X-Git-Url: http://git.droids-corp.org/?a=blobdiff_plain;f=lib%2Flibrte_vhost%2Fvirtio-net.c;h=1785695b119ec4b8b333199b06540e5b4d213a49;hb=99c12dcca65d;hp=9dea69c47d2b2933813940f120793dad96c04b21;hpb=8f972312b8f469af5cf6f20de48719468b1c7cda;p=dpdk.git

diff --git a/lib/librte_vhost/virtio-net.c b/lib/librte_vhost/virtio-net.c
index 9dea69c47d..1785695b11 100644
--- a/lib/librte_vhost/virtio-net.c
+++ b/lib/librte_vhost/virtio-net.c
@@ -36,42 +36,48 @@
 #include <stddef.h>
 #include <stdint.h>
 #include <stdlib.h>
+#include <assert.h>
 #include <sys/mman.h>
 #include <unistd.h>
-#include <sys/ioctl.h>
-
+#ifdef RTE_LIBRTE_VHOST_NUMA
+#include <numaif.h>
+#endif
 
 #include <sys/socket.h>
-#include <linux/if_tun.h>
-#include <linux/if.h>
 
 #include <rte_ethdev.h>
 #include <rte_log.h>
 #include <rte_string_fns.h>
 #include <rte_memory.h>
+#include <rte_malloc.h>
 #include <rte_virtio_net.h>
 
-#include "vhost_cuse/eventfd_copy.h"
 #include "vhost-net.h"
-#include "virtio-net.h"
 
-/*
- * Device linked list structure for configuration.
- */
-struct virtio_net_config_ll {
-	struct virtio_net dev;			/* Virtio device.*/
-	struct virtio_net_config_ll *next;	/* Next dev on linked list.*/
-};
+#define MAX_VHOST_DEVICE	1024
+static struct virtio_net *vhost_devices[MAX_VHOST_DEVICE];
 
 /* device ops to add/remove device to/from data core. */
 struct virtio_net_device_ops const *notify_ops;
-/* root address of the linked list of managed virtio devices */
-static struct virtio_net_config_ll *ll_root;
+
+#define VHOST_USER_F_PROTOCOL_FEATURES	30
 
 /* Features supported by this lib. */
 #define VHOST_SUPPORTED_FEATURES ((1ULL << VIRTIO_NET_F_MRG_RXBUF) | \
 				(1ULL << VIRTIO_NET_F_CTRL_VQ) | \
-				(1ULL << VIRTIO_NET_F_CTRL_RX))
+				(1ULL << VIRTIO_NET_F_CTRL_RX) | \
+				(1ULL << VIRTIO_NET_F_GUEST_ANNOUNCE) | \
+				(VHOST_SUPPORTS_MQ)            | \
+				(1ULL << VIRTIO_F_VERSION_1)   | \
+				(1ULL << VHOST_F_LOG_ALL)      | \
+				(1ULL << VHOST_USER_F_PROTOCOL_FEATURES) | \
+				(1ULL << VIRTIO_NET_F_HOST_TSO4) | \
+				(1ULL << VIRTIO_NET_F_HOST_TSO6) | \
+				(1ULL << VIRTIO_NET_F_CSUM)    | \
+				(1ULL << VIRTIO_NET_F_GUEST_CSUM) | \
+				(1ULL << VIRTIO_NET_F_GUEST_TSO4) | \
+				(1ULL << VIRTIO_NET_F_GUEST_TSO6))
+
 static uint64_t VHOST_FEATURES = VHOST_SUPPORTED_FEATURES;
 
 
@@ -101,174 +107,144 @@ qva_to_vva(struct virtio_net *dev, uint64_t qemu_va)
 	return vhost_va;
 }
 
+struct virtio_net *
+get_device(int vid)
+{
+	struct virtio_net *dev = vhost_devices[vid];
+
+	if (unlikely(!dev)) {
+		RTE_LOG(ERR, VHOST_CONFIG,
+			"(%d) device not found.\n", vid);
+	}
+
+	return dev;
+}
+
+static void
+cleanup_vq(struct vhost_virtqueue *vq, int destroy)
+{
+	if ((vq->callfd >= 0) && (destroy != 0))
+		close(vq->callfd);
+	if (vq->kickfd >= 0)
+		close(vq->kickfd);
+}
 
 /*
- * Retrieves an entry from the devices configuration linked list.
+ * Unmap any memory, close any file descriptors and
+ * free any memory owned by a device.
  */
-static struct virtio_net_config_ll *
-get_config_ll_entry(struct vhost_device_ctx ctx)
+static void
+cleanup_device(struct virtio_net *dev, int destroy)
 {
-	struct virtio_net_config_ll *ll_dev = ll_root;
+	uint32_t i;
 
-	/* Loop through linked list until the device_fh is found. */
-	while (ll_dev != NULL) {
-		if (ll_dev->dev.device_fh == ctx.fh)
-			return ll_dev;
-		ll_dev = ll_dev->next;
-	}
+	vhost_backend_cleanup(dev);
 
-	return NULL;
+	for (i = 0; i < dev->virt_qp_nb; i++) {
+		cleanup_vq(dev->virtqueue[i * VIRTIO_QNUM + VIRTIO_RXQ], destroy);
+		cleanup_vq(dev->virtqueue[i * VIRTIO_QNUM + VIRTIO_TXQ], destroy);
+	}
 }
 
