aboutsummaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorPierre Pfister <ppfister@cisco.com>2016-09-21 08:04:59 +0100
committerDamjan Marion <dmarion.lists@gmail.com>2016-11-04 12:31:11 +0000
commite21c52861d7c503bef7fc464f23bbc7891e150d7 (patch)
tree0b47f4e64326e649e0e02cd780f49f072c787140
parentf524f0add51e603118184fcd42f5b8eeafa11462 (diff)
vhost-user: multiqueue support
This patch adds multi-queue support to non-dpdk's vhost-user driver. Waiting for a unified way to manage threads, this patch defines a way to assign threads to interfaces that is specific to vhost. Change-Id: I86298788b1a4e886c5431f187dc17175d12c7a8b Signed-off-by: Pierre Pfister <ppfister@cisco.com>
-rw-r--r--vnet/vnet/devices/virtio/vhost-user.c765
-rw-r--r--vnet/vnet/devices/virtio/vhost-user.h85
2 files changed, 651 insertions, 199 deletions
diff --git a/vnet/vnet/devices/virtio/vhost-user.c b/vnet/vnet/devices/virtio/vhost-user.c
index b55547cd16c..7137a22a2d9 100644
--- a/vnet/vnet/devices/virtio/vhost-user.c
+++ b/vnet/vnet/devices/virtio/vhost-user.c
@@ -279,6 +279,167 @@ unmap_all_mem_regions (vhost_user_intf_t * vui)
vui->nregions = 0;
}
+static void
+vhost_user_tx_thread_placement (vhost_user_intf_t * vui)
+{
+ //Let's try to assign one queue to each thread
+ u32 qid = 0;
+ u32 cpu_index = 0;
+ vui->use_tx_spinlock = 0;
+ while (1)
+ {
+ for (qid = 0; qid < VHOST_VRING_MAX_N / 2; qid++)
+ {
+ vhost_user_vring_t *rxvq = &vui->vrings[VHOST_VRING_IDX_RX (qid)];
+ if (!rxvq->started || !rxvq->enabled)
+ continue;
+
+ vui->per_cpu_tx_qid[cpu_index] = qid;
+ cpu_index++;
+ if (cpu_index == vlib_get_thread_main ()->n_vlib_mains)
+ return;
+ }
+ //We need to loop, meaning the spinlock has to be used
+ vui->use_tx_spinlock = 1;
+ if (cpu_index == 0)
+ {
+ //Could not find a single valid one
+ for (cpu_index = 0;
+ cpu_index < vlib_get_thread_main ()->n_vlib_mains; cpu_index++)
+ {
+ vui->per_cpu_tx_qid[cpu_index] = 0;
+ }
+ return;
+ }
+ }
+}
+
+static void
+vhost_user_rx_thread_placement ()
+{
+ vhost_user_main_t *vum = &vhost_user_main;
+ vhost_user_intf_t *vui;
+ vhost_cpu_t *vhc;
+ u32 *workers = 0;
+
+ //Let's list all workers cpu indexes
+ u32 i;
+ for (i = vum->input_cpu_first_index;
+ i < vum->input_cpu_first_index + vum->input_cpu_count; i++)
+ {
+ vlib_node_set_state (vlib_mains ? vlib_mains[i] : &vlib_global_main,
+ vhost_user_input_node.index,
+ VLIB_NODE_STATE_DISABLED);
+ vec_add1 (workers, i);
+ }
+
+ vec_foreach (vhc, vum->cpus)
+ {
+ vec_reset_length (vhc->rx_queues);
+ }
+
+ i = 0;
+ vec_foreach (vui, vum->vhost_user_interfaces)
+ {
+ if (!vui->active)
+ continue;
+
+ u32 *vui_workers = vec_len (vui->workers) ? vui->workers : workers;
+ u32 qid;
+ for (qid = 0; qid < VHOST_VRING_MAX_N / 2; qid++)
+ {
+ vhost_user_vring_t *txvq = &vui->vrings[VHOST_VRING_IDX_TX (qid)];
+ if (!txvq->started)
+ continue;
+
+ i %= vec_len (vui_workers);
+ u32 cpu_index = vui_workers[i];
+ i++;
+ vhc = &vum->cpus[cpu_index];
+
+ vhost_iface_and_queue_t iaq = {
+ .qid = qid,
+ .vhost_iface_index = vui - vum->vhost_user_interfaces,
+ };
+ vec_add1 (vhc->rx_queues, iaq);
+ vlib_node_set_state (vlib_mains ? vlib_mains[cpu_index] :
+ &vlib_global_main, vhost_user_input_node.index,
+ VLIB_NODE_STATE_POLLING);
+ }
+ }
+}
+
+static int
+vhost_user_thread_placement (u32 sw_if_index, u32 worker_thread_index, u8 del)
+{
+ vhost_user_main_t *vum = &vhost_user_main;
+ vhost_user_intf_t *vui;
+ vnet_hw_interface_t *hw;
+
+ if (worker_thread_index < vum->input_cpu_first_index ||
+ worker_thread_index >=
+ vum->input_cpu_first_index + vum->input_cpu_count)
+ return -1;
+
+ if (!(hw = vnet_get_sup_hw_interface (vnet_get_main (), sw_if_index)))
+ return -2;
+
+ vui = vec_elt_at_index (vum->vhost_user_interfaces, hw->dev_instance);
+ u32 found = ~0, *w;
+ vec_foreach (w, vui->workers)
+ {
+ if (*w == worker_thread_index)
+ {
+ found = w - vui->workers;
+ break;
+ }
+ }
+
+ if (del)
+ {
+ if (found == ~0)
+ return -3;
+ vec_del1 (vui->workers, found);
+ }
+ else if (found == ~0)
+ {
+ vec_add1 (vui->workers, worker_thread_index);
+ }
+
+ vhost_user_rx_thread_placement ();
+ return 0;
+}
+
+/** @brief Returns whether at least one TX and one RX vring are enabled */
+int
+vhost_user_intf_ready (vhost_user_intf_t * vui)
+{
+ int i, found[2] = { }; //RX + TX
+
+ for (i = 0; i < VHOST_VRING_MAX_N; i++)
+ if (vui->vrings[i].started && vui->vrings[i].enabled)
+ found[i & 1] = 1;
+
+ return found[0] && found[1];
+}
+
+static void
+vhost_user_update_iface_state (vhost_user_intf_t * vui)
+{
+ /* if we have pointers to descriptor table, go up */
+ int is_up = vhost_user_intf_ready (vui);
+ if (is_up != vui->is_up)
+ {
+ DBG_SOCK ("interface %d %s", vui->sw_if_index,
+ is_up ? "ready" : "down");
+ vnet_hw_interface_set_flags (vnet_get_main (), vui->hw_if_index,
+ is_up ? VNET_HW_INTERFACE_FLAG_LINK_UP :
+ 0);
+ vui->is_up = is_up;
+ }
+ vhost_user_rx_thread_placement ();
+ vhost_user_tx_thread_placement (vui);
+}
static clib_error_t *
vhost_user_callfd_read_ready (unix_file_t * uf)
@@ -289,6 +450,97 @@ vhost_user_callfd_read_ready (unix_file_t * uf)
return 0;
}
+static clib_error_t *
+vhost_user_kickfd_read_ready (unix_file_t * uf)
+{
+ __attribute__ ((unused)) int n;
+ u8 buff[8];
+ vhost_user_intf_t *vui =
+ vec_elt_at_index (vhost_user_main.vhost_user_interfaces,
+ uf->private_data >> 8);
+ u32 qid = uf->private_data & 0xff;
+ n = read (uf->file_descriptor, ((char *) &buff), 8);
+ DBG_SOCK ("if %d KICK queue %d", uf->private_data >> 8, qid);
+
+ vlib_worker_thread_barrier_sync (vlib_get_main ());
+ vui->vrings[qid].started = 1;
+ vhost_user_update_iface_state (vui);
+ vlib_worker_thread_barrier_release (vlib_get_main ());
+ return 0;
+}
+
+/**
+ * @brief Try once to lock the vring
+ * @return 0 on success, non-zero on failure.
