* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
+#include <rte_atomic.h>
+#include <rte_common.h>
#include <rte_mbuf.h>
#include <rte_ethdev.h>
#include <rte_malloc.h>
#include <rte_vdev.h>
#include <rte_kvargs.h>
+#include <rte_net.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <sys/socket.h>
#include <sys/ioctl.h>
#include <sys/mman.h>
+#include <errno.h>
+#include <signal.h>
+#include <stdint.h>
#include <unistd.h>
-#include <poll.h>
#include <arpa/inet.h>
#include <linux/if.h>
#include <linux/if_tun.h>
#define ETH_TAP_IFACE_ARG "iface"
#define ETH_TAP_SPEED_ARG "speed"
+#ifdef IFF_MULTI_QUEUE
#define RTE_PMD_TAP_MAX_QUEUES 16
+#else
+#define RTE_PMD_TAP_MAX_QUEUES 1
+#endif
static struct rte_vdev_driver pmd_tap_drv;
static int tap_unit;
+static volatile uint32_t tap_trigger; /* Rx trigger */
+
static struct rte_eth_link pmd_link = {
.link_speed = ETH_SPEED_NUM_10G,
.link_duplex = ETH_LINK_FULL_DUPLEX,
struct rx_queue {
struct rte_mempool *mp; /* Mempool for RX packets */
+ uint32_t trigger_seen; /* Last seen Rx trigger value */
uint16_t in_port; /* Port ID */
int fd;
struct ether_addr eth_addr; /* Mac address of the device port */
int if_index; /* IF_INDEX for the port */
- int fds[RTE_PMD_TAP_MAX_QUEUES]; /* List of all file descriptors */
+ int ioctl_sock; /* socket for ioctl calls */
struct rx_queue rxq[RTE_PMD_TAP_MAX_QUEUES]; /* List of RX queues */
struct tx_queue txq[RTE_PMD_TAP_MAX_QUEUES]; /* List of TX queues */
};
+static void
+tap_trigger_cb(int sig __rte_unused)
+{
+ /* Valid trigger values are nonzero */
+ tap_trigger = (tap_trigger + 1) | 0x80000000;
+}
+
+static int
+tap_ioctl(struct pmd_internals *pmd, unsigned long request,
+ struct ifreq *ifr, int set);
+
/* Tun/Tap allocation routine
*
* name is the number of the interface to use, unless NULL to take the host
* supplied name.
*/
static int
-tun_alloc(char *name)
+tun_alloc(struct pmd_internals *pmd, uint16_t qid)
{
struct ifreq ifr;
+#ifdef IFF_MULTI_QUEUE
unsigned int features;
+#endif
int fd;
memset(&ifr, 0, sizeof(struct ifreq));
ifr.ifr_flags = IFF_TAP | IFF_NO_PI;
- if (name && name[0])
- strncpy(ifr.ifr_name, name, IFNAMSIZ);
+ snprintf(ifr.ifr_name, IFNAMSIZ, "%s", pmd->name);
+
+ RTE_LOG(DEBUG, PMD, "ifr_name '%s'\n", ifr.ifr_name);
fd = open(TUN_TAP_DEV_PATH, O_RDWR);
if (fd < 0) {
goto error;
}
- /* Grab the TUN features to verify we can work */
+#ifdef IFF_MULTI_QUEUE
+ /* Grab the TUN features to verify we can work multi-queue */
if (ioctl(fd, TUNGETFEATURES, &features) < 0) {
- RTE_LOG(ERR, PMD, "Unable to get TUN/TAP features\n");
+ RTE_LOG(ERR, PMD, "TAP unable to get TUN/TAP features\n");
goto error;
}
- RTE_LOG(DEBUG, PMD, "TUN/TAP Features %08x\n", features);
+ RTE_LOG(DEBUG, PMD, " TAP Features %08x\n", features);
- if (!(features & IFF_MULTI_QUEUE) && (RTE_PMD_TAP_MAX_QUEUES > 1)) {
