Hi all,
this series adds the multi-transports support to vsock, following
this proposal:
https://www.spinics.net/lists/netdev/msg575792.html
With the multi-transports support, we can use vsock with nested VMs
(using also different hypervisors) loading both guest->host and
host->guest transports at the same time.
Before this series, vmci-transport supported this behavior but only
using VMware hypervisor on L0, L1, etc.
The first 8 patches are cleanups and preparations, maybe some of
these can go regardless of this series.
Patch 9 changes the hvs_remote_addr_init(). setting the
VMADDR_CID_HOST as remote CID instead of VMADDR_CID_ANY to make
the choice of transport to be used work properly.
@Dexuan Could this change break anything?
Patch 10 adds multi-transports support.
RFC:
- I'd like to move MODULE_ALIAS_NETPROTO(PF_VSOCK) to af_vsock.c.
@Jorgen could this break the VMware products?
- DGRAM sockets are handled as before, I don't know if make sense
work on it now, or when another transport will support DGRAM.
The big issues here is that we cannot link 1-1 a socket to
transport as for stream sockets since DGRAM is not
connection-oriented.
Patches 11 and 12 maybe can be merged with patch 10.
Patch 11 maybe is tricky, but it allows to have vmci_transport and
vhost_vsock loaded at the same time and it also alleviates the
problem of having MODULE_ALIAS_NETPROTO(PF_VSOCK) in vmci_transport.c
Patch 12 prevents the transport modules unloading while sockets are
assigned to them.
Patch 13 fixes an issue in the bind() logic discoverable only with
the new multi-transport support.
I've tested this series with nested KVM (vsock-transport [L0,L1],
virtio-transport[L1,L2]) and with VMware (L0) + KVM (L1)
(vmci-transport [L0,L1], vhost-transport [L1], virtio-transport[L2]).
@Dexuan please can you test on HyperV that I didn't break anything
even without nested VMs?
I'll try to setup a Windows host where to test the nested VMs.
Thanks in advance for your comments and suggestions,
Stefano
Stefano Garzarella (13):
vsock/vmci: remove unused VSOCK_DEFAULT_CONNECT_TIMEOUT
vsock: remove vm_sockets_get_local_cid()
vsock: remove include/linux/vm_sockets.h file
vsock: add 'transport' member in the struct vsock_sock
vsock/virtio: add transport parameter to the
virtio_transport_reset_no_sock()
vsock: add 'struct vsock_sock *' param to vsock_core_get_transport()
vsock: handle buffer_size sockopts in the core
vsock: move vsock_insert_unbound() in the vsock_create()
hv_sock: set VMADDR_CID_HOST in the hvs_remote_addr_init()
vsock: add multi-transports support
vsock: add 'transport_hg' to handle g2h\h2g transports
vsock: prevent transport modules unloading
vsock: fix bind() behaviour taking care of CID
drivers/vhost/vsock.c | 96 +++---
include/linux/virtio_vsock.h | 18 +-
include/linux/vm_sockets.h | 15 -
include/net/af_vsock.h | 35 ++-
include/net/vsock_addr.h | 2 +-
net/vmw_vsock/af_vsock.c | 374 ++++++++++++++++++------
net/vmw_vsock/hyperv_transport.c | 68 ++---
net/vmw_vsock/virtio_transport.c | 177 ++++++-----
net/vmw_vsock/virtio_transport_common.c | 127 +++-----
net/vmw_vsock/vmci_transport.c | 123 +++-----
net/vmw_vsock/vmci_transport.h | 3 -
net/vmw_vsock/vmci_transport_notify.h | 1 -
12 files changed, 555 insertions(+), 484 deletions(-)
delete mode 100644 include/linux/vm_sockets.h
--
2.21.0
The VSOCK_DEFAULT_CONNECT_TIMEOUT definition was introduced with
commit d021c344051af ("VSOCK: Introduce VM Sockets"), but it is
never used in the net/vmw_vsock/vmci_transport.c.
VSOCK_DEFAULT_CONNECT_TIMEOUT is used and defined in
net/vmw_vsock/af_vsock.c
Signed-off-by: Stefano Garzarella <[email protected]>
---
net/vmw_vsock/vmci_transport.c | 5 -----
1 file changed, 5 deletions(-)
diff --git a/net/vmw_vsock/vmci_transport.c b/net/vmw_vsock/vmci_transport.c
index 8c9c4ed90fa7..f8e3131ac480 100644
--- a/net/vmw_vsock/vmci_transport.c
+++ b/net/vmw_vsock/vmci_transport.c
@@ -78,11 +78,6 @@ static int PROTOCOL_OVERRIDE = -1;
#define VMCI_TRANSPORT_DEFAULT_QP_SIZE 262144
#define VMCI_TRANSPORT_DEFAULT_QP_SIZE_MAX 262144
-/* The default peer timeout indicates how long we will wait for a peer response
- * to a control message.
- */
-#define VSOCK_DEFAULT_CONNECT_TIMEOUT (2 * HZ)
-
/* Helper function to convert from a VMCI error code to a VSock error code. */
static s32 vmci_transport_error_to_vsock_error(s32 vmci_error)
--
2.21.0
vm_sockets_get_local_cid() is only used in virtio_transport_common.c.
We can replace it calling the virtio_transport_get_ops() and
using the get_local_cid() callback registered by the transport.
Signed-off-by: Stefano Garzarella <[email protected]>
---
include/linux/vm_sockets.h | 2 --
net/vmw_vsock/af_vsock.c | 10 ----------
net/vmw_vsock/virtio_transport_common.c | 2 +-
3 files changed, 1 insertion(+), 13 deletions(-)
diff --git a/include/linux/vm_sockets.h b/include/linux/vm_sockets.h
index 33f1a2ecd905..7dd899ccb920 100644
--- a/include/linux/vm_sockets.h
+++ b/include/linux/vm_sockets.h
@@ -10,6 +10,4 @@
#include <uapi/linux/vm_sockets.h>
-int vm_sockets_get_local_cid(void);
-
#endif /* _VM_SOCKETS_H */
diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c
index ab47bf3ab66e..f609434b2794 100644
--- a/net/vmw_vsock/af_vsock.c
+++ b/net/vmw_vsock/af_vsock.c
@@ -129,16 +129,6 @@ static struct proto vsock_proto = {
static const struct vsock_transport *transport;
static DEFINE_MUTEX(vsock_register_mutex);
-/**** EXPORTS ****/
-
-/* Get the ID of the local context. This is transport dependent. */
-
-int vm_sockets_get_local_cid(void)
-{
- return transport->get_local_cid();
-}
-EXPORT_SYMBOL_GPL(vm_sockets_get_local_cid);
-
/**** UTILS ****/
/* Each bound VSocket is stored in the bind hash table and each connected
diff --git a/net/vmw_vsock/virtio_transport_common.c b/net/vmw_vsock/virtio_transport_common.c
index 5bb70c692b1e..ed1ad5289164 100644
--- a/net/vmw_vsock/virtio_transport_common.c
+++ b/net/vmw_vsock/virtio_transport_common.c
@@ -168,7 +168,7 @@ static int virtio_transport_send_pkt_info(struct vsock_sock *vsk,
struct virtio_vsock_pkt *pkt;
u32 pkt_len = info->pkt_len;
- src_cid = vm_sockets_get_local_cid();
+ src_cid = virtio_transport_get_ops()->transport.get_local_cid();
src_port = vsk->local_addr.svm_port;
if (!info->remote_cid) {
dst_cid = vsk->remote_addr.svm_cid;
--
2.21.0
This header file now only includes the "uapi/linux/vm_sockets.h".
We can include directly it when needed.
Signed-off-by: Stefano Garzarella <[email protected]>
---
include/linux/vm_sockets.h | 13 -------------
include/net/af_vsock.h | 2 +-
include/net/vsock_addr.h | 2 +-
net/vmw_vsock/vmci_transport_notify.h | 1 -
4 files changed, 2 insertions(+), 16 deletions(-)
delete mode 100644 include/linux/vm_sockets.h
diff --git a/include/linux/vm_sockets.h b/include/linux/vm_sockets.h
deleted file mode 100644
index 7dd899ccb920..000000000000
--- a/include/linux/vm_sockets.h
+++ /dev/null
@@ -1,13 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0-only */
-/*
- * VMware vSockets Driver
- *
- * Copyright (C) 2007-2013 VMware, Inc. All rights reserved.
- */
-
-#ifndef _VM_SOCKETS_H
-#define _VM_SOCKETS_H
-
-#include <uapi/linux/vm_sockets.h>
-
-#endif /* _VM_SOCKETS_H */
diff --git a/include/net/af_vsock.h b/include/net/af_vsock.h
index 80ea0f93d3f7..c660402b10f2 100644
--- a/include/net/af_vsock.h
+++ b/include/net/af_vsock.h
@@ -10,7 +10,7 @@
#include <linux/kernel.h>
#include <linux/workqueue.h>
-#include <linux/vm_sockets.h>
+#include <uapi/linux/vm_sockets.h>
#include "vsock_addr.h"
diff --git a/include/net/vsock_addr.h b/include/net/vsock_addr.h
index 57d2db5c4bdf..cf8cc140d68d 100644
--- a/include/net/vsock_addr.h
+++ b/include/net/vsock_addr.h
@@ -8,7 +8,7 @@
#ifndef _VSOCK_ADDR_H_
#define _VSOCK_ADDR_H_
-#include <linux/vm_sockets.h>
+#include <uapi/linux/vm_sockets.h>
void vsock_addr_init(struct sockaddr_vm *addr, u32 cid, u32 port);
int vsock_addr_validate(const struct sockaddr_vm *addr);
diff --git a/net/vmw_vsock/vmci_transport_notify.h b/net/vmw_vsock/vmci_transport_notify.h
index 7843f08d4290..a1aa5a998c0e 100644
--- a/net/vmw_vsock/vmci_transport_notify.h
+++ b/net/vmw_vsock/vmci_transport_notify.h
@@ -11,7 +11,6 @@
#include <linux/types.h>
#include <linux/vmw_vmci_defs.h>
#include <linux/vmw_vmci_api.h>
-#include <linux/vm_sockets.h>
#include "vmci_transport.h"
--
2.21.0
As a preparation to support multiple transports, this patch adds
the 'transport' member at the 'struct vsock_sock'.
This new field is initialized during the creation in the
__vsock_create() function.
This patch also renames the global 'transport' pointer to
'transport_single', since for now we're only supporting a single
transport registered at run-time.
Signed-off-by: Stefano Garzarella <[email protected]>
---
include/net/af_vsock.h | 1 +
net/vmw_vsock/af_vsock.c | 56 +++++++++++++++++++++++++++-------------
2 files changed, 39 insertions(+), 18 deletions(-)
diff --git a/include/net/af_vsock.h b/include/net/af_vsock.h
index c660402b10f2..a5e1e134261d 100644
--- a/include/net/af_vsock.h
+++ b/include/net/af_vsock.h
@@ -27,6 +27,7 @@ extern spinlock_t vsock_table_lock;
struct vsock_sock {
/* sk must be the first member. */
struct sock sk;
+ const struct vsock_transport *transport;
struct sockaddr_vm local_addr;
struct sockaddr_vm remote_addr;
/* Links for the global tables of bound and connected sockets. */
diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c
index f609434b2794..81ee2561c76f 100644
--- a/net/vmw_vsock/af_vsock.c
+++ b/net/vmw_vsock/af_vsock.c
@@ -126,7 +126,7 @@ static struct proto vsock_proto = {
*/
#define VSOCK_DEFAULT_CONNECT_TIMEOUT (2 * HZ)
-static const struct vsock_transport *transport;
+static const struct vsock_transport *transport_single;
static DEFINE_MUTEX(vsock_register_mutex);
/**** UTILS ****/
@@ -408,7 +408,9 @@ static bool vsock_is_pending(struct sock *sk)
static int vsock_send_shutdown(struct sock *sk, int mode)
{
- return transport->shutdown(vsock_sk(sk), mode);
+ struct vsock_sock *vsk = vsock_sk(sk);
+
+ return vsk->transport->shutdown(vsk, mode);
}
static void vsock_pending_work(struct work_struct *work)
@@ -518,7 +520,7 @@ static int __vsock_bind_stream(struct vsock_sock *vsk,
static int __vsock_bind_dgram(struct vsock_sock *vsk,
struct sockaddr_vm *addr)
{
- return transport->dgram_bind(vsk, addr);
+ return vsk->transport->dgram_bind(vsk, addr);
}
static int __vsock_bind(struct sock *sk, struct sockaddr_vm *addr)
@@ -536,7 +538,7 @@ static int __vsock_bind(struct sock *sk, struct sockaddr_vm *addr)
* like AF_INET prevents binding to a non-local IP address (in most
* cases), we only allow binding to the local CID.
*/
- cid = transport->get_local_cid();
+ cid = vsk->transport->get_local_cid();
if (addr->svm_cid != cid && addr->svm_cid != VMADDR_CID_ANY)
return -EADDRNOTAVAIL;
@@ -586,6 +588,7 @@ struct sock *__vsock_create(struct net *net,
sk->sk_type = type;
vsk = vsock_sk(sk);
+ vsk->transport = transport_single;
vsock_addr_init(&vsk->local_addr, VMADDR_CID_ANY, VMADDR_PORT_ANY);
vsock_addr_init(&vsk->remote_addr, VMADDR_CID_ANY, VMADDR_PORT_ANY);
@@ -616,7 +619,7 @@ struct sock *__vsock_create(struct net *net,
vsk->connect_timeout = VSOCK_DEFAULT_CONNECT_TIMEOUT;
}
- if (transport->init(vsk, psk) < 0) {
+ if (vsk->transport->init(vsk, psk) < 0) {
sk_free(sk);
return NULL;
}
@@ -638,7 +641,7 @@ static void __vsock_release(struct sock *sk)
vsk = vsock_sk(sk);
pending = NULL; /* Compiler warning. */
- transport->release(vsk);
+ vsk->transport->release(vsk);
lock_sock(sk);
sock_orphan(sk);
@@ -662,7 +665,7 @@ static void vsock_sk_destruct(struct sock *sk)
{
struct vsock_sock *vsk = vsock_sk(sk);
- transport->destruct(vsk);
+ vsk->transport->destruct(vsk);
/* When clearing these addresses, there's no need to set the family and
* possibly register the address family with the kernel.
@@ -686,13 +689,13 @@ static int vsock_queue_rcv_skb(struct sock *sk, struct sk_buff *skb)
s64 vsock_stream_has_data(struct vsock_sock *vsk)
{
- return transport->stream_has_data(vsk);
+ return vsk->transport->stream_has_data(vsk);
}
EXPORT_SYMBOL_GPL(vsock_stream_has_data);
s64 vsock_stream_has_space(struct vsock_sock *vsk)
{
- return transport->stream_has_space(vsk);
+ return vsk->transport->stream_has_space(vsk);
}
EXPORT_SYMBOL_GPL(vsock_stream_has_space);
@@ -861,6 +864,7 @@ static __poll_t vsock_poll(struct file *file, struct socket *sock,
mask |= EPOLLOUT | EPOLLWRNORM | EPOLLWRBAND;
} else if (sock->type == SOCK_STREAM) {
+ const struct vsock_transport *transport = vsk->transport;
lock_sock(sk);
/* Listening sockets that have connections in their accept
@@ -936,6 +940,7 @@ static int vsock_dgram_sendmsg(struct socket *sock, struct msghdr *msg,
struct sock *sk;
struct vsock_sock *vsk;
struct sockaddr_vm *remote_addr;
+ const struct vsock_transport *transport;
if (msg->msg_flags & MSG_OOB)
return -EOPNOTSUPP;
@@ -944,6 +949,7 @@ static int vsock_dgram_sendmsg(struct socket *sock, struct msghdr *msg,
err = 0;
sk = sock->sk;
vsk = vsock_sk(sk);
+ transport = vsk->transport;
lock_sock(sk);
@@ -1028,8 +1034,8 @@ static int vsock_dgram_connect(struct socket *sock,
if (err)
goto out;
- if (!transport->dgram_allow(remote_addr->svm_cid,
- remote_addr->svm_port)) {
+ if (!vsk->transport->dgram_allow(remote_addr->svm_cid,
+ remote_addr->svm_port)) {
err = -EINVAL;
goto out;
}
@@ -1045,7 +1051,9 @@ static int vsock_dgram_connect(struct socket *sock,
static int vsock_dgram_recvmsg(struct socket *sock, struct msghdr *msg,
size_t len, int flags)
{
- return transport->dgram_dequeue(vsock_sk(sock->sk), msg, len, flags);
+ struct vsock_sock *vsk = vsock_sk(sock->sk);
+
+ return vsk->transport->dgram_dequeue(vsk, msg, len, flags);
}
static const struct proto_ops vsock_dgram_ops = {
@@ -1071,6 +1079,8 @@ static const struct proto_ops vsock_dgram_ops = {
static int vsock_transport_cancel_pkt(struct vsock_sock *vsk)
{
+ const struct vsock_transport *transport = vsk->transport;
+
if (!transport->cancel_pkt)
return -EOPNOTSUPP;
@@ -1107,6 +1117,7 @@ static int vsock_stream_connect(struct socket *sock, struct sockaddr *addr,
int err;
struct sock *sk;
struct vsock_sock *vsk;
+ const struct vsock_transport *transport;
struct sockaddr_vm *remote_addr;
long timeout;
DEFINE_WAIT(wait);
@@ -1114,6 +1125,7 @@ static int vsock_stream_connect(struct socket *sock, struct sockaddr *addr,
err = 0;
sk = sock->sk;
vsk = vsock_sk(sk);
+ transport = vsk->transport;
lock_sock(sk);
@@ -1357,6 +1369,7 @@ static int vsock_stream_setsockopt(struct socket *sock,
int err;
struct sock *sk;
struct vsock_sock *vsk;
+ const struct vsock_transport *transport;
u64 val;
if (level != AF_VSOCK)
@@ -1377,6 +1390,7 @@ static int vsock_stream_setsockopt(struct socket *sock,
err = 0;
sk = sock->sk;
vsk = vsock_sk(sk);
+ transport = vsk->transport;
lock_sock(sk);
@@ -1434,6 +1448,7 @@ static int vsock_stream_getsockopt(struct socket *sock,
int len;
struct sock *sk;
struct vsock_sock *vsk;
+ const struct vsock_transport *transport;
u64 val;
if (level != AF_VSOCK)
@@ -1457,6 +1472,7 @@ static int vsock_stream_getsockopt(struct socket *sock,
err = 0;
sk = sock->sk;
vsk = vsock_sk(sk);
+ transport = vsk->transport;
switch (optname) {
case SO_VM_SOCKETS_BUFFER_SIZE:
@@ -1501,6 +1517,7 @@ static int vsock_stream_sendmsg(struct socket *sock, struct msghdr *msg,
{
struct sock *sk;
struct vsock_sock *vsk;
+ const struct vsock_transport *transport;
ssize_t total_written;
long timeout;
int err;
@@ -1509,6 +1526,7 @@ static int vsock_stream_sendmsg(struct socket *sock, struct msghdr *msg,
sk = sock->sk;
vsk = vsock_sk(sk);
+ transport = vsk->transport;
total_written = 0;
err = 0;
@@ -1640,6 +1658,7 @@ vsock_stream_recvmsg(struct socket *sock, struct msghdr *msg, size_t len,
{
struct sock *sk;
struct vsock_sock *vsk;
+ const struct vsock_transport *transport;
int err;
size_t target;
ssize_t copied;
@@ -1650,6 +1669,7 @@ vsock_stream_recvmsg(struct socket *sock, struct msghdr *msg, size_t len,
sk = sock->sk;
vsk = vsock_sk(sk);
+ transport = vsk->transport;
err = 0;
lock_sock(sk);
@@ -1864,7 +1884,7 @@ static long vsock_dev_do_ioctl(struct file *filp,
switch (cmd) {
case IOCTL_VM_SOCKETS_GET_LOCAL_CID:
- if (put_user(transport->get_local_cid(), p) != 0)
+ if (put_user(transport_single->get_local_cid(), p) != 0)
retval = -EFAULT;
break;
@@ -1911,7 +1931,7 @@ int __vsock_core_init(const struct vsock_transport *t, struct module *owner)
if (err)
return err;
- if (transport) {
+ if (transport_single) {
err = -EBUSY;
goto err_busy;
}
@@ -1920,7 +1940,7 @@ int __vsock_core_init(const struct vsock_transport *t, struct module *owner)
* unload while there are open sockets.
