#include <rte_malloc.h>
#include <rte_log.h>
+#include "iotlb.h"
#include "vhost.h"
#include "vhost_user.h"
[VHOST_USER_SET_VRING_ENABLE] = "VHOST_USER_SET_VRING_ENABLE",
[VHOST_USER_SEND_RARP] = "VHOST_USER_SEND_RARP",
[VHOST_USER_NET_SET_MTU] = "VHOST_USER_NET_SET_MTU",
+ [VHOST_USER_SET_SLAVE_REQ_FD] = "VHOST_USER_SET_SLAVE_REQ_FD",
+ [VHOST_USER_IOTLB_MSG] = "VHOST_USER_IOTLB_MSG",
};
static uint64_t
rte_free(dev->mem);
dev->mem = NULL;
}
+
+ free(dev->guest_pages);
+ dev->guest_pages = NULL;
+
if (dev->log_addr) {
munmap((void *)(uintptr_t)dev->log_addr, dev->log_size);
dev->log_addr = 0;
}
+
+ if (dev->slave_req_fd >= 0) {
+ close(dev->slave_req_fd);
+ dev->slave_req_fd = -1;
+ }
}
/*
uint64_t vhost_features = 0;
rte_vhost_driver_get_features(dev->ifname, &vhost_features);
- if (features & ~vhost_features)
+ if (features & ~vhost_features) {
+ RTE_LOG(ERR, VHOST_CONFIG,
+ "(%d) received invalid negotiated features.\n",
+ dev->vid);
return -1;
+ }
if ((dev->flags & VIRTIO_DEV_RUNNING) && dev->features != features) {
if (dev->notify_ops->features_changed)
*/
static int
vhost_user_set_vring_num(struct virtio_net *dev,
- struct vhost_vring_state *state)
+ VhostUserMsg *msg)
{
- struct vhost_virtqueue *vq = dev->virtqueue[state->index];
+ struct vhost_virtqueue *vq = dev->virtqueue[msg->payload.state.index];
- vq->size = state->num;
+ vq->size = msg->payload.state.num;
if (dev->dequeue_zero_copy) {
vq->nr_zmbuf = 0;
return -1;
}
+ vq->batch_copy_elems = rte_malloc(NULL,
+ vq->size * sizeof(struct batch_copy_elem),
+ RTE_CACHE_LINE_SIZE);
+ if (!vq->batch_copy_elems) {
+ RTE_LOG(ERR, VHOST_CONFIG,
+ "failed to allocate memory for batching copy.\n");
+ return -1;
+ }
+
return 0;
}
struct vhost_virtqueue *old_vq, *vq;
int ret;
- enum {VIRTIO_RXQ, VIRTIO_TXQ, VIRTIO_QNUM};
-
old_dev = dev;
vq = old_vq = dev->virtqueue[index];
if (!vq)
return dev;
- memcpy(vq, old_vq, sizeof(*vq) * VIRTIO_QNUM);
+ memcpy(vq, old_vq, sizeof(*vq));
rte_free(old_vq);
}
}
#endif
-/*
- * Converts QEMU virtual address to Vhost virtual address. This function is
- * used to convert the ring addresses to our address space.
- */
+/* Converts QEMU virtual address to Vhost virtual address. */
static uint64_t
qva_to_vva(struct virtio_net *dev, uint64_t qva)
{
return 0;
}
+
+/*
+ * Converts ring address to Vhost virtual address.
+ * If IOMMU is enabled, the ring address is a guest IO virtual address,
+ * else it is a QEMU virtual address.
+ */
+static uint64_t
+ring_addr_to_vva(struct virtio_net *dev, struct vhost_virtqueue *vq,
+ uint64_t ra, uint64_t size)
+{
+ if (dev->features & (1ULL << VIRTIO_F_IOMMU_PLATFORM)) {
+ uint64_t vva;
+
+ vva = vhost_user_iotlb_cache_find(vq, ra,
+ &size, VHOST_ACCESS_RW);
+ if (!vva)
+ vhost_user_iotlb_miss(dev, ra, VHOST_ACCESS_RW);
+
+ return vva;
+ }
+
+ return qva_to_vva(dev, ra);
+}
+
/*
* The virtio device sends us the desc, used and avail ring addresses.
* This function then converts these to our address space.