+ */
+static inline int
+vhost_user_vring_try_lock (vhost_user_intf_t * vui, u32 qid)
+{
+ return __sync_lock_test_and_set (vui->vring_locks[qid], 1);
+}
+
+/**
+ * @brief Spin until the vring is successfully locked
+ */
+static inline void
+vhost_user_vring_lock (vhost_user_intf_t * vui, u32 qid)
+{
+ while (vhost_user_vring_try_lock (vui, qid))
+ ;
+}
+
+/**
+ * @brief Unlock the vring lock
+ */
+static inline void
+vhost_user_vring_unlock (vhost_user_intf_t * vui, u32 qid)
+{
+ *vui->vring_locks[qid] = 0;
+}
+
+static inline void
+vhost_user_vring_init (vhost_user_intf_t * vui, u32 qid)
+{
+ vhost_user_vring_t *vring = &vui->vrings[qid];
+ memset (vring, 0, sizeof (*vring));
+ vring->kickfd = -1;
+ vring->callfd = -1;
+ vring->errfd = -1;
+
+ /*
+ * We have a bug with some qemu 2.5, and this may be a fix.
+ * Feel like interpretation holy text, but this is from vhost-user.txt.
+ * "
+ * One queue pair is enabled initially. More queues are enabled
+ * dynamically, by sending message VHOST_USER_SET_VRING_ENABLE.
+ * "
+ * Don't know who's right, but this is what DPDK does.
+ */
+ if (qid == 0 || qid == 1)
+ vring->enabled = 1;
+}
+
+static inline void
+vhost_user_vring_close (vhost_user_intf_t * vui, u32 qid)
+{
+ vhost_user_vring_t *vring = &vui->vrings[qid];
+ if (vring->kickfd != -1)
+ {
+ unix_file_t *uf = pool_elt_at_index (unix_main.file_pool,
+ vring->kickfd_idx);
+ unix_file_del (&unix_main, uf);
+ }
+ if (vring->callfd != -1)
+ {
+ unix_file_t *uf = pool_elt_at_index (unix_main.file_pool,
+ vring->callfd_idx);
+ unix_file_del (&unix_main, uf);
+ }
+ if (vring->errfd != -1)
+ close (vring->errfd);
+ vhost_user_vring_init (vui, qid);
+}
+
static inline void
vhost_user_if_disconnect (vhost_user_intf_t * vui)
{
@@ -310,26 +562,9 @@ vhost_user_if_disconnect (vhost_user_intf_t * vui)
hash_unset (vum->vhost_user_interface_index_by_listener_fd, vui->unix_fd);
vui->unix_fd = -1;
vui->is_up = 0;
- for (q = 0; q < vui->num_vrings; q++)
- {
- if (vui->vrings[q].callfd > -1)
- {
- unix_file_t *uf = pool_elt_at_index (unix_main.file_pool,
- vui->vrings[q].callfd_idx);
- unix_file_del (&unix_main, uf);
- }
- if (vui->vrings[q].kickfd > -1)
- close (vui->vrings[q].kickfd);
-
- vui->vrings[q].callfd = -1;
- vui->vrings[q].kickfd = -1;
- vui->vrings[q].desc = NULL;
- vui->vrings[q].avail = NULL;
- vui->vrings[q].used = NULL;
- vui->vrings[q].log_guest_addr = 0;
- vui->vrings[q].log_used = 0;
- }
+ for (q = 0; q < VHOST_VRING_MAX_N; q++)
+ vhost_user_vring_close (vui, q);
unmap_all_mem_regions (vui);
DBG_SOCK ("interface ifindex %d disconnected", vui->sw_if_index);
@@ -411,11 +646,26 @@ vhost_user_socket_read (unix_file_t * uf)
n = recvmsg (uf->file_descriptor, &mh, 0);
+ /* Stop workers to avoid end of the world */
+ vlib_worker_thread_barrier_sync (vlib_get_main ());
+
if (n != VHOST_USER_MSG_HDR_SZ)
- goto close_socket;
+ {
+ if (n == -1)
+ {
+ DBG_SOCK ("recvmsg returned error %d %s", errno, strerror (errno));
+ }
+ else
+ {
+ DBG_SOCK ("n (%d) != VHOST_USER_MSG_HDR_SZ (%d)",
+ n, VHOST_USER_MSG_HDR_SZ);
+ }
+ goto close_socket;
+ }
if (mh.msg_flags & MSG_CTRUNC)
{
+ DBG_SOCK ("MSG_CTRUNC is set");
goto close_socket;
}
@@ -438,27 +688,39 @@ vhost_user_socket_read (unix_file_t * uf)
}
{
- int rv __attribute__ ((unused));
- /* $$$$ pay attention to rv */
- rv = read (uf->file_descriptor, ((char *) &msg) + n, msg.size);
+ int rv;
+ rv =
+ read (uf->file_descriptor, ((char *) &msg) + VHOST_USER_MSG_HDR_SZ,
+ msg.size);
+ if (rv < 0)
+ {
+ DBG_SOCK ("read failed %s", strerror (errno));
+ goto close_socket;
+ }
+ else if (rv != msg.size)
+ {
+ DBG_SOCK ("message too short (read %dB should be %dB)", rv, msg.size);
+ goto close_socket;
+ }
}
switch (msg.request)
{
case VHOST_USER_GET_FEATURES:
- DBG_SOCK ("if %d msg VHOST_USER_GET_FEATURES", vui->hw_if_index);
-
msg.flags |= 4;
- msg.u64 = (1 << FEAT_VIRTIO_NET_F_MRG_RXBUF) |
- (1 << FEAT_VIRTIO_F_ANY_LAYOUT) |
- (1 << FEAT_VIRTIO_F_INDIRECT_DESC) |
- (1 << FEAT_VHOST_F_LOG_ALL) |
- (1 << FEAT_VIRTIO_NET_F_GUEST_ANNOUNCE) |
- (1 << FEAT_VHOST_USER_F_PROTOCOL_FEATURES) |
- (1UL << FEAT_VIRTIO_F_VERSION_1);
+ msg.u64 = (1ULL << FEAT_VIRTIO_NET_F_MRG_RXBUF) |
+ (1ULL << FEAT_VIRTIO_NET_F_CTRL_VQ) |
+ (1ULL << FEAT_VIRTIO_F_ANY_LAYOUT) |
+ (1ULL << FEAT_VIRTIO_F_INDIRECT_DESC) |
+ (1ULL << FEAT_VHOST_F_LOG_ALL) |
+ (1ULL << FEAT_VIRTIO_NET_F_GUEST_ANNOUNCE) |
+ (1ULL << FEAT_VIRTIO_NET_F_MQ) |