*/
vsock_proto.owner = owner;
- transport = t;
+ transport_single = t;
vsock_device.minor = MISC_DYNAMIC_MINOR;
err = misc_register(&vsock_device);
@@ -1950,7 +1970,7 @@ int __vsock_core_init(const struct vsock_transport *t, struct module *owner)
err_deregister_misc:
misc_deregister(&vsock_device);
err_reset_transport:
- transport = NULL;
+ transport_single = NULL;
err_busy:
mutex_unlock(&vsock_register_mutex);
return err;
@@ -1967,7 +1987,7 @@ void vsock_core_exit(void)
/* We do not want the assignment below re-ordered. */
mb();
- transport = NULL;
+ transport_single = NULL;
mutex_unlock(&vsock_register_mutex);
}
@@ -1978,7 +1998,7 @@ const struct vsock_transport *vsock_core_get_transport(void)
/* vsock_register_mutex not taken since only the transport uses this
* function and only while registered.
*/
- return transport;
+ return transport_single;
}
EXPORT_SYMBOL_GPL(vsock_core_get_transport);
--
2.21.0
We are going to add 'struct vsock_sock *' parameter to
virtio_transport_get_ops().
In some cases, like in the virtio_transport_reset_no_sock(),
we don't have any socket assigned to the packet received,
so we can't use the virtio_transport_get_ops().
In order to allow virtio_transport_reset_no_sock() to use the
'.send_pkt' callback from the 'vhost_transport' or 'virtio_transport',
we add the 'struct virtio_transport *' to it and to its caller:
virtio_transport_recv_pkt().
We moved the 'vhost_transport' and 'virtio_transport' definition,
to pass their address to the virtio_transport_recv_pkt().
Signed-off-by: Stefano Garzarella <[email protected]>
---
drivers/vhost/vsock.c | 94 +++++++-------
include/linux/virtio_vsock.h | 3 +-
net/vmw_vsock/virtio_transport.c | 160 ++++++++++++------------
net/vmw_vsock/virtio_transport_common.c | 12 +-
4 files changed, 135 insertions(+), 134 deletions(-)
diff --git a/drivers/vhost/vsock.c b/drivers/vhost/vsock.c
index 9f57736fe15e..92ab3852c954 100644
--- a/drivers/vhost/vsock.c
+++ b/drivers/vhost/vsock.c
@@ -384,6 +384,52 @@ static bool vhost_vsock_more_replies(struct vhost_vsock *vsock)
return val < vq->num;
}
+static struct virtio_transport vhost_transport = {
+ .transport = {
+ .get_local_cid = vhost_transport_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 = vhost_transport_cancel_pkt,
+
+ .dgram_enqueue = virtio_transport_dgram_enqueue,
+ .dgram_dequeue = virtio_transport_dgram_dequeue,
+ .dgram_bind = virtio_transport_dgram_bind,
+ .dgram_allow = virtio_transport_dgram_allow,
+
+ .stream_enqueue = virtio_transport_stream_enqueue,
+ .stream_dequeue = virtio_transport_stream_dequeue,
+ .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,
+
+ .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,
+
+ .set_buffer_size = virtio_transport_set_buffer_size,
+ .set_min_buffer_size = virtio_transport_set_min_buffer_size,
+ .set_max_buffer_size = virtio_transport_set_max_buffer_size,
+ .get_buffer_size = virtio_transport_get_buffer_size,
+ .get_min_buffer_size = virtio_transport_get_min_buffer_size,
+ .get_max_buffer_size = virtio_transport_get_max_buffer_size,
+ },
+
+ .send_pkt = vhost_transport_send_pkt,
+};
+
static void vhost_vsock_handle_tx_kick(struct vhost_work *work)
{
struct vhost_virtqueue *vq = container_of(work, struct vhost_virtqueue,
@@ -438,7 +484,7 @@ static void vhost_vsock_handle_tx_kick(struct vhost_work *work)
/* Only accept correctly addressed packets */
if (le64_to_cpu(pkt->hdr.src_cid) == vsock->guest_cid)
- virtio_transport_recv_pkt(pkt);
+ virtio_transport_recv_pkt(&vhost_transport, pkt);
else
virtio_transport_free_pkt(pkt);
@@ -786,52 +832,6 @@ static struct miscdevice vhost_vsock_misc = {
.fops = &vhost_vsock_fops,
};
-static struct virtio_transport vhost_transport = {
- .transport = {
- .get_local_cid = vhost_transport_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 = vhost_transport_cancel_pkt,
-
- .dgram_enqueue = virtio_transport_dgram_enqueue,
- .dgram_dequeue = virtio_transport_dgram_dequeue,
- .dgram_bind = virtio_transport_dgram_bind,
- .dgram_allow = virtio_transport_dgram_allow,
-
- .stream_enqueue = virtio_transport_stream_enqueue,
- .stream_dequeue = virtio_transport_stream_dequeue,
- .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,
-
- .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,
-
- .set_buffer_size = virtio_transport_set_buffer_size,
- .set_min_buffer_size = virtio_transport_set_min_buffer_size,
- .set_max_buffer_size = virtio_transport_set_max_buffer_size,
- .get_buffer_size = virtio_transport_get_buffer_size,
- .get_min_buffer_size = virtio_transport_get_min_buffer_size,
- .get_max_buffer_size = virtio_transport_get_max_buffer_size,
- },
-
- .send_pkt = vhost_transport_send_pkt,
-};
-
static int __init vhost_vsock_init(void)
{
int ret;
diff --git a/include/linux/virtio_vsock.h b/include/linux/virtio_vsock.h
index 4c7781f4b29b..96d8132acbd7 100644
--- a/include/linux/virtio_vsock.h
+++ b/include/linux/virtio_vsock.h
@@ -151,7 +151,8 @@ virtio_transport_dgram_enqueue(struct vsock_sock *vsk,
void virtio_transport_destruct(struct vsock_sock *vsk);
-void virtio_transport_recv_pkt(struct virtio_vsock_pkt *pkt);
+void virtio_transport_recv_pkt(struct virtio_transport *t,
+ struct virtio_vsock_pkt *pkt);
void virtio_transport_free_pkt(struct virtio_vsock_pkt *pkt);
void virtio_transport_inc_tx_pkt(struct virtio_vsock_sock *vvs, struct virtio_vsock_pkt *pkt);
u32 virtio_transport_get_credit(struct virtio_vsock_sock *vvs, u32 wanted);
diff --git a/net/vmw_vsock/virtio_transport.c b/net/vmw_vsock/virtio_transport.c
index 082a30936690..3756f0857946 100644
--- a/net/vmw_vsock/virtio_transport.c
+++ b/net/vmw_vsock/virtio_transport.c
@@ -86,33 +86,6 @@ static u32 virtio_transport_get_local_cid(void)
return ret;
}
-static void virtio_transport_loopback_work(struct work_struct *work)
-{
- struct virtio_vsock *vsock =
- container_of(work, struct virtio_vsock, loopback_work);
- LIST_HEAD(pkts);
-
- spin_lock_bh(&vsock->loopback_list_lock);
- list_splice_init(&vsock->loopback_list, &pkts);
- spin_unlock_bh(&vsock->loopback_list_lock);
-
- mutex_lock(&vsock->rx_lock);
-
- if (!vsock->rx_run)
- goto out;
-
- while (!list_empty(&pkts)) {
- struct virtio_vsock_pkt *pkt;
-
- pkt = list_first_entry(&pkts, struct virtio_vsock_pkt, list);
- list_del_init(&pkt->list);
-
- virtio_transport_recv_pkt(pkt);
- }
-out:
- mutex_unlock(&vsock->rx_lock);
-}
-
static int virtio_transport_send_pkt_loopback(struct virtio_vsock *vsock,
struct virtio_vsock_pkt *pkt)
{
@@ -370,59 +343,6 @@ static bool virtio_transport_more_replies(struct virtio_vsock *vsock)
return val < virtqueue_get_vring_size(vq);
}
-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;
-
- vq = vsock->vqs[VSOCK_VQ_RX];
-
- mutex_lock(&vsock->rx_lock);
-
- if (!vsock->rx_run)
- goto out;
-
- do {
- virtqueue_disable_cb(vq);
- for (;;) {
- struct virtio_vsock_pkt *pkt;
- unsigned int len;
-
- if (!virtio_transport_more_replies(vsock)) {
- /* Stop rx until the device processes already
- * pending replies. Leave rx virtqueue
- * callbacks disabled.
- */
- goto out;
- }
-
- pkt = virtqueue_get_buf(vq, &len);
- if (!pkt) {
- break;
- }
-
- vsock->rx_buf_nr--;
-
- /* Drop short/long packets */
- if (unlikely(len < sizeof(pkt->hdr) ||
- len > sizeof(pkt->hdr) + pkt->len)) {
- virtio_transport_free_pkt(pkt);
- continue;
- }
-
- pkt->len = len - sizeof(pkt->hdr);
- virtio_transport_deliver_tap_pkt(pkt);
- virtio_transport_recv_pkt(pkt);
- }
- } while (!virtqueue_enable_cb(vq));
-
-out:
- if (vsock->rx_buf_nr < vsock->rx_buf_max_nr / 2)
- virtio_vsock_rx_fill(vsock);
- mutex_unlock(&vsock->rx_lock);
-}
-
/* event_lock must be held */
static int virtio_vsock_event_fill_one(struct virtio_vsock *vsock,
struct virtio_vsock_event *event)
@@ -586,6 +506,86 @@ static struct virtio_transport virtio_transport = {
.send_pkt = virtio_transport_send_pkt,
};
+static void virtio_transport_loopback_work(struct work_struct *work)
+{
+ struct virtio_vsock *vsock =
+ container_of(work, struct virtio_vsock, loopback_work);
+ LIST_HEAD(pkts);
+
+ spin_lock_bh(&vsock->loopback_list_lock);
+ list_splice_init(&vsock->loopback_list, &pkts);
+ spin_unlock_bh(&vsock->loopback_list_lock);
+
+ mutex_lock(&vsock->rx_lock);
+
+ if (!vsock->rx_run)
+ goto out;
+
+ while (!list_empty(&pkts)) {
+ struct virtio_vsock_pkt *pkt;
+
+ pkt = list_first_entry(&pkts, struct virtio_vsock_pkt, list);
+ list_del_init(&pkt->list);
+
+ virtio_transport_recv_pkt(&virtio_transport, pkt);
+ }
+out:
+ mutex_unlock(&vsock->rx_lock);
+}
+
+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;
+
+ vq = vsock->vqs[VSOCK_VQ_RX];
+
+ mutex_lock(&vsock->rx_lock);
+
+ if (!vsock->rx_run)
+ goto out;
+
+ do {
+ virtqueue_disable_cb(vq);
+ for (;;) {
+ struct virtio_vsock_pkt *pkt;
+ unsigned int len;
+
+ if (!virtio_transport_more_replies(vsock)) {
+ /* Stop rx until the device processes already
+ * pending replies. Leave rx virtqueue
+ * callbacks disabled.
+ */
+ goto out;
+ }
+
+ pkt = virtqueue_get_buf(vq, &len);
+ if (!pkt) {
+ break;
+ }
+
+ vsock->rx_buf_nr--;
+
+ /* Drop short/long packets */
+ if (unlikely(len < sizeof(pkt->hdr) ||
+ len > sizeof(pkt->hdr) + pkt->len)) {
+ virtio_transport_free_pkt(pkt);
+ continue;
+ }
+
+ pkt->len = len - sizeof(pkt->hdr);
+ virtio_transport_deliver_tap_pkt(pkt);
+ virtio_transport_recv_pkt(&virtio_transport, pkt);
+ }
+ } while (!virtqueue_enable_cb(vq));
+
+out:
+ if (vsock->rx_buf_nr < vsock->rx_buf_max_nr / 2)
+ virtio_vsock_rx_fill(vsock);
+ mutex_unlock(&vsock->rx_lock);
+}
+
static int virtio_vsock_probe(struct virtio_device *vdev)
{
vq_callback_t *callbacks[] = {
diff --git a/net/vmw_vsock/virtio_transport_common.c b/net/vmw_vsock/virtio_transport_common.c
index ed1ad5289164..382536b69029 100644
--- a/net/vmw_vsock/virtio_transport_common.c
+++ b/net/vmw_vsock/virtio_transport_common.c
@@ -689,9 +689,9 @@ static int virtio_transport_reset(struct vsock_sock *vsk,
/* Normally packets are associated with a socket. There may be no socket if an
* attempt was made to connect to a socket that does not exist.
*/
-static int virtio_transport_reset_no_sock(struct virtio_vsock_pkt *pkt)
+static int virtio_transport_reset_no_sock(const struct virtio_transport *t,
+ struct virtio_vsock_pkt *pkt)
{
- const struct virtio_transport *t;
struct virtio_vsock_pkt *reply;
struct virtio_vsock_pkt_info info = {
.op = VIRTIO_VSOCK_OP_RST,
@@ -711,7 +711,6 @@ static int virtio_transport_reset_no_sock(struct virtio_vsock_pkt *pkt)
if (!reply)
return -ENOMEM;
- t = virtio_transport_get_ops();
if (!t) {
virtio_transport_free_pkt(reply);
return -ENOTCONN;
@@ -1047,7 +1046,8 @@ static bool virtio_transport_space_update(struct sock *sk,
/* We are under the virtio-vsock's vsock->rx_lock or vhost-vsock's vq->mutex
* lock.
*/
-void virtio_transport_recv_pkt(struct virtio_vsock_pkt *pkt)
+void virtio_transport_recv_pkt(struct virtio_transport *t,
+ struct virtio_vsock_pkt *pkt)
{
struct sockaddr_vm src, dst;
struct vsock_sock *vsk;
@@ -1069,7 +1069,7 @@ void virtio_transport_recv_pkt(struct virtio_vsock_pkt *pkt)
le32_to_cpu(pkt->hdr.fwd_cnt));
if (le16_to_cpu(pkt->hdr.type) != VIRTIO_VSOCK_TYPE_STREAM) {
- (void)virtio_transport_reset_no_sock(pkt);
+ (void)virtio_transport_reset_no_sock(t, pkt);
goto free_pkt;
}
@@ -1080,7 +1080,7 @@ void virtio_transport_recv_pkt(struct virtio_vsock_pkt *pkt)
if (!sk) {
sk = vsock_find_bound_socket(&dst);
if (!sk) {
- (void)virtio_transport_reset_no_sock(pkt);
+ (void)virtio_transport_reset_no_sock(t, pkt);
goto free_pkt;
}
}
--
2.21.0
vsock_insert_unbound() was called only when 'sock' parameter of
__vsock_create() was not null. This only happened when
__vsock_create() was called by vsock_create().
In order to simplify the multi-transports support, this patch
moves vsock_insert_unbound() at the end of vsock_create().