*/
static int
-vhost_user_set_vring_addr(struct virtio_net *dev, struct vhost_vring_addr *addr)
+vhost_user_set_vring_addr(struct virtio_net *dev, VhostUserMsg *msg)
{
struct vhost_virtqueue *vq;
+ struct vhost_vring_addr *addr = &msg->payload.addr;
if (dev->mem == NULL)
return -1;
/* addr->index refers to the queue index. The txq 1, rxq is 0. */
- vq = dev->virtqueue[addr->index];
+ vq = dev->virtqueue[msg->payload.addr.index];
+
+ /*
+ * Rings addresses should not be interpreted as long as the ring is not
+ * started and enabled
+ */
+ memcpy(&vq->ring_addrs, addr, sizeof(*addr));
+
+ vring_invalidate(dev, vq);
+
+ return 0;
+}
+
+static struct virtio_net *
+translate_ring_addresses(struct virtio_net *dev, int vq_index)
+{
+ struct vhost_virtqueue *vq = dev->virtqueue[vq_index];
+ struct vhost_vring_addr *addr = &vq->ring_addrs;
/* The addresses are converted from QEMU virtual to Vhost virtual. */
- vq->desc = (struct vring_desc *)(uintptr_t)qva_to_vva(dev,
- addr->desc_user_addr);
+ if (vq->desc && vq->avail && vq->used)
+ return dev;
+
+ vq->desc = (struct vring_desc *)(uintptr_t)ring_addr_to_vva(dev,
+ vq, addr->desc_user_addr, sizeof(struct vring_desc));
if (vq->desc == 0) {
- RTE_LOG(ERR, VHOST_CONFIG,
+ RTE_LOG(DEBUG, VHOST_CONFIG,
"(%d) failed to find desc ring address.\n",
dev->vid);
- return -1;
+ return dev;
}
- dev = numa_realloc(dev, addr->index);
- vq = dev->virtqueue[addr->index];
+ dev = numa_realloc(dev, vq_index);
+ vq = dev->virtqueue[vq_index];
- vq->avail = (struct vring_avail *)(uintptr_t)qva_to_vva(dev,
- addr->avail_user_addr);
+ vq->avail = (struct vring_avail *)(uintptr_t)ring_addr_to_vva(dev,
+ vq, addr->avail_user_addr, sizeof(struct vring_avail));
if (vq->avail == 0) {
- RTE_LOG(ERR, VHOST_CONFIG,
+ RTE_LOG(DEBUG, VHOST_CONFIG,
"(%d) failed to find avail ring address.\n",
dev->vid);
- return -1;
+ return dev;
}
- vq->used = (struct vring_used *)(uintptr_t)qva_to_vva(dev,
- addr->used_user_addr);
+ vq->used = (struct vring_used *)(uintptr_t)ring_addr_to_vva(dev,
+ vq, addr->used_user_addr, sizeof(struct vring_used));
if (vq->used == 0) {
- RTE_LOG(ERR, VHOST_CONFIG,
+ RTE_LOG(DEBUG, VHOST_CONFIG,
"(%d) failed to find used ring address.\n",
dev->vid);
- return -1;
+ return dev;
}
if (vq->last_used_idx != vq->used->idx) {
LOG_DEBUG(VHOST_CONFIG, "(%d) log_guest_addr: %" PRIx64 "\n",
dev->vid, vq->log_guest_addr);
- return 0;
+ return dev;
}
/*
*/
static int
vhost_user_set_vring_base(struct virtio_net *dev,
- struct vhost_vring_state *state)
+ VhostUserMsg *msg)
{
- dev->virtqueue[state->index]->last_used_idx = state->num;
- dev->virtqueue[state->index]->last_avail_idx = state->num;
+ dev->virtqueue[msg->payload.state.index]->last_used_idx =
+ msg->payload.state.num;
+ dev->virtqueue[msg->payload.state.index]->last_avail_idx =
+ msg->payload.state.num;
return 0;
}
static int
vq_is_ready(struct vhost_virtqueue *vq)
{
- return vq && vq->desc &&
+ return vq && vq->desc && vq->avail && vq->used &&
vq->kickfd != VIRTIO_UNINITIALIZED_EVENTFD &&
vq->callfd != VIRTIO_UNINITIALIZED_EVENTFD;
}
}
static void
-vhost_user_set_vring_kick(struct virtio_net *dev, struct VhostUserMsg *pmsg)
+vhost_user_set_vring_kick(struct virtio_net **pdev, struct VhostUserMsg *pmsg)
{
struct vhost_vring_file file;
struct vhost_virtqueue *vq;
+ struct virtio_net *dev = *pdev;
file.index = pmsg->payload.u64 & VHOST_USER_VRING_IDX_MASK;
if (pmsg->payload.u64 & VHOST_USER_VRING_NOFD_MASK)
RTE_LOG(INFO, VHOST_CONFIG,
"vring kick idx:%d file:%d\n", file.index, file.fd);
+ /*
+ * Interpret ring addresses only when ring is started and enabled.
