@@ -59,6 +59,7 @@ SRCS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio_rxtx_simple_neon.c
endif
ifeq ($(CONFIG_RTE_VIRTIO_USER),y)
+SRCS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio_user/vhost.c
SRCS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio_user/vhost_user.c
SRCS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio_user/virtio_user_dev.c
SRCS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio_user_ethdev.c
new file mode 100644
@@ -0,0 +1,162 @@
+/*-
+ * BSD LICENSE
+ *
+ * Copyright(c) 2010-2016 Intel Corporation. All rights reserved.
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * * Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ * * Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in
+ * the documentation and/or other materials provided with the
+ * distribution.
+ * * Neither the name of Intel Corporation nor the names of its
+ * contributors may be used to endorse or promote products derived
+ * from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <unistd.h>
+
+#include "vhost.h"
+
+#define VHOST_MAX_DEVICES 8
+
+static struct vhost_internal internals[VHOST_MAX_DEVICES];
+
+static inline int
+vhost_internal_valid_vid(int vid)
+{
+ if (vid < 0 || vid >= VHOST_MAX_DEVICES)
+ return 0;
+
+ if (!internals[vid].ops)
+ return 0;
+
+ return 1;
+}
+
+static int
+vhost_internal_alloc(void)
+{
+ int i;
+
+ for (i = 0; i < VHOST_MAX_DEVICES; ++i)
+ if (internals[i].ops == NULL)
+ break;
+
+ if (i >= VHOST_MAX_DEVICES)
+ return -1;
+
+ internals[i].features = 0;
+
+ internals[i].vhostfd = -1;
+
+ return 0;
+}
+
+static void
+vhost_internal_free(int id)
+{
+ internals[id].ops = NULL;
+
+ if (internals[id].vhostfd >= 0)
+ close(internals[id].vhostfd);
+}
+
+
+static int
+is_vhost_user_by_type(const char *path)
+{
+ struct stat sb;
+
+ if (stat(path, &sb) == -1)
+ return 0;
+
+ return S_ISSOCK(sb.st_mode);
+}
+
+int
+vhost_setup(const char *path)
+{
+ int ret = -1;
+ int vid = vhost_internal_alloc();
+
+ if (vid < 0) {
+ PMD_DRV_LOG(ERR, "exceeding %d devices", VHOST_MAX_DEVICES);
+ return -1;
+ }
+
+ if (is_vhost_user_by_type(path)) {
+ ret = vhost_ops_user.setup(&internals[vid], path);
+ internals[vid].ops = &vhost_ops_user;
+ }
+
+ if (ret < 0)
+ vhost_internal_free(vid);
+
+ return vid;
+}
+
+static const char * const vhost_msg_strings[] = {
+ [VHOST_USER_SET_OWNER] = "VHOST_USER_SET_OWNER",
+ [VHOST_USER_RESET_OWNER] = "VHOST_USER_RESET_OWNER",
+ [VHOST_USER_SET_FEATURES] = "VHOST_USER_SET_FEATURES",
+ [VHOST_USER_GET_FEATURES] = "VHOST_USER_GET_FEATURES",
+ [VHOST_USER_SET_VRING_CALL] = "VHOST_USER_SET_VRING_CALL",
+ [VHOST_USER_SET_VRING_NUM] = "VHOST_USER_SET_VRING_NUM",
+ [VHOST_USER_SET_VRING_BASE] = "VHOST_USER_SET_VRING_BASE",
+ [VHOST_USER_GET_VRING_BASE] = "VHOST_USER_GET_VRING_BASE",
+ [VHOST_USER_SET_VRING_ADDR] = "VHOST_USER_SET_VRING_ADDR",
+ [VHOST_USER_SET_VRING_KICK] = "VHOST_USER_SET_VRING_KICK",
+ [VHOST_USER_SET_MEM_TABLE] = "VHOST_USER_SET_MEM_TABLE",
