virtio: fast TAP interfaces with vhost-net backend

Change-Id: Ided667356d5c6fb9648eb34685aabd6b16a598b7
Signed-off-by: Damjan Marion <damarion@cisco.com>
Signed-off-by: Steven Luong <sluong@cisco.com>

1 files changed, 361 insertions, 0 deletions

diff --git a/src/vnet/devices/virtio/tap.c b/src/vnet/devices/virtio/tap.c
new file mode 100644
index 00000000000..658ba6bfc68
--- /dev/null
+++ b/src/vnet/devices/virtio/tap.c
@@ -0,0 +1,361 @@
+/*
+ *------------------------------------------------------------------
+ * Copyright (c) 2017 Cisco and/or its affiliates.
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at:
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ *------------------------------------------------------------------
+ */
+
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <fcntl.h>
+#include <net/if.h>
+#include <linux/if_tun.h>
+#include <sys/ioctl.h>
+#include <linux/virtio_net.h>
+#include <linux/vhost.h>
+#include <sys/eventfd.h>
+
+#include <linux/netlink.h>
+#include <linux/rtnetlink.h>
+
+#include <vlib/vlib.h>
+#include <vlib/unix/unix.h>
+#include <vnet/ethernet/ethernet.h>
+#include <vnet/devices/virtio/virtio.h>
+#include <vnet/devices/virtio/tap.h>
+
+#define _IOCTL(fd,a,...) \
+  if (ioctl (fd, a, __VA_ARGS__) < 0) \
+    { \
+      err = clib_error_return_unix (0, "ioctl(" #a ")"); \
+      goto error; \
+    }
+
+static u32
+virtio_eth_flag_change (vnet_main_t * vnm, vnet_hw_interface_t * hi,
+			u32 flags)
+{
+  /* nothing for now */
+  return 0;
+}
+
+
+clib_error_t *
+clib_netlink_set_if_attr (int ifindex, unsigned short rta_type, void *data,
+			  int data_len)
+{
+  clib_error_t *err = 0;
+  int sock;
+  struct sockaddr_nl ra = { 0 };
+  struct
+  {
+    struct nlmsghdr nh;
+    struct ifinfomsg ifmsg;
+    char attrbuf[512];
+  } req;
+  struct rtattr *rta;
+
+  memset (&req, 0, sizeof (req));
+  if ((sock = socket (AF_NETLINK, SOCK_RAW, NETLINK_ROUTE)) == -1)
+    {
+      err = clib_error_return_unix (0, "socket(AF_NETLINK)");
+      goto error;
+    }
+
+  ra.nl_family = AF_NETLINK;
+  ra.nl_pid = getpid ();
+
+  if ((bind (sock, (struct sockaddr *) &ra, sizeof (ra))) == -1)
+    return clib_error_return_unix (0, "bind");
+
+  req.nh.nlmsg_len = NLMSG_LENGTH (sizeof (struct ifinfomsg));
+  req.nh.nlmsg_flags = NLM_F_REQUEST;
+  req.nh.nlmsg_type = RTM_SETLINK;
+  req.ifmsg.ifi_family = AF_UNSPEC;
+  req.ifmsg.ifi_index = ifindex;
+  req.ifmsg.ifi_change = 0xffffffff;
+  rta = (struct rtattr *) (((char *) &req) + NLMSG_ALIGN (req.nh.nlmsg_len));
+  rta->rta_type = rta_type;
+  rta->rta_len = RTA_LENGTH (data_len);
+  req.nh.nlmsg_len = NLMSG_ALIGN (req.nh.nlmsg_len) + RTA_LENGTH (data_len);
+  memcpy (RTA_DATA (rta), data, data_len);
+
+  if ((send (sock, &req, req.nh.nlmsg_len, 0)) == -1)
+    err = clib_error_return_unix (0, "send");
+
+error:
+  return err;
+}
+
+clib_error_t *
+clib_netlink_set_if_mtu (int ifindex, int mtu)
+{
+  clib_error_t *err;
+
+  err = clib_netlink_set_if_attr (ifindex, IFLA_MTU, &mtu, sizeof (int));
+  return err;
+}
+
+clib_error_t *
+clib_netlink_set_if_namespace (int ifindex, char *net_ns)
+{
+  clib_error_t *err;
+  int ns_fd;
+  u8 *s;
+  s = format (0, "/var/run/netns/%s%c", net_ns, 0);
+  ns_fd = open ((char *) s, O_RDONLY);
+  vec_free (s);
+  if (ns_fd == -1)
