatibility between endpoints with mixed versions of this
+	 * function.
+	 *
+	 * The notify_limit is used to delay notifications in the case where
+	 * flow control is enabled. Below the test is expressed in terms of
+	 * free space in the queue: if free_space > ConsumeSize -
+	 * write_notify_window then notify An alternate way of expressing this
+	 * is to rewrite the expression to use the data ready in the receive
+	 * queue: if write_notify_window > bufferReady then notify as
+	 * free_space == ConsumeSize - bufferReady.
+	 */
+	retval = vmci_qpair_consume_free_space(vmci_trans(vsk)->qpair) >
+		notify_limit;
+#ifdef VSOCK_OPTIMIZATION_FLOW_CONTROL
+	if (retval) {
+		/*
+		 * Once we notify the peer, we reset the detected flag so the
+		 * next wait will again cause a decrease in the window size.
+		 */
+
+		PKT_FIELD(vsk, peer_waiting_write_detected) = false;
+	}
+#endif
+	return retval;
+#else
+	return true;
+#endif
+}
+
+static bool vmci_transport_notify_waiting_read(struct vsock_sock *vsk)
+{
+#if defined(VSOCK_OPTIMIZATION_WAITING_NOTIFY)
+	if (!PKT_FIELD(vsk, peer_waiting_read))
+		return false;
+
+	/* For now we ignore the wait information and just see if there is any
+	 * data for our peer to read.  Note that improving this function to be
+	 * more intelligent will not require a protocol change and will retain
+	 * compatibility between endpoints with mixed versions of this
+	 * function.
+	 */
+	return vmci_qpair_produce_buf_ready(vmci_trans(vsk)->qpair) > 0;
+#else
+	return true;
+#endif
+}
+
+static void
+vmci_transport_handle_waiting_read(struct sock *sk,
+				   struct vmci_transport_packet *pkt,
+				   bool bottom_half,
+				   struct sockaddr_vm *dst,
+				   struct sockaddr_vm *src)
+{
+#if defined(VSOCK_OPTIMIZATION_WAITING_NOTIFY)
+	struct vsock_sock *vsk;
+
+	vsk = vsock_sk(sk);
+
+	PKT_FIELD(vsk, peer_waiting_read) = true;
+	memcpy(&PKT_FIELD(vsk, peer_waiting_read_info), &pkt->u.wait,
+	       sizeof(PKT_FIELD(vsk, peer_waiting_read_info)));
+
+	if (vmci_transport_notify_waiting_read(vsk)) {
+		bool sent;
+
+		if (bottom_half)
+			sent = vmci_transport_send_wrote_bh(dst, src) > 0;
+		else
+			sent = vmci_transport_send_wrote(sk) > 0;
+
+		if (sent)
+			PKT_FIELD(vsk, peer_waiting_read) = false;
+	}
+#endif
+}
+
+static void
+vmci_transport_handle_waiting_write(struct sock *sk,
+				    struct vmci_transport_packet *pkt,
+				    bool bottom_half,
+				    struct sockaddr_vm *dst,
+				    struct sockaddr_vm *src)
+{
+#if defined(VSOCK_OPTIMIZATION_WAITING_NOTIFY)
+	struct vsock_sock *vsk;
+
+	vsk = vsock_sk(sk);
+
+	PKT_FIELD(vsk, peer_waiting_write) = true;
+	memcpy(&PKT_FIELD(vsk, peer_waiting_write_info), &pkt->u.wait,
+	       sizeof(PKT_FIELD(vsk, peer_waiting_write_info)));
+
+	if (vmci_transport_notify_waiting_write(vsk)) {
+		bool sent;
+
+		if (bottom_half)
+			sent = vmci_transport_send_read_bh(dst, src) > 0;
+		else
+			sent = vmci_transport_send_read(sk) > 0;
+
+		if (sent)
+			PKT_FIELD(vsk, peer_waiting_write) = false;
+	}
+#endif
+}
+
+static void
+vmci_transport_handle_read(struct sock *sk,
+			   struct vmci_transport_packet *pkt,
+			   bool bottom_half,
+			   struct sockaddr_vm *dst, struct sockaddr_vm *src)
+{
+#if defined(VSOCK_OPTIMIZATION_WAITING_NOTIFY)
+	struct vsock_sock *vsk;
+
+	vsk = vsock_sk(sk);
+	PKT_FIELD(vsk, sent_waiting_write) = false;
+#endif
+
+	sk->sk_write_space(sk);
+}
+
+static bool send_waiting_read(struct sock *sk, u64 room_needed)
+{
+#if defined(VSOCK_OPTIMIZATION_WAITING_NOTIFY)
+	struct vsock_sock *vsk;
+	struct vmci_transport_waiting_info waiting_info;
+	u64 tail;
+	u64 head;
+	u64 room_left;
+	bool ret;
+
+	vsk = vsock_sk(sk);
+
+	if (PKT_FIELD(vsk, sent_waiting_read))
+		return true;
+
+	if (PKT_FIELD(vsk, write_notify_window) <
+			vmci_trans(vsk)->consume_size)
+		PKT_FIELD(vsk, write_notify_window) =
+		    min(PKT_FIELD(vsk, write_notify_window) + PAGE_SIZE,
+			vmci_trans(vsk)->consume_size);
+
+	vmci_qpair_get_consume_indexes(vmci_trans(vsk)->qpair, &tail, &head);
+	room_left = vmci_trans(vsk)->consume_size - head;
+	if (room_needed >= room_left) {
+		waiting_info.offset = room_needed - room_left;
+		waiting_info.generation =
+		    PKT_FIELD(vsk, consume_q_generation) + 1;
+	} else {
+		waiting_info.offset = head + room_needed;
+		waiting_info.generation = PKT_FIELD(vsk, consume_q_generation);
+	}
+
+	ret = vmci_transport_send_waiting_read(sk, &waiting_info) > 0;
+	if (ret)
+		PKT_FIELD(vsk, sent_waiting_read) = true;
+
+	return ret;
+#else
+	return true;
+#endif
+}
+
+static bool send_waiting_write(struct sock *sk, u64 room_needed)
+{
+#if defined(VSOCK_OPTIMIZATION_WAITING_NOTIFY)
+	struct vsock_sock *vsk;
+	struct vmci_transport_waiting_info waiting_info;
+	u64 tail;
+	u64 head;
+	u64 room_left;
+	bool ret;
+
+	vsk = vsock_sk(sk);
+
+	if (PKT_FIELD(vsk, sent_waiting_write))
+		return true;
+
+	vmci_qpair_get_produce_indexes(vmci_trans(vsk)->qpair, &tail, &head);
+	room_left = vmci_trans(vsk)->produce_size - tail;
+	if (room_needed + 1 >= room_left) {
+		/* Wraps around to current generation. */
+		waiting_info.offset = room_needed + 1 - room_left;
+		waiting_info.generation = PKT_FIELD(vsk, produce_q_generation);
+	} else {
+		waiting_info.offset = tail + room_needed + 1;
+		waiting_info.generation =
+		    PKT_FIELD(vsk, produce_q_generation) - 1;
+	}
+
+	ret = vmci_transport_send_waiting_write(sk, &waiting_info) > 0;
+	if (ret)
+		PKT_FIELD(vsk, sent_waiting_write) = true;
+
+	return ret;
+#else
+	return true;
+#endif
+}
+
+static int vmci_transport_send_read_notification(struct sock *sk)
+{
+	struct vsock_sock *vsk;
+	bool sent_read;
+	unsigned int retries;
+	int err;
+
+	vsk = vsock_sk(sk);
+	sent_read = false;
+	retries = 0;
+	err = 0;
+
+	if (vmci_transport_notify_waiting_write(vsk)) {
+		/* Notify the peer that we have read, retrying the send on
+		 * failure up to our maximum value.  XXX For now we just log
+		 * the failure, but later we should schedule a work item to
+		 * handle the resend until it succeeds.  That would require
+		 * keeping track of work items in the vsk and cleaning them up
+		 * upon socket close.