 /*
- * Searches the configuration core linked list and
- * retrieves the device if it exists.
+ * Release virtqueues and device memory.
  */
-struct virtio_net *
-get_device(struct vhost_device_ctx ctx)
+static void
+free_device(struct virtio_net *dev)
 {
-	struct virtio_net_config_ll *ll_dev;
+	uint32_t i;
 
-	ll_dev = get_config_ll_entry(ctx);
+	for (i = 0; i < dev->virt_qp_nb; i++)
+		rte_free(dev->virtqueue[i * VIRTIO_QNUM]);
 
-	if (ll_dev)
-		return &ll_dev->dev;
+	rte_free(dev);
+}
+
+static void
+init_vring_queue(struct vhost_virtqueue *vq, int qp_idx)
+{
+	memset(vq, 0, sizeof(struct vhost_virtqueue));
 
-	RTE_LOG(ERR, VHOST_CONFIG,
-		"(%"PRIu64") Device not found in linked list.\n", ctx.fh);
-	return NULL;
+	vq->kickfd = VIRTIO_UNINITIALIZED_EVENTFD;
+	vq->callfd = VIRTIO_UNINITIALIZED_EVENTFD;
+
+	/* Backends are set to -1 indicating an inactive device. */
+	vq->backend = -1;
+
+	/* always set the default vq pair to enabled */
+	if (qp_idx == 0)
+		vq->enabled = 1;
 }
 
-/*
- * Add entry containing a device to the device configuration linked list.
- */
 static void
-add_config_ll_entry(struct virtio_net_config_ll *new_ll_dev)
-{
-	struct virtio_net_config_ll *ll_dev = ll_root;
-
-	/* If ll_dev == NULL then this is the first device so go to else */
-	if (ll_dev) {
-		/* If the 1st device_fh != 0 then we insert our device here. */
-		if (ll_dev->dev.device_fh != 0) {
-			new_ll_dev->dev.device_fh = 0;
-			new_ll_dev->next = ll_dev;
-			ll_root = new_ll_dev;
-		} else {
-			/*
-			 * Increment through the ll until we find un unused
-			 * device_fh. Insert the device at that entry.
-			 */
-			while ((ll_dev->next != NULL) &&
-				(ll_dev->dev.device_fh ==
-					(ll_dev->next->dev.device_fh - 1)))
-				ll_dev = ll_dev->next;
-
-			new_ll_dev->dev.device_fh = ll_dev->dev.device_fh + 1;
-			new_ll_dev->next = ll_dev->next;
-			ll_dev->next = new_ll_dev;
-		}
-	} else {
-		ll_root = new_ll_dev;
-		ll_root->dev.device_fh = 0;
-	}
+init_vring_queue_pair(struct virtio_net *dev, uint32_t qp_idx)
+{
+	uint32_t base_idx = qp_idx * VIRTIO_QNUM;
 
+	init_vring_queue(dev->virtqueue[base_idx + VIRTIO_RXQ], qp_idx);
+	init_vring_queue(dev->virtqueue[base_idx + VIRTIO_TXQ], qp_idx);
 }
 
-/*
- * Unmap any memory, close any file descriptors and
- * free any memory owned by a device.
- */
 static void
-cleanup_device(struct virtio_net *dev)
+reset_vring_queue(struct vhost_virtqueue *vq, int qp_idx)
 {
-	/* Unmap QEMU memory file if mapped. */
-	if (dev->mem) {
-		munmap((void *)(uintptr_t)dev->mem->mapped_address,
-			(size_t)dev->mem->mapped_size);
-		free(dev->mem);
-	}
+	int callfd;
 
-	/* Close any event notifiers opened by device. */
-	if (dev->virtqueue[VIRTIO_RXQ]->callfd)
-		close((int)dev->virtqueue[VIRTIO_RXQ]->callfd);
-	if (dev->virtqueue[VIRTIO_RXQ]->kickfd)
-		close((int)dev->virtqueue[VIRTIO_RXQ]->kickfd);
-	if (dev->virtqueue[VIRTIO_TXQ]->callfd)
-		close((int)dev->virtqueue[VIRTIO_TXQ]->callfd);
-	if (dev->virtqueue[VIRTIO_TXQ]->kickfd)
-		close((int)dev->virtqueue[VIRTIO_TXQ]->kickfd);
+	callfd = vq->callfd;
+	init_vring_queue(vq, qp_idx);
+	vq->callfd = callfd;
 }
 
-/*
- * Release virtqueues and device memory.
- */
 static void
-free_device(struct virtio_net_config_ll *ll_dev)
+reset_vring_queue_pair(struct virtio_net *dev, uint32_t qp_idx)
 {
-	/* Free any malloc'd memory */
-	free(ll_dev->dev.virtqueue[VIRTIO_RXQ]);
-	free(ll_dev->dev.virtqueue[VIRTIO_TXQ]);
-	free(ll_dev);
+	uint32_t base_idx = qp_idx * VIRTIO_QNUM;
+
+	reset_vring_queue(dev->virtqueue[base_idx + VIRTIO_RXQ], qp_idx);
+	reset_vring_queue(dev->virtqueue[base_idx + VIRTIO_TXQ], qp_idx);
 }
 