- RTE_LOG(DEBUG, PMD, "TUN/TAP device only one queue\n");
- goto error;
- } else if ((features & IFF_ONE_QUEUE) &&
- (RTE_PMD_TAP_MAX_QUEUES == 1)) {
- ifr.ifr_flags |= IFF_ONE_QUEUE;
- RTE_LOG(DEBUG, PMD, "Single queue only support\n");
- } else {
- ifr.ifr_flags |= IFF_MULTI_QUEUE;
- RTE_LOG(DEBUG, PMD, "Multi-queue support for %d queues\n",
+ if (features & IFF_MULTI_QUEUE) {
+ RTE_LOG(DEBUG, PMD, " Multi-queue support for %d queues\n",
RTE_PMD_TAP_MAX_QUEUES);
+ ifr.ifr_flags |= IFF_MULTI_QUEUE;
+ } else
+#endif
+ {
+ ifr.ifr_flags |= IFF_ONE_QUEUE;
+ RTE_LOG(DEBUG, PMD, " Single queue only support\n");
}
- /* Set the TUN/TAP configuration and get the name if needed */
+ /* Set the TUN/TAP configuration and set the name if needed */
if (ioctl(fd, TUNSETIFF, (void *)&ifr) < 0) {
- RTE_LOG(ERR, PMD, "Unable to set TUNSETIFF for %s\n",
+ RTE_LOG(WARNING, PMD,
+ "Unable to set TUNSETIFF for %s\n",
ifr.ifr_name);
perror("TUNSETIFF");
goto error;
/* Always set the file descriptor to non-blocking */
if (fcntl(fd, F_SETFL, O_NONBLOCK) < 0) {
- RTE_LOG(ERR, PMD, "Unable to set to nonblocking\n");
+ RTE_LOG(WARNING, PMD,
+ "Unable to set %s to nonblocking\n",
+ ifr.ifr_name);
perror("F_SETFL, NONBLOCK");
goto error;
}
- /* If the name is different that new name as default */
- if (name && strcmp(name, ifr.ifr_name))
- snprintf(name, RTE_ETH_NAME_MAX_LEN - 1, "%s", ifr.ifr_name);
+ /* Set up trigger to optimize empty Rx bursts */
+ errno = 0;
+ do {
+ struct sigaction sa;
+ int flags = fcntl(fd, F_GETFL);
+
+ if (flags == -1 || sigaction(SIGIO, NULL, &sa) == -1)
+ break;
+ if (sa.sa_handler != tap_trigger_cb) {
+ /*
+ * Make sure SIGIO is not already taken. This is done
+ * as late as possible to leave the application a
+ * chance to set up its own signal handler first.
+ */
+ if (sa.sa_handler != SIG_IGN &&
+ sa.sa_handler != SIG_DFL) {
+ errno = EBUSY;
+ break;
+ }
+ sa = (struct sigaction){
+ .sa_flags = SA_RESTART,
+ .sa_handler = tap_trigger_cb,
+ };
+ if (sigaction(SIGIO, &sa, NULL) == -1)
+ break;
+ }
+ /* Enable SIGIO on file descriptor */
+ fcntl(fd, F_SETFL, flags | O_ASYNC);
+ fcntl(fd, F_SETOWN, getpid());
+ } while (0);
+ if (errno) {
+ /* Disable trigger globally in case of error */
+ tap_trigger = 0;
+ RTE_LOG(WARNING, PMD, "Rx trigger disabled: %s\n",
+ strerror(errno));
+ }
+
+ if (qid == 0) {
+ struct ifreq ifr;
+
+ if (tap_ioctl(pmd, SIOCGIFHWADDR, &ifr, 0) < 0)
+ goto error;
+ rte_memcpy(&pmd->eth_addr, ifr.ifr_hwaddr.sa_data,
+ ETHER_ADDR_LEN);
+ }
return fd;
struct rx_queue *rxq = queue;
uint16_t num_rx;
unsigned long num_rx_bytes = 0;
+ uint32_t trigger = tap_trigger;
+ if (trigger == rxq->trigger_seen)
+ return 0;
+ if (trigger)
+ rxq->trigger_seen = trigger;
+ rte_compiler_barrier();
for (num_rx = 0; num_rx < nb_pkts; ) {