+ [VHOST_USER_SET_VRING_ENABLE] = "VHOST_USER_SET_VRING_ENABLE",
+ NULL,
+};
+
+int
+vhost_call(int vid, enum vhost_user_request req, void *arg)
+{
+ RTE_SET_USED(vhost_msg_strings);
+ PMD_DRV_LOG(INFO, "%s", vhost_msg_strings[req]);
+
+ if (!vhost_internal_valid_vid(vid))
+ return -1;
+
+ return internals[vid].ops->control(&internals[vid], req, arg);
+}
+
+int
+vhost_enable_queue_pair(int vid, uint16_t pair_idx, int enable)
+{
+ if (!vhost_internal_valid_vid(vid))
+ return -1;
+
+ return internals[vid].ops->enable_qp(&internals[vid], pair_idx, enable);
+}
+
+void
+vhost_close(int vid)
+{
+ if (!vhost_internal_valid_vid(vid))
+ return;
+
+ vhost_internal_free(vid);
+}
@@ -42,8 +42,6 @@
#include "../virtio_logs.h"
#include "../virtqueue.h"
-#define VHOST_MEMORY_MAX_NREGIONS 8
-
struct vhost_vring_state {
unsigned int index;
unsigned int num;
@@ -105,42 +103,38 @@ struct vhost_memory_region {
uint64_t mmap_offset;
};
-struct vhost_memory {
- uint32_t nregions;
- uint32_t padding;
- struct vhost_memory_region regions[VHOST_MEMORY_MAX_NREGIONS];
+struct vhost_ops;
+
+struct vhost_internal {
+ /* for both vhost-user and vhost-kernel */
+ uint64_t features;
+ struct vhost_ops *ops;
+
+ /* for vhost-user */
+ int vhostfd;
+
+ /* for vhost-kernel */
+};
+
+typedef int (*vhost_setup_t)(struct vhost_internal *internal,
+ const char *path);
+typedef int (*vhost_control_t)(struct vhost_internal *internal,
+ enum vhost_user_request req,
+ void *arg);
+typedef int (*vhost_enable_qp_t)(struct vhost_internal *internal,
+ uint16_t pair_idx,
+ int enable);
+
+struct vhost_ops {
+ vhost_setup_t setup;
+ vhost_control_t control;
+ vhost_enable_qp_t enable_qp;
};
-struct vhost_user_msg {
- enum vhost_user_request request;
-
-#define VHOST_USER_VERSION_MASK 0x3
-#define VHOST_USER_REPLY_MASK (0x1 << 2)
- uint32_t flags;
- uint32_t size; /* the following payload size */
- union {
-#define VHOST_USER_VRING_IDX_MASK 0xff
-#define VHOST_USER_VRING_NOFD_MASK (0x1 << 8)
- uint64_t u64;
- struct vhost_vring_state state;
- struct vhost_vring_addr addr;
- struct vhost_memory memory;
- } payload;
- int fds[VHOST_MEMORY_MAX_NREGIONS];
-} __attribute((packed));
-
-#define VHOST_USER_HDR_SIZE offsetof(struct vhost_user_msg, payload.u64)
-#define VHOST_USER_PAYLOAD_SIZE \
- (sizeof(struct vhost_user_msg) - VHOST_USER_HDR_SIZE)
-
-/* The version of the protocol we support */
-#define VHOST_USER_VERSION 0x1
-
-#define VHOST_USER_F_PROTOCOL_FEATURES 30
-#define VHOST_USER_MQ (1ULL << VHOST_USER_F_PROTOCOL_FEATURES)
-
-int vhost_user_sock(int vhostfd, enum vhost_user_request req, void *arg);
-int vhost_user_setup(const char *path);
-int vhost_user_enable_queue_pair(int vhostfd, uint16_t pair_idx, int enable);
+struct vhost_ops vhost_ops_user;
+int vhost_setup(const char *path);
+int vhost_call(int vid, enum vhost_user_request req, void *arg);
+int vhost_enable_queue_pair(int vid, uint16_t pair_idx, int enable);
+void vhost_close(int vid);
#endif
@@ -42,6 +42,38 @@
#include "vhost.h"