+ (1ULL << FEAT_VHOST_USER_F_PROTOCOL_FEATURES) |
+ (1ULL << FEAT_VIRTIO_F_VERSION_1);
msg.u64 &= vui->feature_mask;
-
msg.size = sizeof (msg.u64);
+ DBG_SOCK ("if %d msg VHOST_USER_GET_FEATURES - reply 0x%016llx",
+ vui->hw_if_index, msg.u64);
break;
case VHOST_USER_SET_FEATURES:
@@ -467,7 +729,9 @@ vhost_user_socket_read (unix_file_t * uf)
vui->features = msg.u64;
- if (vui->features & (1 << FEAT_VIRTIO_NET_F_MRG_RXBUF))
+ if (vui->features &
+ ((1 << FEAT_VIRTIO_NET_F_MRG_RXBUF) |
+ (1ULL << FEAT_VIRTIO_F_VERSION_1)))
vui->virtio_net_hdr_sz = 12;
else
vui->virtio_net_hdr_sz = 10;
@@ -479,16 +743,8 @@ vhost_user_socket_read (unix_file_t * uf)
vnet_hw_interface_set_flags (vnm, vui->hw_if_index, 0);
vui->is_up = 0;
- for (q = 0; q < 2; q++)
- {
- vui->vrings[q].desc = 0;
- vui->vrings[q].avail = 0;
- vui->vrings[q].used = 0;
- vui->vrings[q].log_guest_addr = 0;
- vui->vrings[q].log_used = 0;
- }
-
- DBG_SOCK ("interface %d disconnected", vui->sw_if_index);
+ /*for (q = 0; q < VHOST_VRING_MAX_N; q++)
+ vhost_user_vring_close(&vui->vrings[q]); */
break;
@@ -552,7 +808,7 @@ vhost_user_socket_read (unix_file_t * uf)
if ((msg.state.num > 32768) || /* maximum ring size is 32768 */
(msg.state.num == 0) || /* it cannot be zero */
- (msg.state.num % 2)) /* must be power of 2 */
+ ((msg.state.num - 1) & msg.state.num)) /* must be power of 2 */
goto close_socket;
vui->vrings[msg.state.index].qsz = msg.state.num;
break;
@@ -561,6 +817,20 @@ vhost_user_socket_read (unix_file_t * uf)
DBG_SOCK ("if %d msg VHOST_USER_SET_VRING_ADDR idx %d",
vui->hw_if_index, msg.state.index);
+ if (msg.state.index >= VHOST_VRING_MAX_N)
+ {
+ DBG_SOCK ("invalid vring index VHOST_USER_SET_VRING_ADDR:"
+ " %d >= %d", msg.state.index, VHOST_VRING_MAX_N);
+ goto close_socket;
+ }
+
+ if (msg.size < sizeof (msg.addr))
+ {
+ DBG_SOCK ("vhost message is too short (%d < %d)",
+ msg.size, sizeof (msg.addr));
+ goto close_socket;
+ }
+
vui->vrings[msg.state.index].desc = (vring_desc_t *)
map_user_mem (vui, msg.addr.desc_user_addr);
vui->vrings[msg.state.index].used = (vring_used_t *)
@@ -583,7 +853,6 @@ vhost_user_socket_read (unix_file_t * uf)
/* Spec says: If VHOST_USER_F_PROTOCOL_FEATURES has not been negotiated,
the ring is initialized in an enabled state. */
-
if (!(vui->features & (1 << FEAT_VHOST_USER_F_PROTOCOL_FEATURES)))
{
vui->vrings[msg.state.index].enabled = 1;
@@ -594,7 +863,7 @@ vhost_user_socket_read (unix_file_t * uf)
vui->vrings[msg.state.index].used->idx;
/* tell driver that we don't want interrupts */
- vui->vrings[msg.state.index].used->flags |= 1;
+ vui->vrings[msg.state.index].used->flags = VRING_USED_F_NO_NOTIFY;
break;
case VHOST_USER_SET_OWNER:
@@ -611,18 +880,22 @@ vhost_user_socket_read (unix_file_t * uf)
q = (u8) (msg.u64 & 0xFF);
+ /* if there is old fd, delete and close it */
+ if (vui->vrings[q].callfd != -1)
+ {
+ unix_file_t *uf = pool_elt_at_index (unix_main.file_pool,
+ vui->vrings[q].callfd_idx);
+ unix_file_del (&unix_main, uf);
+ }
+
if (!(msg.u64 & 0x100))
{
if (number_of_fds != 1)
- goto close_socket;
-
- /* if there is old fd, delete it */
- if (vui->vrings[q].callfd > -1)
{
- unix_file_t *uf = pool_elt_at_index (unix_main.file_pool,
- vui->vrings[q].callfd_idx);
- unix_file_del (&unix_main, uf);
+ DBG_SOCK ("More than one fd received !");
+ goto close_socket;
}
+
vui->vrings[q].callfd = fds[0];
template.read_function = vhost_user_callfd_read_ready;
template.file_descriptor = fds[0];
@@ -638,18 +911,39 @@ vhost_user_socket_read (unix_file_t * uf)
q = (u8) (msg.u64 & 0xFF);
+ if (vui->vrings[q].kickfd != -1)
+ {
+ unix_file_t *uf = pool_elt_at_index (unix_main.file_pool,
+ vui->vrings[q].kickfd);
+ unix_file_del (&unix_main, uf);
+ }
+
if (!(msg.u64 & 0x100))
{
if (number_of_fds != 1)
- goto close_socket;
+ {
+ DBG_SOCK ("More than one fd received !");
+ goto close_socket;
+ }
if (vui->vrings[q].kickfd > -1)
close (vui->vrings[q].kickfd);
vui->vrings[q].kickfd = fds[0];
+ template.read_function = vhost_user_kickfd_read_ready;
+ template.file_descriptor = fds[0];
+ template.private_data =
+ ((vui - vhost_user_main.vhost_user_interfaces) << 8) + q;
+ vui->vrings[q].kickfd_idx = unix_file_add (&unix_main, &template);
}
else
- vui->vrings[q].kickfd = -1;
+ {
+ vui->vrings[q].kickfd = -1;
+ vui->vrings[q].started = 1;
+ }
+
+ //TODO: When kickfd is specified, 'started' is set when the first kick
+ //is received.