/* allocate the next mbuf */
mbuf = rte_pktmbuf_alloc(rxq->mp);
if (unlikely(!mbuf)) {
- RTE_LOG(WARNING, PMD, "Unable to allocate mbuf\n");
+ RTE_LOG(WARNING, PMD, "TAP unable to allocate mbuf\n");
break;
}
mbuf->data_len = len;
mbuf->pkt_len = len;
mbuf->port = rxq->in_port;
+ mbuf->packet_type = rte_net_get_ptype(mbuf, NULL,
+ RTE_PTYPE_ALL_MASK);
/* account for the receive frame */
bufs[num_rx++] = mbuf;
{
struct rte_mbuf *mbuf;
struct tx_queue *txq = queue;
- struct pollfd pfd;
uint16_t num_tx = 0;
unsigned long num_tx_bytes = 0;
int i, n;
if (unlikely(nb_pkts == 0))
return 0;
- pfd.events = POLLOUT;
- pfd.fd = txq->fd;
for (i = 0; i < nb_pkts; i++) {
- n = poll(&pfd, 1, 0);
-
+ /* copy the tx frame data */
+ mbuf = bufs[num_tx];
+ n = write(txq->fd,
+ rte_pktmbuf_mtod(mbuf, void *),
+ rte_pktmbuf_pkt_len(mbuf));
if (n <= 0)
break;
- if (pfd.revents & POLLOUT) {
- /* copy the tx frame data */
- mbuf = bufs[num_tx];
- n = write(pfd.fd, rte_pktmbuf_mtod(mbuf, void*),
- rte_pktmbuf_pkt_len(mbuf));
- if (n <= 0)
- break;
-
- num_tx++;
- num_tx_bytes += mbuf->pkt_len;
- rte_pktmbuf_free(mbuf);
- }
+ num_tx++;
+ num_tx_bytes += mbuf->pkt_len;
+ rte_pktmbuf_free(mbuf);
}
txq->stats.opackets += num_tx;
}
static int
-tap_dev_start(struct rte_eth_dev *dev)
+tap_ioctl(struct pmd_internals *pmd, unsigned long request,
+ struct ifreq *ifr, int set)
+{
+ short req_flags = ifr->ifr_flags;
+
+ snprintf(ifr->ifr_name, IFNAMSIZ, "%s", pmd->name);
+ switch (request) {
+ case SIOCSIFFLAGS:
+ /* fetch current flags to leave other flags untouched */
+ if (ioctl(pmd->ioctl_sock, SIOCGIFFLAGS, ifr) < 0)
+ goto error;
+ if (set)
+ ifr->ifr_flags |= req_flags;
+ else
+ ifr->ifr_flags &= ~req_flags;
+ break;
+ case SIOCGIFHWADDR:
+ case SIOCSIFHWADDR:
+ case SIOCSIFMTU:
+ break;
+ default:
+ RTE_LOG(WARNING, PMD, "%s: ioctl() called with wrong arg\n",
+ pmd->name);
+ return -EINVAL;
+ }
+ if (ioctl(pmd->ioctl_sock, request, ifr) < 0)
+ goto error;
+ return 0;
+
+error:
+ RTE_LOG(ERR, PMD, "%s: ioctl(%lu) failed with error: %s\n",
+ ifr->ifr_name, request, strerror(errno));
+ return -errno;
+}
+
+static int
+tap_link_set_down(struct rte_eth_dev *dev)
+{
+ struct pmd_internals *pmd = dev->data->dev_private;
+ struct ifreq ifr = { .ifr_flags = IFF_UP };
+
+ dev->data->dev_link.link_status = ETH_LINK_DOWN;
+ return tap_ioctl(pmd, SIOCSIFFLAGS, &ifr, 0);
+}
+
+static int
+tap_link_set_up(struct rte_eth_dev *dev)
{
- /* Force the Link up */
+ struct pmd_internals *pmd = dev->data->dev_private;
+ struct ifreq ifr = { .ifr_flags = IFF_UP };
+
dev->data->dev_link.link_status = ETH_LINK_UP;
+ return tap_ioctl(pmd, SIOCSIFFLAGS, &ifr, 1);
+}
- return 0;
+static int
+tap_dev_start(struct rte_eth_dev *dev)
+{
+ return tap_link_set_up(dev);
}
/* This function gets called when the current port gets stopped.