+		 */
+		while (!(vsk->peer_shutdown & RCV_SHUTDOWN) &&
+		       !sent_read &&
+		       retries < VMCI_TRANSPORT_MAX_DGRAM_RESENDS) {
+			err = vmci_transport_send_read(sk);
+			if (err >= 0)
+				sent_read = true;
+
+			retries++;
+		}
+
+		if (retries >= VMCI_TRANSPORT_MAX_DGRAM_RESENDS)
+			pr_err("%p unable to send read notify to peer\n", sk);
+		else
+#if defined(VSOCK_OPTIMIZATION_WAITING_NOTIFY)
+			PKT_FIELD(vsk, peer_waiting_write) = false;
+#endif
+
+	}
+	return err;
+}
+
+static void
+vmci_transport_handle_wrote(struct sock *sk,
+			    struct vmci_transport_packet *pkt,
+			    bool bottom_half,
+			    struct sockaddr_vm *dst, struct sockaddr_vm *src)
+{
+#if defined(VSOCK_OPTIMIZATION_WAITING_NOTIFY)
+	struct vsock_sock *vsk = vsock_sk(sk);
+	PKT_FIELD(vsk, sent_waiting_read) = false;
+#endif
+	vsock_data_ready(sk);
+}
+
+static void vmci_transport_notify_pkt_socket_init(struct sock *sk)
+{
+	struct vsock_sock *vsk = vsock_sk(sk);
+
+	PKT_FIELD(vsk, write_notify_window) = PAGE_SIZE;
+	PKT_FIELD(vsk, write_notify_min_window) = PAGE_SIZE;
+	PKT_FIELD(vsk, peer_waiting_read) = false;
+	PKT_FIELD(vsk, peer_waiting_write) = false;
+	PKT_FIELD(vsk, peer_waiting_write_detected) = false;
+	PKT_FIELD(vsk, sent_waiting_read) = false;
+	PKT_FIELD(vsk, sent_waiting_write) = false;
+	PKT_FIELD(vsk, produce_q_generation) = 0;
+	PKT_FIELD(vsk, consume_q_generation) = 0;
+
+	memset(&PKT_FIELD(vsk, peer_waiting_read_info), 0,
+	       sizeof(PKT_FIELD(vsk, peer_waiting_read_info)));
+	memset(&PKT_FIELD(vsk, peer_waiting_write_info), 0,
+	       sizeof(PKT_FIELD(vsk, peer_waiting_write_info)));
+}
+
+static void vmci_transport_notify_pkt_socket_destruct(struct vsock_sock *vsk)
+{
+}
+
+static int
+vmci_transport_notify_pkt_poll_in(struct sock *sk,
+				  size_t target, bool *data_ready_now)
+{
+	struct vsock_sock *vsk = vsock_sk(sk);
+
+	if (vsock_stream_has_data(vsk) >= target) {
+		*data_ready_now = true;
+	} else {
+		/* 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))
+				return -1;
+
+		}
+		*data_ready_now = false;
+	}
+
+	return 0;
+}
+
+static int
+vmci_transport_notify_pkt_poll_out(struct sock *sk,
+				   size_t target, bool *space_avail_now)
+{
+	s64 produce_q_free_space;
+	struct vsock_sock *vsk = vsock_sk(sk);
+
+	produce_q_free_space = vsock_stream_has_space(vsk);
+	if (produce_q_free_space > 0) {
+		*space_avail_now = true;
+		return 0;
+	} else if (produce_q_free_space == 0) {
+		/* This is a connected socket but we can't currently send data.
+		 * Notify the peer that we are waiting if the queue is full. We
+		 * only send a waiting write if the queue is full because
+		 * otherwise we end up in an infinite WAITING_WRITE, READ,
+		 * WAITING_WRITE, READ, etc. loop. Treat failing to send the
+		 * notification as a socket error, passing that back through
+		 * the mask.
+		 */
+		if (!send_waiting_write(sk, 1))
+			return -1;
+
+		*space_avail_now = false;
+	}
+
+	return 0;
+}
+
+static int
+vmci_transport_notify_pkt_recv_init(
+			struct sock *sk,
+			size_t target,
+			struct vmci_transport_recv_notify_data *data)
+{
+	struct vsock_sock *vsk = vsock_sk(sk);
+
+#ifdef VSOCK_OPTIMIZATION_WAITING_NOTIFY
+	data->consume_head = 0;
+	data->produce_tail = 0;
+#ifdef VSOCK_OPTIMIZATION_FLOW_CONTROL
+	data->notify_on_block = false;
+
+	if (PKT_FIELD(vsk, write_notify_min_window) < target + 1) {
+		PKT_FIELD(vsk, write_notify_min_window) = target + 1;
+		if (PKT_FIELD(vsk, write_notify_window) <
+		    PKT_FIELD(vsk, write_notify_min_window)) {
+			/* If the current window is smaller than the new
+			 * minimal window size, we need to reevaluate whether
+			 * we need to notify the sender. If the number of ready
+			 * bytes are smaller than the new window, we need to
+			 * send a notification to the sender before we block.