+    return clib_error_return (0, "namespace '%s' doesn't exist", net_ns);
+
+  err =
+    clib_netlink_set_if_attr (ifindex, IFLA_NET_NS_FD, &ns_fd, sizeof (int));
+  close (ns_fd);
+  return err;
+}
+
+int
+tap_create_if (vlib_main_t * vm, tap_create_if_args_t * args)
+{
+  vnet_main_t *vnm = vnet_get_main ();
+  virtio_main_t *vim = &virtio_main;
+  vnet_sw_interface_t *sw;
+  vnet_hw_interface_t *hw;
+  int i;
+  clib_error_t *err = 0;
+  struct ifreq ifr;
+  size_t hdrsz;
+  struct vhost_memory *vhost_mem = 0;
+  virtio_if_t *vif = 0;
+  int rv = 0;
+
+  memset (&ifr, 0, sizeof (ifr));
+  pool_get (vim->interfaces, vif);
+  vif->dev_instance = vif - vim->interfaces;
+  vif->tap_fd = -1;
+
+  if ((vif->fd = open ("/dev/vhost-net", O_RDWR | O_NONBLOCK)) < 0)
+    {
+      rv = VNET_API_ERROR_SYSCALL_ERROR_1;
+      goto error;
+    }
+
+  _IOCTL (vif->fd, VHOST_GET_FEATURES, &vif->remote_features);
+
+  if ((vif->remote_features & (1ULL << VIRTIO_NET_F_MRG_RXBUF)) == 0)
+    {
+      rv = VNET_API_ERROR_UNSUPPORTED;
+      goto error;
+    }
+
+  if ((vif->remote_features & (1ULL << VIRTIO_RING_F_INDIRECT_DESC)) == 0)
+    {
+      rv = VNET_API_ERROR_UNSUPPORTED;
+      goto error;
+    }
+
+  if ((vif->remote_features & (1ULL << VIRTIO_F_VERSION_1)) == 0)
+    {
+      rv = VNET_API_ERROR_UNSUPPORTED;
+      goto error;
+    }
+
+  vif->features |= 1ULL << VIRTIO_NET_F_MRG_RXBUF;
+  vif->features |= 1ULL << VIRTIO_F_VERSION_1;
+  vif->features |= 1ULL << VIRTIO_RING_F_INDIRECT_DESC;
+
+  _IOCTL (vif->fd, VHOST_SET_FEATURES, &vif->features);
+
+  if ((vif->tap_fd = open ("/dev/net/tun", O_RDWR | O_NONBLOCK)) < 0)
+    {
+      rv = VNET_API_ERROR_SYSCALL_ERROR_2;
+      goto error;
+    }
+
+  ifr.ifr_flags = IFF_TAP | IFF_NO_PI | IFF_ONE_QUEUE | IFF_VNET_HDR;
+  strncpy (ifr.ifr_ifrn.ifrn_name, (char *) args->name, IF_NAMESIZE);
+  _IOCTL (vif->tap_fd, TUNSETIFF, (void *) &ifr);
+
+  vif->ifindex = if_nametoindex ((char *) args->name);
+
+  unsigned int offload = 0;
+  hdrsz = sizeof (struct virtio_net_hdr_v1);
+  _IOCTL (vif->tap_fd, TUNSETOFFLOAD, offload);
+  _IOCTL (vif->tap_fd, TUNSETVNETHDRSZ, &hdrsz);
+  _IOCTL (vif->fd, VHOST_SET_OWNER, 0);
+
+  if (args->net_ns)
+    {
+      err = clib_netlink_set_if_namespace (vif->ifindex,
+					   (char *) args->net_ns);
+      if (err)
+	{
+	  rv = VNET_API_ERROR_NAMESPACE_CREATE;
+	  goto error;
+	}
+    }
+
+  /* Set vhost memory table */
+  i = sizeof (struct vhost_memory) + sizeof (struct vhost_memory_region);
+  vhost_mem = clib_mem_alloc (i);
+  memset (vhost_mem, 0, i);
+  vhost_mem->nregions = 1;
+  vhost_mem->regions[0].memory_size = (1ULL << 47) - 4096;
+  _IOCTL (vif->fd, VHOST_SET_MEM_TABLE, vhost_mem);
+
+  if ((err = virtio_vring_init (vm, vif, 0, args->rx_ring_sz)))
+    {
+      rv = VNET_API_ERROR_VIRTIO_INIT;
+      goto error;
+    }
+
+  if ((err = virtio_vring_init (vm, vif, 1, args->tx_ring_sz)))
+    {
+      rv = VNET_API_ERROR_VIRTIO_INIT;
+      goto error;
+    }
+
+  if (!args->hw_addr_set)
+    {
+      f64 now = vlib_time_now (vm);
+      u32 rnd;
+      rnd = (u32) (now * 1e6);
+      rnd = random_u32 (&rnd);
+
+      memcpy (args->hw_addr + 2, &rnd, sizeof (rnd));
+      args->hw_addr[0] = 2;
+      args->hw_addr[1] = 0xfe;
+    }
+  vif->name = args->name;