-/*
- * Remove an entry from the device configuration linked list.
- */
-static struct virtio_net_config_ll *
-rm_config_ll_entry(struct virtio_net_config_ll *ll_dev,
-	struct virtio_net_config_ll *ll_dev_last)
-{
-	/* First remove the device and then clean it up. */
-	if (ll_dev == ll_root) {
-		ll_root = ll_dev->next;
-		cleanup_device(&ll_dev->dev);
-		free_device(ll_dev);
-		return ll_root;
-	} else {
-		if (likely(ll_dev_last != NULL)) {
-			ll_dev_last->next = ll_dev->next;
-			cleanup_device(&ll_dev->dev);
-			free_device(ll_dev);
-			return ll_dev_last->next;
-		} else {
-			cleanup_device(&ll_dev->dev);
-			free_device(ll_dev);
-			RTE_LOG(ERR, VHOST_CONFIG,
-				"Remove entry from config_ll failed\n");
-			return NULL;
-		}
+static int
+alloc_vring_queue_pair(struct virtio_net *dev, uint32_t qp_idx)
+{
+	struct vhost_virtqueue *virtqueue = NULL;
+	uint32_t virt_rx_q_idx = qp_idx * VIRTIO_QNUM + VIRTIO_RXQ;
+	uint32_t virt_tx_q_idx = qp_idx * VIRTIO_QNUM + VIRTIO_TXQ;
+
+	virtqueue = rte_malloc(NULL,
+			       sizeof(struct vhost_virtqueue) * VIRTIO_QNUM, 0);
+	if (virtqueue == NULL) {
+		RTE_LOG(ERR, VHOST_CONFIG,
+			"Failed to allocate memory for virt qp:%d.\n", qp_idx);
+		return -1;
 	}
+
+	dev->virtqueue[virt_rx_q_idx] = virtqueue;
+	dev->virtqueue[virt_tx_q_idx] = virtqueue + VIRTIO_TXQ;
+
+	init_vring_queue_pair(dev, qp_idx);
+
+	dev->virt_qp_nb += 1;
+
+	return 0;
 }
 
 /*
- *  Initialise all variables in device structure.
+ * Reset some variables in device structure, while keeping few
+ * others untouched, such as vid, ifname, virt_qp_nb: they
+ * should be same unless the device is removed.
  */
 static void
-init_device(struct virtio_net *dev)
+reset_device(struct virtio_net *dev)
 {
-	uint64_t vq_offset;
-
-	/*
-	 * Virtqueues have already been malloced so
-	 * we don't want to set them to NULL.
-	 */
-	vq_offset = offsetof(struct virtio_net, mem);
-
-	/* Set everything to 0. */
-	memset((void *)(uintptr_t)((uint64_t)(uintptr_t)dev + vq_offset), 0,
-		(sizeof(struct virtio_net) - (size_t)vq_offset));
-	memset(dev->virtqueue[VIRTIO_RXQ], 0, sizeof(struct vhost_virtqueue));
-	memset(dev->virtqueue[VIRTIO_TXQ], 0, sizeof(struct vhost_virtqueue));
+	uint32_t i;
 
-	dev->virtqueue[VIRTIO_RXQ]->kickfd = (eventfd_t)-1;
-	dev->virtqueue[VIRTIO_RXQ]->callfd = (eventfd_t)-1;
-	dev->virtqueue[VIRTIO_TXQ]->kickfd = (eventfd_t)-1;
-	dev->virtqueue[VIRTIO_TXQ]->callfd = (eventfd_t)-1;
+	dev->features = 0;
+	dev->protocol_features = 0;
+	dev->flags = 0;
 
-	/* Backends are set to -1 indicating an inactive device. */
-	dev->virtqueue[VIRTIO_RXQ]->backend = VIRTIO_DEV_STOPPED;
-	dev->virtqueue[VIRTIO_TXQ]->backend = VIRTIO_DEV_STOPPED;
+	for (i = 0; i < dev->virt_qp_nb; i++)
+		reset_vring_queue_pair(dev, i);
 }
 
 /*
@@ -276,98 +252,87 @@ init_device(struct virtio_net *dev)
  * initialised and a new entry is added to the device configuration linked
  * list.
  */
-static int
-new_device(struct vhost_device_ctx ctx)
+int
+vhost_new_device(void)
 {
-	struct virtio_net_config_ll *new_ll_dev;
-	struct vhost_virtqueue *virtqueue_rx, *virtqueue_tx;
+	struct virtio_net *dev;
+	int i;
 
-	/* Setup device and virtqueues. */
-	new_ll_dev = malloc(sizeof(struct virtio_net_config_ll));
-	if (new_ll_dev == NULL) {
+	dev = rte_zmalloc(NULL, sizeof(struct virtio_net), 0);
+	if (dev == NULL) {
 		RTE_LOG(ERR, VHOST_CONFIG,
-			"(%"PRIu64") Failed to allocate memory for dev.\n",
-			ctx.fh);
+			"Failed to allocate memory for new dev.\n");
 		return -1;
 	}
 