Signed-off-by: Stefano Garzarella <[email protected]>
---
net/vmw_vsock/af_vsock.c | 13 +++++++++----
1 file changed, 9 insertions(+), 4 deletions(-)
diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c
index dee69d7ee148..95e6db21e7e1 100644
--- a/net/vmw_vsock/af_vsock.c
+++ b/net/vmw_vsock/af_vsock.c
@@ -634,9 +634,6 @@ struct sock *__vsock_create(struct net *net,
return NULL;
}
- if (sock)
- vsock_insert_unbound(vsk);
-
return sk;
}
EXPORT_SYMBOL_GPL(__vsock_create);
@@ -1875,6 +1872,8 @@ static const struct proto_ops vsock_stream_ops = {
static int vsock_create(struct net *net, struct socket *sock,
int protocol, int kern)
{
+ struct sock *sk;
+
if (!sock)
return -EINVAL;
@@ -1894,7 +1893,13 @@ static int vsock_create(struct net *net, struct socket *sock,
sock->state = SS_UNCONNECTED;
- return __vsock_create(net, sock, NULL, GFP_KERNEL, 0, kern) ? 0 : -ENOMEM;
+ sk = __vsock_create(net, sock, NULL, GFP_KERNEL, 0, kern);
+ if (!sk)
+ return -ENOMEM;
+
+ vsock_insert_unbound(vsock_sk(sk));
+
+ return 0;
}
static const struct net_proto_family vsock_family_ops = {
--
2.21.0
Remote peer is always the host, so we set VMADDR_CID_HOST as
remote CID instead of VMADDR_CID_ANY.
Signed-off-by: Stefano Garzarella <[email protected]>
---
net/vmw_vsock/hyperv_transport.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/net/vmw_vsock/hyperv_transport.c b/net/vmw_vsock/hyperv_transport.c
index 4f47af2054dd..306310794522 100644
--- a/net/vmw_vsock/hyperv_transport.c
+++ b/net/vmw_vsock/hyperv_transport.c
@@ -186,7 +186,8 @@ static void hvs_remote_addr_init(struct sockaddr_vm *remote,
static u32 host_ephemeral_port = MIN_HOST_EPHEMERAL_PORT;
struct sock *sk;
- vsock_addr_init(remote, VMADDR_CID_ANY, VMADDR_PORT_ANY);
+ /* Remote peer is always the host */
+ vsock_addr_init(remote, VMADDR_CID_HOST, VMADDR_PORT_ANY);
while (1) {
/* Wrap around ? */
--
2.21.0
This patch adds the support of multiple transports in the
VSOCK core.
With the multi-transports support, we can use vsock with nested VMs
(using also different hypervisors) loading both guest->host and
host->guest transports at the same time.
Major changes:
- vsock core module can be loaded interdependently of the transports
- each 'struct virtio_transport' has a new feature fields
(H2G, G2H, DGRAM) to identify which directions the transport can
handle and if it's support DGRAM (only vmci)
- vsock_core_init() and vsock_core_exit() are renamed to
vsock_core_register() and vsock_core_unregister()
- each stream socket is assigned to a transport when the remote CID
is set (during the connect() or when we receive a connection request
on a listener socket).
The remote CID is used to decide which transport to use:
- remote CID > VMADDR_CID_HOST will use host->guest transport
- remote CID <= VMADDR_CID_HOST will use guest->host transport
- listener sockets are not bound to any transports since no transport
operations are done on it. In this way we can create a listener
socket, also if the transports are not loaded or with VMADDR_CID_ANY
to listen on all transports.
- DGRAM sockets are handled as before, since only the vmci_transport
provides this feature.
Signed-off-by: Stefano Garzarella <[email protected]>
---
RFC:
- I'd like to move MODULE_ALIAS_NETPROTO(PF_VSOCK) to af_vsock.c.
@Jorgen could this break the VMware products?
- DGRAM sockets are handled as before, I don't know if make sense work
on it now, or when another transport will support DGRAM. The big
issues here is that we cannot link 1-1 a socket to transport as
for stream sockets since DGRAM is not connection-oriented.
---
drivers/vhost/vsock.c | 6 +-
include/net/af_vsock.h | 15 +-
net/vmw_vsock/af_vsock.c | 240 ++++++++++++++++++------
net/vmw_vsock/hyperv_transport.c | 28 ++-
net/vmw_vsock/virtio_transport.c | 8 +-
net/vmw_vsock/virtio_transport_common.c | 28 ++-
net/vmw_vsock/vmci_transport.c | 31 ++-
7 files changed, 275 insertions(+), 81 deletions(-)
diff --git a/drivers/vhost/vsock.c b/drivers/vhost/vsock.c
index 6d7e4f022748..375af01a5b64 100644
--- a/drivers/vhost/vsock.c
+++ b/drivers/vhost/vsock.c
@@ -386,6 +386,8 @@ static bool vhost_vsock_more_replies(struct vhost_vsock *vsock)
static struct virtio_transport vhost_transport = {
.transport = {
+ .features = VSOCK_TRANSPORT_F_H2G,
+
.get_local_cid = vhost_transport_get_local_cid,
.init = virtio_transport_do_socket_init,
@@ -831,7 +833,7 @@ static int __init vhost_vsock_init(void)
{
int ret;
- ret = vsock_core_init(&vhost_transport.transport);
+ ret = vsock_core_register(&vhost_transport.transport);
if (ret < 0)
return ret;
return misc_register(&vhost_vsock_misc);
@@ -840,7 +842,7 @@ static int __init vhost_vsock_init(void)
static void __exit vhost_vsock_exit(void)
{
misc_deregister(&vhost_vsock_misc);
- vsock_core_exit();
+ vsock_core_unregister(&vhost_transport.transport);
};
module_init(vhost_vsock_init);
diff --git a/include/net/af_vsock.h b/include/net/af_vsock.h
index 86f8f463e01a..2a081d19e20d 100644
--- a/include/net/af_vsock.h
+++ b/include/net/af_vsock.h
@@ -94,7 +94,13 @@ struct vsock_transport_send_notify_data {
u64 data2; /* Transport-defined. */
};
+#define VSOCK_TRANSPORT_F_H2G 0x00000001
+#define VSOCK_TRANSPORT_F_G2H 0x00000002
+#define VSOCK_TRANSPORT_F_DGRAM 0x00000004
+
struct vsock_transport {
+ uint64_t features;
+
/* Initialize/tear-down socket. */
int (*init)(struct vsock_sock *, struct vsock_sock *);
void (*destruct)(struct vsock_sock *);
@@ -156,12 +162,8 @@ struct vsock_transport {
/**** CORE ****/
-int __vsock_core_init(const struct vsock_transport *t, struct module *owner);
-static inline int vsock_core_init(const struct vsock_transport *t)
-{
- return __vsock_core_init(t, THIS_MODULE);
-}
-void vsock_core_exit(void);
+int vsock_core_register(const struct vsock_transport *t);
+void vsock_core_unregister(const struct vsock_transport *t);
/* The transport may downcast this to access transport-specific functions */
const struct vsock_transport *vsock_core_get_transport(struct vsock_sock *vsk);
@@ -192,6 +194,7 @@ struct sock *vsock_find_connected_socket(struct sockaddr_vm *src,
struct sockaddr_vm *dst);
void vsock_remove_sock(struct vsock_sock *vsk);
void vsock_for_each_connected_socket(void (*fn)(struct sock *sk));
+int vsock_assign_transport(struct vsock_sock *vsk, struct vsock_sock *psk);
/**** TAP ****/
diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c
index 95e6db21e7e1..c52203fe52c4 100644
--- a/net/vmw_vsock/af_vsock.c
+++ b/net/vmw_vsock/af_vsock.c
@@ -130,7 +130,12 @@ static struct proto vsock_proto = {
#define VSOCK_DEFAULT_BUFFER_MAX_SIZE (1024 * 256)
#define VSOCK_DEFAULT_BUFFER_MIN_SIZE 128
-static const struct vsock_transport *transport_single;
+/* Transport used for host->guest communication */
+static const struct vsock_transport *transport_h2g;
+/* Transport used for guest->host communication */
+static const struct vsock_transport *transport_g2h;
+/* Transport used for DGRAM communication */
+static const struct vsock_transport *transport_dgram;
static DEFINE_MUTEX(vsock_register_mutex);
/**** UTILS ****/
@@ -182,7 +187,7 @@ static int vsock_auto_bind(struct vsock_sock *vsk)
return __vsock_bind(sk, &local_addr);
}
-static int __init vsock_init_tables(void)
+static void vsock_init_tables(void)
{
int i;
@@ -191,7 +196,6 @@ static int __init vsock_init_tables(void)
for (i = 0; i < ARRAY_SIZE(vsock_connected_table); i++)
INIT_LIST_HEAD(&vsock_connected_table[i]);
- return 0;
}
static void __vsock_insert_bound(struct list_head *list,
@@ -376,6 +380,55 @@ void vsock_enqueue_accept(struct sock *listener, struct sock *connected)
}
EXPORT_SYMBOL_GPL(vsock_enqueue_accept);
+/* Assign a transport to a socket and call the .init transport callback.
+ *
+ * Note: for stream socket this must be called when vsk->remote_addr is set
+ * (e.g. during the connect() or when a connection request on a listener
+ * socket is received).
+ * The vsk->remote_addr is used to decide which transport to use:
+ * - remote CID > VMADDR_CID_HOST will use host->guest transport
+ * - remote CID <= VMADDR_CID_HOST will use guest->host transport
+ */
+int vsock_assign_transport(struct vsock_sock *vsk, struct vsock_sock *psk)
+{
+ struct sock *sk = sk_vsock(vsk);
+ /* RFC-TODO: should vsk->transport be already assigned?
+ * How to handle?
+ */
+ WARN_ON(vsk->transport);
+
+ switch (sk->sk_type) {
+ case SOCK_DGRAM:
+ vsk->transport = transport_dgram;
+ break;
+ case SOCK_STREAM:
+ if (vsk->remote_addr.svm_cid > VMADDR_CID_HOST)
+ vsk->transport = transport_h2g;
+ else
+ vsk->transport = transport_g2h;
+ break;
+ default:
+ return -ESOCKTNOSUPPORT;
+ }
+
+ if (!vsk->transport)
+ return -ENODEV;
+
+ return vsk->transport->init(vsk, psk);
+}
+EXPORT_SYMBOL_GPL(vsock_assign_transport);
+
+static bool vsock_find_cid(unsigned int cid)
+{
+ if (transport_g2h && cid == transport_g2h->get_local_cid())
+ return true;
+
+ if (transport_h2g && cid == VMADDR_CID_HOST)
+ return true;
+
+ return false;
+}
+
static struct sock *vsock_dequeue_accept(struct sock *listener)
{
struct vsock_sock *vlistener;
@@ -414,6 +467,9 @@ static int vsock_send_shutdown(struct sock *sk, int mode)
{
struct vsock_sock *vsk = vsock_sk(sk);
+ if (!vsk->transport)
+ return -ENODEV;
+
return vsk->transport->shutdown(vsk, mode);
}
@@ -530,7 +586,6 @@ static int __vsock_bind_dgram(struct vsock_sock *vsk,
static int __vsock_bind(struct sock *sk, struct sockaddr_vm *addr)
{
struct vsock_sock *vsk = vsock_sk(sk);
- u32 cid;
int retval;
/* First ensure this socket isn't already bound. */
@@ -540,10 +595,9 @@ static int __vsock_bind(struct sock *sk, struct sockaddr_vm *addr)
/* Now bind to the provided address or select appropriate values if
* none are provided (VMADDR_CID_ANY and VMADDR_PORT_ANY). Note that
* like AF_INET prevents binding to a non-local IP address (in most
- * cases), we only allow binding to the local CID.
+ * cases), we only allow binding to a local CID.
*/
- cid = vsk->transport->get_local_cid();
- if (addr->svm_cid != cid && addr->svm_cid != VMADDR_CID_ANY)
+ if (addr->svm_cid != VMADDR_CID_ANY && !vsock_find_cid(addr->svm_cid))
return -EADDRNOTAVAIL;
switch (sk->sk_socket->type) {
@@ -592,7 +646,6 @@ struct sock *__vsock_create(struct net *net,
sk->sk_type = type;
vsk = vsock_sk(sk);
- vsk->transport = transport_single;
vsock_addr_init(&vsk->local_addr, VMADDR_CID_ANY, VMADDR_PORT_ANY);
vsock_addr_init(&vsk->remote_addr, VMADDR_CID_ANY, VMADDR_PORT_ANY);
@@ -629,11 +682,6 @@ struct sock *__vsock_create(struct net *net,
vsk->buffer_max_size = VSOCK_DEFAULT_BUFFER_MAX_SIZE;
}
- if (vsk->transport->init(vsk, psk) < 0) {
- sk_free(sk);
- return NULL;
- }
-
return sk;
}
EXPORT_SYMBOL_GPL(__vsock_create);
@@ -648,7 +696,10 @@ static void __vsock_release(struct sock *sk)
vsk = vsock_sk(sk);
pending = NULL; /* Compiler warning. */
- vsk->transport->release(vsk);
+ if (vsk->transport)
+ vsk->transport->release(vsk);
+ else if (sk->sk_type == SOCK_STREAM)
+ vsock_remove_sock(vsk);
lock_sock(sk);
sock_orphan(sk);
@@ -672,7 +723,8 @@ static void vsock_sk_destruct(struct sock *sk)
{
struct vsock_sock *vsk = vsock_sk(sk);
- vsk->transport->destruct(vsk);
+ if (vsk->transport)
+ vsk->transport->destruct(vsk);
/* When clearing these addresses, there's no need to set the family and
* possibly register the address family with the kernel.
@@ -882,7 +934,7 @@ static __poll_t vsock_poll(struct file *file, struct socket *sock,
mask |= EPOLLIN | EPOLLRDNORM;
/* If there is something in the queue then we can read. */
- if (transport->stream_is_active(vsk) &&
+ if (transport && transport->stream_is_active(vsk) &&
!(sk->sk_shutdown & RCV_SHUTDOWN)) {
bool data_ready_now = false;
int ret = transport->notify_poll_in(
@@ -1132,7 +1184,6 @@ static int vsock_stream_connect(struct socket *sock, struct sockaddr *addr,
err = 0;
sk = sock->sk;
vsk = vsock_sk(sk);
- transport = vsk->transport;
lock_sock(sk);
@@ -1160,19 +1211,26 @@ static int vsock_stream_connect(struct socket *sock, struct sockaddr *addr,
goto out;
}
+ /* Set the remote address that we are connecting to. */
+ memcpy(&vsk->remote_addr, remote_addr,
+ sizeof(vsk->remote_addr));
+
+ err = vsock_assign_transport(vsk, NULL);
+ if (err)
+ goto out;
+
+ transport = vsk->transport;
+
/* The hypervisor and well-known contexts do not have socket
* endpoints.
*/
- if (!transport->stream_allow(remote_addr->svm_cid,
+ if (!transport ||
+ !transport->stream_allow(remote_addr->svm_cid,
remote_addr->svm_port)) {
err = -ENETUNREACH;
goto out;
}
- /* Set the remote address that we are connecting to. */
- memcpy(&vsk->remote_addr, remote_addr,
- sizeof(vsk->remote_addr));
-
err = vsock_auto_bind(vsk);
if (err)
goto out;
@@ -1572,7 +1630,7 @@ static int vsock_stream_sendmsg(struct socket *sock, struct msghdr *msg,
goto out;
}
- if (sk->sk_state != TCP_ESTABLISHED ||
+ if (!transport || sk->sk_state != TCP_ESTABLISHED ||
!vsock_addr_bound(&vsk->local_addr)) {
err = -ENOTCONN;
goto out;
@@ -1698,7 +1756,7 @@ vsock_stream_recvmsg(struct socket *sock, struct msghdr *msg, size_t len,
lock_sock(sk);
- if (sk->sk_state != TCP_ESTABLISHED) {
+ if (!transport || sk->sk_state != TCP_ESTABLISHED) {
/* Recvmsg is supposed to return 0 if a peer performs an
* orderly shutdown. Differentiate between that case and when a
* peer has not connected or a local shutdown occured with the
@@ -1872,7 +1930,9 @@ static const struct proto_ops vsock_stream_ops = {
static int vsock_create(struct net *net, struct socket *sock,
int protocol, int kern)
{
+ struct vsock_sock *vsk;
struct sock *sk;
+ int ret;
if (!sock)
return -EINVAL;
@@ -1897,7 +1957,20 @@ static int vsock_create(struct net *net, struct socket *sock,
if (!sk)
return -ENOMEM;
- vsock_insert_unbound(vsock_sk(sk));
+ vsk = vsock_sk(sk);
+
+ /* RFC-TODO: for dgram we still support only one transport, and
+ * we assign it during the sock creation.
+ */
+ if (sock->type == SOCK_DGRAM) {
+ ret = vsock_assign_transport(vsk, NULL);
+ if (ret < 0) {
+ sock_put(sk);
+ return ret;
+ }
+ }
+
+ vsock_insert_unbound(vsk);
return 0;
}
@@ -1912,11 +1985,20 @@ static long vsock_dev_do_ioctl(struct file *filp,
unsigned int cmd, void __user *ptr)
{
u32 __user *p = ptr;
+ u32 cid = VMADDR_CID_ANY;
int retval = 0;
switch (cmd) {
case IOCTL_VM_SOCKETS_GET_LOCAL_CID:
- if (put_user(transport_single->get_local_cid(), p) != 0)
+ /* To be compatible with the VMCI behavior, we prioritize the
+ * guest CID instead of well-know host CID (VMADDR_CID_HOST).
+ */
+ if (transport_g2h)
+ cid = transport_g2h->get_local_cid();
+ else if (transport_h2g)
+ cid = transport_h2g->get_local_cid();
+
+ if (put_user(cid, p) != 0)
retval = -EFAULT;
break;
@@ -1956,24 +2038,13 @@ static struct miscdevice vsock_device = {
.fops = &vsock_device_ops,
};
-int __vsock_core_init(const struct vsock_transport *t, struct module *owner)
+static int __init vsock_init(void)
{
- int err = mutex_lock_interruptible(&vsock_register_mutex);
+ int err = 0;
- if (err)
- return err;
-
- if (transport_single) {
- err = -EBUSY;
- goto err_busy;
- }
-
- /* Transport must be the owner of the protocol so that it can't
- * unload while there are open sockets.