+  args->name = 0;
+  vif->net_ns = args->net_ns;
+  args->net_ns = 0;
+  err = ethernet_register_interface (vnm, virtio_device_class.index,
+				     vif->dev_instance, args->hw_addr,
+				     &vif->hw_if_index,
+				     virtio_eth_flag_change);
+  if (err)
+    rv = VNET_API_ERROR_INVALID_REGISTRATION;
+
+  sw = vnet_get_hw_sw_interface (vnm, vif->hw_if_index);
+  vif->sw_if_index = sw->sw_if_index;
+  args->sw_if_index = vif->sw_if_index;
+  hw = vnet_get_hw_interface (vnm, vif->hw_if_index);
+  hw->flags |= VNET_HW_INTERFACE_FLAG_SUPPORTS_INT_MODE;
+  vnet_hw_interface_set_input_node (vnm, vif->hw_if_index,
+				    virtio_input_node.index);
+  vnet_hw_interface_assign_rx_thread (vnm, vif->hw_if_index, 0, ~0);
+  vnet_hw_interface_set_rx_mode (vnm, vif->hw_if_index, 0,
+				 VNET_HW_INTERFACE_RX_MODE_DEFAULT);
+  vif->per_interface_next_index = ~0;
+  vif->type = VIRTIO_IF_TYPE_TAP;
+  vif->flags |= VIRTIO_IF_FLAG_ADMIN_UP;
+  vnet_hw_interface_set_flags (vnm, vif->hw_if_index,
+			       VNET_HW_INTERFACE_FLAG_LINK_UP);
+  goto done;
+
+error:
+  if (vif->tap_fd != -1)
+    close (vif->tap_fd);
+  if (vif->fd != -1)
+    close (vif->fd);
+  vec_foreach_index (i, vif->vrings) virtio_vring_free (vif, i);
+  memset (vif, 0, sizeof (virtio_if_t));
+  pool_put (vim->interfaces, vif);
+
+done:
+  if (vhost_mem)
+    clib_mem_free (vhost_mem);
+
+  return rv;
+}
+
+int
+tap_delete_if (vlib_main_t * vm, u32 sw_if_index)
+{
+  vnet_main_t *vnm = vnet_get_main ();
+  virtio_main_t *mm = &virtio_main;
+  int i;
+  virtio_if_t *vif;
+  vnet_hw_interface_t *hw;
+
+  hw = vnet_get_sup_hw_interface (vnm, sw_if_index);
+  if (hw == NULL || virtio_device_class.index != hw->dev_class_index)
+    return VNET_API_ERROR_INVALID_SW_IF_INDEX;
+
+  vif = pool_elt_at_index (mm->interfaces, hw->dev_instance);
+
+  /* bring down the interface */
+  vnet_hw_interface_set_flags (vnm, vif->hw_if_index, 0);
+  vnet_sw_interface_set_flags (vnm, vif->sw_if_index, 0);
+
+  ethernet_delete_interface (vnm, vif->hw_if_index);
+  vif->hw_if_index = ~0;
+
+  if (vif->tap_fd != -1)
+    close (vif->tap_fd);
+  if (vif->fd != -1)
+    close (vif->fd);
+
+  vec_foreach_index (i, vif->vrings) virtio_vring_free (vif, i);
+  vec_free (vif->vrings);
+
+  memset (vif, 0, sizeof (*vif));
+  pool_put (mm->interfaces, vif);
+
+  return 0;
+}
+
+int
+tap_dump_ifs (tap_interface_details_t ** out_tapids)
+{
+  vnet_main_t *vnm = vnet_get_main ();
+  virtio_main_t *mm = &virtio_main;
+  virtio_if_t *vif;
+  vnet_hw_interface_t *hi;
+  tap_interface_details_t *r_tapids = NULL;
+  tap_interface_details_t *tapid = NULL;
+
+  /* *INDENT-OFF* */
+  pool_foreach (vif, mm->interfaces,
+    vec_add2(r_tapids, tapid, 1);
+    memset (tapid, 0, sizeof (*tapid));
+    tapid->sw_if_index = vif->sw_if_index;
+    hi = vnet_get_hw_interface (vnm, vif->hw_if_index);
+    clib_memcpy(tapid->dev_name, hi->name,
+		MIN (ARRAY_LEN (tapid->dev_name) - 1,
+		     strlen ((const char *) hi->name)));
+  );
+  /* *INDENT-ON* */
+
+  *out_tapids = r_tapids;
+
+  return 0;
+}
+
+static clib_error_t *
+tap_init (vlib_main_t * vm)
+{
+
+  return 0;
+}
+
+VLIB_INIT_FUNCTION (tap_init);
+
+/*
+ * fd.io coding-style-patch-verification: ON
+ *
+ * Local Variables:
+ * eval: (c-set-style "gnu")
+ * End:
+ */