-	virtqueue_rx = malloc(sizeof(struct vhost_virtqueue));
-	if (virtqueue_rx == NULL) {
-		free(new_ll_dev);
-		RTE_LOG(ERR, VHOST_CONFIG,
-			"(%"PRIu64") Failed to allocate memory for rxq.\n",
-			ctx.fh);
-		return -1;
+	for (i = 0; i < MAX_VHOST_DEVICE; i++) {
+		if (vhost_devices[i] == NULL)
+			break;
 	}
-
-	virtqueue_tx = malloc(sizeof(struct vhost_virtqueue));
-	if (virtqueue_tx == NULL) {
-		free(virtqueue_rx);
-		free(new_ll_dev);
+	if (i == MAX_VHOST_DEVICE) {
 		RTE_LOG(ERR, VHOST_CONFIG,
-			"(%"PRIu64") Failed to allocate memory for txq.\n",
-			ctx.fh);
+			"Failed to find a free slot for new device.\n");
 		return -1;
 	}
 
-	new_ll_dev->dev.virtqueue[VIRTIO_RXQ] = virtqueue_rx;
-	new_ll_dev->dev.virtqueue[VIRTIO_TXQ] = virtqueue_tx;
-
-	/* Initialise device and virtqueues. */
-	init_device(&new_ll_dev->dev);
+	vhost_devices[i] = dev;
+	dev->vid = i;
 
-	new_ll_dev->next = NULL;
-
-	/* Add entry to device configuration linked list. */
-	add_config_ll_entry(new_ll_dev);
-
-	return new_ll_dev->dev.device_fh;
+	return i;
 }
 
 /*
  * Function is called from the CUSE release function. This function will
  * cleanup the device and remove it from device configuration linked list.
  */
-static void
-destroy_device(struct vhost_device_ctx ctx)
-{
-	struct virtio_net_config_ll *ll_dev_cur_ctx, *ll_dev_last = NULL;
-	struct virtio_net_config_ll *ll_dev_cur = ll_root;
-
-	/* Find the linked list entry for the device to be removed. */
-	ll_dev_cur_ctx = get_config_ll_entry(ctx);
-	while (ll_dev_cur != NULL) {
-		/*
-		 * If the device is found or
-		 * a device that doesn't exist is found then it is removed.
-		 */
-		if (ll_dev_cur == ll_dev_cur_ctx) {
-			/*
-			 * If the device is running on a data core then call
-			 * the function to remove it from the data core.
-			 */
-			if ((ll_dev_cur->dev.flags & VIRTIO_DEV_RUNNING))
-				notify_ops->destroy_device(&(ll_dev_cur->dev));
-			ll_dev_cur = rm_config_ll_entry(ll_dev_cur,
-					ll_dev_last);
-		} else {
-			ll_dev_last = ll_dev_cur;
-			ll_dev_cur = ll_dev_cur->next;
-		}
+void
+vhost_destroy_device(int vid)
+{
+	struct virtio_net *dev = get_device(vid);
+
+	if (dev == NULL)
+		return;
+
+	if (dev->flags & VIRTIO_DEV_RUNNING) {
+		dev->flags &= ~VIRTIO_DEV_RUNNING;
+		notify_ops->destroy_device(vid);
 	}
+
+	cleanup_device(dev, 1);
+	free_device(dev);
+
+	vhost_devices[vid] = NULL;
 }
 
+void
+vhost_set_ifname(int vid, const char *if_name, unsigned int if_len)
+{
+	struct virtio_net *dev;
+	unsigned int len;
+
+	dev = get_device(vid);
+	if (dev == NULL)
+		return;
+
+	len = if_len > sizeof(dev->ifname) ?
+		sizeof(dev->ifname) : if_len;
+
+	strncpy(dev->ifname, if_name, len);
+	dev->ifname[sizeof(dev->ifname) - 1] = '\0';
+}
+
+
 /*
  * Called from CUSE IOCTL: VHOST_SET_OWNER
  * This function just returns success at the moment unless
  * the device hasn't been initialised.
  */
-static int
-set_owner(struct vhost_device_ctx ctx)
+int
+vhost_set_owner(int vid)
 {
 	struct virtio_net *dev;
 
-	dev = get_device(ctx);
+	dev = get_device(vid);
 	if (dev == NULL)
 		return -1;
 
@@ -377,16 +342,22 @@ set_owner(struct vhost_device_ctx ctx)
 /*
  * Called from CUSE IOCTL: VHOST_RESET_OWNER
  */
-static int
-reset_owner(struct vhost_device_ctx ctx)
+int
+vhost_reset_owner(int vid)
 {
-	struct virtio_net_config_ll *ll_dev;
+	struct virtio_net *dev;
 
-	ll_dev = get_config_ll_entry(ctx);
+	dev = get_device(vid);
+	if (dev == NULL)
+		return -1;
 
-	cleanup_device(&ll_dev->dev);
-	init_device(&ll_dev->dev);
+	if (dev->flags & VIRTIO_DEV_RUNNING) {
+		dev->flags &= ~VIRTIO_DEV_RUNNING;
+		notify_ops->destroy_device(vid);
+	}
 
+	cleanup_device(dev, 0);
+	reset_device(dev);
 	return 0;
 }
 
@@ -394,12 +365,12 @@ reset_owner(struct vhost_device_ctx ctx)
  * Called from CUSE IOCTL: VHOST_GET_FEATURES
  * The features that we support are requested.
  */
-static int
-get_features(struct vhost_device_ctx ctx, uint64_t *pu)
+int
+vhost_get_features(int vid, uint64_t *pu)
 {
 	struct virtio_net *dev;
 