+/* The version of the protocol we support */
+#define VHOST_USER_VERSION 0x1
+#define VHOST_MEMORY_MAX_NREGIONS 8
+
+struct vhost_memory {
+ uint32_t nregions;
+ uint32_t padding;
+ struct vhost_memory_region regions[VHOST_MEMORY_MAX_NREGIONS];
+};
+
+struct vhost_user_msg {
+ enum vhost_user_request request;
+
+#define VHOST_USER_VERSION_MASK 0x3
+#define VHOST_USER_REPLY_MASK (0x1 << 2)
+ uint32_t flags;
+ uint32_t size; /* the following payload size */
+ union {
+#define VHOST_USER_VRING_IDX_MASK 0xff
+#define VHOST_USER_VRING_NOFD_MASK (0x1 << 8)
+ uint64_t u64;
+ struct vhost_vring_state state;
+ struct vhost_vring_addr addr;
+ struct vhost_memory memory;
+ } payload;
+ int fds[VHOST_MEMORY_MAX_NREGIONS];
+} __attribute((packed));
+
+#define VHOST_USER_HDR_SIZE offsetof(struct vhost_user_msg, payload.u64)
+#define VHOST_USER_PAYLOAD_SIZE \
+ (sizeof(struct vhost_user_msg) - VHOST_USER_HDR_SIZE)
+
static int
vhost_user_write(int fd, void *buf, int len, int *fds, int fd_num)
{
@@ -223,24 +255,10 @@ prepare_vhost_memory_user(struct vhost_user_msg *msg, int fds[])
static struct vhost_user_msg m;
-static const char * const vhost_msg_strings[] = {
- [VHOST_USER_SET_OWNER] = "VHOST_USER_SET_OWNER",
- [VHOST_USER_RESET_OWNER] = "VHOST_USER_RESET_OWNER",
- [VHOST_USER_SET_FEATURES] = "VHOST_USER_SET_FEATURES",
- [VHOST_USER_GET_FEATURES] = "VHOST_USER_GET_FEATURES",
- [VHOST_USER_SET_VRING_CALL] = "VHOST_USER_SET_VRING_CALL",
- [VHOST_USER_SET_VRING_NUM] = "VHOST_USER_SET_VRING_NUM",
- [VHOST_USER_SET_VRING_BASE] = "VHOST_USER_SET_VRING_BASE",
- [VHOST_USER_GET_VRING_BASE] = "VHOST_USER_GET_VRING_BASE",
- [VHOST_USER_SET_VRING_ADDR] = "VHOST_USER_SET_VRING_ADDR",
- [VHOST_USER_SET_VRING_KICK] = "VHOST_USER_SET_VRING_KICK",
- [VHOST_USER_SET_MEM_TABLE] = "VHOST_USER_SET_MEM_TABLE",
- [VHOST_USER_SET_VRING_ENABLE] = "VHOST_USER_SET_VRING_ENABLE",
- NULL,
-};
-
-int
-vhost_user_sock(int vhostfd, enum vhost_user_request req, void *arg)
+static int
+vhost_user_sock(struct vhost_internal *internal,
+ enum vhost_user_request req,
+ void *arg)
{
struct vhost_user_msg msg;
struct vhost_vring_file *file = 0;
@@ -248,11 +266,9 @@ vhost_user_sock(int vhostfd, enum vhost_user_request req, void *arg)
int fds[VHOST_MEMORY_MAX_NREGIONS];
int fd_num = 0;
int i, len;
+ int vhostfd = internal->vhostfd;
RTE_SET_USED(m);
- RTE_SET_USED(vhost_msg_strings);
-
- PMD_DRV_LOG(INFO, "%s", vhost_msg_strings[req]);
msg.request = req;
msg.flags = VHOST_USER_VERSION;
@@ -323,8 +339,8 @@ vhost_user_sock(int vhostfd, enum vhost_user_request req, void *arg)
len = VHOST_USER_HDR_SIZE + msg.size;
if (vhost_user_write(vhostfd, &msg, len, fds, fd_num) < 0) {
- PMD_DRV_LOG(ERR, "%s failed: %s",
- vhost_msg_strings[req], strerror(errno));
+ PMD_DRV_LOG(ERR, "fail to write vhost user msg: %s",
+ strerror(errno));
return -1;
}
@@ -378,8 +394,8 @@ vhost_user_sock(int vhostfd, enum vhost_user_request req, void *arg)
* - (-1) if fail to set up;
* - (>=0) if successful, and it is the fd to vhostfd.