- */
- vsock_proto.owner = owner;
- transport_single = t;
+ vsock_init_tables();
+ vsock_proto.owner = THIS_MODULE;
vsock_device.minor = MISC_DYNAMIC_MINOR;
err = misc_register(&vsock_device);
if (err) {
@@ -1994,7 +2065,6 @@ int __vsock_core_init(const struct vsock_transport *t, struct module *owner)
goto err_unregister_proto;
}
- mutex_unlock(&vsock_register_mutex);
return 0;
err_unregister_proto:
@@ -2002,28 +2072,15 @@ int __vsock_core_init(const struct vsock_transport *t, struct module *owner)
err_deregister_misc:
misc_deregister(&vsock_device);
err_reset_transport:
- transport_single = NULL;
-err_busy:
- mutex_unlock(&vsock_register_mutex);
return err;
}
-EXPORT_SYMBOL_GPL(__vsock_core_init);
-void vsock_core_exit(void)
+static void __exit vsock_exit(void)
{
- mutex_lock(&vsock_register_mutex);
-
misc_deregister(&vsock_device);
sock_unregister(AF_VSOCK);
proto_unregister(&vsock_proto);
-
- /* We do not want the assignment below re-ordered. */
- mb();
- transport_single = NULL;
-
- mutex_unlock(&vsock_register_mutex);
}
-EXPORT_SYMBOL_GPL(vsock_core_exit);
const struct vsock_transport *vsock_core_get_transport(struct vsock_sock *vsk)
{
@@ -2034,12 +2091,77 @@ const struct vsock_transport *vsock_core_get_transport(struct vsock_sock *vsk)
}
EXPORT_SYMBOL_GPL(vsock_core_get_transport);
-static void __exit vsock_exit(void)
+int vsock_core_register(const struct vsock_transport *t)
{
- /* Do nothing. This function makes this module removable. */
+ const struct vsock_transport *t_h2g, *t_g2h, *t_dgram;
+ int err = mutex_lock_interruptible(&vsock_register_mutex);
+
+ if (err)
+ return err;
+
+ t_h2g = transport_h2g;
+ t_g2h = transport_g2h;
+ t_dgram = transport_dgram;
+
+ /* RFC-TODO: vmci transport offer both H2G and G2H features in the
+ * same transport. We are able to set the G2H feature only if we are
+ * in a VMware guest, but we are not able to do the same for the host.
+ */
+ if (t->features & VSOCK_TRANSPORT_F_H2G) {
+ if (t_h2g) {
+ err = -EBUSY;
+ goto err_busy;
+ }
+ t_h2g = t;
+ }
+
+ if (t->features & VSOCK_TRANSPORT_F_G2H) {
+ if (t_g2h) {
+ err = -EBUSY;
+ goto err_busy;
+ }
+ t_g2h = t;
+ }
+
+ if (t->features & VSOCK_TRANSPORT_F_DGRAM) {
+ if (t_dgram) {
+ err = -EBUSY;
+ goto err_busy;
+ }
+ t_dgram = t;
+ }
+
+ transport_h2g = t_h2g;
+ transport_g2h = t_g2h;
+ transport_dgram = t_dgram;
+
+err_busy:
+ mutex_unlock(&vsock_register_mutex);
+ return err;
+}
+EXPORT_SYMBOL_GPL(vsock_core_register);
+
+void vsock_core_unregister(const struct vsock_transport *t)
+{
+ mutex_lock(&vsock_register_mutex);
+
+ /* RFC-TODO: maybe we should check if there are open sockets
+ * assigned to that transport and avoid the unregistration
+ */
+ if (transport_h2g == t)
+ transport_h2g = NULL;
+
+ if (transport_g2h == t)
+ transport_g2h = NULL;
+
+ if (transport_dgram == t)
+ transport_dgram = NULL;
+
+ mutex_unlock(&vsock_register_mutex);
}
+EXPORT_SYMBOL_GPL(vsock_core_unregister);
-module_init(vsock_init_tables);
+module_init(vsock_init);
module_exit(vsock_exit);
MODULE_AUTHOR("VMware, Inc.");
diff --git a/net/vmw_vsock/hyperv_transport.c b/net/vmw_vsock/hyperv_transport.c
index 306310794522..94e6fc905a77 100644
--- a/net/vmw_vsock/hyperv_transport.c
+++ b/net/vmw_vsock/hyperv_transport.c
@@ -163,6 +163,8 @@ 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);
@@ -366,6 +368,18 @@ static void hvs_open_connection(struct vmbus_channel *chan)
new->sk_state = TCP_SYN_SENT;
vnew = vsock_sk(new);
+
+ hvs_addr_init(&vnew->local_addr, if_type);
+ hvs_remote_addr_init(&vnew->remote_addr, &vnew->local_addr);
+
+ 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 {
@@ -429,9 +443,6 @@ static void hvs_open_connection(struct vmbus_channel *chan)
new->sk_state = TCP_ESTABLISHED;
sk->sk_ack_backlog++;
- hvs_addr_init(&vnew->local_addr, if_type);
- hvs_remote_addr_init(&vnew->remote_addr, &vnew->local_addr);
-
hvs_new->vm_srv_id = *if_type;
hvs_new->host_srv_id = *if_instance;
@@ -845,6 +856,8 @@ int hvs_notify_send_post_enqueue(struct vsock_sock *vsk, ssize_t written,
}
static struct vsock_transport hvs_transport = {
+ .features = VSOCK_TRANSPORT_F_G2H,
+
.get_local_cid = hvs_get_local_cid,
.init = hvs_sock_init,
@@ -879,6 +892,11 @@ static struct vsock_transport hvs_transport = {
};
+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)
{
@@ -927,7 +945,7 @@ static int __init hvs_init(void)
if (ret != 0)
return ret;
- ret = vsock_core_init(&hvs_transport);
+ ret = vsock_core_register(&hvs_transport);
if (ret) {
vmbus_driver_unregister(&hvs_drv);
return ret;
@@ -938,7 +956,7 @@ static int __init hvs_init(void)
static void __exit hvs_exit(void)
{
- vsock_core_exit();
+ vsock_core_unregister(&hvs_transport);
vmbus_driver_unregister(&hvs_drv);
}
diff --git a/net/vmw_vsock/virtio_transport.c b/net/vmw_vsock/virtio_transport.c
index fb1fc7760e8c..0ff037ef7f8e 100644
--- a/net/vmw_vsock/virtio_transport.c
+++ b/net/vmw_vsock/virtio_transport.c
@@ -462,6 +462,8 @@ static void virtio_vsock_rx_done(struct virtqueue *vq)
static struct virtio_transport virtio_transport = {
.transport = {
+ .features = VSOCK_TRANSPORT_F_G2H,
+
.get_local_cid = virtio_transport_get_local_cid,
.init = virtio_transport_do_socket_init,
@@ -770,7 +772,7 @@ static int __init virtio_vsock_init(void)
if (!virtio_vsock_workqueue)
return -ENOMEM;
- ret = vsock_core_init(&virtio_transport.transport);
+ ret = vsock_core_register(&virtio_transport.transport);
if (ret)
goto out_wq;
@@ -781,7 +783,7 @@ static int __init virtio_vsock_init(void)
return 0;
out_vci:
- vsock_core_exit();
+ vsock_core_unregister(&virtio_transport.transport);
out_wq:
destroy_workqueue(virtio_vsock_workqueue);
return ret;
@@ -790,7 +792,7 @@ static int __init virtio_vsock_init(void)
static void __exit virtio_vsock_exit(void)
{
unregister_virtio_driver(&virtio_vsock_driver);
- vsock_core_exit();
+ vsock_core_unregister(&virtio_transport.transport);
destroy_workqueue(virtio_vsock_workqueue);
}
diff --git a/net/vmw_vsock/virtio_transport_common.c b/net/vmw_vsock/virtio_transport_common.c
index bac9e7430a2e..ebb4701310a4 100644
--- a/net/vmw_vsock/virtio_transport_common.c
+++ b/net/vmw_vsock/virtio_transport_common.c
@@ -400,7 +400,7 @@ int virtio_transport_do_socket_init(struct vsock_sock *vsk,
vsk->trans = vvs;
vvs->vsk = vsk;
- if (psk) {
+ if (psk && psk->trans) {
struct virtio_vsock_sock *ptrans = psk->trans;
vvs->peer_buf_alloc = ptrans->peer_buf_alloc;
@@ -927,11 +927,13 @@ virtio_transport_send_response(struct vsock_sock *vsk,
/* Handle server socket */
static int
-virtio_transport_recv_listen(struct sock *sk, struct virtio_vsock_pkt *pkt)
+virtio_transport_recv_listen(struct sock *sk, struct virtio_vsock_pkt *pkt,
+ struct virtio_transport *t)
{
struct vsock_sock *vsk = vsock_sk(sk);
struct vsock_sock *vchild;
struct sock *child;
+ int ret;
if (le16_to_cpu(pkt->hdr.op) != VIRTIO_VSOCK_OP_REQUEST) {
virtio_transport_reset(vsk, pkt);
@@ -962,6 +964,17 @@ virtio_transport_recv_listen(struct sock *sk, struct virtio_vsock_pkt *pkt)
vsock_addr_init(&vchild->remote_addr, le64_to_cpu(pkt->hdr.src_cid),
le32_to_cpu(pkt->hdr.src_port));
+ ret = vsock_assign_transport(vchild, vsk);
+ /* Transport assigned (looking at remote_addr) must be the same
+ * where we received the request.
+ */
+ if (ret || vchild->transport != &t->transport) {
+ release_sock(child);
+ virtio_transport_reset(vsk, pkt);
+ sock_put(child);
+ return ret;
+ }
+
vsock_insert_connected(vchild);
vsock_enqueue_accept(sk, child);
virtio_transport_send_response(vchild, pkt);
@@ -979,6 +992,14 @@ static bool virtio_transport_space_update(struct sock *sk,
struct virtio_vsock_sock *vvs = vsk->trans;
bool space_available;
+ /* Listener sockets are not associated with any transport, so we are
+ * not able to take the state to see if there is space available in the
+ * remote peer, but since they are only used to receive requests, we
+ * can assume that there is always space available in the other peer.
+ */
+ if (!vvs)
+ return true;
+
/* buf_alloc and fwd_cnt is always included in the hdr */
spin_lock_bh(&vvs->tx_lock);
vvs->peer_buf_alloc = le32_to_cpu(pkt->hdr.buf_alloc);
@@ -1044,7 +1065,7 @@ void virtio_transport_recv_pkt(struct virtio_transport *t,
switch (sk->sk_state) {
case TCP_LISTEN:
- virtio_transport_recv_listen(sk, pkt);
+ virtio_transport_recv_listen(sk, pkt, t);
virtio_transport_free_pkt(pkt);
break;
case TCP_SYN_SENT:
@@ -1062,6 +1083,7 @@ void virtio_transport_recv_pkt(struct virtio_transport *t,
virtio_transport_free_pkt(pkt);
break;
}
+
release_sock(sk);
/* Release refcnt obtained when we fetched this socket out of the
diff --git a/net/vmw_vsock/vmci_transport.c b/net/vmw_vsock/vmci_transport.c
index 8290d37b6587..52e63952d0d4 100644
--- a/net/vmw_vsock/vmci_transport.c
+++ b/net/vmw_vsock/vmci_transport.c
@@ -57,6 +57,7 @@ static bool vmci_transport_old_proto_override(bool *old_pkt_proto);
static u16 vmci_transport_new_proto_supported_versions(void);
static bool vmci_transport_proto_to_notify_struct(struct sock *sk, u16 *proto,
bool old_pkt_proto);
+static bool vmci_check_transport(struct vsock_sock *vsk);
struct vmci_transport_recv_pkt_info {
struct work_struct work;
@@ -1018,6 +1019,15 @@ static int vmci_transport_recv_listen(struct sock *sk,
vsock_addr_init(&vpending->remote_addr, pkt->dg.src.context,
pkt->src_port);
+ err = vsock_assign_transport(vpending, vsock_sk(sk));
+ /* Transport assigned (looking at remote_addr) must be the same
+ * where we received the request.
+ */
+ if (err || !vmci_check_transport(vpending)) {
+ sock_put(pending);
+ return err;
+ }
+
/* If the proposed size fits within our min/max, accept it. Otherwise
* propose our own size.
*/
@@ -2009,7 +2019,8 @@ static u32 vmci_transport_get_local_cid(void)
return vmci_get_context_id();
}
-static const struct vsock_transport vmci_transport = {
+static struct vsock_transport vmci_transport = {
+ .features = VSOCK_TRANSPORT_F_DGRAM | VSOCK_TRANSPORT_F_H2G,
.init = vmci_transport_socket_init,
.destruct = vmci_transport_destruct,
.release = vmci_transport_release,
@@ -2039,10 +2050,24 @@ static const struct vsock_transport vmci_transport = {
.get_local_cid = vmci_transport_get_local_cid,
};
+static bool vmci_check_transport(struct vsock_sock *vsk)
+{
+ return vsk->transport == &vmci_transport;
+}
+
static int __init vmci_transport_init(void)
{
+ int cid;
int err;
+ cid = vmci_get_context_id();
+
+ if (cid == VMCI_INVALID_ID)
+ return -EINVAL;
+
+ if (cid != VMCI_HOST_CONTEXT_ID)
+ vmci_transport.features |= VSOCK_TRANSPORT_F_G2H;
+
/* Create the datagram handle that we will use to send and receive all
* VSocket control messages for this context.
*/
@@ -2066,7 +2091,7 @@ static int __init vmci_transport_init(void)
goto err_destroy_stream_handle;
}
- err = vsock_core_init(&vmci_transport);
+ err = vsock_core_register(&vmci_transport);
if (err < 0)
goto err_unsubscribe;
@@ -2097,7 +2122,7 @@ static void __exit vmci_transport_exit(void)
vmci_transport_qp_resumed_sub_id = VMCI_INVALID_ID;
}
- vsock_core_exit();
+ vsock_core_unregister(&vmci_transport);
}
module_exit(vmci_transport_exit);
--
2.21.0
Since now the 'struct vsock_sock' object contains a pointer to
the transport, this patch adds a parameter to the
vsock_core_get_transport() to return the right transport
assigned to the socket.
This patch modifies also the virtio_transport_get_ops(), that
uses the vsock_core_get_transport(), adding the
'struct vsock_sock *' parameter.
Signed-off-by: Stefano Garzarella <[email protected]>
---
include/net/af_vsock.h | 2 +-
net/vmw_vsock/af_vsock.c | 4 ++--
net/vmw_vsock/virtio_transport_common.c | 9 +++++----
3 files changed, 8 insertions(+), 7 deletions(-)
diff --git a/include/net/af_vsock.h b/include/net/af_vsock.h
index a5e1e134261d..2ca67d048de4 100644
--- a/include/net/af_vsock.h
+++ b/include/net/af_vsock.h
@@ -166,7 +166,7 @@ static inline int vsock_core_init(const struct vsock_transport *t)
void vsock_core_exit(void);
/* The transport may downcast this to access transport-specific functions */
-const struct vsock_transport *vsock_core_get_transport(void);
+const struct vsock_transport *vsock_core_get_transport(struct vsock_sock *vsk);
/**** UTILS ****/
diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c
index 81ee2561c76f..f7540a3ac64e 100644
--- a/net/vmw_vsock/af_vsock.c
+++ b/net/vmw_vsock/af_vsock.c
@@ -1993,12 +1993,12 @@ void vsock_core_exit(void)
}
EXPORT_SYMBOL_GPL(vsock_core_exit);
-const struct vsock_transport *vsock_core_get_transport(void)
+const struct vsock_transport *vsock_core_get_transport(struct vsock_sock *vsk)
{
/* vsock_register_mutex not taken since only the transport uses this
* function and only while registered.
*/
- return transport_single;
+ return vsk->transport;
}
EXPORT_SYMBOL_GPL(vsock_core_get_transport);
diff --git a/net/vmw_vsock/virtio_transport_common.c b/net/vmw_vsock/virtio_transport_common.c
index 382536b69029..fc046c071178 100644
--- a/net/vmw_vsock/virtio_transport_common.c
+++ b/net/vmw_vsock/virtio_transport_common.c
@@ -29,9 +29,10 @@
/* Threshold for detecting small packets to copy */
#define GOOD_COPY_LEN 128
-static const struct virtio_transport *virtio_transport_get_ops(void)
+static const struct virtio_transport *
+virtio_transport_get_ops(struct vsock_sock *vsk)
{
- const struct vsock_transport *t = vsock_core_get_transport();
+ const struct vsock_transport *t = vsock_core_get_transport(vsk);
return container_of(t, struct virtio_transport, transport);
}
@@ -168,7 +169,7 @@ static int virtio_transport_send_pkt_info(struct vsock_sock *vsk,
struct virtio_vsock_pkt *pkt;
u32 pkt_len = info->pkt_len;
- src_cid = virtio_transport_get_ops()->transport.get_local_cid();
+ src_cid = virtio_transport_get_ops(vsk)->transport.get_local_cid();
src_port = vsk->local_addr.svm_port;
if (!info->remote_cid) {
dst_cid = vsk->remote_addr.svm_cid;
@@ -201,7 +202,7 @@ static int virtio_transport_send_pkt_info(struct vsock_sock *vsk,
virtio_transport_inc_tx_pkt(vvs, pkt);
- return virtio_transport_get_ops()->send_pkt(pkt);
+ return virtio_transport_get_ops(vsk)->send_pkt(pkt);
}
static void virtio_transport_inc_rx_pkt(struct virtio_vsock_sock *vvs,
--
2.21.0
VMCI transport provides both g2h and h2g behaviors in a single
transport.
We are able to set (or not) the g2h behavior, detecting if we
are in a VMware guest (or not), but the h2g feature is always set.
This prevents to load other h2g transports while we are in a
VMware guest.
This patch adds a new 'transport_hg' to handle this case, reducing
the priority of transports that provide both g2h and h2g
behaviors. A transport that has g2h and h2g features, can be
bypassed by a transport that has only the h2g feature.