-	dev = get_device(ctx);
+	dev = get_device(vid);
 	if (dev == NULL)
 		return -1;
 
@@ -412,38 +383,30 @@ get_features(struct vhost_device_ctx ctx, uint64_t *pu)
  * Called from CUSE IOCTL: VHOST_SET_FEATURES
  * We receive the negotiated features supported by us and the virtio device.
  */
-static int
-set_features(struct vhost_device_ctx ctx, uint64_t *pu)
+int
+vhost_set_features(int vid, uint64_t *pu)
 {
 	struct virtio_net *dev;
 
-	dev = get_device(ctx);
+	dev = get_device(vid);
 	if (dev == NULL)
 		return -1;
 	if (*pu & ~VHOST_FEATURES)
 		return -1;
 
-	/* Store the negotiated feature list for the device. */
 	dev->features = *pu;
-
-	/* Set the vhost_hlen depending on if VIRTIO_NET_F_MRG_RXBUF is set. */
-	if (dev->features & (1 << VIRTIO_NET_F_MRG_RXBUF)) {
-		LOG_DEBUG(VHOST_CONFIG,
-			"(%"PRIu64") Mergeable RX buffers enabled\n",
-			dev->device_fh);
-		dev->virtqueue[VIRTIO_RXQ]->vhost_hlen =
-			sizeof(struct virtio_net_hdr_mrg_rxbuf);
-		dev->virtqueue[VIRTIO_TXQ]->vhost_hlen =
-			sizeof(struct virtio_net_hdr_mrg_rxbuf);
+	if (dev->features &
+		((1 << VIRTIO_NET_F_MRG_RXBUF) | (1ULL << VIRTIO_F_VERSION_1))) {
+		dev->vhost_hlen = sizeof(struct virtio_net_hdr_mrg_rxbuf);
 	} else {
-		LOG_DEBUG(VHOST_CONFIG,
-			"(%"PRIu64") Mergeable RX buffers disabled\n",
-			dev->device_fh);
-		dev->virtqueue[VIRTIO_RXQ]->vhost_hlen =
-			sizeof(struct virtio_net_hdr);
-		dev->virtqueue[VIRTIO_TXQ]->vhost_hlen =
-			sizeof(struct virtio_net_hdr);
+		dev->vhost_hlen = sizeof(struct virtio_net_hdr);
 	}
+	LOG_DEBUG(VHOST_CONFIG,
+		"(%d) mergeable RX buffers %s, virtio 1 %s\n",
+		dev->vid,
+		(dev->features & (1 << VIRTIO_NET_F_MRG_RXBUF)) ? "on" : "off",
+		(dev->features & (1ULL << VIRTIO_F_VERSION_1)) ? "on" : "off");
+
 	return 0;
 }
 
@@ -451,12 +414,12 @@ set_features(struct vhost_device_ctx ctx, uint64_t *pu)
  * Called from CUSE IOCTL: VHOST_SET_VRING_NUM
  * The virtio device sends us the size of the descriptor ring.
  */
-static int
-set_vring_num(struct vhost_device_ctx ctx, struct vhost_vring_state *state)
+int
+vhost_set_vring_num(int vid, struct vhost_vring_state *state)
 {
 	struct virtio_net *dev;
 
-	dev = get_device(ctx);
+	dev = get_device(vid);
 	if (dev == NULL)
 		return -1;
 
@@ -466,19 +429,102 @@ set_vring_num(struct vhost_device_ctx ctx, struct vhost_vring_state *state)
 	return 0;
 }
 