+			 */
+
+			PKT_FIELD(vsk, write_notify_window) =
+			    PKT_FIELD(vsk, write_notify_min_window);
+			data->notify_on_block = true;
+		}
+	}
+#endif
+#endif
+
+	return 0;
+}
+
+static int
+vmci_transport_notify_pkt_recv_pre_block(
+				struct sock *sk,
+				size_t target,
+				struct vmci_transport_recv_notify_data *data)
+{
+	int err = 0;
+
+	/* Notify our peer that we are waiting for data to read. */
+	if (!send_waiting_read(sk, target)) {
+		err = -EHOSTUNREACH;
+		return err;
+	}
+#ifdef VSOCK_OPTIMIZATION_FLOW_CONTROL
+	if (data->notify_on_block) {
+		err = vmci_transport_send_read_notification(sk);
+		if (err < 0)
+			return err;
+
+		data->notify_on_block = false;
+	}
+#endif
+
+	return err;
+}
+
+static int
+vmci_transport_notify_pkt_recv_pre_dequeue(
+				struct sock *sk,
+				size_t target,
+				struct vmci_transport_recv_notify_data *data)
+{
+	struct vsock_sock *vsk = vsock_sk(sk);
+
+	/* Now consume up to len bytes from the queue.  Note that since we have
+	 * the socket locked we should copy at least ready bytes.
+	 */
+#if defined(VSOCK_OPTIMIZATION_WAITING_NOTIFY)
+	vmci_qpair_get_consume_indexes(vmci_trans(vsk)->qpair,
+				       &data->produce_tail,
+				       &data->consume_head);
+#endif
+
+	return 0;
+}
+
+static int
+vmci_transport_notify_pkt_recv_post_dequeue(
+				struct sock *sk,
+				size_t target,
+				ssize_t copied,
+				bool data_read,
+				struct vmci_transport_recv_notify_data *data)
+{
+	struct vsock_sock *vsk;
+	int err;
+
+	vsk = vsock_sk(sk);
+	err = 0;
+
+	if (data_read) {
+#if defined(VSOCK_OPTIMIZATION_WAITING_NOTIFY)
+		/* Detect a wrap-around to maintain queue generation.  Note
+		 * that this is safe since we hold the socket lock across the
+		 * two queue pair operations.
+		 */
+		if (copied >=
+			vmci_trans(vsk)->consume_size - data->consume_head)
+			PKT_FIELD(vsk, consume_q_generation)++;
+#endif
+
+		err = vmci_transport_send_read_notification(sk);
+		if (err < 0)
+			return err;
+
+	}
+	return err;
+}
+
+static int
+vmci_transport_notify_pkt_send_init(
+			struct sock *sk,
+			struct vmci_transport_send_notify_data *data)
+{
+#ifdef VSOCK_OPTIMIZATION_WAITING_NOTIFY
+	data->consume_head = 0;
+	data->produce_tail = 0;
+#endif
+
+	return 0;
+}
+
+static int
+vmci_transport_notify_pkt_send_pre_block(
+				struct sock *sk,
+				struct vmci_transport_send_notify_data *data)
+{
+	/* Notify our peer that we are waiting for room to write. */
+	if (!send_waiting_write(sk, 1))
+		return -EHOSTUNREACH;
+
+	return 0;
+}
+
+static int
+vmci_transport_notify_pkt_send_pre_enqueue(
+				struct sock *sk,
+				struct vmci_transport_send_notify_data *data)
+{
+	struct vsock_sock *vsk = vsock_sk(sk);
+
+#if defined(VSOCK_OPTIMIZATION_WAITING_NOTIFY)
+	vmci_qpair_get_produce_indexes(vmci_trans(vsk)->qpair,
+				       &data->produce_tail,
+				       &data->consume_head);
+#endif
+
+	return 0;
+}
+
+static int
+vmci_transport_notify_pkt_send_post_enqueue(
+				struct sock *sk,
+				ssize_t written,
+				struct vmci_transport_send_notify_data *data)
+{
+	int err = 0;
+	struct vsock_sock *vsk;
+	bool sent_wrote = false;
+	int retries = 0;
+
+	vsk = vsock_sk(sk);
+
+#if defined(VSOCK_OPTIMIZATION_WAITING_NOTIFY)
+	/* Detect a wrap-around to maintain queue generation.  Note that this
+	 * is safe since we hold the socket lock across the two queue pair
+	 * operations.
+	 */
+	if (written >= vmci_trans(vsk)->produce_size - data->produce_tail)
+		PKT_FIELD(vsk, produce_q_generation)++;
+
+#endif
+
+	if (vmci_transport_notify_waiting_read(vsk)) {
+		/* Notify the peer that we have written, retrying the send on
+		 * failure up to our maximum value. See the XXX comment for the
+		 * corresponding piece of code in StreamRecvmsg() for potential
+		 * improvements.