Signed-off-by: Stefano Garzarella <[email protected]>
---
net/vmw_vsock/af_vsock.c | 29 ++++++++++++++++++++++++-----
1 file changed, 24 insertions(+), 5 deletions(-)
diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c
index c52203fe52c4..c5f46b8242ce 100644
--- a/net/vmw_vsock/af_vsock.c
+++ b/net/vmw_vsock/af_vsock.c
@@ -134,6 +134,8 @@ static struct proto vsock_proto = {
static const struct vsock_transport *transport_h2g;
/* Transport used for guest->host communication */
static const struct vsock_transport *transport_g2h;
+/* Transport used for both host->guest and guest->host communication */
+static const struct vsock_transport *transport_hg;
/* Transport used for DGRAM communication */
static const struct vsock_transport *transport_dgram;
static DEFINE_MUTEX(vsock_register_mutex);
@@ -402,10 +404,13 @@ int vsock_assign_transport(struct vsock_sock *vsk, struct vsock_sock *psk)
vsk->transport = transport_dgram;
break;
case SOCK_STREAM:
- if (vsk->remote_addr.svm_cid > VMADDR_CID_HOST)
+ if (vsk->remote_addr.svm_cid > VMADDR_CID_HOST) {
vsk->transport = transport_h2g;
- else
+ if (!vsk->transport)
+ vsk->transport = transport_hg;
+ } else {
vsk->transport = transport_g2h;
+ }
break;
default:
return -ESOCKTNOSUPPORT;
@@ -423,7 +428,7 @@ static bool vsock_find_cid(unsigned int cid)
if (transport_g2h && cid == transport_g2h->get_local_cid())
return true;
- if (transport_h2g && cid == VMADDR_CID_HOST)
+ if ((transport_h2g || transport_hg) && cid == VMADDR_CID_HOST)
return true;
return false;
@@ -1997,6 +2002,8 @@ static long vsock_dev_do_ioctl(struct file *filp,
cid = transport_g2h->get_local_cid();
else if (transport_h2g)
cid = transport_h2g->get_local_cid();
+ else if (transport_hg)
+ cid = transport_hg->get_local_cid();
if (put_user(cid, p) != 0)
retval = -EFAULT;
@@ -2093,13 +2100,14 @@ EXPORT_SYMBOL_GPL(vsock_core_get_transport);
int vsock_core_register(const struct vsock_transport *t)
{
- const struct vsock_transport *t_h2g, *t_g2h, *t_dgram;
+ const struct vsock_transport *t_h2g, *t_hg, *t_g2h, *t_dgram;
int err = mutex_lock_interruptible(&vsock_register_mutex);
if (err)
return err;
t_h2g = transport_h2g;
+ t_hg = transport_hg;
t_g2h = transport_g2h;
t_dgram = transport_dgram;
@@ -2107,7 +2115,14 @@ int vsock_core_register(const struct vsock_transport *t)
* same transport. We are able to set the G2H feature only if we are
* in a VMware guest, but we are not able to do the same for the host.
*/
- if (t->features & VSOCK_TRANSPORT_F_H2G) {
+ if ((t->features & VSOCK_TRANSPORT_F_H2G) &&
+ (t->features & VSOCK_TRANSPORT_F_G2H)) {
+ if (t_hg) {
+ err = -EBUSY;
+ goto err_busy;
+ }
+ t_hg = t;
+ } else if (t->features & VSOCK_TRANSPORT_F_H2G) {
if (t_h2g) {
err = -EBUSY;
goto err_busy;
@@ -2132,6 +2147,7 @@ int vsock_core_register(const struct vsock_transport *t)
}
transport_h2g = t_h2g;
+ transport_hg = t_hg;
transport_g2h = t_g2h;
transport_dgram = t_dgram;
@@ -2151,6 +2167,9 @@ void vsock_core_unregister(const struct vsock_transport *t)
if (transport_h2g == t)
transport_h2g = NULL;
+ if (transport_hg == t)
+ transport_hg = NULL;
+
if (transport_g2h == t)
transport_g2h = NULL;
--
2.21.0
This patch adds 'module' member in the 'struct vsock_transport'
in order to get/put the transport module. This prevents the
module unloading while sockets are assigned to it.
We increase the module refcnt when a socket is assigned to a
transport, and we decrease the module refcnt when the socket
is destructed.
Signed-off-by: Stefano Garzarella <[email protected]>
---
drivers/vhost/vsock.c | 1 +
include/net/af_vsock.h | 1 +
net/vmw_vsock/af_vsock.c | 23 +++++++++++++++++------
net/vmw_vsock/hyperv_transport.c | 1 +
net/vmw_vsock/virtio_transport.c | 1 +
net/vmw_vsock/vmci_transport.c | 1 +
6 files changed, 22 insertions(+), 6 deletions(-)
diff --git a/drivers/vhost/vsock.c b/drivers/vhost/vsock.c
index 375af01a5b64..6d7a8fc9eb63 100644
--- a/drivers/vhost/vsock.c
+++ b/drivers/vhost/vsock.c
@@ -387,6 +387,7 @@ static bool vhost_vsock_more_replies(struct vhost_vsock *vsock)
static struct virtio_transport vhost_transport = {
.transport = {
.features = VSOCK_TRANSPORT_F_H2G,
+ .module = THIS_MODULE,
.get_local_cid = vhost_transport_get_local_cid,
diff --git a/include/net/af_vsock.h b/include/net/af_vsock.h
index 2a081d19e20d..f10fa918bf23 100644
--- a/include/net/af_vsock.h
+++ b/include/net/af_vsock.h
@@ -100,6 +100,7 @@ struct vsock_transport_send_notify_data {
struct vsock_transport {
uint64_t features;
+ struct module *module;
/* Initialize/tear-down socket. */
int (*init)(struct vsock_sock *, struct vsock_sock *);
diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c
index c5f46b8242ce..750b62711b01 100644
--- a/net/vmw_vsock/af_vsock.c
+++ b/net/vmw_vsock/af_vsock.c
@@ -416,13 +416,28 @@ int vsock_assign_transport(struct vsock_sock *vsk, struct vsock_sock *psk)
return -ESOCKTNOSUPPORT;
}
- if (!vsk->transport)
+ /* We increase the module refcnt to prevent the tranport unloading
+ * while there are open sockets assigned to it.
+ */
+ if (!vsk->transport || !try_module_get(vsk->transport->module)) {
+ vsk->transport = NULL;
return -ENODEV;
+ }
return vsk->transport->init(vsk, psk);
}
EXPORT_SYMBOL_GPL(vsock_assign_transport);
+static void vsock_deassign_transport(struct vsock_sock *vsk)
+{
+ if (!vsk->transport)
+ return;
+
+ vsk->transport->destruct(vsk);
+ module_put(vsk->transport->module);
+ vsk->transport = NULL;
+}
+
static bool vsock_find_cid(unsigned int cid)
{
if (transport_g2h && cid == transport_g2h->get_local_cid())
@@ -728,8 +743,7 @@ static void vsock_sk_destruct(struct sock *sk)
{
struct vsock_sock *vsk = vsock_sk(sk);
- if (vsk->transport)
- vsk->transport->destruct(vsk);
+ vsock_deassign_transport(vsk);
/* When clearing these addresses, there's no need to set the family and
* possibly register the address family with the kernel.
@@ -2161,9 +2175,6 @@ void vsock_core_unregister(const struct vsock_transport *t)
{
mutex_lock(&vsock_register_mutex);
- /* RFC-TODO: maybe we should check if there are open sockets
- * assigned to that transport and avoid the unregistration
- */
if (transport_h2g == t)
transport_h2g = NULL;
diff --git a/net/vmw_vsock/hyperv_transport.c b/net/vmw_vsock/hyperv_transport.c
index 94e6fc905a77..bd4f3c222904 100644
--- a/net/vmw_vsock/hyperv_transport.c
+++ b/net/vmw_vsock/hyperv_transport.c
@@ -857,6 +857,7 @@ int hvs_notify_send_post_enqueue(struct vsock_sock *vsk, ssize_t written,
static struct vsock_transport hvs_transport = {
.features = VSOCK_TRANSPORT_F_G2H,
+ .module = THIS_MODULE,
.get_local_cid = hvs_get_local_cid,
diff --git a/net/vmw_vsock/virtio_transport.c b/net/vmw_vsock/virtio_transport.c
index 0ff037ef7f8e..439fe01e6691 100644
--- a/net/vmw_vsock/virtio_transport.c
+++ b/net/vmw_vsock/virtio_transport.c
@@ -463,6 +463,7 @@ static void virtio_vsock_rx_done(struct virtqueue *vq)
static struct virtio_transport virtio_transport = {
.transport = {
.features = VSOCK_TRANSPORT_F_G2H,
+ .module = THIS_MODULE,
.get_local_cid = virtio_transport_get_local_cid,
diff --git a/net/vmw_vsock/vmci_transport.c b/net/vmw_vsock/vmci_transport.c
index 52e63952d0d4..900392686c03 100644
--- a/net/vmw_vsock/vmci_transport.c
+++ b/net/vmw_vsock/vmci_transport.c
@@ -2021,6 +2021,7 @@ static u32 vmci_transport_get_local_cid(void)
static struct vsock_transport vmci_transport = {
.features = VSOCK_TRANSPORT_F_DGRAM | VSOCK_TRANSPORT_F_H2G,
+ .module = THIS_MODULE,
.init = vmci_transport_socket_init,
.destruct = vmci_transport_destruct,
.release = vmci_transport_release,
--
2.21.0
When we are looking for a socket bound to a specific address,
we also have to take into account the CID.
This patch is useful with multi-transports support because it
allows the binding of the same port with different CID, and
it prevents a connection to a wrong socket bound to the same
port, but with different CID.
Signed-off-by: Stefano Garzarella <[email protected]>
---
net/vmw_vsock/af_vsock.c | 10 ++++++++--
1 file changed, 8 insertions(+), 2 deletions(-)
diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c
index 750b62711b01..bffa6fa7b8e5 100644
--- a/net/vmw_vsock/af_vsock.c
+++ b/net/vmw_vsock/af_vsock.c
@@ -230,10 +230,16 @@ static struct sock *__vsock_find_bound_socket(struct sockaddr_vm *addr)
{
struct vsock_sock *vsk;
- list_for_each_entry(vsk, vsock_bound_sockets(addr), bound_table)
- if (addr->svm_port == vsk->local_addr.svm_port)
+ list_for_each_entry(vsk, vsock_bound_sockets(addr), bound_table) {
+ if (vsock_addr_equals_addr(addr, &vsk->local_addr))
return sk_vsock(vsk);
+ if (addr->svm_port == vsk->local_addr.svm_port &&
+ (vsk->local_addr.svm_cid == VMADDR_CID_ANY ||
+ addr->svm_cid == VMADDR_CID_ANY))
+ return sk_vsock(vsk);
+ }
+
return NULL;
}
--
2.21.0
virtio_transport and vmci_transport handle the buffer_size
sockopts in a very similar way.
In order to support multiple transports, this patch moves this
handling in the core to allow the user to change the options
also if the socket is not yet assigned to any transport.
This patch also adds the '.notify_buffer_size' callback in the
'struct virtio_transport' in order to inform the transport,
when the buffer_size is changed by the user. It is also useful
to limit the 'buffer_size' requested (e.g. virtio transports).
Signed-off-by: Stefano Garzarella <[email protected]>
---
drivers/vhost/vsock.c | 7 +-
include/linux/virtio_vsock.h | 15 +----
include/net/af_vsock.h | 14 ++--
net/vmw_vsock/af_vsock.c | 43 ++++++++++---
net/vmw_vsock/hyperv_transport.c | 36 -----------
net/vmw_vsock/virtio_transport.c | 8 +--
net/vmw_vsock/virtio_transport_common.c | 78 ++++------------------
net/vmw_vsock/vmci_transport.c | 86 +++----------------------
net/vmw_vsock/vmci_transport.h | 3 -
9 files changed, 64 insertions(+), 226 deletions(-)
diff --git a/drivers/vhost/vsock.c b/drivers/vhost/vsock.c
index 92ab3852c954..6d7e4f022748 100644
--- a/drivers/vhost/vsock.c
+++ b/drivers/vhost/vsock.c
@@ -418,13 +418,8 @@ static struct virtio_transport vhost_transport = {
.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,
- .set_buffer_size = virtio_transport_set_buffer_size,
- .set_min_buffer_size = virtio_transport_set_min_buffer_size,
- .set_max_buffer_size = virtio_transport_set_max_buffer_size,
- .get_buffer_size = virtio_transport_get_buffer_size,
- .get_min_buffer_size = virtio_transport_get_min_buffer_size,
- .get_max_buffer_size = virtio_transport_get_max_buffer_size,
},
.send_pkt = vhost_transport_send_pkt,
diff --git a/include/linux/virtio_vsock.h b/include/linux/virtio_vsock.h
index 96d8132acbd7..ab02d119fe79 100644
--- a/include/linux/virtio_vsock.h
+++ b/include/linux/virtio_vsock.h
@@ -7,9 +7,6 @@
#include <net/sock.h>
#include <net/af_vsock.h>
-#define VIRTIO_VSOCK_DEFAULT_MIN_BUF_SIZE 128
-#define VIRTIO_VSOCK_DEFAULT_BUF_SIZE (1024 * 256)
-#define VIRTIO_VSOCK_DEFAULT_MAX_BUF_SIZE (1024 * 256)
#define VIRTIO_VSOCK_DEFAULT_RX_BUF_SIZE (1024 * 4)
#define VIRTIO_VSOCK_MAX_BUF_SIZE 0xFFFFFFFFUL
#define VIRTIO_VSOCK_MAX_PKT_BUF_SIZE (1024 * 64)
@@ -25,11 +22,6 @@ enum {
struct virtio_vsock_sock {
struct vsock_sock *vsk;
- /* Protected by lock_sock(sk_vsock(trans->vsk)) */
- u32 buf_size;
- u32 buf_size_min;
- u32 buf_size_max;
-
spinlock_t tx_lock;
spinlock_t rx_lock;
@@ -93,12 +85,6 @@ s64 virtio_transport_stream_has_space(struct vsock_sock *vsk);
int virtio_transport_do_socket_init(struct vsock_sock *vsk,
struct vsock_sock *psk);
-u64 virtio_transport_get_buffer_size(struct vsock_sock *vsk);
-u64 virtio_transport_get_min_buffer_size(struct vsock_sock *vsk);
-u64 virtio_transport_get_max_buffer_size(struct vsock_sock *vsk);
-void virtio_transport_set_buffer_size(struct vsock_sock *vsk, u64 val);
-void virtio_transport_set_min_buffer_size(struct vsock_sock *vsk, u64 val);
-void virtio_transport_set_max_buffer_size(struct vsock_sock *vs, u64 val);
int
virtio_transport_notify_poll_in(struct vsock_sock *vsk,
size_t target,
@@ -125,6 +111,7 @@ int virtio_transport_notify_send_pre_enqueue(struct vsock_sock *vsk,
struct vsock_transport_send_notify_data *data);
int virtio_transport_notify_send_post_enqueue(struct vsock_sock *vsk,
ssize_t written, struct vsock_transport_send_notify_data *data);
+int virtio_transport_notify_buffer_size(struct vsock_sock *vsk, u64 *val);
u64 virtio_transport_stream_rcvhiwat(struct vsock_sock *vsk);
bool virtio_transport_stream_is_active(struct vsock_sock *vsk);
diff --git a/include/net/af_vsock.h b/include/net/af_vsock.h
index 2ca67d048de4..86f8f463e01a 100644
--- a/include/net/af_vsock.h
+++ b/include/net/af_vsock.h
@@ -65,6 +65,11 @@ struct vsock_sock {
bool sent_request;
bool ignore_connecting_rst;
+ /* Protected by lock_sock(sk) */
+ u64 buffer_size;
+ u64 buffer_min_size;
+ u64 buffer_max_size;
+
/* Private to transport. */
void *trans;
};
@@ -140,18 +145,11 @@ struct vsock_transport {
struct vsock_transport_send_notify_data *);
int (*notify_send_post_enqueue)(struct vsock_sock *, ssize_t,
struct vsock_transport_send_notify_data *);
+ int (*notify_buffer_size)(struct vsock_sock *, u64 *);
/* Shutdown. */
int (*shutdown)(struct vsock_sock *, int);
- /* Buffer sizes. */
- void (*set_buffer_size)(struct vsock_sock *, u64);
- void (*set_min_buffer_size)(struct vsock_sock *, u64);
- void (*set_max_buffer_size)(struct vsock_sock *, u64);
- u64 (*get_buffer_size)(struct vsock_sock *);
- u64 (*get_min_buffer_size)(struct vsock_sock *);
- u64 (*get_max_buffer_size)(struct vsock_sock *);
-
/* Addressing. */
u32 (*get_local_cid)(void);
};
diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c
index f7540a3ac64e..dee69d7ee148 100644
--- a/net/vmw_vsock/af_vsock.c
+++ b/net/vmw_vsock/af_vsock.c
@@ -126,6 +126,10 @@ static struct proto vsock_proto = {
*/
#define VSOCK_DEFAULT_CONNECT_TIMEOUT (2 * HZ)
+#define VSOCK_DEFAULT_BUFFER_SIZE (1024 * 256)
+#define VSOCK_DEFAULT_BUFFER_MAX_SIZE (1024 * 256)
+#define VSOCK_DEFAULT_BUFFER_MIN_SIZE 128
+
static const struct vsock_transport *transport_single;
static DEFINE_MUTEX(vsock_register_mutex);
@@ -613,10 +617,16 @@ struct sock *__vsock_create(struct net *net,
vsk->trusted = psk->trusted;
vsk->owner = get_cred(psk->owner);
vsk->connect_timeout = psk->connect_timeout;
+ vsk->buffer_size = psk->buffer_size;
+ vsk->buffer_min_size = psk->buffer_min_size;
+ vsk->buffer_max_size = psk->buffer_max_size;
} else {
vsk->trusted = capable(CAP_NET_ADMIN);
vsk->owner = get_current_cred();
vsk->connect_timeout = VSOCK_DEFAULT_CONNECT_TIMEOUT;
+ vsk->buffer_size = VSOCK_DEFAULT_BUFFER_SIZE;
+ vsk->buffer_min_size = VSOCK_DEFAULT_BUFFER_MIN_SIZE;
+ vsk->buffer_max_size = VSOCK_DEFAULT_BUFFER_MAX_SIZE;
}
if (vsk->transport->init(vsk, psk) < 0) {