+/*
+ * Reallocate virtio_dev and vhost_virtqueue data structure to make them on the
+ * same numa node as the memory of vring descriptor.
+ */
+#ifdef RTE_LIBRTE_VHOST_NUMA
+static struct virtio_net*
+numa_realloc(struct virtio_net *dev, int index)
+{
+	int oldnode, newnode;
+	struct virtio_net *old_dev;
+	struct vhost_virtqueue *old_vq, *vq;
+	int ret;
+
+	/*
+	 * vq is allocated on pairs, we should try to do realloc
+	 * on first queue of one queue pair only.
+	 */
+	if (index % VIRTIO_QNUM != 0)
+		return dev;
+
+	old_dev = dev;
+	vq = old_vq = dev->virtqueue[index];
+
+	ret = get_mempolicy(&newnode, NULL, 0, old_vq->desc,
+			    MPOL_F_NODE | MPOL_F_ADDR);
+
+	/* check if we need to reallocate vq */
+	ret |= get_mempolicy(&oldnode, NULL, 0, old_vq,
+			     MPOL_F_NODE | MPOL_F_ADDR);
+	if (ret) {
+		RTE_LOG(ERR, VHOST_CONFIG,
+			"Unable to get vq numa information.\n");
+		return dev;
+	}
+	if (oldnode != newnode) {
+		RTE_LOG(INFO, VHOST_CONFIG,
+			"reallocate vq from %d to %d node\n", oldnode, newnode);
+		vq = rte_malloc_socket(NULL, sizeof(*vq) * VIRTIO_QNUM, 0,
+				       newnode);
+		if (!vq)
+			return dev;
+
+		memcpy(vq, old_vq, sizeof(*vq) * VIRTIO_QNUM);
+		rte_free(old_vq);
+	}
+
+	/* check if we need to reallocate dev */
+	ret = get_mempolicy(&oldnode, NULL, 0, old_dev,
+			    MPOL_F_NODE | MPOL_F_ADDR);
+	if (ret) {
+		RTE_LOG(ERR, VHOST_CONFIG,
+			"Unable to get dev numa information.\n");
+		goto out;
+	}
+	if (oldnode != newnode) {
+		RTE_LOG(INFO, VHOST_CONFIG,
+			"reallocate dev from %d to %d node\n",
+			oldnode, newnode);
+		dev = rte_malloc_socket(NULL, sizeof(*dev), 0, newnode);
+		if (!dev) {
+			dev = old_dev;
+			goto out;
+		}
+
+		memcpy(dev, old_dev, sizeof(*dev));
+		rte_free(old_dev);
+	}
+
+out:
+	dev->virtqueue[index] = vq;
+	dev->virtqueue[index + 1] = vq + 1;
+	vhost_devices[dev->vid] = dev;
+
+	return dev;
+}
+#else
+static struct virtio_net*
+numa_realloc(struct virtio_net *dev, int index __rte_unused)
+{
+	return dev;
+}
+#endif
+
 /*
  * Called from CUSE IOCTL: VHOST_SET_VRING_ADDR
  * The virtio device sends us the desc, used and avail ring addresses.
  * This function then converts these to our address space.
  */
-static int
-set_vring_addr(struct vhost_device_ctx ctx, struct vhost_vring_addr *addr)
+int
+vhost_set_vring_addr(int vid, struct vhost_vring_addr *addr)
 {
 	struct virtio_net *dev;
 	struct vhost_virtqueue *vq;
 
-	dev = get_device(ctx);
-	if (dev == NULL)
+	dev = get_device(vid);
+	if ((dev == NULL) || (dev->mem == NULL))
 		return -1;
 
 	/* addr->index refers to the queue index. The txq 1, rxq is 0. */
@@ -489,17 +535,20 @@ set_vring_addr(struct vhost_device_ctx ctx, struct vhost_vring_addr *addr)
 			addr->desc_user_addr);
 	if (vq->desc == 0) {
 		RTE_LOG(ERR, VHOST_CONFIG,
-			"(%"PRIu64") Failed to find desc ring address.\n",
-			dev->device_fh);
+			"(%d) failed to find desc ring address.\n",
+			dev->vid);
 		return -1;
 	}
 
+	dev = numa_realloc(dev, addr->index);
+	vq = dev->virtqueue[addr->index];
+
 	vq->avail = (struct vring_avail *)(uintptr_t)qva_to_vva(dev,
 			addr->avail_user_addr);
 	if (vq->avail == 0) {
 		RTE_LOG(ERR, VHOST_CONFIG,
-			"(%"PRIu64") Failed to find avail ring address.\n",
-			dev->device_fh);
+			"(%d) failed to find avail ring address.\n",
+			dev->vid);
 		return -1;
 	}
 
@@ -507,17 +556,29 @@ set_vring_addr(struct vhost_device_ctx ctx, struct vhost_vring_addr *addr)
 			addr->used_user_addr);
 	if (vq->used == 0) {
 		RTE_LOG(ERR, VHOST_CONFIG,
-			"(%"PRIu64") Failed to find used ring address.\n",
-			dev->device_fh);
+			"(%d) failed to find used ring address.\n",
+			dev->vid);
 		return -1;
 	}
 
-	LOG_DEBUG(VHOST_CONFIG, "(%"PRIu64") mapped address desc: %p\n",
-			dev->device_fh, vq->desc);
-	LOG_DEBUG(VHOST_CONFIG, "(%"PRIu64") mapped address avail: %p\n",
-			dev->device_fh, vq->avail);
-	LOG_DEBUG(VHOST_CONFIG, "(%"PRIu64") mapped address used: %p\n",
-			dev->device_fh, vq->used);
+	if (vq->last_used_idx != vq->used->idx) {
+		RTE_LOG(WARNING, VHOST_CONFIG,
+			"last_used_idx (%u) and vq->used->idx (%u) mismatches; "
+			"some packets maybe resent for Tx and dropped for Rx\n",
+			vq->last_used_idx, vq->used->idx);
+		vq->last_used_idx     = vq->used->idx;
+	}
+
+	vq->log_guest_addr = addr->log_guest_addr;
+
+	LOG_DEBUG(VHOST_CONFIG, "(%d) mapped address desc: %p\n",
+			dev->vid, vq->desc);
+	LOG_DEBUG(VHOST_CONFIG, "(%d) mapped address avail: %p\n",
+			dev->vid, vq->avail);
+	LOG_DEBUG(VHOST_CONFIG, "(%d) mapped address used: %p\n",
+			dev->vid, vq->used);
+	LOG_DEBUG(VHOST_CONFIG, "(%d) log_guest_addr: %" PRIx64 "\n",
+			dev->vid, vq->log_guest_addr);
 
 	return 0;
 }
@@ -526,18 +587,17 @@ set_vring_addr(struct vhost_device_ctx ctx, struct vhost_vring_addr *addr)
  * Called from CUSE IOCTL: VHOST_SET_VRING_BASE
  * The virtio device sends us the available ring last used index.
  */
-static int
-set_vring_base(struct vhost_device_ctx ctx, struct vhost_vring_state *state)
+int
+vhost_set_vring_base(int vid, struct vhost_vring_state *state)
 {
 	struct virtio_net *dev;
 