break;
case VHOST_USER_SET_VRING_ERR:
@@ -658,17 +952,19 @@ vhost_user_socket_read (unix_file_t * uf)
q = (u8) (msg.u64 & 0xFF);
+ if (vui->vrings[q].errfd != -1)
+ close (vui->vrings[q].errfd);
+
if (!(msg.u64 & 0x100))
{
if (number_of_fds != 1)
goto close_socket;
- fd = fds[0];
+ vui->vrings[q].errfd = fds[0];
}
else
- fd = -1;
+ vui->vrings[q].errfd = -1;
- vui->vrings[q].errfd = fd;
break;
case VHOST_USER_SET_VRING_BASE:
@@ -682,8 +978,15 @@ vhost_user_socket_read (unix_file_t * uf)
DBG_SOCK ("if %d msg VHOST_USER_GET_VRING_BASE idx %d num %d",
vui->hw_if_index, msg.state.index, msg.state.num);
+ if (msg.state.index >= VHOST_VRING_MAX_N)
+ {
+ DBG_SOCK ("invalid vring index VHOST_USER_GET_VRING_BASE:"
+ " %d >= %d", msg.state.index, VHOST_VRING_MAX_N);
+ goto close_socket;
+ }
+
/* Spec says: Client must [...] stop ring upon receiving VHOST_USER_GET_VRING_BASE. */
- vui->vrings[msg.state.index].enabled = 0;
+ vhost_user_vring_close (vui, msg.state.index);
msg.state.num = vui->vrings[msg.state.index].last_avail_idx;
msg.flags |= 4;
@@ -753,7 +1056,8 @@ vhost_user_socket_read (unix_file_t * uf)
vui->hw_if_index);
msg.flags |= 4;
- msg.u64 = (1 << VHOST_USER_PROTOCOL_F_LOG_SHMFD);
+ msg.u64 = (1 << VHOST_USER_PROTOCOL_F_LOG_SHMFD) |
+ (1 << VHOST_USER_PROTOCOL_F_MQ);
msg.size = sizeof (msg.u64);
break;
@@ -765,9 +1069,24 @@ vhost_user_socket_read (unix_file_t * uf)
break;
+ case VHOST_USER_GET_QUEUE_NUM:
+ DBG_SOCK ("if %d msg VHOST_USER_GET_QUEUE_NUM", vui->hw_if_index);
+ msg.flags |= 4;
+ msg.u64 = VHOST_VRING_MAX_N;
+ msg.size = sizeof (msg.u64);
+ break;
+
case VHOST_USER_SET_VRING_ENABLE:
- DBG_SOCK ("if %d VHOST_USER_SET_VRING_ENABLE, enable: %d",
- vui->hw_if_index, msg.state.num);
+ DBG_SOCK ("if %d VHOST_USER_SET_VRING_ENABLE: %s queue %d",
+ vui->hw_if_index, msg.state.num ? "enable" : "disable",
+ msg.state.index);
+ if (msg.state.index >= VHOST_VRING_MAX_N)
+ {
+ DBG_SOCK ("invalid vring index VHOST_USER_SET_VRING_ENABLE:"
+ " %d >= %d", msg.state.index, VHOST_VRING_MAX_N);
+ goto close_socket;
+ }
+
vui->vrings[msg.state.index].enabled = msg.state.num;
break;
@@ -777,39 +1096,33 @@ vhost_user_socket_read (unix_file_t * uf)
goto close_socket;
}
- /* if we have pointers to descriptor table, go up */
- if (!vui->is_up &&
- vui->vrings[VHOST_NET_VRING_IDX_TX].desc &&
- vui->vrings[VHOST_NET_VRING_IDX_RX].desc)
- {
-
- DBG_SOCK ("interface %d connected", vui->sw_if_index);
-
- vnet_hw_interface_set_flags (vnm, vui->hw_if_index,
- VNET_HW_INTERFACE_FLAG_LINK_UP);
- vui->is_up = 1;
-
- }
-
/* if we need to reply */
if (msg.flags & 4)
{
n =
send (uf->file_descriptor, &msg, VHOST_USER_MSG_HDR_SZ + msg.size, 0);
if (n != (msg.size + VHOST_USER_MSG_HDR_SZ))
- goto close_socket;
+ {
+ DBG_SOCK ("could not send message response");
+ goto close_socket;
+ }
}
+ vhost_user_update_iface_state (vui);
+ vlib_worker_thread_barrier_release (vlib_get_main ());
return 0;
close_socket:
vhost_user_if_disconnect (vui);
+ vhost_user_update_iface_state (vui);
+ vlib_worker_thread_barrier_release (vlib_get_main ());
return 0;
}
static clib_error_t *
vhost_user_socket_error (unix_file_t * uf)
{
+ vlib_main_t *vm = vlib_get_main ();
vhost_user_main_t *vum = &vhost_user_main;
vhost_user_intf_t *vui;
uword *p;
@@ -824,7 +1137,11 @@ vhost_user_socket_error (unix_file_t * uf)
else
vui = vec_elt_at_index (vum->vhost_user_interfaces, p[0]);
+ DBG_SOCK ("socket error on if %d", vui->sw_if_index);
+ vlib_worker_thread_barrier_sync (vm);
vhost_user_if_disconnect (vui);
+ vhost_user_rx_thread_placement ();
+ vlib_worker_thread_barrier_release (vm);
return 0;
}
@@ -856,6 +1173,7 @@ vhost_user_socksvr_accept_ready (unix_file_t * uf)
if (client_fd < 0)
return clib_error_return_unix (0, "accept");
+ DBG_SOCK ("New client socket for vhost interface %d", vui->sw_if_index);
template.read_function = vhost_user_socket_read;
template.error_function = vhost_user_socket_error;
template.file_descriptor = client_fd;
@@ -892,6 +1210,7 @@ vhost_user_init (vlib_main_t * vm)
vec_validate_aligned (vum->rx_buffers, tm->n_vlib_mains - 1,
CLIB_CACHE_LINE_BYTES);
+ vec_validate (vum->cpus, tm->n_vlib_mains - 1);
/* find out which cpus will be used for input */
vum->input_cpu_first_index = 0;
@@ -921,8 +1240,8 @@ VLIB_MAIN_LOOP_EXIT_FUNCTION (vhost_user_exit);
typedef struct
{
- u16 virtqueue;
- u16 device_index;
+ u16 qid; /** The interface queue index (Not the virtio vring idx) */
+ u16 device_index; /** The device index */
#if VHOST_USER_COPY_TX_HDR == 1
virtio_net_hdr_t hdr;
#endif
@@ -945,8 +1264,8 @@ format_vhost_user_input_trace (u8 * s, va_list * va)
uword indent = format_get_indent (s);
#endif
- s = format (s, "%U virtqueue %d",
- format_vnet_sw_interface_name, vnm, sw, t->virtqueue);
+ s = format (s, "%U queue %d",
+ format_vnet_sw_interface_name, vnm, sw, t->qid);
#if VHOST_USER_COPY_TX_HDR == 1
s = format (s, "\n%Uvirtio_net_hdr flags 0x%02x gso_type %u hdr_len %u",
@@ -960,15 +1279,17 @@ format_vhost_user_input_trace (u8 * s, va_list * va)
void
vhost_user_rx_trace (vlib_main_t * vm,
vlib_node_runtime_t * node,
- vhost_user_intf_t * vui, i16 virtqueue)
+ vhost_user_intf_t * vui, u16 qid)
{
u32 *b, n_left;
vhost_user_main_t *vum = &vhost_user_main;
u32 next_index = VNET_DEVICE_INPUT_NEXT_ETHERNET_INPUT;
+ u32 cpu_index = os_get_cpu_number ();
- n_left = vec_len (vui->d_trace_buffers);
- b = vui->d_trace_buffers;
+ vhost_cpu_t *vhc = vec_elt_at_index (vum->cpus, cpu_index);
+ n_left = vec_len (vhc->trace_buffers);
+ b = vhc->trace_buffers;
while (n_left >= 1)
{
@@ -982,7 +1303,7 @@ vhost_user_rx_trace (vlib_main_t * vm,
b0 = vlib_get_buffer (vm, bi0);
vlib_trace_buffer (vm, node, next_index, b0, /* follow_chain */ 0);