@@ -1360,6 +1370,23 @@ static int vsock_listen(struct socket *sock, int backlog)
return err;
}
+static void vsock_update_buffer_size(struct vsock_sock *vsk,
+ const struct vsock_transport *transport,
+ u64 val)
+{
+ if (val > vsk->buffer_max_size)
+ val = vsk->buffer_max_size;
+
+ if (val < vsk->buffer_min_size)
+ val = vsk->buffer_min_size;
+
+ if (val != vsk->buffer_size &&
+ transport && transport->notify_buffer_size)
+ transport->notify_buffer_size(vsk, &val);
+
+ vsk->buffer_size = val;
+}
+
static int vsock_stream_setsockopt(struct socket *sock,
int level,
int optname,
@@ -1397,17 +1424,19 @@ static int vsock_stream_setsockopt(struct socket *sock,
switch (optname) {
case SO_VM_SOCKETS_BUFFER_SIZE:
COPY_IN(val);
- transport->set_buffer_size(vsk, val);
+ vsock_update_buffer_size(vsk, transport, val);
break;
case SO_VM_SOCKETS_BUFFER_MAX_SIZE:
COPY_IN(val);
- transport->set_max_buffer_size(vsk, val);
+ vsk->buffer_max_size = val;
+ vsock_update_buffer_size(vsk, transport, vsk->buffer_size);
break;
case SO_VM_SOCKETS_BUFFER_MIN_SIZE:
COPY_IN(val);
- transport->set_min_buffer_size(vsk, val);
+ vsk->buffer_min_size = val;
+ vsock_update_buffer_size(vsk, transport, vsk->buffer_size);
break;
case SO_VM_SOCKETS_CONNECT_TIMEOUT: {
@@ -1448,7 +1477,6 @@ static int vsock_stream_getsockopt(struct socket *sock,
int len;
struct sock *sk;
struct vsock_sock *vsk;
- const struct vsock_transport *transport;
u64 val;
if (level != AF_VSOCK)
@@ -1472,21 +1500,20 @@ static int vsock_stream_getsockopt(struct socket *sock,
err = 0;
sk = sock->sk;
vsk = vsock_sk(sk);
- transport = vsk->transport;
switch (optname) {
case SO_VM_SOCKETS_BUFFER_SIZE:
- val = transport->get_buffer_size(vsk);
+ val = vsk->buffer_size;
COPY_OUT(val);
break;
case SO_VM_SOCKETS_BUFFER_MAX_SIZE:
- val = transport->get_max_buffer_size(vsk);
+ val = vsk->buffer_max_size;
COPY_OUT(val);
break;
case SO_VM_SOCKETS_BUFFER_MIN_SIZE:
- val = transport->get_min_buffer_size(vsk);
+ val = vsk->buffer_min_size;
COPY_OUT(val);
break;
diff --git a/net/vmw_vsock/hyperv_transport.c b/net/vmw_vsock/hyperv_transport.c
index 261521d286d6..4f47af2054dd 100644
--- a/net/vmw_vsock/hyperv_transport.c
+++ b/net/vmw_vsock/hyperv_transport.c
@@ -843,36 +843,6 @@ int hvs_notify_send_post_enqueue(struct vsock_sock *vsk, ssize_t written,
return 0;
}
-static void hvs_set_buffer_size(struct vsock_sock *vsk, u64 val)
-{
- /* Ignored. */
-}
-
-static void hvs_set_min_buffer_size(struct vsock_sock *vsk, u64 val)
-{
- /* Ignored. */
-}
-
-static void hvs_set_max_buffer_size(struct vsock_sock *vsk, u64 val)
-{
- /* Ignored. */
-}
-
-static u64 hvs_get_buffer_size(struct vsock_sock *vsk)
-{
- return -ENOPROTOOPT;
-}
-
-static u64 hvs_get_min_buffer_size(struct vsock_sock *vsk)
-{
- return -ENOPROTOOPT;
-}
-
-static u64 hvs_get_max_buffer_size(struct vsock_sock *vsk)
-{
- return -ENOPROTOOPT;
-}
-
static struct vsock_transport hvs_transport = {
.get_local_cid = hvs_get_local_cid,
@@ -906,12 +876,6 @@ static struct vsock_transport hvs_transport = {
.notify_send_pre_enqueue = hvs_notify_send_pre_enqueue,
.notify_send_post_enqueue = hvs_notify_send_post_enqueue,
- .set_buffer_size = hvs_set_buffer_size,
- .set_min_buffer_size = hvs_set_min_buffer_size,
- .set_max_buffer_size = hvs_set_max_buffer_size,
- .get_buffer_size = hvs_get_buffer_size,
- .get_min_buffer_size = hvs_get_min_buffer_size,
- .get_max_buffer_size = hvs_get_max_buffer_size,
};
static int hvs_probe(struct hv_device *hdev,
diff --git a/net/vmw_vsock/virtio_transport.c b/net/vmw_vsock/virtio_transport.c
index 3756f0857946..fb1fc7760e8c 100644
--- a/net/vmw_vsock/virtio_transport.c
+++ b/net/vmw_vsock/virtio_transport.c
@@ -494,13 +494,7 @@ static struct virtio_transport virtio_transport = {
.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,
-
- .set_buffer_size = virtio_transport_set_buffer_size,
- .set_min_buffer_size = virtio_transport_set_min_buffer_size,
- .set_max_buffer_size = virtio_transport_set_max_buffer_size,
- .get_buffer_size = virtio_transport_get_buffer_size,
- .get_min_buffer_size = virtio_transport_get_min_buffer_size,
- .get_max_buffer_size = virtio_transport_get_max_buffer_size,
+ .notify_buffer_size = virtio_transport_notify_buffer_size,
},
.send_pkt = virtio_transport_send_pkt,
diff --git a/net/vmw_vsock/virtio_transport_common.c b/net/vmw_vsock/virtio_transport_common.c
index fc046c071178..bac9e7430a2e 100644
--- a/net/vmw_vsock/virtio_transport_common.c
+++ b/net/vmw_vsock/virtio_transport_common.c
@@ -403,17 +403,13 @@ int virtio_transport_do_socket_init(struct vsock_sock *vsk,
if (psk) {
struct virtio_vsock_sock *ptrans = psk->trans;
- vvs->buf_size = ptrans->buf_size;
- vvs->buf_size_min = ptrans->buf_size_min;
- vvs->buf_size_max = ptrans->buf_size_max;
vvs->peer_buf_alloc = ptrans->peer_buf_alloc;
- } else {
- vvs->buf_size = VIRTIO_VSOCK_DEFAULT_BUF_SIZE;
- vvs->buf_size_min = VIRTIO_VSOCK_DEFAULT_MIN_BUF_SIZE;
- vvs->buf_size_max = VIRTIO_VSOCK_DEFAULT_MAX_BUF_SIZE;
}
- vvs->buf_alloc = vvs->buf_size;
+ if (vsk->buffer_size > VIRTIO_VSOCK_MAX_BUF_SIZE)
+ vsk->buffer_size = VIRTIO_VSOCK_MAX_BUF_SIZE;
+
+ vvs->buf_alloc = vsk->buffer_size;
spin_lock_init(&vvs->rx_lock);
spin_lock_init(&vvs->tx_lock);
@@ -423,68 +419,18 @@ int virtio_transport_do_socket_init(struct vsock_sock *vsk,
}
EXPORT_SYMBOL_GPL(virtio_transport_do_socket_init);
-u64 virtio_transport_get_buffer_size(struct vsock_sock *vsk)
-{
- struct virtio_vsock_sock *vvs = vsk->trans;
-
- return vvs->buf_size;
-}
-EXPORT_SYMBOL_GPL(virtio_transport_get_buffer_size);
-
-u64 virtio_transport_get_min_buffer_size(struct vsock_sock *vsk)
-{
- struct virtio_vsock_sock *vvs = vsk->trans;
-
- return vvs->buf_size_min;
-}
-EXPORT_SYMBOL_GPL(virtio_transport_get_min_buffer_size);
-
-u64 virtio_transport_get_max_buffer_size(struct vsock_sock *vsk)
-{
- struct virtio_vsock_sock *vvs = vsk->trans;
-
- return vvs->buf_size_max;
-}
-EXPORT_SYMBOL_GPL(virtio_transport_get_max_buffer_size);
-
-void virtio_transport_set_buffer_size(struct vsock_sock *vsk, u64 val)
-{
- struct virtio_vsock_sock *vvs = vsk->trans;
-
- if (val > VIRTIO_VSOCK_MAX_BUF_SIZE)
- val = VIRTIO_VSOCK_MAX_BUF_SIZE;
- if (val < vvs->buf_size_min)
- vvs->buf_size_min = val;
- if (val > vvs->buf_size_max)
- vvs->buf_size_max = val;
- vvs->buf_size = val;
- vvs->buf_alloc = val;
-}
-EXPORT_SYMBOL_GPL(virtio_transport_set_buffer_size);
-
-void virtio_transport_set_min_buffer_size(struct vsock_sock *vsk, u64 val)
+int virtio_transport_notify_buffer_size(struct vsock_sock *vsk, u64 *val)
{
struct virtio_vsock_sock *vvs = vsk->trans;
- if (val > VIRTIO_VSOCK_MAX_BUF_SIZE)
- val = VIRTIO_VSOCK_MAX_BUF_SIZE;
- if (val > vvs->buf_size)
- vvs->buf_size = val;
- vvs->buf_size_min = val;
-}
-EXPORT_SYMBOL_GPL(virtio_transport_set_min_buffer_size);
+ if (*val > VIRTIO_VSOCK_MAX_BUF_SIZE)
+ *val = VIRTIO_VSOCK_MAX_BUF_SIZE;
-void virtio_transport_set_max_buffer_size(struct vsock_sock *vsk, u64 val)
-{
- struct virtio_vsock_sock *vvs = vsk->trans;
+ vvs->buf_alloc = *val;
- if (val > VIRTIO_VSOCK_MAX_BUF_SIZE)
- val = VIRTIO_VSOCK_MAX_BUF_SIZE;
- if (val < vvs->buf_size)
- vvs->buf_size = val;
- vvs->buf_size_max = val;
+ return 0;
}
-EXPORT_SYMBOL_GPL(virtio_transport_set_max_buffer_size);
+EXPORT_SYMBOL_GPL(virtio_transport_notify_buffer_size);
int
virtio_transport_notify_poll_in(struct vsock_sock *vsk,
@@ -576,9 +522,7 @@ EXPORT_SYMBOL_GPL(virtio_transport_notify_send_post_enqueue);
u64 virtio_transport_stream_rcvhiwat(struct vsock_sock *vsk)
{
- struct virtio_vsock_sock *vvs = vsk->trans;
-
- return vvs->buf_size;
+ return vsk->buffer_size;
}
EXPORT_SYMBOL_GPL(virtio_transport_stream_rcvhiwat);
diff --git a/net/vmw_vsock/vmci_transport.c b/net/vmw_vsock/vmci_transport.c
index f8e3131ac480..8290d37b6587 100644
--- a/net/vmw_vsock/vmci_transport.c
+++ b/net/vmw_vsock/vmci_transport.c
@@ -74,10 +74,6 @@ static u32 vmci_transport_qp_resumed_sub_id = VMCI_INVALID_ID;
static int PROTOCOL_OVERRIDE = -1;
-#define VMCI_TRANSPORT_DEFAULT_QP_SIZE_MIN 128
-#define VMCI_TRANSPORT_DEFAULT_QP_SIZE 262144
-#define VMCI_TRANSPORT_DEFAULT_QP_SIZE_MAX 262144
-
/* Helper function to convert from a VMCI error code to a VSock error code. */
static s32 vmci_transport_error_to_vsock_error(s32 vmci_error)
@@ -1025,11 +1021,11 @@ static int vmci_transport_recv_listen(struct sock *sk,
/* If the proposed size fits within our min/max, accept it. Otherwise
* propose our own size.
*/
- if (pkt->u.size >= vmci_trans(vpending)->queue_pair_min_size &&
- pkt->u.size <= vmci_trans(vpending)->queue_pair_max_size) {
+ if (pkt->u.size >= vpending->buffer_min_size &&
+ pkt->u.size <= vpending->buffer_max_size) {
qp_size = pkt->u.size;
} else {
- qp_size = vmci_trans(vpending)->queue_pair_size;
+ qp_size = vpending->buffer_size;
}
/* Figure out if we are using old or new requests based on the
@@ -1098,7 +1094,7 @@ static int vmci_transport_recv_listen(struct sock *sk,
pending->sk_state = TCP_SYN_SENT;
vmci_trans(vpending)->produce_size =
vmci_trans(vpending)->consume_size = qp_size;
- vmci_trans(vpending)->queue_pair_size = qp_size;
+ vpending->buffer_size = qp_size;
vmci_trans(vpending)->notify_ops->process_request(pending);
@@ -1392,8 +1388,8 @@ static int vmci_transport_recv_connecting_client_negotiate(
vsk->ignore_connecting_rst = false;
/* Verify that we're OK with the proposed queue pair size */
- if (pkt->u.size < vmci_trans(vsk)->queue_pair_min_size ||
- pkt->u.size > vmci_trans(vsk)->queue_pair_max_size) {
+ if (pkt->u.size < vsk->buffer_min_size ||
+ pkt->u.size > vsk->buffer_max_size) {
err = -EINVAL;
goto destroy;
}
@@ -1498,8 +1494,7 @@ vmci_transport_recv_connecting_client_invalid(struct sock *sk,
vsk->sent_request = false;
vsk->ignore_connecting_rst = true;
- err = vmci_transport_send_conn_request(
- sk, vmci_trans(vsk)->queue_pair_size);
+ err = vmci_transport_send_conn_request(sk, vsk->buffer_size);
if (err < 0)
err = vmci_transport_error_to_vsock_error(err);
else
@@ -1583,21 +1578,6 @@ static int vmci_transport_socket_init(struct vsock_sock *vsk,
INIT_LIST_HEAD(&vmci_trans(vsk)->elem);
vmci_trans(vsk)->sk = &vsk->sk;
spin_lock_init(&vmci_trans(vsk)->lock);
- if (psk) {
- vmci_trans(vsk)->queue_pair_size =
- vmci_trans(psk)->queue_pair_size;
- vmci_trans(vsk)->queue_pair_min_size =
- vmci_trans(psk)->queue_pair_min_size;
- vmci_trans(vsk)->queue_pair_max_size =
- vmci_trans(psk)->queue_pair_max_size;
- } else {
- vmci_trans(vsk)->queue_pair_size =
- VMCI_TRANSPORT_DEFAULT_QP_SIZE;
- vmci_trans(vsk)->queue_pair_min_size =
- VMCI_TRANSPORT_DEFAULT_QP_SIZE_MIN;
- vmci_trans(vsk)->queue_pair_max_size =
- VMCI_TRANSPORT_DEFAULT_QP_SIZE_MAX;
- }
return 0;
}
@@ -1813,8 +1793,7 @@ static int vmci_transport_connect(struct vsock_sock *vsk)
if (vmci_transport_old_proto_override(&old_pkt_proto) &&
old_pkt_proto) {
- err = vmci_transport_send_conn_request(
- sk, vmci_trans(vsk)->queue_pair_size);
+ err = vmci_transport_send_conn_request(sk, vsk->buffer_size);
if (err < 0) {
sk->sk_state = TCP_CLOSE;
return err;
@@ -1822,8 +1801,7 @@ static int vmci_transport_connect(struct vsock_sock *vsk)
} else {
int supported_proto_versions =
vmci_transport_new_proto_supported_versions();
- err = vmci_transport_send_conn_request2(
- sk, vmci_trans(vsk)->queue_pair_size,
+ err = vmci_transport_send_conn_request2(sk, vsk->buffer_size,
supported_proto_versions);
if (err < 0) {
sk->sk_state = TCP_CLOSE;
@@ -1876,46 +1854,6 @@ static bool vmci_transport_stream_is_active(struct vsock_sock *vsk)
return !vmci_handle_is_invalid(vmci_trans(vsk)->qp_handle);
}
-static u64 vmci_transport_get_buffer_size(struct vsock_sock *vsk)
-{
- return vmci_trans(vsk)->queue_pair_size;
-}
-
-static u64 vmci_transport_get_min_buffer_size(struct vsock_sock *vsk)
-{
- return vmci_trans(vsk)->queue_pair_min_size;
-}
-
-static u64 vmci_transport_get_max_buffer_size(struct vsock_sock *vsk)
-{
- return vmci_trans(vsk)->queue_pair_max_size;
-}
-
-static void vmci_transport_set_buffer_size(struct vsock_sock *vsk, u64 val)
-{
- if (val < vmci_trans(vsk)->queue_pair_min_size)
- vmci_trans(vsk)->queue_pair_min_size = val;
- if (val > vmci_trans(vsk)->queue_pair_max_size)
- vmci_trans(vsk)->queue_pair_max_size = val;
- vmci_trans(vsk)->queue_pair_size = val;
-}
-
-static void vmci_transport_set_min_buffer_size(struct vsock_sock *vsk,
- u64 val)
-{
- if (val > vmci_trans(vsk)->queue_pair_size)
- vmci_trans(vsk)->queue_pair_size = val;
- vmci_trans(vsk)->queue_pair_min_size = val;
-}
-
-static void vmci_transport_set_max_buffer_size(struct vsock_sock *vsk,
- u64 val)
-{
- if (val < vmci_trans(vsk)->queue_pair_size)
- vmci_trans(vsk)->queue_pair_size = val;
- vmci_trans(vsk)->queue_pair_max_size = val;
-}
-
static int vmci_transport_notify_poll_in(
struct vsock_sock *vsk,
size_t target,
@@ -2098,12 +2036,6 @@ static const struct vsock_transport vmci_transport = {
.notify_send_pre_enqueue = vmci_transport_notify_send_pre_enqueue,
.notify_send_post_enqueue = vmci_transport_notify_send_post_enqueue,
.shutdown = vmci_transport_shutdown,
- .set_buffer_size = vmci_transport_set_buffer_size,
- .set_min_buffer_size = vmci_transport_set_min_buffer_size,
- .set_max_buffer_size = vmci_transport_set_max_buffer_size,
- .get_buffer_size = vmci_transport_get_buffer_size,
- .get_min_buffer_size = vmci_transport_get_min_buffer_size,
- .get_max_buffer_size = vmci_transport_get_max_buffer_size,
.get_local_cid = vmci_transport_get_local_cid,
};
diff --git a/net/vmw_vsock/vmci_transport.h b/net/vmw_vsock/vmci_transport.h
index 1ca1e8640b31..b7b072194282 100644
--- a/net/vmw_vsock/vmci_transport.h
+++ b/net/vmw_vsock/vmci_transport.h
@@ -108,9 +108,6 @@ struct vmci_transport {
struct vmci_qp *qpair;
u64 produce_size;
u64 consume_size;
- u64 queue_pair_size;
- u64 queue_pair_min_size;
- u64 queue_pair_max_size;
u32 detach_sub_id;
union vmci_transport_notify notify;
const struct vmci_transport_notify_ops *notify_ops;
--
2.21.0
> From: Stefano Garzarella <[email protected]>
> Sent: Friday, September 27, 2019 4:27 AM
> To: [email protected]
> Cc: [email protected]; KY Srinivasan <[email protected]>; Stefan
> Hajnoczi <[email protected]>; Sasha Levin <[email protected]>;
> [email protected]; [email protected]; David S. Miller
> <[email protected]>; [email protected]; Stephen
> Hemminger <[email protected]>; Jason Wang
> <[email protected]>; Michael S. Tsirkin <[email protected]>; Haiyang
> Zhang <[email protected]>; Dexuan Cui <[email protected]>;
> Jorgen Hansen <[email protected]>
> Subject: [RFC PATCH 09/13] hv_sock: set VMADDR_CID_HOST in the
> hvs_remote_addr_init()
>
> Remote peer is always the host, so we set VMADDR_CID_HOST as
> remote CID instead of VMADDR_CID_ANY.