-	dev = get_device(ctx);
+	dev = get_device(vid);
 	if (dev == NULL)
 		return -1;
 
 	/* State->index refers to the queue index. The txq is 1, rxq is 0. */
 	dev->virtqueue[state->index]->last_used_idx = state->num;
-	dev->virtqueue[state->index]->last_used_idx_res = state->num;
 
 	return 0;
 }
@@ -546,13 +606,13 @@ set_vring_base(struct vhost_device_ctx ctx, struct vhost_vring_state *state)
  * Called from CUSE IOCTL: VHOST_GET_VRING_BASE
  * We send the virtio device our available ring last used index.
  */
-static int
-get_vring_base(struct vhost_device_ctx ctx, uint32_t index,
+int
+vhost_get_vring_base(int vid, uint32_t index,
 	struct vhost_vring_state *state)
 {
 	struct virtio_net *dev;
 
-	dev = get_device(ctx);
+	dev = get_device(vid);
 	if (dev == NULL)
 		return -1;
 
@@ -569,23 +629,34 @@ get_vring_base(struct vhost_device_ctx ctx, uint32_t index,
  * The virtio device sends an eventfd to interrupt the guest. This fd gets
  * copied into our process space.
  */
-static int
-set_vring_call(struct vhost_device_ctx ctx, struct vhost_vring_file *file)
+int
+vhost_set_vring_call(int vid, struct vhost_vring_file *file)
 {
 	struct virtio_net *dev;
 	struct vhost_virtqueue *vq;
+	uint32_t cur_qp_idx = file->index / VIRTIO_QNUM;
 
-	dev = get_device(ctx);
+	dev = get_device(vid);
 	if (dev == NULL)
 		return -1;
 
+	/*
+	 * FIXME: VHOST_SET_VRING_CALL is the first per-vring message
+	 * we get, so we do vring queue pair allocation here.
+	 */
+	if (cur_qp_idx + 1 > dev->virt_qp_nb) {
+		if (alloc_vring_queue_pair(dev, cur_qp_idx) < 0)
+			return -1;
+	}
+
 	/* file->index refers to the queue index. The txq is 1, rxq is 0. */
 	vq = dev->virtqueue[file->index];
+	assert(vq != NULL);
 
-	if ((int)vq->kickfd >= 0)
-		close((int)vq->kickfd);
+	if (vq->callfd >= 0)
+		close(vq->callfd);
 
-	vq->kickfd = file->fd;
+	vq->callfd = file->fd;
 
 	return 0;
 }
@@ -595,60 +666,23 @@ set_vring_call(struct vhost_device_ctx ctx, struct vhost_vring_file *file)
  * The virtio device sends an eventfd that it can use to notify us.
  * This fd gets copied into our process space.
  */
-static int
-set_vring_kick(struct vhost_device_ctx ctx, struct vhost_vring_file *file)
+int
+vhost_set_vring_kick(int vid, struct vhost_vring_file *file)
 {
 	struct virtio_net *dev;
 	struct vhost_virtqueue *vq;
 
-	dev = get_device(ctx);
+	dev = get_device(vid);
 	if (dev == NULL)
 		return -1;
 
 	/* file->index refers to the queue index. The txq is 1, rxq is 0. */
 	vq = dev->virtqueue[file->index];
 
-	if ((int)vq->callfd >= 0)
-		close((int)vq->callfd);
+	if (vq->kickfd >= 0)
+		close(vq->kickfd);
 
-	vq->callfd = file->fd;
-
-	return 0;
-}
-
-/*
- * Function to get the tap device name from the provided file descriptor and
- * save it in the device structure.
- */
-static int
-get_ifname(struct virtio_net *dev, int tap_fd, int pid)
-{
-	int fd_tap;
-	struct ifreq ifr;
-	uint32_t size, ifr_size;
-	int ret;
-
-    fd_tap = eventfd_copy(tap_fd, pid);
-    if (fd_tap < 0)
-        return -1;
-
-	ret = ioctl(fd_tap, TUNGETIFF, &ifr);
-
-	if (close(fd_tap) < 0)
-		RTE_LOG(ERR, VHOST_CONFIG,
-			"(%"PRIu64") fd close failed\n",
-			dev->device_fh);
-
-	if (ret >= 0) {
-		ifr_size = strnlen(ifr.ifr_name, sizeof(ifr.ifr_name));
-		size = ifr_size > sizeof(dev->ifname) ?
-				sizeof(dev->ifname) : ifr_size;
-
-		strncpy(dev->ifname, ifr.ifr_name, size);
-	} else
-		RTE_LOG(ERR, VHOST_CONFIG,
-			"(%"PRIu64") TUNGETIFF ioctl failed\n",
-			dev->device_fh);
+	vq->kickfd = file->fd;
 
 	return 0;
 }
@@ -662,12 +696,12 @@ get_ifname(struct virtio_net *dev, int tap_fd, int pid)
  * At that point we remove the device from the data core.
  * The device will still exist in the device configuration linked list.
  */
-static int
-set_backend(struct vhost_device_ctx ctx, struct vhost_vring_file *file)
+int
+vhost_set_backend(int vid, struct vhost_vring_file *file)
 {
 	struct virtio_net *dev;
 