static void
tap_dev_stop(struct rte_eth_dev *dev)
{
- int i;
- struct pmd_internals *internals = dev->data->dev_private;
-
- for (i = 0; i < internals->nb_queues; i++)
- if (internals->fds[i] != -1)
- close(internals->fds[i]);
-
- dev->data->dev_link.link_status = ETH_LINK_DOWN;
+ tap_link_set_down(dev);
}
static int
return 0;
}
+static uint32_t
+tap_dev_speed_capa(void)
+{
+ uint32_t speed = pmd_link.link_speed;
+ uint32_t capa = 0;
+
+ if (speed >= ETH_SPEED_NUM_10M)
+ capa |= ETH_LINK_SPEED_10M;
+ if (speed >= ETH_SPEED_NUM_100M)
+ capa |= ETH_LINK_SPEED_100M;
+ if (speed >= ETH_SPEED_NUM_1G)
+ capa |= ETH_LINK_SPEED_1G;
+ if (speed >= ETH_SPEED_NUM_5G)
+ capa |= ETH_LINK_SPEED_2_5G;
+ if (speed >= ETH_SPEED_NUM_5G)
+ capa |= ETH_LINK_SPEED_5G;
+ if (speed >= ETH_SPEED_NUM_10G)
+ capa |= ETH_LINK_SPEED_10G;
+ if (speed >= ETH_SPEED_NUM_20G)
+ capa |= ETH_LINK_SPEED_20G;
+ if (speed >= ETH_SPEED_NUM_25G)
+ capa |= ETH_LINK_SPEED_25G;
+ if (speed >= ETH_SPEED_NUM_40G)
+ capa |= ETH_LINK_SPEED_40G;
+ if (speed >= ETH_SPEED_NUM_50G)
+ capa |= ETH_LINK_SPEED_50G;
+ if (speed >= ETH_SPEED_NUM_56G)
+ capa |= ETH_LINK_SPEED_56G;
+ if (speed >= ETH_SPEED_NUM_100G)
+ capa |= ETH_LINK_SPEED_100G;
+
+ return capa;
+}
+
static void
tap_dev_info(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info)
{
dev_info->max_tx_queues = internals->nb_queues;
dev_info->min_rx_bufsize = 0;
dev_info->pci_dev = NULL;
+ dev_info->speed_capa = tap_dev_speed_capa();
}
static void
tap_stats->q_ibytes[i] = pmd->rxq[i].stats.ibytes;
rx_total += tap_stats->q_ipackets[i];
rx_bytes_total += tap_stats->q_ibytes[i];
- }
- for (i = 0; i < imax; i++) {
tap_stats->q_opackets[i] = pmd->txq[i].stats.opackets;
tap_stats->q_errors[i] = pmd->txq[i].stats.errs;
tap_stats->q_obytes[i] = pmd->txq[i].stats.obytes;
for (i = 0; i < pmd->nb_queues; i++) {
pmd->rxq[i].stats.ipackets = 0;
pmd->rxq[i].stats.ibytes = 0;
- }
- for (i = 0; i < pmd->nb_queues; i++) {
pmd->txq[i].stats.opackets = 0;
pmd->txq[i].stats.errs = 0;
pmd->txq[i].stats.obytes = 0;
static void
tap_dev_close(struct rte_eth_dev *dev __rte_unused)
{
+ int i;
+ struct pmd_internals *internals = dev->data->dev_private;
+
+ tap_link_set_down(dev);
+
+ for (i = 0; i < internals->nb_queues; i++) {
+ if (internals->rxq[i].fd != -1)
+ close(internals->rxq[i].fd);
+ internals->rxq[i].fd = -1;
+ internals->txq[i].fd = -1;
+ }
}
static void
return 0;
}
+static void
+tap_promisc_enable(struct rte_eth_dev *dev)
+{
+ struct pmd_internals *pmd = dev->data->dev_private;
+ struct ifreq ifr = { .ifr_flags = IFF_PROMISC };
+
+ dev->data->promiscuous = 1;
+ tap_ioctl(pmd, SIOCSIFFLAGS, &ifr, 1);
+}
+
+static void
+tap_promisc_disable(struct rte_eth_dev *dev)
+{
+ struct pmd_internals *pmd = dev->data->dev_private;
+ struct ifreq ifr = { .ifr_flags = IFF_PROMISC };
+
+ dev->data->promiscuous = 0;