>
> Signed-off-by: Stefano Garzarella <[email protected]>
> ---
> net/vmw_vsock/hyperv_transport.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/net/vmw_vsock/hyperv_transport.c
> b/net/vmw_vsock/hyperv_transport.c
> index 4f47af2054dd..306310794522 100644
> --- a/net/vmw_vsock/hyperv_transport.c
> +++ b/net/vmw_vsock/hyperv_transport.c
> @@ -186,7 +186,8 @@ static void hvs_remote_addr_init(struct sockaddr_vm
> *remote,
> static u32 host_ephemeral_port = MIN_HOST_EPHEMERAL_PORT;
> struct sock *sk;
>
> - vsock_addr_init(remote, VMADDR_CID_ANY, VMADDR_PORT_ANY);
> + /* Remote peer is always the host */
> + vsock_addr_init(remote, VMADDR_CID_HOST, VMADDR_PORT_ANY);
>
> while (1) {
> /* Wrap around ? */
> --
Looks good to me, since hv_sock doesn't really use the CID in the
transport layer.
Reviewed-by: Dexuan Cui <[email protected]>
> From: Stefano Garzarella <[email protected]>
> Sent: Friday, September 27, 2019 4:27 AM
> To: [email protected]
>
> virtio_transport and vmci_transport handle the buffer_size
> sockopts in a very similar way.
>
> In order to support multiple transports, this patch moves this
> handling in the core to allow the user to change the options
> also if the socket is not yet assigned to any transport.
>
> This patch also adds the '.notify_buffer_size' callback in the
> 'struct virtio_transport' in order to inform the transport,
> when the buffer_size is changed by the user. It is also useful
> to limit the 'buffer_size' requested (e.g. virtio transports).
>
> Signed-off-by: Stefano Garzarella <[email protected]>
> ---
> drivers/vhost/vsock.c | 7 +-
> include/linux/virtio_vsock.h | 15 +----
> include/net/af_vsock.h | 14 ++--
> net/vmw_vsock/af_vsock.c | 43 ++++++++++---
> net/vmw_vsock/hyperv_transport.c | 36 -----------
> net/vmw_vsock/virtio_transport.c | 8 +--
> net/vmw_vsock/virtio_transport_common.c | 78 ++++------------------
> net/vmw_vsock/vmci_transport.c | 86 +++----------------------
> net/vmw_vsock/vmci_transport.h | 3 -
> 9 files changed, 64 insertions(+), 226 deletions(-)
>
The hv_sock part (hyperv_transport.c) looks good to me.
Acked-by: Dexuan Cui <[email protected]>
> From: Stefano Garzarella <[email protected]>
> Sent: Friday, September 27, 2019 4:27 AM
>
> vsock_insert_unbound() was called only when 'sock' parameter of
> __vsock_create() was not null. This only happened when
> __vsock_create() was called by vsock_create().
>
> In order to simplify the multi-transports support, this patch
> moves vsock_insert_unbound() at the end of vsock_create().
>
> Signed-off-by: Stefano Garzarella <[email protected]>
> ---
> net/vmw_vsock/af_vsock.c | 13 +++++++++----
> 1 file changed, 9 insertions(+), 4 deletions(-)
>
Reviewed-by: Dexuan Cui <[email protected]>
> From: Stefano Garzarella <[email protected]>
> Sent: Friday, September 27, 2019 4:27 AM
> ...
> Patch 9 changes the hvs_remote_addr_init(). setting the
> VMADDR_CID_HOST as remote CID instead of VMADDR_CID_ANY to make
> the choice of transport to be used work properly.
> @Dexuan Could this change break anything?
This patch looks good to me.
> @Dexuan please can you test on HyperV that I didn't break anything
> even without nested VMs?
I did some quick tests with the 13 patches in a Linux VM (this is not
a nested VM) on Hyper-V and it looks nothing is broken. :-)
> I'll try to setup a Windows host where to test the nested VMs
I suppose you're going to run a Linux VM on a Hyper-V host,
and the Linux VM itself runs KVM/VmWare so it can create its own child
VMs. IMO this is similar to the test "nested KVM ( ..., virtio-transport[L1,L2]"
you have done.
.
Thanks!
Dexuan
On Fri, Oct 04, 2019 at 12:04:46AM +0000, Dexuan Cui wrote:
> > From: Stefano Garzarella <[email protected]>
> > Sent: Friday, September 27, 2019 4:27 AM
> > ...
> > Patch 9 changes the hvs_remote_addr_init(). setting the
> > VMADDR_CID_HOST as remote CID instead of VMADDR_CID_ANY to make
> > the choice of transport to be used work properly.
> > @Dexuan Could this change break anything?
>
> This patch looks good to me.
>
Thank you very much for your reviews!
> > @Dexuan please can you test on HyperV that I didn't break anything
> > even without nested VMs?
>
> I did some quick tests with the 13 patches in a Linux VM (this is not
> a nested VM) on Hyper-V and it looks nothing is broken. :-)
>
Great :-)
> > I'll try to setup a Windows host where to test the nested VMs
>
> I suppose you're going to run a Linux VM on a Hyper-V host,
> and the Linux VM itself runs KVM/VmWare so it can create its own child
> VMs. IMO this is similar to the test "nested KVM ( ..., virtio-transport[L1,L2]"
> you have done.
Yes, I think so. If the Hyper-V transport works well without nested VM,
it should work the same with a nested KVM/VMware.
Thanks,
Stefano
On Fri, Sep 27, 2019 at 01:27:01PM +0200, Stefano Garzarella wrote:
> VMCI transport provides both g2h and h2g behaviors in a single
> transport.
> We are able to set (or not) the g2h behavior, detecting if we
> are in a VMware guest (or not), but the h2g feature is always set.
> This prevents to load other h2g transports while we are in a
> VMware guest.
>
> This patch adds a new 'transport_hg' to handle this case, reducing
> the priority of transports that provide both g2h and h2g
> behaviors. A transport that has g2h and h2g features, can be
> bypassed by a transport that has only the h2g feature.
>
Since I'm enabling the VSOCK_TRANSPORT_F_G2H in the vmci_transport only
when we run in a VMware guest, this patch doesn't work well if a KVM (or
HyperV) guest application create an AF_VSOCK socket and no transports are
loaded, because in this case the vmci_transport is loaded
(MODULE_ALIAS_NETPROTO(PF_VSOCK)) and it is registered as transport_h2g.
At this point, if we want to run a nested VM using vhost_transport, we
can't load it.
So, I can leave VSOCK_TRANSPORT_F_G2H always set in the vmci_transport
and this should fix this issue.
Or maybe I need to change how the registering works, e.g. handling a list
of transport registered, setting priority or using the last registered
transport.
Any suggestion?
Thanks,
Stefano
On Fri, Sep 27, 2019 at 01:26:51PM +0200, Stefano Garzarella wrote:
> The VSOCK_DEFAULT_CONNECT_TIMEOUT definition was introduced with
> commit d021c344051af ("VSOCK: Introduce VM Sockets"), but it is
> never used in the net/vmw_vsock/vmci_transport.c.
>
> VSOCK_DEFAULT_CONNECT_TIMEOUT is used and defined in
> net/vmw_vsock/af_vsock.c
>
> Signed-off-by: Stefano Garzarella <[email protected]>
> ---
> net/vmw_vsock/vmci_transport.c | 5 -----
> 1 file changed, 5 deletions(-)
Reviewed-by: Stefan Hajnoczi <[email protected]>
On Fri, Sep 27, 2019 at 01:26:52PM +0200, Stefano Garzarella wrote:
> vm_sockets_get_local_cid() is only used in virtio_transport_common.c.
> We can replace it calling the virtio_transport_get_ops() and
> using the get_local_cid() callback registered by the transport.
>
> Signed-off-by: Stefano Garzarella <[email protected]>
> ---
> include/linux/vm_sockets.h | 2 --
> net/vmw_vsock/af_vsock.c | 10 ----------
> net/vmw_vsock/virtio_transport_common.c | 2 +-
> 3 files changed, 1 insertion(+), 13 deletions(-)
Reviewed-by: Stefan Hajnoczi <[email protected]>
On Fri, Sep 27, 2019 at 01:26:53PM +0200, Stefano Garzarella wrote:
> This header file now only includes the "uapi/linux/vm_sockets.h".
> We can include directly it when needed.
>
> Signed-off-by: Stefano Garzarella <[email protected]>
> ---
> include/linux/vm_sockets.h | 13 -------------
> include/net/af_vsock.h | 2 +-
> include/net/vsock_addr.h | 2 +-
> net/vmw_vsock/vmci_transport_notify.h | 1 -
> 4 files changed, 2 insertions(+), 16 deletions(-)
> delete mode 100644 include/linux/vm_sockets.h
Reviewed-by: Stefan Hajnoczi <[email protected]>
On Fri, Sep 27, 2019 at 01:26:54PM +0200, Stefano Garzarella wrote:
> As a preparation to support multiple transports, this patch adds
> the 'transport' member at the 'struct vsock_sock'.
> This new field is initialized during the creation in the
> __vsock_create() function.
>
> This patch also renames the global 'transport' pointer to
> 'transport_single', since for now we're only supporting a single
> transport registered at run-time.
>
> Signed-off-by: Stefano Garzarella <[email protected]>
> ---
> include/net/af_vsock.h | 1 +
> net/vmw_vsock/af_vsock.c | 56 +++++++++++++++++++++++++++-------------
> 2 files changed, 39 insertions(+), 18 deletions(-)
Reviewed-by: Stefan Hajnoczi <[email protected]>
On Fri, Sep 27, 2019 at 01:26:56PM +0200, Stefano Garzarella wrote:
> -const struct vsock_transport *vsock_core_get_transport(void)
> +const struct vsock_transport *vsock_core_get_transport(struct vsock_sock *vsk)
> {
> /* vsock_register_mutex not taken since only the transport uses this
> * function and only while registered.
> */
> - return transport_single;
This comment is about protecting transport_single. It no longer applies
when using vsk->transport. Please drop it.
Otherwise:
Reviewed-by: Stefan Hajnoczi <[email protected]>
On Fri, Sep 27, 2019 at 01:26:55PM +0200, Stefano Garzarella wrote:
> We are going to add 'struct vsock_sock *' parameter to
> virtio_transport_get_ops().
>
> In some cases, like in the virtio_transport_reset_no_sock(),
> we don't have any socket assigned to the packet received,
> so we can't use the virtio_transport_get_ops().
>
> In order to allow virtio_transport_reset_no_sock() to use the
> '.send_pkt' callback from the 'vhost_transport' or 'virtio_transport',
> we add the 'struct virtio_transport *' to it and to its caller:
> virtio_transport_recv_pkt().
>
> We moved the 'vhost_transport' and 'virtio_transport' definition,
> to pass their address to the virtio_transport_recv_pkt().
>
> Signed-off-by: Stefano Garzarella <[email protected]>
> ---
> drivers/vhost/vsock.c | 94 +++++++-------
> include/linux/virtio_vsock.h | 3 +-
> net/vmw_vsock/virtio_transport.c | 160 ++++++++++++------------
> net/vmw_vsock/virtio_transport_common.c | 12 +-
> 4 files changed, 135 insertions(+), 134 deletions(-)
Reviewed-by: Stefan Hajnoczi <[email protected]>
On Fri, Sep 27, 2019 at 01:26:57PM +0200, Stefano Garzarella wrote:
> @@ -140,18 +145,11 @@ struct vsock_transport {
> struct vsock_transport_send_notify_data *);
> int (*notify_send_post_enqueue)(struct vsock_sock *, ssize_t,
> struct vsock_transport_send_notify_data *);
> + int (*notify_buffer_size)(struct vsock_sock *, u64 *);
Is ->notify_buffer_size() called under lock_sock(sk)? If yes, please
document it.
> +static void vsock_update_buffer_size(struct vsock_sock *vsk,
> + const struct vsock_transport *transport,
> + u64 val)
> +{
> + if (val > vsk->buffer_max_size)
> + val = vsk->buffer_max_size;
> +
> + if (val < vsk->buffer_min_size)
> + val = vsk->buffer_min_size;
> +
> + if (val != vsk->buffer_size &&
> + transport && transport->notify_buffer_size)
> + transport->notify_buffer_size(vsk, &val);
Why does this function return an int if we don't check the return value?
> diff --git a/net/vmw_vsock/virtio_transport_common.c b/net/vmw_vsock/virtio_transport_common.c
> index fc046c071178..bac9e7430a2e 100644
> --- a/net/vmw_vsock/virtio_transport_common.c
> +++ b/net/vmw_vsock/virtio_transport_common.c
> @@ -403,17 +403,13 @@ int virtio_transport_do_socket_init(struct vsock_sock *vsk,
> if (psk) {
> struct virtio_vsock_sock *ptrans = psk->trans;
>
> - vvs->buf_size = ptrans->buf_size;
> - vvs->buf_size_min = ptrans->buf_size_min;
> - vvs->buf_size_max = ptrans->buf_size_max;
> vvs->peer_buf_alloc = ptrans->peer_buf_alloc;
> - } else {
> - vvs->buf_size = VIRTIO_VSOCK_DEFAULT_BUF_SIZE;
> - vvs->buf_size_min = VIRTIO_VSOCK_DEFAULT_MIN_BUF_SIZE;
> - vvs->buf_size_max = VIRTIO_VSOCK_DEFAULT_MAX_BUF_SIZE;
> }
>
> - vvs->buf_alloc = vvs->buf_size;
> + if (vsk->buffer_size > VIRTIO_VSOCK_MAX_BUF_SIZE)
> + vsk->buffer_size = VIRTIO_VSOCK_MAX_BUF_SIZE;
Hmm...this could be outside the [min, max] range. I'm not sure how much
it matters.
Another issue is that this patch drops the VIRTIO_VSOCK_MAX_BUF_SIZE
limit that used to be enforced by virtio_transport_set_buffer_size().
Now the limit is only applied at socket init time. If the buffer size
is changed later then VIRTIO_VSOCK_MAX_BUF_SIZE can be exceeded. If
that doesn't matter, why even bother with VIRTIO_VSOCK_MAX_BUF_SIZE
here?
On Fri, Sep 27, 2019 at 01:26:58PM +0200, Stefano Garzarella wrote:
> vsock_insert_unbound() was called only when 'sock' parameter of
> __vsock_create() was not null. This only happened when
> __vsock_create() was called by vsock_create().
>
> In order to simplify the multi-transports support, this patch
> moves vsock_insert_unbound() at the end of vsock_create().
>
> Signed-off-by: Stefano Garzarella <[email protected]>
> ---
> net/vmw_vsock/af_vsock.c | 13 +++++++++----
> 1 file changed, 9 insertions(+), 4 deletions(-)
Maybe transports shouldn't call __vsock_create() directly. They always
pass NULL as the parent socket, so we could have a more specific
function that transports call without a parent sock argument. This
would eliminate any concern over moving vsock_insert_unbound() out of
this function. In any case, I've checked the code and this patch is
correct.
Reviewed-by: Stefan Hajnoczi <[email protected]>
On Fri, Sep 27, 2019 at 01:26:59PM +0200, Stefano Garzarella wrote:
> Remote peer is always the host, so we set VMADDR_CID_HOST as
> remote CID instead of VMADDR_CID_ANY.
>
> Signed-off-by: Stefano Garzarella <[email protected]>
> ---
> net/vmw_vsock/hyperv_transport.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
Reviewed-by: Stefan Hajnoczi <[email protected]>
On Fri, Sep 27, 2019 at 01:27:00PM +0200, Stefano Garzarella wrote:
> RFC:
> - I'd like to move MODULE_ALIAS_NETPROTO(PF_VSOCK) to af_vsock.c.