-	dev = get_device(ctx);
+	dev = get_device(vid);
 	if (dev == NULL)
 		return -1;
 
@@ -679,63 +713,105 @@ set_backend(struct vhost_device_ctx ctx, struct vhost_vring_file *file)
 	 * we add the device.
 	 */
 	if (!(dev->flags & VIRTIO_DEV_RUNNING)) {
-		if (((int)dev->virtqueue[VIRTIO_TXQ]->backend != VIRTIO_DEV_STOPPED) &&
-			((int)dev->virtqueue[VIRTIO_RXQ]->backend != VIRTIO_DEV_STOPPED)) {
-			get_ifname(dev, file->fd, ctx.pid);
-			return notify_ops->new_device(dev);
+		if (dev->virtqueue[VIRTIO_TXQ]->backend != VIRTIO_DEV_STOPPED &&
+		    dev->virtqueue[VIRTIO_RXQ]->backend != VIRTIO_DEV_STOPPED) {
+			if (notify_ops->new_device(vid) < 0)
+				return -1;
+			dev->flags |= VIRTIO_DEV_RUNNING;
 		}
-	/* Otherwise we remove it. */
-	} else
-		if (file->fd == VIRTIO_DEV_STOPPED)
-			notify_ops->destroy_device(dev);
+	} else if (file->fd == VIRTIO_DEV_STOPPED) {
+		dev->flags &= ~VIRTIO_DEV_RUNNING;
+		notify_ops->destroy_device(vid);
+	}
+
 	return 0;
 }
 
-/*
- * Function pointers are set for the device operations to allow CUSE to call
- * functions when an IOCTL, device_add or device_release is received.
- */
-static const struct vhost_net_device_ops vhost_device_ops = {
-	.new_device = new_device,
-	.destroy_device = destroy_device,
+int
+rte_vhost_get_numa_node(int vid)
+{
+#ifdef RTE_LIBRTE_VHOST_NUMA
+	struct virtio_net *dev = get_device(vid);
+	int numa_node;
+	int ret;
+
+	if (dev == NULL)
+		return -1;
+
+	ret = get_mempolicy(&numa_node, NULL, 0, dev,
+			    MPOL_F_NODE | MPOL_F_ADDR);
+	if (ret < 0) {
+		RTE_LOG(ERR, VHOST_CONFIG,
+			"(%d) failed to query numa node: %d\n", vid, ret);
+		return -1;
+	}
+
+	return numa_node;
+#else
+	RTE_SET_USED(vid);
+	return -1;
+#endif
+}
 
-	.get_features = get_features,
-	.set_features = set_features,
+uint32_t
+rte_vhost_get_queue_num(int vid)
+{
+	struct virtio_net *dev = get_device(vid);
+
+	if (dev == NULL)
+		return 0;
 
-	.set_vring_num = set_vring_num,
-	.set_vring_addr = set_vring_addr,
-	.set_vring_base = set_vring_base,
-	.get_vring_base = get_vring_base,
+	return dev->virt_qp_nb;
+}
+
+int
+rte_vhost_get_ifname(int vid, char *buf, size_t len)
+{
+	struct virtio_net *dev = get_device(vid);
 
-	.set_vring_kick = set_vring_kick,
-	.set_vring_call = set_vring_call,
+	if (dev == NULL)
+		return -1;
 
-	.set_backend = set_backend,
+	len = RTE_MIN(len, sizeof(dev->ifname));
 
-	.set_owner = set_owner,
-	.reset_owner = reset_owner,
-};
+	strncpy(buf, dev->ifname, len);
+	buf[len - 1] = '\0';
 
-/*
- * Called by main to setup callbacks when registering CUSE device.
- */
-struct vhost_net_device_ops const *
-get_virtio_net_callbacks(void)
+	return 0;
+}
+
+uint16_t
+rte_vhost_avail_entries(int vid, uint16_t queue_id)
 {
-	return &vhost_device_ops;
+	struct virtio_net *dev;
+	struct vhost_virtqueue *vq;
+
+	dev = get_device(vid);
+	if (!dev)
+		return 0;
+
+	vq = dev->virtqueue[queue_id];
+	if (!vq->enabled)
+		return 0;
+
+	return *(volatile uint16_t *)&vq->avail->idx - vq->last_used_idx;
 }
 
-int rte_vhost_enable_guest_notification(struct virtio_net *dev,
-	uint16_t queue_id, int enable)
+int
+rte_vhost_enable_guest_notification(int vid, uint16_t queue_id, int enable)
 {
+	struct virtio_net *dev = get_device(vid);
+
+	if (dev == NULL)
+		return -1;
+
 	if (enable) {
 		RTE_LOG(ERR, VHOST_CONFIG,
 			"guest notification isn't supported.\n");
 		return -1;
 	}
 
-	dev->virtqueue[queue_id]->used->flags =
-		enable ? 0 : VRING_USED_F_NO_NOTIFY;
+	dev->virtqueue[queue_id]->used->flags = VRING_USED_F_NO_NOTIFY;
 	return 0;
 }