t0 = vlib_add_trace (vm, node, b0, sizeof (t0[0]));
- t0->virtqueue = virtqueue;
+ t0->qid = qid;
t0->device_index = vui - vum->vhost_user_interfaces;
#if VHOST_USER_COPY_TX_HDR == 1
clib_memcpy (&t0->hdr, b0->pre_data, sizeof (virtio_net_hdr_t));
@@ -1008,10 +1329,11 @@ vhost_user_send_call (vlib_main_t * vm, vhost_user_vring_t * vq)
static u32
vhost_user_if_input (vlib_main_t * vm,
vhost_user_main_t * vum,
- vhost_user_intf_t * vui, vlib_node_runtime_t * node)
+ vhost_user_intf_t * vui,
+ u16 qid, vlib_node_runtime_t * node)
{
- vhost_user_vring_t *txvq = &vui->vrings[VHOST_NET_VRING_IDX_TX];
- vhost_user_vring_t *rxvq = &vui->vrings[VHOST_NET_VRING_IDX_RX];
+ vhost_user_vring_t *txvq = &vui->vrings[VHOST_VRING_IDX_TX (qid)];
+ vhost_user_vring_t *rxvq = &vui->vrings[VHOST_VRING_IDX_RX (qid)];
uword n_rx_packets = 0, n_rx_bytes = 0;
uword n_left;
u32 n_left_to_next, *to_next;
@@ -1020,26 +1342,19 @@ vhost_user_if_input (vlib_main_t * vm,
uword n_trace = vlib_get_trace_count (vm, node);
u16 qsz_mask;
u32 cpu_index, rx_len, drops, flush;
+ vhost_cpu_t *vhc;
f64 now = vlib_time_now (vm);
u32 map_guest_hint_desc = 0;
u32 map_guest_hint_indirect = 0;
u32 *map_guest_hint_p = &map_guest_hint_desc;
- vec_reset_length (vui->d_trace_buffers);
-
- /* no descriptor ptr - bail out */
- if (PREDICT_FALSE (!txvq->desc || !txvq->avail || !txvq->enabled))
- return 0;
-
- /* do we have pending intterupts ? */
+ /* do we have pending interrupts ? */
if ((txvq->n_since_last_int) && (txvq->int_deadline < now))
vhost_user_send_call (vm, txvq);
if ((rxvq->n_since_last_int) && (rxvq->int_deadline < now))
vhost_user_send_call (vm, rxvq);
- /* only bit 0 of avail.flags is used so we don't want to deal with this
- interface if any other bit is set */
if (PREDICT_FALSE (txvq->avail->flags & 0xFFFE))
return 0;
@@ -1049,16 +1364,16 @@ vhost_user_if_input (vlib_main_t * vm,
if (PREDICT_FALSE (n_left == 0))
return 0;
- if (PREDICT_FALSE (n_left == txvq->qsz))
+ if (PREDICT_FALSE (!vui->admin_up || !(txvq->enabled)))
{
- //Informational error logging when VPP is not receiving packets fast enough
- vlib_error_count (vm, node->node_index,
- VHOST_USER_INPUT_FUNC_ERROR_FULL_RX_QUEUE, 1);
- }
+ /*
+ * Discard input packet if interface is admin down or vring is not
+ * enabled.
+ * "For example, for a networking device, in the disabled state
+ * client must not supply any new RX packets, but must process
+ * and discard any TX packets."
+ */
- if (PREDICT_FALSE (!vui->admin_up))
- {
- /* if intf is admin down, just drop all packets waiting in the ring */
txvq->last_avail_idx = txvq->last_used_idx = txvq->avail->idx;
CLIB_MEMORY_BARRIER ();
txvq->used->idx = txvq->last_used_idx;
@@ -1067,8 +1382,16 @@ vhost_user_if_input (vlib_main_t * vm,
return 0;
}
+ if (PREDICT_FALSE (n_left == txvq->qsz))
+ {
+ //Informational error logging when VPP is not receiving packets fast enough
+ vlib_error_count (vm, node->node_index,
+ VHOST_USER_INPUT_FUNC_ERROR_FULL_RX_QUEUE, 1);
+ }
+
qsz_mask = txvq->qsz - 1;
cpu_index = os_get_cpu_number ();
+ vhc = vec_elt_at_index (vum->cpus, cpu_index);
drops = 0;
flush = 0;
@@ -1265,7 +1588,7 @@ vhost_user_if_input (vlib_main_t * vm,
b_head->error = node->errors[error];
if (PREDICT_FALSE (n_trace > n_rx_packets))
- vec_add1 (vui->d_trace_buffers, bi_head);
+ vec_add1 (vhc->trace_buffers, bi_head);
if (PREDICT_FALSE (error))
{
@@ -1293,10 +1616,18 @@ vhost_user_if_input (vlib_main_t * vm,
to_next, n_left_to_next,
bi_head, next0);
n_left--;
+ if (PREDICT_FALSE (!n_left))
+ {
+ // I NEED SOME MORE !
+ u32 remain = (u16) (txvq->avail->idx - txvq->last_avail_idx);
+ remain = (remain > VLIB_FRAME_SIZE - n_rx_packets) ?
+ VLIB_FRAME_SIZE - n_rx_packets : remain;
+ remain = (remain > rx_len) ? rx_len : remain;
+ n_left = remain;
+ }
}
vlib_put_next_frame (vm, node, next_index, n_left_to_next);
-
}
if (PREDICT_TRUE (vum->rx_buffers[cpu_index] != 0))
@@ -1307,13 +1638,6 @@ vhost_user_if_input (vlib_main_t * vm,
txvq->used->idx = txvq->last_used_idx;
vhost_user_log_dirty_ring (vui, txvq, idx);
- if (PREDICT_FALSE (vec_len (vui->d_trace_buffers) > 0))
- {
- vhost_user_rx_trace (vm, node, vui, VHOST_NET_VRING_IDX_TX);
- vlib_set_trace_count (vm, node,
- n_trace - vec_len (vui->d_trace_buffers));
- }
-
/* interrupt (call) handling */
if ((txvq->callfd > -1) && !(txvq->avail->flags & 1))
{
@@ -1323,6 +1647,13 @@ vhost_user_if_input (vlib_main_t * vm,
vhost_user_send_call (vm, txvq);
}
+ if (PREDICT_FALSE (vec_len (vhc->trace_buffers) > 0))
+ {
+ vhost_user_rx_trace (vm, node, vui, qid);
+ vlib_set_trace_count (vm, node, n_trace - vec_len (vhc->trace_buffers));
+ vec_reset_length (vhc->trace_buffers);
+ }
+
if (PREDICT_FALSE (drops))
{
vlib_increment_simple_counter
@@ -1345,21 +1676,20 @@ vhost_user_input (vlib_main_t * vm,
vlib_node_runtime_t * node, vlib_frame_t * f)
{
vhost_user_main_t *vum = &vhost_user_main;
- u32 cpu_index = os_get_cpu_number ();
- vhost_user_intf_t *vui;
uword n_rx_packets = 0;
- int i;
+ u32 cpu_index = os_get_cpu_number ();
- for (i = 0; i < vec_len (vum->vhost_user_interfaces); i++)
- {
- vui = vec_elt_at_index (vum->vhost_user_interfaces, i);
- if (vui->is_up)
- {
- if ((i % vum->input_cpu_count) ==
- (cpu_index - vum->input_cpu_first_index))
- n_rx_packets += vhost_user_if_input (vm, vum, vui, node);
- }
- }
+
+ vhost_iface_and_queue_t *vhiq;
+ vec_foreach (vhiq, vum->cpus[cpu_index].rx_queues)
+ {
+ vhost_user_intf_t *vui =
+ &vum->vhost_user_interfaces[vhiq->vhost_iface_index];
+ n_rx_packets += vhost_user_if_input (vm, vum, vui, vhiq->qid, node);
+ }
+
+ //TODO: One call might return more than 256 packets here.