> @Jorgen could this break the VMware products?
What will cause the vmw_vsock_vmci_transport.ko module to be loaded
after you remove MODULE_ALIAS_NETPROTO(PF_VSOCK)? Perhaps
drivers/misc/vmw_vmci/vmci_guest.c:vmci_guest_probe_device() could do
something when the guest driver loads. There would need to be something
equivalent for the host side too.
This will solve another issue too. Today the VMCI transport can be
loaded if an application creates an AF_VSOCK socket during early boot
before the virtio transport has been probed. This happens because the
VMCI transport uses MODULE_ALIAS_NETPROTO(PF_VSOCK) *and* it does not
probe whether this system is actually a VMware guest.
If we instead load the core af_vsock.ko module and transports are only
loaded based on hardware feature probing (e.g. the presence of VMware
guest mode, a virtio PCI adapter, etc) then transports will be
well-behaved.
> - DGRAM sockets are handled as before, I don't know if make sense work
> on it now, or when another transport will support DGRAM. The big
> issues here is that we cannot link 1-1 a socket to transport as
> for stream sockets since DGRAM is not connection-oriented.
Let's ignore DGRAM for now since only VMCI supports it and we therefore
do not require multi-transport support.
> diff --git a/include/net/af_vsock.h b/include/net/af_vsock.h
> index 86f8f463e01a..2a081d19e20d 100644
> --- a/include/net/af_vsock.h
> +++ b/include/net/af_vsock.h
> @@ -94,7 +94,13 @@ struct vsock_transport_send_notify_data {
> u64 data2; /* Transport-defined. */
> };
>
> +#define VSOCK_TRANSPORT_F_H2G 0x00000001
> +#define VSOCK_TRANSPORT_F_G2H 0x00000002
> +#define VSOCK_TRANSPORT_F_DGRAM 0x00000004
Documentation comments, please.
> +void vsock_core_unregister(const struct vsock_transport *t)
> +{
> + mutex_lock(&vsock_register_mutex);
> +
> + /* RFC-TODO: maybe we should check if there are open sockets
> + * assigned to that transport and avoid the unregistration
> + */
If unregister() is only called from module_exit() functions then holding
a reference to the transport module would be enough to prevent this
case. The transport could only be removed once all sockets have been
destroyed (and dropped their transport module reference).
On Fri, Sep 27, 2019 at 01:27:01PM +0200, Stefano Garzarella wrote:
> VMCI transport provides both g2h and h2g behaviors in a single
> transport.
> We are able to set (or not) the g2h behavior, detecting if we
> are in a VMware guest (or not), but the h2g feature is always set.
> This prevents to load other h2g transports while we are in a
> VMware guest.
In the vhost_vsock.ko case we only register the h2g transport when
userspace has loaded the module (by opening /dev/vhost-vsock).
VMCI has something kind of similar: /dev/vmci and the
vmci_host_active_users counter. Maybe we can use this instead of
introducing the transport_hg concept?
On Fri, Sep 27, 2019 at 01:27:02PM +0200, Stefano Garzarella wrote:
> diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c
> index c5f46b8242ce..750b62711b01 100644
> --- a/net/vmw_vsock/af_vsock.c
> +++ b/net/vmw_vsock/af_vsock.c
> @@ -416,13 +416,28 @@ int vsock_assign_transport(struct vsock_sock *vsk, struct vsock_sock *psk)
> return -ESOCKTNOSUPPORT;
> }
>
> - if (!vsk->transport)
> + /* We increase the module refcnt to prevent the tranport unloading
s/tranport/transport/
Otherwise:
Reviewed-by: Stefan Hajnoczi <[email protected]>
On Fri, Sep 27, 2019 at 01:27:03PM +0200, Stefano Garzarella wrote:
> When we are looking for a socket bound to a specific address,
> we also have to take into account the CID.
>
> This patch is useful with multi-transports support because it
> allows the binding of the same port with different CID, and
> it prevents a connection to a wrong socket bound to the same
> port, but with different CID.
>
> Signed-off-by: Stefano Garzarella <[email protected]>
> ---
> net/vmw_vsock/af_vsock.c | 10 ++++++++--
> 1 file changed, 8 insertions(+), 2 deletions(-)
Reviewed-by: Stefan Hajnoczi <[email protected]>
On Fri, Sep 27, 2019 at 01:26:50PM +0200, Stefano Garzarella wrote:
> Hi all,
> this series adds the multi-transports support to vsock, following
> this proposal:
> https://www.spinics.net/lists/netdev/msg575792.html
Nice series! I have left a few comments but overall it looks promising.
Stefan
On Wed, Oct 09, 2019 at 12:54:33PM +0100, Stefan Hajnoczi wrote:
> On Fri, Sep 27, 2019 at 01:26:56PM +0200, Stefano Garzarella wrote:
> > -const struct vsock_transport *vsock_core_get_transport(void)
> > +const struct vsock_transport *vsock_core_get_transport(struct vsock_sock *vsk)
> > {
> > /* vsock_register_mutex not taken since only the transport uses this
> > * function and only while registered.
> > */
> > - return transport_single;
>
> This comment is about protecting transport_single. It no longer applies
> when using vsk->transport. Please drop it.
Right, dropped.
>
> Otherwise:
>
> Reviewed-by: Stefan Hajnoczi <[email protected]>
Thanks,
Stefano
On Wed, Oct 09, 2019 at 02:29:52PM +0100, Stefan Hajnoczi wrote:
> On Fri, Sep 27, 2019 at 01:26:50PM +0200, Stefano Garzarella wrote:
> > Hi all,
> > this series adds the multi-transports support to vsock, following
> > this proposal:
> > https://www.spinics.net/lists/netdev/msg575792.html
>
> Nice series! I have left a few comments but overall it looks promising.
Thank you very much for the comments!
I'll follow them and respin.
Cheers,
Stefano
On Wed, Oct 09, 2019 at 01:30:26PM +0100, Stefan Hajnoczi wrote:
> On Fri, Sep 27, 2019 at 01:26:57PM +0200, Stefano Garzarella wrote:
> > @@ -140,18 +145,11 @@ struct vsock_transport {
> > struct vsock_transport_send_notify_data *);
> > int (*notify_send_post_enqueue)(struct vsock_sock *, ssize_t,
> > struct vsock_transport_send_notify_data *);
> > + int (*notify_buffer_size)(struct vsock_sock *, u64 *);
>
> Is ->notify_buffer_size() called under lock_sock(sk)? If yes, please
> document it.
Yes, it is. I'll document it!
>
> > +static void vsock_update_buffer_size(struct vsock_sock *vsk,
> > + const struct vsock_transport *transport,
> > + u64 val)
> > +{
> > + if (val > vsk->buffer_max_size)
> > + val = vsk->buffer_max_size;
> > +
> > + if (val < vsk->buffer_min_size)
> > + val = vsk->buffer_min_size;
> > +
> > + if (val != vsk->buffer_size &&
> > + transport && transport->notify_buffer_size)
> > + transport->notify_buffer_size(vsk, &val);
>
> Why does this function return an int if we don't check the return value?
>
Copy and past :-(
I'll fix it returning void since I don't think it can fail.
> > diff --git a/net/vmw_vsock/virtio_transport_common.c b/net/vmw_vsock/virtio_transport_common.c
> > index fc046c071178..bac9e7430a2e 100644
> > --- a/net/vmw_vsock/virtio_transport_common.c
> > +++ b/net/vmw_vsock/virtio_transport_common.c
> > @@ -403,17 +403,13 @@ int virtio_transport_do_socket_init(struct vsock_sock *vsk,
> > if (psk) {
> > struct virtio_vsock_sock *ptrans = psk->trans;
> >
> > - vvs->buf_size = ptrans->buf_size;
> > - vvs->buf_size_min = ptrans->buf_size_min;
> > - vvs->buf_size_max = ptrans->buf_size_max;
> > vvs->peer_buf_alloc = ptrans->peer_buf_alloc;
> > - } else {
> > - vvs->buf_size = VIRTIO_VSOCK_DEFAULT_BUF_SIZE;
> > - vvs->buf_size_min = VIRTIO_VSOCK_DEFAULT_MIN_BUF_SIZE;
> > - vvs->buf_size_max = VIRTIO_VSOCK_DEFAULT_MAX_BUF_SIZE;
> > }
> >
> > - vvs->buf_alloc = vvs->buf_size;
> > + if (vsk->buffer_size > VIRTIO_VSOCK_MAX_BUF_SIZE)
> > + vsk->buffer_size = VIRTIO_VSOCK_MAX_BUF_SIZE;
>
> Hmm...this could be outside the [min, max] range. I'm not sure how much
> it matters.
The core guarantees that vsk->buffer_size is <= of the max, so since we are
lowering it, the max should be respected. For the min you are right,
but I think this limit is stricter than the min set by the user.
>
> Another issue is that this patch drops the VIRTIO_VSOCK_MAX_BUF_SIZE
> limit that used to be enforced by virtio_transport_set_buffer_size().
> Now the limit is only applied at socket init time. If the buffer size
> is changed later then VIRTIO_VSOCK_MAX_BUF_SIZE can be exceeded. If
> that doesn't matter, why even bother with VIRTIO_VSOCK_MAX_BUF_SIZE
> here?
>
The .notify_buffer_size() should avoid this issue, since it allows the
transport to limit the buffer size requested after the initialization.
But again the min set by the user can not be respected and in the
previous implementation we forced it to VIRTIO_VSOCK_MAX_BUF_SIZE.
Now we don't limit the min, but we guarantee only that vsk->buffer_size
is lower than VIRTIO_VSOCK_MAX_BUF_SIZE.
Can that be an acceptable compromise?
Thanks,
Stefano
On Wed, Oct 09, 2019 at 01:34:23PM +0100, Stefan Hajnoczi wrote:
> On Fri, Sep 27, 2019 at 01:26:58PM +0200, Stefano Garzarella wrote:
> > vsock_insert_unbound() was called only when 'sock' parameter of
> > __vsock_create() was not null. This only happened when
> > __vsock_create() was called by vsock_create().
> >
> > In order to simplify the multi-transports support, this patch
> > moves vsock_insert_unbound() at the end of vsock_create().
> >
> > Signed-off-by: Stefano Garzarella <[email protected]>
> > ---
> > net/vmw_vsock/af_vsock.c | 13 +++++++++----
> > 1 file changed, 9 insertions(+), 4 deletions(-)
>
> Maybe transports shouldn't call __vsock_create() directly. They always
> pass NULL as the parent socket, so we could have a more specific
> function that transports call without a parent sock argument. This
> would eliminate any concern over moving vsock_insert_unbound() out of
> this function. In any case, I've checked the code and this patch is
> correct.
Yes, I agree with you, I can add a new patch to do this cleaning.
>
> Reviewed-by: Stefan Hajnoczi <[email protected]>
Thanks,
Stefano
On Wed, Oct 09, 2019 at 02:11:23PM +0100, Stefan Hajnoczi wrote:
> On Fri, Sep 27, 2019 at 01:27:00PM +0200, Stefano Garzarella wrote:
> > RFC:
> > - I'd like to move MODULE_ALIAS_NETPROTO(PF_VSOCK) to af_vsock.c.
> > @Jorgen could this break the VMware products?
>
> What will cause the vmw_vsock_vmci_transport.ko module to be loaded
> after you remove MODULE_ALIAS_NETPROTO(PF_VSOCK)? Perhaps
> drivers/misc/vmw_vmci/vmci_guest.c:vmci_guest_probe_device() could do
> something when the guest driver loads.
Good idea, maybe we can call some function provided by vmci_transport
to register it as a guest (I'll remove the type from the transport
and I add it as a parameter of vsock_core_register())
> There would need to be something
> equivalent for the host side too.
Maybe in the vmci_host_do_init_context().
>
> This will solve another issue too. Today the VMCI transport can be
> loaded if an application creates an AF_VSOCK socket during early boot
> before the virtio transport has been probed. This happens because the
> VMCI transport uses MODULE_ALIAS_NETPROTO(PF_VSOCK) *and* it does not
> probe whether this system is actually a VMware guest.
>
> If we instead load the core af_vsock.ko module and transports are only
> loaded based on hardware feature probing (e.g. the presence of VMware
> guest mode, a virtio PCI adapter, etc) then transports will be
> well-behaved.
Yes, I completely agree with you. I'll try to follow your suggestion,
>
> > - DGRAM sockets are handled as before, I don't know if make sense work
> > on it now, or when another transport will support DGRAM. The big
> > issues here is that we cannot link 1-1 a socket to transport as
> > for stream sockets since DGRAM is not connection-oriented.
>
> Let's ignore DGRAM for now since only VMCI supports it and we therefore
> do not require multi-transpor) support.
Okay :)
>
> > diff --git a/include/net/af_vsock.h b/include/net/af_vsock.h
> > index 86f8f463e01a..2a081d19e20d 100644
> > --- a/include/net/af_vsock.h
> > +++ b/include/net/af_vsock.h
> > @@ -94,7 +94,13 @@ struct vsock_transport_send_notify_data {
> > u64 data2; /* Transport-defined. */
> > };
> >
> > +#define VSOCK_TRANSPORT_F_H2G 0x00000001
> > +#define VSOCK_TRANSPORT_F_G2H 0x00000002
> > +#define VSOCK_TRANSPORT_F_DGRAM 0x00000004
>
> Documentation comments, please.
I'll fix!
>
> > +void vsock_core_unregister(const struct vsock_transport *t)
> > +{
> > + mutex_lock(&vsock_register_mutex);
> > +
> > + /* RFC-TODO: maybe we should check if there are open sockets
> > + * assigned to that transport and avoid the unregistration
> > + */
>
> If unregister() is only called from module_exit() functions then holding
> a reference to the transport module would be enough to prevent this
> case. The transport could only be removed once all sockets have been
> destroyed (and dropped their transport module reference).
Yes. I did this in
"[RFC PATCH 12/13] vsock: prevent transport modules unloading".
Maybe I can merge it in this patch...
Thanks,
Stefano
On Wed, Oct 09, 2019 at 02:16:43PM +0100, Stefan Hajnoczi wrote:
> On Fri, Sep 27, 2019 at 01:27:01PM +0200, Stefano Garzarella wrote:
> > VMCI transport provides both g2h and h2g behaviors in a single
> > transport.
> > We are able to set (or not) the g2h behavior, detecting if we
> > are in a VMware guest (or not), but the h2g feature is always set.
> > This prevents to load other h2g transports while we are in a
> > VMware guest.
>
> In the vhost_vsock.ko case we only register the h2g transport when
> userspace has loaded the module (by opening /dev/vhost-vsock).
>
> VMCI has something kind of similar: /dev/vmci and the
> vmci_host_active_users counter. Maybe we can use this instead of
> introducing the transport_hg concept?
Yes, maybe we can register the host in the vmci_host_do_init_context().
I also don't like a lot the transport_hg concept, so I'll try to found
an alternative.
Thanks,
Stefano
On Thu, Oct 10, 2019 at 11:32:54AM +0200, Stefano Garzarella wrote:
> On Wed, Oct 09, 2019 at 01:30:26PM +0100, Stefan Hajnoczi wrote:
> > On Fri, Sep 27, 2019 at 01:26:57PM +0200, Stefano Garzarella wrote:
> > Another issue is that this patch drops the VIRTIO_VSOCK_MAX_BUF_SIZE
> > limit that used to be enforced by virtio_transport_set_buffer_size().
> > Now the limit is only applied at socket init time. If the buffer size
> > is changed later then VIRTIO_VSOCK_MAX_BUF_SIZE can be exceeded. If
> > that doesn't matter, why even bother with VIRTIO_VSOCK_MAX_BUF_SIZE
> > here?
> >
>
> The .notify_buffer_size() should avoid this issue, since it allows the
> transport to limit the buffer size requested after the initialization.
>
> But again the min set by the user can not be respected and in the
> previous implementation we forced it to VIRTIO_VSOCK_MAX_BUF_SIZE.
>
> Now we don't limit the min, but we guarantee only that vsk->buffer_size
> is lower than VIRTIO_VSOCK_MAX_BUF_SIZE.
>
> Can that be an acceptable compromise?
I think so.
Setting buffer sizes was never tested or used much by userspace
applications that I'm aware of. We should probably include tests for
changing buffer sizes in the test suite.
Stefan
On Fri, Oct 11, 2019 at 09:27:14AM +0100, Stefan Hajnoczi wrote:
> On Thu, Oct 10, 2019 at 11:32:54AM +0200, Stefano Garzarella wrote:
> > On Wed, Oct 09, 2019 at 01:30:26PM +0100, Stefan Hajnoczi wrote:
> > > On Fri, Sep 27, 2019 at 01:26:57PM +0200, Stefano Garzarella wrote:
> > > Another issue is that this patch drops the VIRTIO_VSOCK_MAX_BUF_SIZE
> > > limit that used to be enforced by virtio_transport_set_buffer_size().
> > > Now the limit is only applied at socket init time. If the buffer size
> > > is changed later then VIRTIO_VSOCK_MAX_BUF_SIZE can be exceeded. If
> > > that doesn't matter, why even bother with VIRTIO_VSOCK_MAX_BUF_SIZE
> > > here?
> > >
> >
> > The .notify_buffer_size() should avoid this issue, since it allows the
> > transport to limit the buffer size requested after the initialization.
> >
> > But again the min set by the user can not be respected and in the
> > previous implementation we forced it to VIRTIO_VSOCK_MAX_BUF_SIZE.
> >
> > Now we don't limit the min, but we guarantee only that vsk->buffer_size
> > is lower than VIRTIO_VSOCK_MAX_BUF_SIZE.
> >
> > Can that be an acceptable compromise?
>
> I think so.
>
> Setting buffer sizes was never tested or used much by userspace
> applications that I'm aware of. We should probably include tests for
> changing buffer sizes in the test suite.
Good idea! We should add a test to check if min/max are respected,
playing a bit with these sockopt.
I'll do it in the test series!
Thanks,
Stefano