+		 */
+		while (!(vsk->peer_shutdown & RCV_SHUTDOWN) &&
+		       !sent_wrote &&
+		       retries < VMCI_TRANSPORT_MAX_DGRAM_RESENDS) {
+			err = vmci_transport_send_wrote(sk);
+			if (err >= 0)
+				sent_wrote = true;
+
+			retries++;
+		}
+
+		if (retries >= VMCI_TRANSPORT_MAX_DGRAM_RESENDS) {
+			pr_err("%p unable to send wrote notify to peer\n", sk);
+			return err;
+		} else {
+#if defined(VSOCK_OPTIMIZATION_WAITING_NOTIFY)
+			PKT_FIELD(vsk, peer_waiting_read) = false;
+#endif
+		}
+	}
+	return err;
+}
+
+static void
+vmci_transport_notify_pkt_handle_pkt(
+			struct sock *sk,
+			struct vmci_transport_packet *pkt,
+			bool bottom_half,
+			struct sockaddr_vm *dst,
+			struct sockaddr_vm *src, bool *pkt_processed)
+{
+	bool processed = false;
+
+	switch (pkt->type) {
+	case VMCI_TRANSPORT_PACKET_TYPE_WROTE:
+		vmci_transport_handle_wrote(sk, pkt, bottom_half, dst, src);
+		processed = true;
+		break;
+	case VMCI_TRANSPORT_PACKET_TYPE_READ:
+		vmci_transport_handle_read(sk, pkt, bottom_half, dst, src);
+		processed = true;
+		break;
+	case VMCI_TRANSPORT_PACKET_TYPE_WAITING_WRITE:
+		vmci_transport_handle_waiting_write(sk, pkt, bottom_half,
+						    dst, src);
+		processed = true;
+		break;
+
+	case VMCI_TRANSPORT_PACKET_TYPE_WAITING_READ:
+		vmci_transport_handle_waiting_read(sk, pkt, bottom_half,
+						   dst, src);
+		processed = true;
+		break;
+	}
+
+	if (pkt_processed)
+		*pkt_processed = processed;
+}
+
+static void vmci_transport_notify_pkt_process_request(struct sock *sk)
+{
+	struct vsock_sock *vsk = vsock_sk(sk);
+
+	PKT_FIELD(vsk, write_notify_window) = vmci_trans(vsk)->consume_size;
+	if (vmci_trans(vsk)->consume_size <
+		PKT_FIELD(vsk, write_notify_min_window))
+		PKT_FIELD(vsk, write_notify_min_window) =
+			vmci_trans(vsk)->consume_size;
+}
+
+static void vmci_transport_notify_pkt_process_negotiate(struct sock *sk)
+{
+	struct vsock_sock *vsk = vsock_sk(sk);
+
+	PKT_FIELD(vsk, write_notify_window) = vmci_trans(vsk)->consume_size;
+	if (vmci_trans(vsk)->consume_size <
+		PKT_FIELD(vsk, write_notify_min_window))
+		PKT_FIELD(vsk, write_notify_min_window) =
+			vmci_trans(vsk)->consume_size;
+}
+
+/* Socket control packet based operations. */
+const struct vmci_transport_notify_ops vmci_transport_notify_pkt_ops = {
+	.socket_init = vmci_transport_notify_pkt_socket_init,
+	.socket_destruct = vmci_transport_notify_pkt_socket_destruct,
+	.poll_in = vmci_transport_notify_pkt_poll_in,
+	.poll_out = vmci_transport_notify_pkt_poll_out,
+	.handle_notify_pkt = vmci_transport_notify_pkt_handle_pkt,
+	.recv_init = vmci_transport_notify_pkt_recv_init,
+	.recv_pre_block = vmci_transport_notify_pkt_recv_pre_block,
+	.recv_pre_dequeue = vmci_transport_notify_pkt_recv_pre_dequeue,
+	.recv_post_dequeue = vmci_transport_notify_pkt_recv_post_dequeue,
+	.send_init = vmci_transport_notify_pkt_send_init,
+	.send_pre_block = vmci_transport_notify_pkt_send_pre_block,
+	.send_pre_enqueue = vmci_transport_notify_pkt_send_pre_enqueue,
+	.send_post_enqueue = vmci_transport_notify_pkt_send_post_enqueue,
+	.process_request = vmci_transport_notify_pkt_process_request,
+	.process_negotiate = vmci_transport_notify_pkt_process_negotiate,
+};

linux-7.2.3/net/vmw_vsock/vmci_transport_notify.h

@@ -0,0 +1,75 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * VMware vSockets Driver
+ *
+ * Copyright (C) 2009-2013 VMware, Inc. All rights reserved.
+ */
+
+#ifndef __VMCI_TRANSPORT_NOTIFY_H__
+#define __VMCI_TRANSPORT_NOTIFY_H__
+
+#include <linux/types.h>
+#include <linux/vmw_vmci_defs.h>
+#include <linux/vmw_vmci_api.h>
+
+#include "vmci_transport.h"
+
+/* Comment this out to compare with old protocol. */
+#define VSOCK_OPTIMIZATION_WAITING_NOTIFY 1
+#if defined(VSOCK_OPTIMIZATION_WAITING_NOTIFY)
+/* Comment this out to remove flow control for "new" protocol */
+#define VSOCK_OPTIMIZATION_FLOW_CONTROL 1
+#endif
+
+#define VMCI_TRANSPORT_MAX_DGRAM_RESENDS       10
+
+struct vmci_transport_recv_notify_data {
+	u64 consume_head;
+	u64 produce_tail;
+	bool notify_on_block;
+};
+
+struct vmci_transport_send_notify_data {
+	u64 consume_head;
+	u64 produce_tail;
+};
+
+/* Socket notification callbacks. */
+struct vmci_transport_notify_ops {
+	void (*socket_init) (struct sock *sk);
+	void (*socket_destruct) (struct vsock_sock *vsk);
+	int (*poll_in) (struct sock *sk, size_t target,
+			  bool *data_ready_now);
+	int (*poll_out) (struct sock *sk, size_t target,
+			   bool *space_avail_now);
+	void (*handle_notify_pkt) (struct sock *sk,
+				   struct vmci_transport_packet *pkt,
+				   bool bottom_half, struct sockaddr_vm *dst,
+				   struct sockaddr_vm *src,
+				   bool *pkt_processed);
+	int (*recv_init) (struct sock *sk, size_t target,
+			  struct vmci_transport_recv_notify_data *data);
+	int (*recv_pre_block) (struct sock *sk, size_t target,
+			       struct vmci_transport_recv_notify_data *data);
+	int (*recv_pre_dequeue) (struct sock *sk, size_t target,
+				 struct vmci_transport_recv_notify_data *data);
+	int (*recv_post_dequeue) (struct sock *sk, size_t target,
+				  ssize_t copied, bool data_read,
+				  struct vmci_transport_recv_notify_data *data);
+	int (*send_init) (struct sock *sk,
+			  struct vmci_transport_send_notify_data *data);
+	int (*send_pre_block) (struct sock *sk,
+			       struct vmci_transport_send_notify_data *data);
+	int (*send_pre_enqueue) (struct sock *sk,
+				 struct vmci_transport_send_notify_data *data);
+	int (*send_post_enqueue) (struct sock *sk, ssize_t written,
+				  struct vmci_transport_send_notify_data *data);
+	void (*process_request) (struct sock *sk);
+	void (*process_negotiate) (struct sock *sk);
+};
+
+extern const struct vmci_transport_notify_ops vmci_transport_notify_pkt_ops;
+extern const
+struct vmci_transport_notify_ops vmci_transport_notify_pkt_q_state_ops;
+
+#endif /* __VMCI_TRANSPORT_NOTIFY_H__ */

linux-7.2.3/net/vmw_vsock/vmci_transport_notify_qstate.c

@@ -0,0 +1,430 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * VMware vSockets Driver
+ *
+ * Copyright (C) 2009-2013 VMware, Inc. All rights reserved.