+ //But this is supposed to be the vector size.
return n_rx_packets;
}
@@ -1396,38 +1726,27 @@ vhost_user_intfc_tx (vlib_main_t * vm,
vnet_interface_output_runtime_t *rd = (void *) node->runtime_data;
vhost_user_intf_t *vui =
vec_elt_at_index (vum->vhost_user_interfaces, rd->dev_instance);
- vhost_user_vring_t *rxvq = &vui->vrings[VHOST_NET_VRING_IDX_RX];
+ u32 qid = ~0;
+ vhost_user_vring_t *rxvq;
u16 qsz_mask;
u8 error = VHOST_USER_TX_FUNC_ERROR_NONE;
-
+ u32 cpu_index = os_get_cpu_number ();
n_left = n_packets = frame->n_vectors;
u32 map_guest_hint_desc = 0;
u32 map_guest_hint_indirect = 0;
u32 *map_guest_hint_p = &map_guest_hint_desc;
- if (PREDICT_FALSE (!vui->is_up))
- goto done2;
-
- if (PREDICT_FALSE
- (!rxvq->desc || !rxvq->avail || vui->sock_errno != 0 || !rxvq->enabled))
+ if (PREDICT_FALSE (!vui->is_up || !vui->admin_up))
{
error = VHOST_USER_TX_FUNC_ERROR_NOT_READY;
- goto done2;
+ goto done3;
}
- if (PREDICT_FALSE (vui->lockp != 0))
- {
- while (__sync_lock_test_and_set (vui->lockp, 1))
- ;
- }
-
- /* only bit 0 of avail.flags is used so we don't want to deal with this
- interface if any other bit is set */
- if (PREDICT_FALSE (rxvq->avail->flags & 0xFFFE))
- {
- error = VHOST_USER_TX_FUNC_ERROR_NOT_READY;
- goto done2;
- }
+ qid =
+ VHOST_VRING_IDX_RX (*vec_elt_at_index (vui->per_cpu_tx_qid, cpu_index));
+ rxvq = &vui->vrings[qid];
+ if (PREDICT_FALSE (vui->use_tx_spinlock))
+ vhost_user_vring_lock (vui, qid);
if (PREDICT_FALSE ((rxvq->avail->idx == rxvq->last_avail_idx)))
{
@@ -1624,6 +1943,7 @@ vhost_user_intfc_tx (vlib_main_t * vm,
bytes_left - bytes_to_copy,
bytes_to_copy);
+ CLIB_PREFETCH (rxvq, sizeof (*rxvq), STORE);
bytes_left -= bytes_to_copy;
buffer_len -= bytes_to_copy;
buffer_addr += bytes_to_copy;
@@ -1657,10 +1977,9 @@ done:
}
done2:
+ vhost_user_vring_unlock (vui, qid);
- if (PREDICT_FALSE (vui->lockp != 0))
- *vui->lockp = 0;
-
+done3:
if (PREDICT_FALSE (n_left && error != VHOST_USER_TX_FUNC_ERROR_NONE))
{
vlib_error_count (vm, node->node_index, error, n_left);
@@ -1780,7 +2099,10 @@ vhost_user_process (vlib_main_t * vm,
getsockopt (vui->unix_fd, SOL_SOCKET, SO_ERROR, &error, &len);
if (retval)
- vhost_user_if_disconnect (vui);
+ {
+ DBG_SOCK ("getsockopt returned %d", retval);
+ vhost_user_if_disconnect (vui);
+ }
}
}
}
@@ -1902,6 +2224,7 @@ vhost_user_vui_new ()
// vui was not retrieved from inactive ifaces - create new
if (!vui)
vec_add2 (vum->vhost_user_interfaces, vui, 1);
+
return vui;
}
@@ -1953,12 +2276,10 @@ vhost_user_vui_init (vnet_main_t * vnm,
{
vnet_sw_interface_t *sw;
sw = vnet_get_hw_sw_interface (vnm, vui->hw_if_index);
- vlib_thread_main_t *tm = vlib_get_thread_main ();
int q;
vui->unix_fd = sockfd;
vui->sw_if_index = sw->sw_if_index;
- vui->num_vrings = 2;
vui->sock_is_server = is_server;
strncpy (vui->sock_filename, sock_filename,
ARRAY_LEN (vui->sock_filename) - 1);
@@ -1969,24 +2290,24 @@ vhost_user_vui_init (vnet_main_t * vnm,
vui->unix_file_index = ~0;
vui->log_base_addr = 0;
- for (q = 0; q < 2; q++)
- {
- vui->vrings[q].enabled = 0;
- vui->vrings[q].callfd = -1;
- vui->vrings[q].kickfd = -1;
- }
+ for (q = 0; q < VHOST_VRING_MAX_N; q++)
+ vhost_user_vring_init (vui, q);
vnet_hw_interface_set_flags (vnm, vui->hw_if_index, 0);
if (sw_if_index)
*sw_if_index = vui->sw_if_index;
- if (tm->n_vlib_mains > 1)
+ for (q = 0; q < VHOST_VRING_MAX_N; q++)
{
- vui->lockp = clib_mem_alloc_aligned (CLIB_CACHE_LINE_BYTES,
- CLIB_CACHE_LINE_BYTES);
- memset ((void *) vui->lockp, 0, CLIB_CACHE_LINE_BYTES);
+ vui->vring_locks[q] = clib_mem_alloc_aligned (CLIB_CACHE_LINE_BYTES,
+ CLIB_CACHE_LINE_BYTES);
+ memset ((void *) vui->vring_locks[q], 0, CLIB_CACHE_LINE_BYTES);
}
+
+ vec_validate (vui->per_cpu_tx_qid,
+ vlib_get_thread_main ()->n_vlib_mains - 1);
+ vhost_user_tx_thread_placement (vui);
}
// register vui and start polling on it
@@ -2117,7 +2438,7 @@ vhost_user_connect_command_fn (vlib_main_t * vm,
u8 *sock_filename = NULL;
u32 sw_if_index;
u8 is_server = 0;
- u64 feature_mask = (u64) ~ 0;
+ u64 feature_mask = (u64) ~ (0ULL);
u8 renumber = 0;
u32 custom_dev_instance = ~0;
u8 hwaddr[6];
@@ -2259,6 +2580,10 @@ show_vhost_user_command_fn (vlib_main_t * vm,
vhost_user_intf_t *vui;
u32 hw_if_index, *hw_if_indices = 0;
vnet_hw_interface_t *hi;
+ vhost_cpu_t *vhc;
+ vhost_iface_and_queue_t *vhiq;
+ u32 ci;
+
int i, j, q;
int show_descr = 0;
struct feat_struct
@@ -2275,13 +2600,23 @@ show_vhost_user_command_fn (vlib_main_t * vm,
{.str = NULL}
};
+#define foreach_protocol_feature \
+ _(VHOST_USER_PROTOCOL_F_MQ) \
+ _(VHOST_USER_PROTOCOL_F_LOG_SHMFD)
+