+ * This is now if protocol features aren't supported.
+ */
+ if (!(dev->features & (1ULL << VHOST_USER_F_PROTOCOL_FEATURES))) {
+ *pdev = dev = translate_ring_addresses(dev, file.index);
+ if (!dev)
+ return;
+ }
+
vq = dev->virtqueue[file.index];
+
+ /*
+ * When VHOST_USER_F_PROTOCOL_FEATURES is not negotiated,
+ * the ring starts already enabled. Otherwise, it is enabled via
+ * the SET_VRING_ENABLE message.
+ */
+ if (!(dev->features & (1ULL << VHOST_USER_F_PROTOCOL_FEATURES)))
+ vq->enabled = 1;
+
if (vq->kickfd >= 0)
close(vq->kickfd);
vq->kickfd = file.fd;
*/
static int
vhost_user_get_vring_base(struct virtio_net *dev,
- struct vhost_vring_state *state)
+ VhostUserMsg *msg)
{
- struct vhost_virtqueue *vq = dev->virtqueue[state->index];
+ struct vhost_virtqueue *vq = dev->virtqueue[msg->payload.state.index];
/* We have to stop the queue (virtio) if it is running. */
if (dev->flags & VIRTIO_DEV_RUNNING) {
dev->flags &= ~VIRTIO_DEV_READY;
/* Here we are safe to get the last used index */
- state->num = vq->last_used_idx;
+ msg->payload.state.num = vq->last_used_idx;
RTE_LOG(INFO, VHOST_CONFIG,
- "vring base idx:%d file:%d\n", state->index, state->num);
+ "vring base idx:%d file:%d\n", msg->payload.state.index,
+ msg->payload.state.num);
/*
* Based on current qemu vhost-user implementation, this message is
* sent and only sent in vhost_vring_stop.
rte_free(vq->shadow_used_ring);
vq->shadow_used_ring = NULL;
+ rte_free(vq->batch_copy_elems);
+ vq->batch_copy_elems = NULL;
+
return 0;
}
* enable the virtio queue pair.
*/
static int
-vhost_user_set_vring_enable(struct virtio_net *dev,
- struct vhost_vring_state *state)
+vhost_user_set_vring_enable(struct virtio_net **pdev,
+ VhostUserMsg *msg)
{
- int enable = (int)state->num;
+ struct virtio_net *dev = *pdev;
+ int enable = (int)msg->payload.state.num;
RTE_LOG(INFO, VHOST_CONFIG,
"set queue enable: %d to qp idx: %d\n",
- enable, state->index);
+ enable, msg->payload.state.index);
+
+ /*
+ * Interpret ring addresses only when ring is started and enabled.
+ * This is now if protocol features are supported.
+ */
+ if (enable && (dev->features &
+ (1ULL << VHOST_USER_F_PROTOCOL_FEATURES))) {
+ dev = translate_ring_addresses(dev, msg->payload.state.index);
+ if (!dev)
+ return -1;
+
+ *pdev = dev;
+ }
if (dev->notify_ops->vring_state_changed)
- dev->notify_ops->vring_state_changed(dev->vid, state->index, enable);
+ dev->notify_ops->vring_state_changed(dev->vid,
+ msg->payload.state.index, enable);
- dev->virtqueue[state->index]->enabled = enable;
+ dev->virtqueue[msg->payload.state.index]->enabled = enable;
return 0;
}
return 0;
}
+static int
+vhost_user_set_req_fd(struct virtio_net *dev, struct VhostUserMsg *msg)
+{
+ int fd = msg->fds[0];
+
+ if (fd < 0) {
+ RTE_LOG(ERR, VHOST_CONFIG,
+ "Invalid file descriptor for slave channel (%d)\n",
+ fd);
+ return -1;
+ }
+
+ dev->slave_req_fd = fd;
+
+ return 0;
+}
+
+static int
+is_vring_iotlb_update(struct vhost_virtqueue *vq, struct vhost_iotlb_msg *imsg)
+{
+ struct vhost_vring_addr *ra;
+ uint64_t start, end;
+
+ start = imsg->iova;
+ end = start + imsg->size;
+
+ ra = &vq->ring_addrs;
+ if (ra->desc_user_addr >= start && ra->desc_user_addr < end)