+ */
+
+#include <linux/types.h>
+#include <linux/socket.h>
+#include <linux/stddef.h>
+#include <net/sock.h>
+
+#include "vmci_transport_notify.h"
+
+#define PKT_FIELD(vsk, field_name) \
+	(vmci_trans(vsk)->notify.pkt_q_state.field_name)
+
+static bool vmci_transport_notify_waiting_write(struct vsock_sock *vsk)
+{
+	bool retval;
+	u64 notify_limit;
+
+	if (!PKT_FIELD(vsk, peer_waiting_write))
+		return false;
+
+	/* When the sender blocks, we take that as a sign that the sender is
+	 * faster than the receiver. To reduce the transmit rate of the sender,
+	 * we delay the sending of the read notification by decreasing the
+	 * write_notify_window. The notification is delayed until the number of
+	 * bytes used in the queue drops below the write_notify_window.
+	 */
+
+	if (!PKT_FIELD(vsk, peer_waiting_write_detected)) {
+		PKT_FIELD(vsk, peer_waiting_write_detected) = true;
+		if (PKT_FIELD(vsk, write_notify_window) < PAGE_SIZE) {
+			PKT_FIELD(vsk, write_notify_window) =
+			    PKT_FIELD(vsk, write_notify_min_window);
+		} else {
+			PKT_FIELD(vsk, write_notify_window) -= PAGE_SIZE;
+			if (PKT_FIELD(vsk, write_notify_window) <
+			    PKT_FIELD(vsk, write_notify_min_window))
+				PKT_FIELD(vsk, write_notify_window) =
+				    PKT_FIELD(vsk, write_notify_min_window);
+
+		}
+	}
+	notify_limit = vmci_trans(vsk)->consume_size -
+		PKT_FIELD(vsk, write_notify_window);
+
+	/* The notify_limit is used to delay notifications in the case where
+	 * flow control is enabled. Below the test is expressed in terms of
+	 * free space in the queue: if free_space > ConsumeSize -
+	 * write_notify_window then notify An alternate way of expressing this
+	 * is to rewrite the expression to use the data ready in the receive
+	 * queue: if write_notify_window > bufferReady then notify as
+	 * free_space == ConsumeSize - bufferReady.
+	 */
+
+	retval = vmci_qpair_consume_free_space(vmci_trans(vsk)->qpair) >
+		notify_limit;
+
+	if (retval) {
+		/* Once we notify the peer, we reset the detected flag so the
+		 * next wait will again cause a decrease in the window size.
+		 */
+
+		PKT_FIELD(vsk, peer_waiting_write_detected) = false;
+	}
+	return retval;
+}
+
+static void
+vmci_transport_handle_read(struct sock *sk,
+			   struct vmci_transport_packet *pkt,
+			   bool bottom_half,
+			   struct sockaddr_vm *dst, struct sockaddr_vm *src)
+{
+	sk->sk_write_space(sk);
+}
+
+static void
+vmci_transport_handle_wrote(struct sock *sk,
+			    struct vmci_transport_packet *pkt,
+			    bool bottom_half,
+			    struct sockaddr_vm *dst, struct sockaddr_vm *src)
+{
+	vsock_data_ready(sk);
+}
+
+static void vsock_block_update_write_window(struct sock *sk)
+{
+	struct vsock_sock *vsk = vsock_sk(sk);
+
+	if (PKT_FIELD(vsk, write_notify_window) < vmci_trans(vsk)->consume_size)
+		PKT_FIELD(vsk, write_notify_window) =
+		    min(PKT_FIELD(vsk, write_notify_window) + PAGE_SIZE,
+			vmci_trans(vsk)->consume_size);
+}
+
+static int vmci_transport_send_read_notification(struct sock *sk)
+{
+	struct vsock_sock *vsk;
+	bool sent_read;
+	unsigned int retries;
+	int err;
+
+	vsk = vsock_sk(sk);
+	sent_read = false;
+	retries = 0;
+	err = 0;
+
+	if (vmci_transport_notify_waiting_write(vsk)) {
+		/* Notify the peer that we have read, retrying the send on
+		 * failure up to our maximum value.  XXX For now we just log
+		 * the failure, but later we should schedule a work item to
+		 * handle the resend until it succeeds.  That would require
+		 * keeping track of work items in the vsk and cleaning them up
+		 * upon socket close.
+		 */
+		while (!(vsk->peer_shutdown & RCV_SHUTDOWN) &&
+		       !sent_read &&
+		       retries < VMCI_TRANSPORT_MAX_DGRAM_RESENDS) {
+			err = vmci_transport_send_read(sk);
+			if (err >= 0)
+				sent_read = true;
+
+			retries++;
+		}
+
+		if (retries >= VMCI_TRANSPORT_MAX_DGRAM_RESENDS && !sent_read)
+			pr_err("%p unable to send read notification to peer\n",
+			       sk);
+		else
+			PKT_FIELD(vsk, peer_waiting_write) = false;
+
+	}
+	return err;
+}
+
+static void vmci_transport_notify_pkt_socket_init(struct sock *sk)
+{
+	struct vsock_sock *vsk = vsock_sk(sk);
+
+	PKT_FIELD(vsk, write_notify_window) = PAGE_SIZE;
+	PKT_FIELD(vsk, write_notify_min_window) = PAGE_SIZE;
+	PKT_FIELD(vsk, peer_waiting_write) = false;
+	PKT_FIELD(vsk, peer_waiting_write_detected) = false;
+}
+
+static void vmci_transport_notify_pkt_socket_destruct(struct vsock_sock *vsk)
+{
+	PKT_FIELD(vsk, write_notify_window) = PAGE_SIZE;
+	PKT_FIELD(vsk, write_notify_min_window) = PAGE_SIZE;
+	PKT_FIELD(vsk, peer_waiting_write) = false;
+	PKT_FIELD(vsk, peer_waiting_write_detected) = false;
+}
+
+static int
+vmci_transport_notify_pkt_poll_in(struct sock *sk,
+				  size_t target, bool *data_ready_now)
+{
+	struct vsock_sock *vsk = vsock_sk(sk);
+
+	if (vsock_stream_has_data(vsk) >= target) {
+		*data_ready_now = true;
+	} else {
+		/* 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);
+		*data_ready_now = false;
+	}
+
+	return 0;
+}
+
+static int
+vmci_transport_notify_pkt_poll_out(struct sock *sk,
+				   size_t target, bool *space_avail_now)
+{
+	s64 produce_q_free_space;
+	struct vsock_sock *vsk = vsock_sk(sk);
+
+	produce_q_free_space = vsock_stream_has_space(vsk);
+	if (produce_q_free_space > 0) {
+		*space_avail_now = true;
+		return 0;
+	} else if (produce_q_free_space == 0) {
+		/* This is a connected socket but we can't currently send data.