+ static struct feat_struct proto_feat_array[] = {
+#define _(s) { .str = #s, .bit = s},
+ foreach_protocol_feature
+#undef _
+ {.str = NULL}
+ };
+
while (unformat_check_input (input) != UNFORMAT_END_OF_INPUT)
{
if (unformat
(input, "%U", unformat_vnet_hw_interface, vnm, &hw_if_index))
{
vec_add1 (hw_if_indices, hw_if_index);
- vlib_cli_output (vm, "add %d", hw_if_index);
}
else if (unformat (input, "descriptors") || unformat (input, "desc"))
show_descr = 1;
@@ -2301,7 +2636,7 @@ show_vhost_user_command_fn (vlib_main_t * vm,
}
}
vlib_cli_output (vm, "Virtio vhost-user interfaces");
- vlib_cli_output (vm, "Global:\n coalesce frames %d time %e\n\n",
+ vlib_cli_output (vm, "Global:\n coalesce frames %d time %e",
vum->coalesce_frames, vum->coalesce_time);
for (i = 0; i < vec_len (hw_if_indices); i++)
@@ -2311,26 +2646,61 @@ show_vhost_user_command_fn (vlib_main_t * vm,
vlib_cli_output (vm, "Interface: %s (ifindex %d)",
hi->name, hw_if_indices[i]);
- vlib_cli_output (vm, "virtio_net_hdr_sz %d\n features (0x%llx): \n",
- vui->virtio_net_hdr_sz, vui->features);
+ vlib_cli_output (vm, "virtio_net_hdr_sz %d\n"
+ " features mask (0x%llx): \n"
+ " features (0x%llx): \n",
+ vui->virtio_net_hdr_sz, vui->feature_mask,
+ vui->features);
feat_entry = (struct feat_struct *) &feat_array;
while (feat_entry->str)
{
- if (vui->features & (1 << feat_entry->bit))
+ if (vui->features & (1ULL << feat_entry->bit))
vlib_cli_output (vm, " %s (%d)", feat_entry->str,
feat_entry->bit);
feat_entry++;
}
- vlib_cli_output (vm, "\n");
+ vlib_cli_output (vm, " protocol features (0x%llx)",
+ vui->protocol_features);
+ feat_entry = (struct feat_struct *) &proto_feat_array;
+ while (feat_entry->str)
+ {
+ if (vui->protocol_features & (1ULL << feat_entry->bit))
+ vlib_cli_output (vm, " %s (%d)", feat_entry->str,
+ feat_entry->bit);
+ feat_entry++;
+ }
+ vlib_cli_output (vm, "\n");
vlib_cli_output (vm, " socket filename %s type %s errno \"%s\"\n\n",
vui->sock_filename,
vui->sock_is_server ? "server" : "client",
strerror (vui->sock_errno));
+ vlib_cli_output (vm, " rx placement: ");
+ vec_foreach (vhc, vum->cpus)
+ {
+ vec_foreach (vhiq, vhc->rx_queues)
+ {
+ if (vhiq->vhost_iface_index == vui - vum->vhost_user_interfaces)
+ vlib_cli_output (vm, " thread %d on vring %d\n",
+ vhc - vum->cpus, VHOST_VRING_IDX_TX (vhiq->qid));
+ }
+ }
+
+ vlib_cli_output (vm, " tx placement: %s\n",
+ vui->use_tx_spinlock ? "spin-lock" : "lock-free");
+
+ vec_foreach_index (ci, vui->per_cpu_tx_qid)
+ {
+ vlib_cli_output (vm, " thread %d on vring %d\n", ci,
+ VHOST_VRING_IDX_RX (vui->per_cpu_tx_qid[ci]));
+ }
+
+ vlib_cli_output (vm, "\n");
+
vlib_cli_output (vm, " Memory regions (total %d)\n", vui->nregions);
if (vui->nregions)
@@ -2351,9 +2721,14 @@ show_vhost_user_command_fn (vlib_main_t * vm,
vui->regions[j].mmap_offset,
pointer_to_uword (vui->region_mmap_addr[j]));
}
- for (q = 0; q < vui->num_vrings; q++)
+ for (q = 0; q < VHOST_VRING_MAX_N; q++)
{
- vlib_cli_output (vm, "\n Virtqueue %d\n", q);
+ if (!vui->vrings[q].started)
+ continue;
+
+ vlib_cli_output (vm, "\n Virtqueue %d (%s%s)\n", q,
+ (q & 1) ? "RX" : "TX",
+ vui->vrings[q].enabled ? "" : " disabled");
vlib_cli_output (vm,
" qsz %d last_avail_idx %d last_used_idx %d\n",
@@ -2465,6 +2840,48 @@ vhost_user_unmap_all (void)
}
}
+static clib_error_t *
+vhost_thread_command_fn (vlib_main_t * vm,
+ unformat_input_t * input, vlib_cli_command_t * cmd)
+{
+ unformat_input_t _line_input, *line_input = &_line_input;
+ u32 worker_thread_index;
+ u32 sw_if_index;
+ u8 del = 0;
+ int rv;
+
+ /* Get a line of input. */
+ if (!unformat_user (input, unformat_line_input, line_input))
+ return 0;
+
+ if (!unformat
+ (line_input, "%U %d", unformat_vnet_sw_interface, vnet_get_main (),
+ &sw_if_index, &worker_thread_index))
+ {
+ unformat_free (line_input);
+ return clib_error_return (0, "unknown input `%U'",
+ format_unformat_error, input);
+ }
+
+ if (unformat (line_input, "del"))
+ del = 1;
+
+ if ((rv =
+ vhost_user_thread_placement (sw_if_index, worker_thread_index, del)))
+ return clib_error_return (0, "vhost_user_thread_placement returned %d",
+ rv);
+ return 0;
+}
+
+
+/* *INDENT-OFF* */
+VLIB_CLI_COMMAND (vhost_user_thread_command, static) = {
+ .path = "vhost thread",
+ .short_help = "vhost thread <iface> <worker-index> [del]",
+ .function = vhost_thread_command_fn,
+};
+/* *INDENT-ON* */
+
/*
* fd.io coding-style-patch-verification: ON
*
diff --git a/vnet/vnet/devices/virtio/vhost-user.h b/vnet/vnet/devices/virtio/vhost-user.h
index 47fad1d6465..69bbda14ab6 100644
--- a/vnet/vnet/devices/virtio/vhost-user.h