+ tap_ioctl(pmd, SIOCSIFFLAGS, &ifr, 0);
+}
+
+static void
+tap_allmulti_enable(struct rte_eth_dev *dev)
+{
+ struct pmd_internals *pmd = dev->data->dev_private;
+ struct ifreq ifr = { .ifr_flags = IFF_ALLMULTI };
+
+ dev->data->all_multicast = 1;
+ tap_ioctl(pmd, SIOCSIFFLAGS, &ifr, 1);
+}
+
+static void
+tap_allmulti_disable(struct rte_eth_dev *dev)
+{
+ struct pmd_internals *pmd = dev->data->dev_private;
+ struct ifreq ifr = { .ifr_flags = IFF_ALLMULTI };
+
+ dev->data->all_multicast = 0;
+ tap_ioctl(pmd, SIOCSIFFLAGS, &ifr, 0);
+}
+
+
+static void
+tap_mac_set(struct rte_eth_dev *dev, struct ether_addr *mac_addr)
+{
+ struct pmd_internals *pmd = dev->data->dev_private;
+ struct ifreq ifr;
+
+ if (is_zero_ether_addr(mac_addr)) {
+ RTE_LOG(ERR, PMD, "%s: can't set an empty MAC address\n",
+ dev->data->name);
+ return;
+ }
+
+ ifr.ifr_hwaddr.sa_family = AF_LOCAL;
+ rte_memcpy(ifr.ifr_hwaddr.sa_data, mac_addr, ETHER_ADDR_LEN);
+ if (tap_ioctl(pmd, SIOCSIFHWADDR, &ifr, 1) < 0)
+ return;
+ rte_memcpy(&pmd->eth_addr, mac_addr, ETHER_ADDR_LEN);
+}
+
static int
tap_setup_queue(struct rte_eth_dev *dev,
struct pmd_internals *internals,
uint16_t qid)
{
+ struct pmd_internals *pmd = dev->data->dev_private;
struct rx_queue *rx = &internals->rxq[qid];
struct tx_queue *tx = &internals->txq[qid];
int fd;
fd = tx->fd;
if (fd < 0) {
RTE_LOG(INFO, PMD, "Add queue to TAP %s for qid %d\n",
- dev->data->name, qid);
- fd = tun_alloc(dev->data->name);
+ pmd->name, qid);
+ fd = tun_alloc(pmd, qid);
if (fd < 0) {
- RTE_LOG(ERR, PMD, "tun_alloc(%s) failed\n",
- dev->data->name);
+ RTE_LOG(ERR, PMD, "tun_alloc(%s, %d) failed\n",
+ pmd->name, qid);
return -1;
}
+ if (qid == 0) {
+ struct ifreq ifr;
+
+ ifr.ifr_mtu = dev->data->mtu;
+ if (tap_ioctl(pmd, SIOCSIFMTU, &ifr, 1) < 0) {
+ close(fd);
+ return -1;
+ }
+ }
}
}
- dev->data->rx_queues[qid] = rx;
- dev->data->tx_queues[qid] = tx;
rx->fd = fd;
tx->fd = fd;
return fd;
}
+static int
+rx_setup_queue(struct rte_eth_dev *dev,
+ struct pmd_internals *internals,
+ uint16_t qid)
+{
+ dev->data->rx_queues[qid] = &internals->rxq[qid];
+
+ return tap_setup_queue(dev, internals, qid);
+}
+
+static int
+tx_setup_queue(struct rte_eth_dev *dev,
+ struct pmd_internals *internals,
+ uint16_t qid)
+{
+ dev->data->tx_queues[qid] = &internals->txq[qid];
+
+ return tap_setup_queue(dev, internals, qid);
+}
+
static int
tap_rx_queue_setup(struct rte_eth_dev *dev,
uint16_t rx_queue_id,
int fd;
if ((rx_queue_id >= internals->nb_queues) || !mp) {
- RTE_LOG(ERR, PMD, "nb_queues %d mp %p\n",
- internals->nb_queues, mp);
+ RTE_LOG(WARNING, PMD,
+ "nb_queues %d too small or mempool NULL\n",
+ internals->nb_queues);
return -1;
}
internals->rxq[rx_queue_id].mp = mp;
+ internals->rxq[rx_queue_id].trigger_seen = 1; /* force initial burst */
internals->rxq[rx_queue_id].in_port = dev->data->port_id;
/* Now get the space available for data in the mbuf */