*/
-int
-vhost_user_setup(const char *path)
+static int
+vhost_user_setup(struct vhost_internal *internal, const char *path)
{
int fd;
int flag;
@@ -404,11 +420,14 @@ vhost_user_setup(const char *path)
return -1;
}
+ internal->vhostfd = fd;
+
return fd;
}
-int
-vhost_user_enable_queue_pair(int vhostfd, uint16_t pair_idx, int enable)
+static int
+vhost_user_enable_queue_pair(struct vhost_internal *internal,
+ uint16_t pair_idx, int enable)
{
int i;
@@ -418,10 +437,16 @@ vhost_user_enable_queue_pair(int vhostfd, uint16_t pair_idx, int enable)
.num = enable,
};
- if (vhost_user_sock(vhostfd,
+ if (vhost_user_sock(internal,
VHOST_USER_SET_VRING_ENABLE, &state))
return -1;
}
return 0;
}
+
+struct vhost_ops vhost_ops_user = {
+ .setup = vhost_user_setup,
+ .control = vhost_user_sock,
+ .enable_qp = vhost_user_enable_queue_pair
+};
@@ -64,7 +64,7 @@ virtio_user_create_queue(struct virtio_user_dev *dev, uint32_t queue_sel)
}
file.index = queue_sel;
file.fd = callfd;
- vhost_user_sock(dev->vhostfd, VHOST_USER_SET_VRING_CALL, &file);
+ vhost_call(dev->vid, VHOST_USER_SET_VRING_CALL, &file);
dev->callfds[queue_sel] = callfd;
return 0;
@@ -88,12 +88,12 @@ virtio_user_kick_queue(struct virtio_user_dev *dev, uint32_t queue_sel)
state.index = queue_sel;
state.num = vring->num;
- vhost_user_sock(dev->vhostfd, VHOST_USER_SET_VRING_NUM, &state);
+ vhost_call(dev->vid, VHOST_USER_SET_VRING_NUM, &state);
state.num = 0; /* no reservation */
- vhost_user_sock(dev->vhostfd, VHOST_USER_SET_VRING_BASE, &state);
+ vhost_call(dev->vid, VHOST_USER_SET_VRING_BASE, &state);
- vhost_user_sock(dev->vhostfd, VHOST_USER_SET_VRING_ADDR, &addr);
+ vhost_call(dev->vid, VHOST_USER_SET_VRING_ADDR, &addr);
/* Of all per virtqueue MSGs, make sure VHOST_USER_SET_VRING_KICK comes
* lastly because vhost depends on this msg to judge if
@@ -106,7 +106,7 @@ virtio_user_kick_queue(struct virtio_user_dev *dev, uint32_t queue_sel)
}
file.index = queue_sel;
file.fd = kickfd;
- vhost_user_sock(dev->vhostfd, VHOST_USER_SET_VRING_KICK, &file);
+ vhost_call(dev->vid, VHOST_USER_SET_VRING_KICK, &file);
dev->kickfds[queue_sel] = kickfd;
return 0;
@@ -146,21 +146,18 @@ virtio_user_start_device(struct virtio_user_dev *dev)
if (virtio_user_queue_setup(dev, virtio_user_create_queue) < 0)
goto error;
- /* Step 1: set features
- * Make sure VHOST_USER_F_PROTOCOL_FEATURES is added if mq is enabled,
- * and VIRTIO_NET_F_MAC is stripped.
+ /* Step 1: set feature
+ * Strip VIRTIO_NET_F_MAC, as MAC address is handled in vdev init.