+		 * Nothing else to do.
+		 */
+		*space_avail_now = false;
+	}
+
+	return 0;
+}
+
+static int
+vmci_transport_notify_pkt_recv_init(
+				struct sock *sk,
+				size_t target,
+				struct vmci_transport_recv_notify_data *data)
+{
+	struct vsock_sock *vsk = vsock_sk(sk);
+
+	data->consume_head = 0;
+	data->produce_tail = 0;
+	data->notify_on_block = false;
+
+	if (PKT_FIELD(vsk, write_notify_min_window) < target + 1) {
+		PKT_FIELD(vsk, write_notify_min_window) = target + 1;
+		if (PKT_FIELD(vsk, write_notify_window) <
+		    PKT_FIELD(vsk, write_notify_min_window)) {
+			/* If the current window is smaller than the new
+			 * minimal window size, we need to reevaluate whether
+			 * we need to notify the sender. If the number of ready
+			 * bytes are smaller than the new window, we need to
+			 * send a notification to the sender before we block.
+			 */
+
+			PKT_FIELD(vsk, write_notify_window) =
+			    PKT_FIELD(vsk, write_notify_min_window);
+			data->notify_on_block = true;
+		}
+	}
+
+	return 0;
+}
+
+static int
+vmci_transport_notify_pkt_recv_pre_block(
+				struct sock *sk,
+				size_t target,
+				struct vmci_transport_recv_notify_data *data)
+{
+	int err = 0;
+
+	vsock_block_update_write_window(sk);
+
+	if (data->notify_on_block) {
+		err = vmci_transport_send_read_notification(sk);
+		if (err < 0)
+			return err;
+		data->notify_on_block = false;
+	}
+
+	return err;
+}
+
+static int
+vmci_transport_notify_pkt_recv_post_dequeue(
+				struct sock *sk,
+				size_t target,
+				ssize_t copied,
+				bool data_read,
+				struct vmci_transport_recv_notify_data *data)
+{
+	struct vsock_sock *vsk;
+	int err;
+	bool was_full = false;
+	u64 free_space;
+
+	vsk = vsock_sk(sk);
+	err = 0;
+
+	if (data_read) {
+		smp_mb();
+
+		free_space =
+			vmci_qpair_consume_free_space(vmci_trans(vsk)->qpair);
+		was_full = free_space == copied;
+
+		if (was_full)
+			PKT_FIELD(vsk, peer_waiting_write) = true;
+
+		err = vmci_transport_send_read_notification(sk);
+		if (err < 0)
+			return err;
+
+		/* See the comment in
+		 * vmci_transport_notify_pkt_send_post_enqueue().
+		 */
+		vsock_data_ready(sk);
+	}
+
+	return err;
+}
+
+static int
+vmci_transport_notify_pkt_send_init(
+				struct sock *sk,
+				struct vmci_transport_send_notify_data *data)
+{
+	data->consume_head = 0;
+	data->produce_tail = 0;
+
+	return 0;
+}
+
+static int
+vmci_transport_notify_pkt_send_post_enqueue(
+				struct sock *sk,
+				ssize_t written,
+				struct vmci_transport_send_notify_data *data)
+{
+	int err = 0;
+	struct vsock_sock *vsk;
+	bool sent_wrote = false;
+	bool was_empty;
+	int retries = 0;
+
+	vsk = vsock_sk(sk);
+
+	smp_mb();
+
+	was_empty =
+		vmci_qpair_produce_buf_ready(vmci_trans(vsk)->qpair) == written;
+	if (was_empty) {
+		while (!(vsk->peer_shutdown & RCV_SHUTDOWN) &&
+		       !sent_wrote &&
+		       retries < VMCI_TRANSPORT_MAX_DGRAM_RESENDS) {
+			err = vmci_transport_send_wrote(sk);
+			if (err >= 0)
+				sent_wrote = true;
+
+			retries++;
+		}
+	}
+
+	if (retries >= VMCI_TRANSPORT_MAX_DGRAM_RESENDS && !sent_wrote) {
+		pr_err("%p unable to send wrote notification to peer\n",
+		       sk);
+		return err;
+	}
+
+	return err;
+}
+
+static void
+vmci_transport_notify_pkt_handle_pkt(
+				struct sock *sk,
+				struct vmci_transport_packet *pkt,
+				bool bottom_half,
+				struct sockaddr_vm *dst,
+				struct sockaddr_vm *src, bool *pkt_processed)
+{
+	bool processed = false;
+
+	switch (pkt->type) {
+	case VMCI_TRANSPORT_PACKET_TYPE_WROTE:
+		vmci_transport_handle_wrote(sk, pkt, bottom_half, dst, src);
+		processed = true;
+		break;
+	case VMCI_TRANSPORT_PACKET_TYPE_READ:
+		vmci_transport_handle_read(sk, pkt, bottom_half, dst, src);
+		processed = true;
+		break;
+	}
+
+	if (pkt_processed)
+		*pkt_processed = processed;
+}
+
+static void vmci_transport_notify_pkt_process_request(struct sock *sk)
+{
+	struct vsock_sock *vsk = vsock_sk(sk);
+
+	PKT_FIELD(vsk, write_notify_window) = vmci_trans(vsk)->consume_size;
+	if (vmci_trans(vsk)->consume_size <
+		PKT_FIELD(vsk, write_notify_min_window))
+		PKT_FIELD(vsk, write_notify_min_window) =
+			vmci_trans(vsk)->consume_size;
+}
+
+static void vmci_transport_notify_pkt_process_negotiate(struct sock *sk)
+{
+	struct vsock_sock *vsk = vsock_sk(sk);
+
+	PKT_FIELD(vsk, write_notify_window) = vmci_trans(vsk)->consume_size;