RTE_PKTMBUF_HEADROOM);
if (buf_size < ETH_FRAME_LEN) {
- RTE_LOG(ERR, PMD,
+ RTE_LOG(WARNING, PMD,
"%s: %d bytes will not fit in mbuf (%d bytes)\n",
dev->data->name, ETH_FRAME_LEN, buf_size);
return -ENOMEM;
}
- fd = tap_setup_queue(dev, internals, rx_queue_id);
+ fd = rx_setup_queue(dev, internals, rx_queue_id);
if (fd == -1)
return -1;
- internals->fds[rx_queue_id] = fd;
- RTE_LOG(INFO, PMD, "RX TAP device name %s, qid %d on fd %d\n",
- dev->data->name, rx_queue_id, internals->rxq[rx_queue_id].fd);
+ RTE_LOG(DEBUG, PMD, " RX TAP device name %s, qid %d on fd %d\n",
+ internals->name, rx_queue_id, internals->rxq[rx_queue_id].fd);
return 0;
}
if (tx_queue_id >= internals->nb_queues)
return -1;
- ret = tap_setup_queue(dev, internals, tx_queue_id);
+ ret = tx_setup_queue(dev, internals, tx_queue_id);
if (ret == -1)
return -1;
- RTE_LOG(INFO, PMD, "TX TAP device name %s, qid %d on fd %d\n",
- dev->data->name, tx_queue_id, internals->txq[tx_queue_id].fd);
+ RTE_LOG(DEBUG, PMD, " TX TAP device name %s, qid %d on fd %d\n",
+ internals->name, tx_queue_id, internals->txq[tx_queue_id].fd);
+
+ return 0;
+}
+
+static int
+tap_mtu_set(struct rte_eth_dev *dev, uint16_t mtu)
+{
+ struct pmd_internals *pmd = dev->data->dev_private;
+ struct ifreq ifr = { .ifr_mtu = mtu };
+ int err = 0;
+
+ err = tap_ioctl(pmd, SIOCSIFMTU, &ifr, 1);
+ if (!err)
+ dev->data->mtu = mtu;
+
+ return err;
+}
+
+static int
+tap_set_mc_addr_list(struct rte_eth_dev *dev __rte_unused,
+ struct ether_addr *mc_addr_set __rte_unused,
+ uint32_t nb_mc_addr __rte_unused)
+{
+ /*
+ * Nothing to do actually: the tap has no filtering whatsoever, every
+ * packet is received.
+ */
+ return 0;
+}
+
+static const uint32_t*
+tap_dev_supported_ptypes_get(struct rte_eth_dev *dev __rte_unused)
+{
+ static const uint32_t ptypes[] = {
+ RTE_PTYPE_INNER_L2_ETHER,
+ RTE_PTYPE_INNER_L2_ETHER_VLAN,
+ RTE_PTYPE_INNER_L2_ETHER_QINQ,
+ RTE_PTYPE_INNER_L3_IPV4,
+ RTE_PTYPE_INNER_L3_IPV4_EXT,
+ RTE_PTYPE_INNER_L3_IPV6,
+ RTE_PTYPE_INNER_L3_IPV6_EXT,
+ RTE_PTYPE_INNER_L4_FRAG,
+ RTE_PTYPE_INNER_L4_UDP,
+ RTE_PTYPE_INNER_L4_TCP,
+ RTE_PTYPE_INNER_L4_SCTP,
+ RTE_PTYPE_L2_ETHER,
+ RTE_PTYPE_L2_ETHER_VLAN,
+ RTE_PTYPE_L2_ETHER_QINQ,
+ RTE_PTYPE_L3_IPV4,
+ RTE_PTYPE_L3_IPV4_EXT,
+ RTE_PTYPE_L3_IPV6_EXT,
+ RTE_PTYPE_L3_IPV6,
+ RTE_PTYPE_L4_FRAG,
+ RTE_PTYPE_L4_UDP,
+ RTE_PTYPE_L4_TCP,
+ RTE_PTYPE_L4_SCTP,
+ };
+
+ return ptypes;
+}
+
+static int
+tap_flow_ctrl_get(struct rte_eth_dev *dev __rte_unused,
+ struct rte_eth_fc_conf *fc_conf)
+{
+ fc_conf->mode = RTE_FC_NONE;
+ return 0;
+}
+static int
+tap_flow_ctrl_set(struct rte_eth_dev *dev __rte_unused,
+ struct rte_eth_fc_conf *fc_conf)
+{
+ if (fc_conf->mode != RTE_FC_NONE)
+ return -ENOTSUP;
return 0;
}
.tx_queue_setup = tap_tx_queue_setup,
.rx_queue_release = tap_rx_queue_release,
.tx_queue_release = tap_tx_queue_release,