+ return 1;
+ if (ra->avail_user_addr >= start && ra->avail_user_addr < end)
+ return 1;
+ if (ra->used_user_addr >= start && ra->used_user_addr < end)
+ return 1;
+
+ return 0;
+}
+
+static int
+is_vring_iotlb_invalidate(struct vhost_virtqueue *vq,
+ struct vhost_iotlb_msg *imsg)
+{
+ uint64_t istart, iend, vstart, vend;
+
+ istart = imsg->iova;
+ iend = istart + imsg->size - 1;
+
+ vstart = (uintptr_t)vq->desc;
+ vend = vstart + sizeof(struct vring_desc) * vq->size - 1;
+ if (vstart <= iend && istart <= vend)
+ return 1;
+
+ vstart = (uintptr_t)vq->avail;
+ vend = vstart + sizeof(struct vring_avail);
+ vend += sizeof(uint16_t) * vq->size - 1;
+ if (vstart <= iend && istart <= vend)
+ return 1;
+
+ vstart = (uintptr_t)vq->used;
+ vend = vstart + sizeof(struct vring_used);
+ vend += sizeof(struct vring_used_elem) * vq->size - 1;
+ if (vstart <= iend && istart <= vend)
+ return 1;
+
+ return 0;
+}
+
+static int
+vhost_user_iotlb_msg(struct virtio_net **pdev, struct VhostUserMsg *msg)
+{
+ struct virtio_net *dev = *pdev;
+ struct vhost_iotlb_msg *imsg = &msg->payload.iotlb;
+ uint16_t i;
+ uint64_t vva;
+
+ switch (imsg->type) {
+ case VHOST_IOTLB_UPDATE:
+ vva = qva_to_vva(dev, imsg->uaddr);
+ if (!vva)
+ return -1;
+
+ for (i = 0; i < dev->nr_vring; i++) {
+ struct vhost_virtqueue *vq = dev->virtqueue[i];
+
+ vhost_user_iotlb_cache_insert(vq, imsg->iova, vva,
+ imsg->size, imsg->perm);
+
+ if (is_vring_iotlb_update(vq, imsg))
+ *pdev = dev = translate_ring_addresses(dev, i);
+ }
+ break;
+ case VHOST_IOTLB_INVALIDATE:
+ for (i = 0; i < dev->nr_vring; i++) {
+ struct vhost_virtqueue *vq = dev->virtqueue[i];
+
+ vhost_user_iotlb_cache_remove(vq, imsg->iova,
+ imsg->size);
+
+ if (is_vring_iotlb_invalidate(vq, imsg))
+ vring_invalidate(dev, vq);
+ }
+ break;
+ default:
+ RTE_LOG(ERR, VHOST_CONFIG, "Invalid IOTLB message type (%d)\n",
+ imsg->type);
+ return -1;
+ }
+
+ return 0;
+}
+
/* return bytes# of read on success or negative val on failure. */
static int
read_vhost_message(int sockfd, struct VhostUserMsg *msg)
static int
send_vhost_message(int sockfd, struct VhostUserMsg *msg)
{
- int ret;
+ if (!msg)
+ return 0;
+
+ return send_fd_message(sockfd, (char *)msg,
+ VHOST_USER_HDR_SIZE + msg->size, NULL, 0);
+}
+static int
+send_vhost_reply(int sockfd, struct VhostUserMsg *msg)
+{
if (!msg)
return 0;
msg->flags |= VHOST_USER_VERSION;
msg->flags |= VHOST_USER_REPLY_MASK;
- ret = send_fd_message(sockfd, (char *)msg,
- VHOST_USER_HDR_SIZE + msg->size, NULL, 0);
-
- return ret;
+ return send_vhost_message(sockfd, msg);
}
/*
{
uint16_t vring_idx;
- switch (msg->request) {
+ switch (msg->request.master) {
case VHOST_USER_SET_VRING_KICK:
case VHOST_USER_SET_VRING_CALL:
case VHOST_USER_SET_VRING_ERR:
}
ret = read_vhost_message(fd, &msg);
- if (ret <= 0 || msg.request >= VHOST_USER_MAX) {
+ if (ret <= 0 || msg.request.master >= VHOST_USER_MAX) {
if (ret < 0)
RTE_LOG(ERR, VHOST_CONFIG,
"vhost read message failed\n");
}
ret = 0;
- RTE_LOG(INFO, VHOST_CONFIG, "read message %s\n",
- vhost_message_str[msg.request]);
+ if (msg.request.master != VHOST_USER_IOTLB_MSG)
+ RTE_LOG(INFO, VHOST_CONFIG, "read message %s\n",
+ vhost_message_str[msg.request.master]);