+	if (vmci_trans(vsk)->consume_size <
+		PKT_FIELD(vsk, write_notify_min_window))
+		PKT_FIELD(vsk, write_notify_min_window) =
+			vmci_trans(vsk)->consume_size;
+}
+
+static int
+vmci_transport_notify_pkt_recv_pre_dequeue(
+				struct sock *sk,
+				size_t target,
+				struct vmci_transport_recv_notify_data *data)
+{
+	return 0; /* NOP for QState. */
+}
+
+static int
+vmci_transport_notify_pkt_send_pre_block(
+				struct sock *sk,
+				struct vmci_transport_send_notify_data *data)
+{
+	return 0; /* NOP for QState. */
+}
+
+static int
+vmci_transport_notify_pkt_send_pre_enqueue(
+				struct sock *sk,
+				struct vmci_transport_send_notify_data *data)
+{
+	return 0; /* NOP for QState. */
+}
+
+/* Socket always on control packet based operations. */
+const struct vmci_transport_notify_ops vmci_transport_notify_pkt_q_state_ops = {
+	.socket_init = vmci_transport_notify_pkt_socket_init,
+	.socket_destruct = vmci_transport_notify_pkt_socket_destruct,
+	.poll_in = vmci_transport_notify_pkt_poll_in,
+	.poll_out = vmci_transport_notify_pkt_poll_out,
+	.handle_notify_pkt = vmci_transport_notify_pkt_handle_pkt,
+	.recv_init = vmci_transport_notify_pkt_recv_init,
+	.recv_pre_block = vmci_transport_notify_pkt_recv_pre_block,
+	.recv_pre_dequeue = vmci_transport_notify_pkt_recv_pre_dequeue,
+	.recv_post_dequeue = vmci_transport_notify_pkt_recv_post_dequeue,
+	.send_init = vmci_transport_notify_pkt_send_init,
+	.send_pre_block = vmci_transport_notify_pkt_send_pre_block,
+	.send_pre_enqueue = vmci_transport_notify_pkt_send_pre_enqueue,
+	.send_post_enqueue = vmci_transport_notify_pkt_send_post_enqueue,
+	.process_request = vmci_transport_notify_pkt_process_request,
+	.process_negotiate = vmci_transport_notify_pkt_process_negotiate,
+};

linux-7.2.3/net/vmw_vsock/vsock_addr.c

@@ -0,0 +1,69 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * VMware vSockets Driver
+ *
+ * Copyright (C) 2007-2012 VMware, Inc. All rights reserved.
+ */
+
+#include <linux/types.h>
+#include <linux/socket.h>
+#include <linux/stddef.h>
+#include <net/sock.h>
+#include <net/vsock_addr.h>
+
+void vsock_addr_init(struct sockaddr_vm *addr, u32 cid, u32 port)
+{
+	memset(addr, 0, sizeof(*addr));
+	addr->svm_family = AF_VSOCK;
+	addr->svm_cid = cid;
+	addr->svm_port = port;
+}
+EXPORT_SYMBOL_GPL(vsock_addr_init);
+
+int vsock_addr_validate(const struct sockaddr_vm *addr)
+{
+	__u8 svm_valid_flags = VMADDR_FLAG_TO_HOST;
+
+	if (!addr)
+		return -EFAULT;
+
+	if (addr->svm_family != AF_VSOCK)
+		return -EAFNOSUPPORT;
+
+	if (addr->svm_flags & ~svm_valid_flags)
+		return -EINVAL;
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(vsock_addr_validate);
+
+bool vsock_addr_bound(const struct sockaddr_vm *addr)
+{
+	return addr->svm_port != VMADDR_PORT_ANY;
+}
+EXPORT_SYMBOL_GPL(vsock_addr_bound);
+
+void vsock_addr_unbind(struct sockaddr_vm *addr)
+{
+	vsock_addr_init(addr, VMADDR_CID_ANY, VMADDR_PORT_ANY);
+}
+EXPORT_SYMBOL_GPL(vsock_addr_unbind);
+
+bool vsock_addr_equals_addr(const struct sockaddr_vm *addr,
+			    const struct sockaddr_vm *other)
+{
+	return addr->svm_cid == other->svm_cid &&
+		addr->svm_port == other->svm_port;
+}
+EXPORT_SYMBOL_GPL(vsock_addr_equals_addr);
+
+int vsock_addr_cast(const struct sockaddr_unsized *addr,
+		    size_t len, struct sockaddr_vm **out_addr)
+{
+	if (len < sizeof(**out_addr))
+		return -EFAULT;
+
+	*out_addr = (struct sockaddr_vm *)addr;
+	return vsock_addr_validate(*out_addr);
+}
+EXPORT_SYMBOL_GPL(vsock_addr_cast);

linux-7.2.3/net/vmw_vsock/vsock_bpf.c

@@ -0,0 +1,175 @@
+// 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)
+{
+	struct sk_psock *psock;
+	struct vsock_sock *vsk;
+	int copied;
+
+	psock = sk_psock_get(sk);
+	if (unlikely(!psock))
+		return __vsock_recvmsg(sk, msg, len, flags);
+
+	lock_sock(sk);
+	vsk = vsock_sk(sk);
+
+	if (WARN_ON_ONCE(!vsk->transport)) {
+		copied = -ENODEV;
+		goto out;
+	}
+
+	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);
+	}
+
+out:
+	release_sock(sk);
+	sk_psock_put(sk, psock);
+
+	return copied;
+}
+
+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);
+}

linux-7.2.3/net/vmw_vsock/vsock_loopback.c

@@ -0,0 +1,189 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* loopback transport for vsock using virtio_transport_common APIs
+ *
+ * Copyright (C) 2013-2019 Red Hat, Inc.