+ .flow_ctrl_get = tap_flow_ctrl_get,
+ .flow_ctrl_set = tap_flow_ctrl_set,
.link_update = tap_link_update,
+ .dev_set_link_up = tap_link_set_up,
+ .dev_set_link_down = tap_link_set_down,
+ .promiscuous_enable = tap_promisc_enable,
+ .promiscuous_disable = tap_promisc_disable,
+ .allmulticast_enable = tap_allmulti_enable,
+ .allmulticast_disable = tap_allmulti_disable,
+ .mac_addr_set = tap_mac_set,
+ .mtu_set = tap_mtu_set,
+ .set_mc_addr_list = tap_set_mc_addr_list,
.stats_get = tap_stats_get,
.stats_reset = tap_stats_reset,
+ .dev_supported_ptypes_get = tap_dev_supported_ptypes_get,
};
-static int
-pmd_mac_address(int fd, struct rte_eth_dev *dev, struct ether_addr *addr)
-{
- struct ifreq ifr;
-
- if ((fd <= 0) || !dev || !addr)
- return -1;
-
- memset(&ifr, 0, sizeof(ifr));
-
- if (ioctl(fd, SIOCGIFHWADDR, &ifr) == -1) {
- RTE_LOG(ERR, PMD, "ioctl failed (SIOCGIFHWADDR) (%s)\n",
- ifr.ifr_name);
- return -1;
- }
-
- /* Set the host based MAC address to this special MAC format */
- ifr.ifr_hwaddr.sa_data[0] = 'T';
- ifr.ifr_hwaddr.sa_data[1] = 'a';
- ifr.ifr_hwaddr.sa_data[2] = 'p';
- ifr.ifr_hwaddr.sa_data[3] = '-';
- ifr.ifr_hwaddr.sa_data[4] = dev->data->port_id;
- ifr.ifr_hwaddr.sa_data[5] = dev->data->numa_node;
- if (ioctl(fd, SIOCSIFHWADDR, &ifr) == -1) {
- RTE_LOG(ERR, PMD, "%s: ioctl failed (SIOCSIFHWADDR) (%s)\n",
- dev->data->name, ifr.ifr_name);
- return -1;
- }
-
- /* Set the local application MAC address, needs to be different then
- * the host based MAC address.
- */
- ifr.ifr_hwaddr.sa_data[0] = 'd';
- ifr.ifr_hwaddr.sa_data[1] = 'n';
- ifr.ifr_hwaddr.sa_data[2] = 'e';
- ifr.ifr_hwaddr.sa_data[3] = 't';
- ifr.ifr_hwaddr.sa_data[4] = dev->data->port_id;
- ifr.ifr_hwaddr.sa_data[5] = dev->data->numa_node;
- rte_memcpy(addr, ifr.ifr_hwaddr.sa_data, ETH_ALEN);
-
- return 0;
-}
-
static int
eth_dev_tap_create(const char *name, char *tap_name)
{
struct rte_eth_dev *dev = NULL;
struct pmd_internals *pmd = NULL;
struct rte_eth_dev_data *data = NULL;
- int i, fd = -1;
+ int i;
- RTE_LOG(INFO, PMD,
- "%s: Create TAP Ethernet device with %d queues on numa %u\n",
- name, RTE_PMD_TAP_MAX_QUEUES, rte_socket_id());
+ RTE_LOG(DEBUG, PMD, " TAP device on numa %u\n", rte_socket_id());
data = rte_zmalloc_socket(tap_name, sizeof(*data), 0, numa_node);
if (!data) {
- RTE_LOG(INFO, PMD, "Failed to allocate data\n");
+ RTE_LOG(ERR, PMD, "TAP Failed to allocate data\n");
goto error_exit;
}
pmd = rte_zmalloc_socket(tap_name, sizeof(*pmd), 0, numa_node);
if (!pmd) {
- RTE_LOG(INFO, PMD, "Unable to allocate internal struct\n");
+ RTE_LOG(ERR, PMD, "TAP Unable to allocate internal struct\n");
goto error_exit;
}
- /* Use the name and not the tap_name */
- dev = rte_eth_dev_allocate(tap_name);
+ /* name in allocation and data->name must be consistent */
+ snprintf(data->name, sizeof(data->name), "%s", name);