*/
features = dev->features;
- if (dev->max_queue_pairs > 1)
- features |= VHOST_USER_MQ;
features &= ~(1ull << VIRTIO_NET_F_MAC);
- ret = vhost_user_sock(dev->vhostfd, VHOST_USER_SET_FEATURES, &features);
+ ret = vhost_call(dev->vid, VHOST_USER_SET_FEATURES, &features);
if (ret < 0)
goto error;
PMD_DRV_LOG(INFO, "set features: %" PRIx64, features);
/* Step 2: share memory regions */
- ret = vhost_user_sock(dev->vhostfd, VHOST_USER_SET_MEM_TABLE, NULL);
+ ret = vhost_call(dev->vid, VHOST_USER_SET_MEM_TABLE, NULL);
if (ret < 0)
goto error;
@@ -171,7 +168,7 @@ virtio_user_start_device(struct virtio_user_dev *dev)
/* Step 4: enable queues
* we enable the 1st queue pair by default.
*/
- vhost_user_enable_queue_pair(dev->vhostfd, 0, 1);
+ vhost_enable_queue_pair(dev->vid, 0, 1);
return 0;
error:
@@ -181,7 +178,7 @@ virtio_user_start_device(struct virtio_user_dev *dev)
int virtio_user_stop_device(struct virtio_user_dev *dev)
{
- return vhost_user_sock(dev->vhostfd, VHOST_USER_RESET_OWNER, NULL);
+ return vhost_call(dev->vid, VHOST_USER_RESET_OWNER, NULL);
}
static inline void
@@ -215,19 +212,18 @@ virtio_user_dev_init(struct virtio_user_dev *dev, char *path, int queues,
dev->queue_size = queue_size;
dev->mac_specified = 0;
parse_mac(dev, mac);
- dev->vhostfd = -1;
- dev->vhostfd = vhost_user_setup(dev->path);
- if (dev->vhostfd < 0) {
+ dev->vid = vhost_setup(dev->path);
+ if (dev->vid < 0) {
PMD_INIT_LOG(ERR, "backend set up fails");
return -1;
}
- if (vhost_user_sock(dev->vhostfd, VHOST_USER_SET_OWNER, NULL) < 0) {
+ if (vhost_call(dev->vid, VHOST_USER_SET_OWNER, NULL) < 0) {
PMD_INIT_LOG(ERR, "set_owner fails: %s", strerror(errno));
return -1;
}
- if (vhost_user_sock(dev->vhostfd, VHOST_USER_GET_FEATURES,
+ if (vhost_call(dev->vid, VHOST_USER_GET_FEATURES,
&dev->features) < 0) {
PMD_INIT_LOG(ERR, "get_features failed: %s", strerror(errno));
return -1;
@@ -253,13 +249,6 @@ virtio_user_dev_init(struct virtio_user_dev *dev, char *path, int queues,
PMD_INIT_LOG(INFO, "vhost does not support ctrl-q");
}
- if (dev->max_queue_pairs > 1) {
- if (!(dev->features & VHOST_USER_MQ)) {
- PMD_INIT_LOG(ERR, "MQ not supported by the backend");
- return -1;
- }
- }
-
return 0;
}
@@ -273,7 +262,7 @@ virtio_user_dev_uninit(struct virtio_user_dev *dev)
close(dev->kickfds[i]);
}
- close(dev->vhostfd);
+ vhost_close(dev->vid);
}
static uint8_t
@@ -289,9 +278,9 @@ virtio_user_handle_mq(struct virtio_user_dev *dev, uint16_t q_pairs)
}
for (i = 0; i < q_pairs; ++i)
- ret |= vhost_user_enable_queue_pair(dev->vhostfd, i, 1);
+ ret |= vhost_enable_queue_pair(dev->vid, i, 1);
for (i = q_pairs; i < dev->max_queue_pairs; ++i)
- ret |= vhost_user_enable_queue_pair(dev->vhostfd, i, 0);
+ ret |= vhost_enable_queue_pair(dev->vid, i, 0);
dev->queue_pairs = q_pairs;
@@ -39,7 +39,7 @@
#include "../virtio_ring.h"
struct virtio_user_dev {
- int vhostfd;
+ int vid;
int callfds[VIRTIO_MAX_VIRTQUEUES * 2 + 1];
int kickfds[VIRTIO_MAX_VIRTQUEUES * 2 + 1];
int mac_specified;