+ else
+ RTE_LOG(DEBUG, VHOST_CONFIG, "read message %s\n",
+ vhost_message_str[msg.request.master]);
ret = vhost_user_check_and_alloc_queue_pair(dev, &msg);
if (ret < 0) {
return -1;
}
- switch (msg.request) {
+ switch (msg.request.master) {
case VHOST_USER_GET_FEATURES:
msg.payload.u64 = vhost_user_get_features(dev);
msg.size = sizeof(msg.payload.u64);
- send_vhost_message(fd, &msg);
+ send_vhost_reply(fd, &msg);
break;
case VHOST_USER_SET_FEATURES:
vhost_user_set_features(dev, msg.payload.u64);
case VHOST_USER_GET_PROTOCOL_FEATURES:
msg.payload.u64 = VHOST_USER_PROTOCOL_FEATURES;
msg.size = sizeof(msg.payload.u64);
- send_vhost_message(fd, &msg);
+ send_vhost_reply(fd, &msg);
break;
case VHOST_USER_SET_PROTOCOL_FEATURES:
vhost_user_set_protocol_features(dev, msg.payload.u64);
/* it needs a reply */
msg.size = sizeof(msg.payload.u64);
- send_vhost_message(fd, &msg);
+ send_vhost_reply(fd, &msg);
break;
case VHOST_USER_SET_LOG_FD:
close(msg.fds[0]);
break;
case VHOST_USER_SET_VRING_NUM:
- vhost_user_set_vring_num(dev, &msg.payload.state);
+ vhost_user_set_vring_num(dev, &msg);
break;
case VHOST_USER_SET_VRING_ADDR:
- vhost_user_set_vring_addr(dev, &msg.payload.addr);
+ vhost_user_set_vring_addr(dev, &msg);
break;
case VHOST_USER_SET_VRING_BASE:
- vhost_user_set_vring_base(dev, &msg.payload.state);
+ vhost_user_set_vring_base(dev, &msg);
break;
case VHOST_USER_GET_VRING_BASE:
- vhost_user_get_vring_base(dev, &msg.payload.state);
+ vhost_user_get_vring_base(dev, &msg);
msg.size = sizeof(msg.payload.state);
- send_vhost_message(fd, &msg);
+ send_vhost_reply(fd, &msg);
break;
case VHOST_USER_SET_VRING_KICK:
- vhost_user_set_vring_kick(dev, &msg);
+ vhost_user_set_vring_kick(&dev, &msg);
break;
case VHOST_USER_SET_VRING_CALL:
vhost_user_set_vring_call(dev, &msg);
case VHOST_USER_GET_QUEUE_NUM:
msg.payload.u64 = VHOST_MAX_QUEUE_PAIRS;
msg.size = sizeof(msg.payload.u64);
- send_vhost_message(fd, &msg);
+ send_vhost_reply(fd, &msg);
break;
case VHOST_USER_SET_VRING_ENABLE:
- vhost_user_set_vring_enable(dev, &msg.payload.state);
+ vhost_user_set_vring_enable(&dev, &msg);
break;
case VHOST_USER_SEND_RARP:
vhost_user_send_rarp(dev, &msg);
ret = vhost_user_net_set_mtu(dev, &msg);
break;
+ case VHOST_USER_SET_SLAVE_REQ_FD:
+ ret = vhost_user_set_req_fd(dev, &msg);
+ break;
+
+ case VHOST_USER_IOTLB_MSG:
+ ret = vhost_user_iotlb_msg(&dev, &msg);
+ break;
+
default:
ret = -1;
break;
if (msg.flags & VHOST_USER_NEED_REPLY) {
msg.payload.u64 = !!ret;
msg.size = sizeof(msg.payload.u64);
- send_vhost_message(fd, &msg);
+ send_vhost_reply(fd, &msg);
}
if (!(dev->flags & VIRTIO_DEV_RUNNING) && virtio_is_ready(dev)) {
return 0;
}
+
+int
+vhost_user_iotlb_miss(struct virtio_net *dev, uint64_t iova, uint8_t perm)
+{
+ int ret;
+ struct VhostUserMsg msg = {
+ .request.slave = VHOST_USER_SLAVE_IOTLB_MSG,
+ .flags = VHOST_USER_VERSION,
+ .size = sizeof(msg.payload.iotlb),
+ .payload.iotlb = {
+ .iova = iova,
+ .perm = perm,
+ .type = VHOST_IOTLB_MISS,
+ },
+ };
+
+ ret = send_vhost_message(dev->slave_req_fd, &msg);
+ if (ret < 0) {
+ RTE_LOG(ERR, VHOST_CONFIG,
+ "Failed to send IOTLB miss message (%d)\n",
+ ret);
+ return ret;
+ }
+
+ return 0;
+}