+++ b/vnet/vnet/devices/virtio/vhost-user.h
@@ -19,9 +19,9 @@
#define VHOST_MEMORY_MAX_NREGIONS 8
#define VHOST_USER_MSG_HDR_SZ 12
#define VHOST_VRING_MAX_SIZE 32768
-#define VHOST_NET_VRING_IDX_RX 0
-#define VHOST_NET_VRING_IDX_TX 1
-#define VHOST_NET_VRING_NUM 2
+#define VHOST_VRING_MAX_N 16 //8TX + 8RX
+#define VHOST_VRING_IDX_RX(qid) (2*qid)
+#define VHOST_VRING_IDX_TX(qid) (2*qid + 1)
#define VIRTQ_DESC_F_NEXT 1
#define VIRTQ_DESC_F_INDIRECT 4
@@ -45,12 +45,16 @@
#define VIRTIO_NET_CTRL_MQ_VQ_PAIRS_MIN 1
#define VIRTIO_NET_CTRL_MQ_VQ_PAIRS_MAX 0x8000
+#define VRING_USED_F_NO_NOTIFY 1
+
#define foreach_virtio_net_feature \
_ (VIRTIO_NET_F_MRG_RXBUF, 15) \
+ _ (VIRTIO_NET_F_CTRL_VQ, 17) \
+ _ (VIRTIO_NET_F_GUEST_ANNOUNCE, 21) \
+ _ (VIRTIO_NET_F_MQ, 22) \
+ _ (VHOST_F_LOG_ALL, 26) \
_ (VIRTIO_F_ANY_LAYOUT, 27) \
_ (VIRTIO_F_INDIRECT_DESC, 28) \
- _ (VHOST_F_LOG_ALL, 26) \
- _ (VIRTIO_NET_F_GUEST_ANNOUNCE, 21) \
_ (VHOST_USER_F_PROTOCOL_FEATURES, 30) \
_ (VIRTIO_F_VERSION_1, 32)
@@ -73,37 +77,38 @@ int vhost_user_modify_if (vnet_main_t * vnm, vlib_main_t * vm,
int vhost_user_delete_if (vnet_main_t * vnm, vlib_main_t * vm,
u32 sw_if_index);
+/* *INDENT-OFF* */
typedef struct vhost_user_memory_region
{
u64 guest_phys_addr;
u64 memory_size;
u64 userspace_addr;
u64 mmap_offset;
-} vhost_user_memory_region_t;
+} __attribute ((packed)) vhost_user_memory_region_t;
typedef struct vhost_user_memory
{
u32 nregions;
u32 padding;
vhost_user_memory_region_t regions[VHOST_MEMORY_MAX_NREGIONS];
-} vhost_user_memory_t;
+} __attribute ((packed)) vhost_user_memory_t;
typedef struct
{
- unsigned int index, num;
-} vhost_vring_state_t;
+ u32 index, num;
+} __attribute ((packed)) vhost_vring_state_t;
typedef struct
{
- unsigned int index, flags;
+ u32 index, flags;
u64 desc_user_addr, used_user_addr, avail_user_addr, log_guest_addr;
-} vhost_vring_addr_t;
+} __attribute ((packed)) vhost_vring_addr_t;
typedef struct vhost_user_log
{
u64 size;
u64 offset;
-} vhost_user_log_t;
+} __attribute ((packed)) vhost_user_log_t;
typedef enum vhost_user_req
{
@@ -130,7 +135,6 @@ typedef enum vhost_user_req
} vhost_user_req_t;
// vring_desc I/O buffer descriptor
-/* *INDENT-OFF* */
typedef struct
{
uint64_t addr; // packet data buffer address
@@ -142,7 +146,7 @@ typedef struct
typedef struct
{
uint16_t flags;
- uint16_t idx;
+ volatile uint16_t idx;
uint16_t ring[VHOST_VRING_MAX_SIZE];
} __attribute ((packed)) vring_avail_t;
@@ -189,27 +193,31 @@ typedef struct vhost_user_msg {
typedef struct
{
- u32 qsz;
+ CLIB_CACHE_LINE_ALIGN_MARK (cacheline0);
+ u16 qsz;
u16 last_avail_idx;
u16 last_used_idx;
+ u16 n_since_last_int;
vring_desc_t *desc;
vring_avail_t *avail;
vring_used_t *used;
- u64 log_guest_addr;
+ f64 int_deadline;
+ u8 started;
+ u8 enabled;
+ u8 log_used;
+ //Put non-runtime in a different cache line
+ CLIB_CACHE_LINE_ALIGN_MARK (cacheline1);
int callfd;
int kickfd;
int errfd;
- u32 enabled;
- u32 log_used;
u32 callfd_idx;
- u32 n_since_last_int;
- f64 int_deadline;
+ u32 kickfd_idx;
+ u64 log_guest_addr;
} vhost_user_vring_t;
typedef struct
{
CLIB_CACHE_LINE_ALIGN_MARK (cacheline0);
- volatile u32 *lockp;
u32 is_up;
u32 admin_up;
u32 unix_fd;
@@ -221,27 +229,51 @@ typedef struct
u32 hw_if_index, sw_if_index;
u8 active;
- u32 nregions;
+ //Feature negotiation
u64 features;
u64 feature_mask;
u64 protocol_features;
- u32 num_vrings;
+
+ //Memory region information
+ u32 nregions;
vhost_user_memory_region_t regions[VHOST_MEMORY_MAX_NREGIONS];
void *region_mmap_addr[VHOST_MEMORY_MAX_NREGIONS];
u64 region_guest_addr_lo[VHOST_MEMORY_MAX_NREGIONS];
u64 region_guest_addr_hi[VHOST_MEMORY_MAX_NREGIONS];
u32 region_mmap_fd[VHOST_MEMORY_MAX_NREGIONS];
- vhost_user_vring_t vrings[2];
+
+ //Virtual rings
+ vhost_user_vring_t vrings[VHOST_VRING_MAX_N];
+ volatile u32 *vring_locks[VHOST_VRING_MAX_N];
+
int virtio_net_hdr_sz;
int is_any_layout;
- u32 *d_trace_buffers;
void *log_base_addr;
u64 log_size;
+
+ /* Whether to use spinlock or per_cpu_tx_qid assignment */
+ u8 use_tx_spinlock;
+ u16 *per_cpu_tx_qid;
+
+ /* Vector of workers for this interface */
+ u32 *workers;
} vhost_user_intf_t;
typedef struct
{
+ u16 vhost_iface_index;
+ u16 qid;
+} vhost_iface_and_queue_t;
+
+typedef struct
+{
+ u32 *trace_buffers;
+ vhost_iface_and_queue_t *rx_queues;
+} vhost_cpu_t;
+
+typedef struct
+{
u32 **rx_buffers;
u32 mtu_bytes;
vhost_user_intf_t *vhost_user_interfaces;
@@ -259,6 +291,9 @@ typedef struct
/* total cpu count */
u32 input_cpu_count;
+
+ /** Per-CPU data for vhost-user */
+ vhost_cpu_t *cpus;
} vhost_user_main_t;
typedef struct