+ dev = rte_eth_dev_allocate(name);
if (!dev) {
- RTE_LOG(INFO, PMD, "Unable to allocate device struct\n");
+ RTE_LOG(ERR, PMD, "TAP Unable to allocate device struct\n");
goto error_exit;
}
pmd->nb_queues = RTE_PMD_TAP_MAX_QUEUES;
+ pmd->ioctl_sock = socket(AF_INET, SOCK_DGRAM, 0);
+ if (pmd->ioctl_sock == -1) {
+ RTE_LOG(ERR, PMD,
+ "TAP Unable to get a socket for management: %s\n",
+ strerror(errno));
+ goto error_exit;
+ }
+
/* Setup some default values */
data->dev_private = pmd;
data->port_id = dev->data->port_id;
+ data->mtu = dev->data->mtu;
data->dev_flags = RTE_ETH_DEV_DETACHABLE;
data->kdrv = RTE_KDRV_NONE;
data->drv_name = pmd_tap_drv.driver.name;
dev->driver = NULL;
dev->rx_pkt_burst = pmd_rx_burst;
dev->tx_pkt_burst = pmd_tx_burst;
- snprintf(dev->data->name, sizeof(dev->data->name), "%s", name);
- /* Create the first Tap device */
- fd = tun_alloc(tap_name);
- if (fd < 0) {
- RTE_LOG(INFO, PMD, "tun_alloc() failed\n");
- goto error_exit;
- }
-
- /* Presetup the fds to -1 as being not working */
+ /* Presetup the fds to -1 as being not valid */
for (i = 0; i < RTE_PMD_TAP_MAX_QUEUES; i++) {
- pmd->fds[i] = -1;
pmd->rxq[i].fd = -1;
pmd->txq[i].fd = -1;
}
- /* Take the TUN/TAP fd and place in the first location */
- pmd->rxq[0].fd = fd;
- pmd->txq[0].fd = fd;
- pmd->fds[0] = fd;
-
- if (pmd_mac_address(fd, dev, &pmd->eth_addr) < 0) {
- RTE_LOG(INFO, PMD, "Unable to get MAC address\n");
- goto error_exit;
- }
-
return 0;
error_exit:
- RTE_PMD_DEBUG_TRACE("Unable to initialize %s\n", name);
+ RTE_LOG(DEBUG, PMD, "TAP Unable to initialize %s\n", name);
rte_free(data);
rte_free(pmd);
snprintf(tap_name, sizeof(tap_name), "%s%d",
DEFAULT_TAP_NAME, tap_unit++);
- RTE_LOG(INFO, PMD, "Initializing pmd_tap for %s as %s\n",
- name, tap_name);
-
if (params && (params[0] != '\0')) {
- RTE_LOG(INFO, PMD, "paramaters (%s)\n", params);
+ RTE_LOG(DEBUG, PMD, "paramaters (%s)\n", params);
kvlist = rte_kvargs_parse(params, valid_arguments);
if (kvlist) {
}
pmd_link.link_speed = speed;
+ RTE_LOG(NOTICE, PMD, "Initializing pmd_tap for %s as %s\n",
+ name, tap_name);
+
ret = eth_dev_tap_create(name, tap_name);
leave:
if (ret == -1) {
- RTE_LOG(INFO, PMD, "Failed to create pmd for %s as %s\n",
+ RTE_LOG(ERR, PMD, "Failed to create pmd for %s as %s\n",
name, tap_name);
tap_unit--; /* Restore the unit number */
}
struct pmd_internals *internals;
int i;
- RTE_LOG(INFO, PMD, "Closing TUN/TAP Ethernet device on numa %u\n",
+ RTE_LOG(DEBUG, PMD, "Closing TUN/TAP Ethernet device on numa %u\n",
rte_socket_id());
/* find the ethdev entry */
internals = eth_dev->data->dev_private;
for (i = 0; i < internals->nb_queues; i++)
- if (internals->fds[i] != -1)
- close(internals->fds[i]);
+ if (internals->rxq[i].fd != -1)
+ close(internals->rxq[i].fd);
+ close(internals->ioctl_sock);
rte_free(eth_dev->data->dev_private);
rte_free(eth_dev->data);