+ * Authors: Asias He <asias@redhat.com>
+ *          Stefan Hajnoczi <stefanha@redhat.com>
+ *          Stefano Garzarella <sgarzare@redhat.com>
+ *
+ */
+#include <linux/spinlock.h>
+#include <linux/module.h>
+#include <linux/list.h>
+#include <linux/virtio_vsock.h>
+
+struct vsock_loopback {
+	struct workqueue_struct *workqueue;
+
+	struct sk_buff_head pkt_queue;
+	struct work_struct pkt_work;
+};
+
+static struct vsock_loopback the_vsock_loopback;
+
+static u32 vsock_loopback_get_local_cid(void)
+{
+	return VMADDR_CID_LOCAL;
+}
+
+static int vsock_loopback_send_pkt(struct sk_buff *skb, struct net *net)
+{
+	struct vsock_loopback *vsock = &the_vsock_loopback;
+	int len = skb->len;
+
+	virtio_vsock_skb_queue_tail(&vsock->pkt_queue, skb);
+	queue_work(vsock->workqueue, &vsock->pkt_work);
+
+	return len;
+}
+
+static int vsock_loopback_cancel_pkt(struct vsock_sock *vsk)
+{
+	struct vsock_loopback *vsock = &the_vsock_loopback;
+
+	virtio_transport_purge_skbs(vsk, &vsock->pkt_queue);
+
+	return 0;
+}
+
+static bool vsock_loopback_seqpacket_allow(struct vsock_sock *vsk,
+					   u32 remote_cid);
+
+static bool vsock_loopback_stream_allow(struct vsock_sock *vsk, u32 cid,
+					u32 port)
+{
+	return true;
+}
+
+static bool vsock_loopback_msgzerocopy_allow(void)
+{
+	return true;
+}
+
+static struct virtio_transport loopback_transport = {
+	.transport = {
+		.module                   = THIS_MODULE,
+
+		.get_local_cid            = vsock_loopback_get_local_cid,
+
+		.init                     = virtio_transport_do_socket_init,
+		.destruct                 = virtio_transport_destruct,
+		.release                  = virtio_transport_release,
+		.connect                  = virtio_transport_connect,
+		.shutdown                 = virtio_transport_shutdown,
+		.cancel_pkt               = vsock_loopback_cancel_pkt,
+
+		.dgram_bind               = virtio_transport_dgram_bind,
+		.dgram_dequeue            = virtio_transport_dgram_dequeue,
+		.dgram_enqueue            = virtio_transport_dgram_enqueue,
+		.dgram_allow              = virtio_transport_dgram_allow,
+
+		.stream_dequeue           = virtio_transport_stream_dequeue,
+		.stream_enqueue           = virtio_transport_stream_enqueue,
+		.stream_has_data          = virtio_transport_stream_has_data,
+		.stream_has_space         = virtio_transport_stream_has_space,
+		.stream_rcvhiwat          = virtio_transport_stream_rcvhiwat,
+		.stream_is_active         = virtio_transport_stream_is_active,
+		.stream_allow             = vsock_loopback_stream_allow,
+
+		.seqpacket_dequeue        = virtio_transport_seqpacket_dequeue,
+		.seqpacket_enqueue        = virtio_transport_seqpacket_enqueue,
+		.seqpacket_allow          = vsock_loopback_seqpacket_allow,
+		.seqpacket_has_data       = virtio_transport_seqpacket_has_data,
+
+		.msgzerocopy_allow        = vsock_loopback_msgzerocopy_allow,
+
+		.notify_poll_in           = virtio_transport_notify_poll_in,
+		.notify_poll_out          = virtio_transport_notify_poll_out,
+		.notify_recv_init         = virtio_transport_notify_recv_init,
+		.notify_recv_pre_block    = virtio_transport_notify_recv_pre_block,
+		.notify_recv_pre_dequeue  = virtio_transport_notify_recv_pre_dequeue,
+		.notify_recv_post_dequeue = virtio_transport_notify_recv_post_dequeue,
+		.notify_send_init         = virtio_transport_notify_send_init,
+		.notify_send_pre_block    = virtio_transport_notify_send_pre_block,
+		.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,
+		.notify_set_rcvlowat      = virtio_transport_notify_set_rcvlowat,
+
+		.unsent_bytes             = virtio_transport_unsent_bytes,
+
+		.read_skb = virtio_transport_read_skb,
+	},
+
+	.send_pkt = vsock_loopback_send_pkt,
+};
+
+static bool
+vsock_loopback_seqpacket_allow(struct vsock_sock *vsk, u32 remote_cid)
+{
+	return vsock_net_mode_global(vsk);
+}
+
+static void vsock_loopback_work(struct work_struct *work)
+{
+	struct vsock_loopback *vsock =
+		container_of(work, struct vsock_loopback, pkt_work);
+	struct sk_buff_head pkts;
+	struct sk_buff *skb;
+
+	skb_queue_head_init(&pkts);
+
+	spin_lock_bh(&vsock->pkt_queue.lock);
+	skb_queue_splice_init(&vsock->pkt_queue, &pkts);
+	spin_unlock_bh(&vsock->pkt_queue.lock);
+
+	while ((skb = __skb_dequeue(&pkts))) {
+		/* Decrement the bytes_unsent counter without deallocating skb
+		 * It is freed by the receiver.
+		 */
+		virtio_transport_consume_skb_sent(skb, false);
+		virtio_transport_deliver_tap_pkt(skb);
+		virtio_transport_recv_pkt(&loopback_transport, skb,
+					  sock_net(skb->sk));
+	}
+}
+
+static int __init vsock_loopback_init(void)
+{
+	struct vsock_loopback *vsock = &the_vsock_loopback;
+	int ret;
+
+	vsock->workqueue = alloc_workqueue("vsock-loopback", WQ_PERCPU, 0);
+	if (!vsock->workqueue)
+		return -ENOMEM;
+
+	skb_queue_head_init(&vsock->pkt_queue);
+	INIT_WORK(&vsock->pkt_work, vsock_loopback_work);
+
+	ret = vsock_core_register(&loopback_transport.transport,
+				  VSOCK_TRANSPORT_F_LOCAL);
+	if (ret)
+		goto out_wq;
+
+	return 0;
+
+out_wq:
+	destroy_workqueue(vsock->workqueue);
+	return ret;
+}
+
+static void __exit vsock_loopback_exit(void)
+{
+	struct vsock_loopback *vsock = &the_vsock_loopback;
+
+	vsock_core_unregister(&loopback_transport.transport);
+
+	flush_work(&vsock->pkt_work);
+
+	virtio_vsock_skb_queue_purge(&vsock->pkt_queue);
+
+	destroy_workqueue(vsock->workqueue);
+}
+
+module_init(vsock_loopback_init);
+module_exit(vsock_loopback_exit);
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR("Stefano Garzarella <sgarzare@redhat.com>");
+MODULE_DESCRIPTION("loopback transport for vsock");
+MODULE_ALIAS_NETPROTO(PF_VSOCK);