examples/distributor: new sample app
authorReshma Pattan <reshma.pattan@intel.com>
Mon, 3 Nov 2014 15:49:44 +0000 (15:49 +0000)
committerThomas Monjalon <thomas.monjalon@6wind.com>
Sun, 16 Nov 2014 21:54:56 +0000 (22:54 +0100)
A new sample app that shows the usage of the distributor library. This
app works as follows:

* An RX thread runs which pulls packets from each ethernet port in turn
  and passes those packets to worker using a distributor component.
* The workers take the packets in turn, and determine the output port
  for those packets using basic l2forwarding doing an xor on the source
  port id.
* The RX thread takes the returned packets from the workers and enqueue
  those packets into an rte_ring structure.
* A TX thread pulls the packets off the rte_ring structure and then
  sends each packet out the output port specified previously by the worker
* Command-line option support provided only for portmask.

Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Signed-off-by: Reshma Pattan <reshma.pattan@intel.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
examples/Makefile
examples/distributor/Makefile [new file with mode: 0644]
examples/distributor/main.c [new file with mode: 0644]
examples/distributor/main.h [new file with mode: 0644]

index 48c4dfc..121dab4 100644 (file)
@@ -66,5 +66,6 @@ DIRS-$(CONFIG_RTE_LIBRTE_VHOST) += vhost
 DIRS-$(CONFIG_RTE_LIBRTE_XEN_DOM0) += vhost_xen
 DIRS-y += vmdq
 DIRS-y += vmdq_dcb
+DIRS-$(CONFIG_RTE_LIBRTE_DISTRIBUTOR) += distributor
 
 include $(RTE_SDK)/mk/rte.extsubdir.mk
diff --git a/examples/distributor/Makefile b/examples/distributor/Makefile
new file mode 100644 (file)
index 0000000..6a5bada
--- /dev/null
@@ -0,0 +1,57 @@
+#   BSD LICENSE
+#
+#   Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
+#   All rights reserved.
+#
+#   Redistribution and use in source and binary forms, with or without
+#   modification, are permitted provided that the following conditions
+#   are met:
+#
+#     * Redistributions of source code must retain the above copyright
+#       notice, this list of conditions and the following disclaimer.
+#     * Redistributions in binary form must reproduce the above copyright
+#       notice, this list of conditions and the following disclaimer in
+#       the documentation and/or other materials provided with the
+#       distribution.
+#     * Neither the name of Intel Corporation nor the names of its
+#       contributors may be used to endorse or promote products derived
+#       from this software without specific prior written permission.
+#
+#   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+#   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+#   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+#   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+#   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+#   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+#   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+#   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+#   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+#   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+#   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+ifeq ($(RTE_SDK),)
+$(error "Please define RTE_SDK environment variable")
+endif
+
+# Default target, can be overriden by command line or environment
+RTE_TARGET ?= x86_64-native-linuxapp-gcc
+
+include $(RTE_SDK)/mk/rte.vars.mk
+
+# binary name
+APP = distributor_app
+
+# all source are stored in SRCS-y
+SRCS-y := main.c
+
+CFLAGS += $(WERROR_FLAGS)
+
+# workaround for a gcc bug with noreturn attribute
+# http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603
+ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y)
+CFLAGS_main.o += -Wno-return-type
+endif
+
+EXTRA_CFLAGS += -O3 -Wfatal-errors
+
+include $(RTE_SDK)/mk/rte.extapp.mk
diff --git a/examples/distributor/main.c b/examples/distributor/main.c
new file mode 100644 (file)
index 0000000..52266f0
--- /dev/null
@@ -0,0 +1,606 @@
+/*-
+ *   BSD LICENSE
+ *
+ *   Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
+ *   All rights reserved.
+ *
+ *   Redistribution and use in source and binary forms, with or without
+ *   modification, are permitted provided that the following conditions
+ *   are met:
+ *
+ *     * Redistributions of source code must retain the above copyright
+ *       notice, this list of conditions and the following disclaimer.
+ *     * Redistributions in binary form must reproduce the above copyright
+ *       notice, this list of conditions and the following disclaimer in
+ *       the documentation and/or other materials provided with the
+ *       distribution.
+ *     * Neither the name of Intel Corporation nor the names of its
+ *       contributors may be used to endorse or promote products derived
+ *       from this software without specific prior written permission.
+ *
+ *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <stdint.h>
+#include <inttypes.h>
+#include <unistd.h>
+#include <signal.h>
+#include <getopt.h>
+
+#include <rte_eal.h>
+#include <rte_ethdev.h>
+#include <rte_cycles.h>
+#include <rte_malloc.h>
+#include <rte_debug.h>
+#include <rte_distributor.h>
+
+#include "main.h"
+
+#define RX_RING_SIZE 256
+#define TX_RING_SIZE 512
+#define NUM_MBUFS ((64*1024)-1)
+#define MBUF_SIZE (2048 + sizeof(struct rte_mbuf) + RTE_PKTMBUF_HEADROOM)
+#define MBUF_CACHE_SIZE 250
+#define BURST_SIZE 32
+#define RTE_RING_SZ 1024
+
+/* uncommnet below line to enable debug logs */
+/* #define DEBUG */
+
+#ifdef DEBUG
+#define LOG_LEVEL RTE_LOG_DEBUG
+#define LOG_DEBUG(log_type, fmt, args...) do { \
+       RTE_LOG(DEBUG, log_type, fmt, ##args)           \
+} while (0)
+#else
+#define LOG_LEVEL RTE_LOG_INFO
+#define LOG_DEBUG(log_type, fmt, args...) do {} while (0)
+#endif
+
+#define RTE_LOGTYPE_DISTRAPP RTE_LOGTYPE_USER1
+
+/* mask of enabled ports */
+static uint32_t enabled_port_mask;
+volatile uint8_t quit_signal;
+volatile uint8_t quit_signal_rx;
+
+static volatile struct app_stats {
+       struct {
+               uint64_t rx_pkts;
+               uint64_t returned_pkts;
+               uint64_t enqueued_pkts;
+       } rx __rte_cache_aligned;
+
+       struct {
+               uint64_t dequeue_pkts;
+               uint64_t tx_pkts;
+       } tx __rte_cache_aligned;
+} app_stats;
+
+static const struct rte_eth_conf port_conf_default = {
+       .rxmode = {
+               .mq_mode = ETH_MQ_RX_RSS,
+               .max_rx_pkt_len = ETHER_MAX_LEN,
+       },
+       .txmode = {
+               .mq_mode = ETH_MQ_TX_NONE,
+       },
+       .rx_adv_conf = {
+                       .rss_conf = {
+                               .rss_hf = ETH_RSS_IPV4 | ETH_RSS_IPV6 |
+                                       ETH_RSS_IPV4_TCP | ETH_RSS_IPV4_UDP |
+                                       ETH_RSS_IPV6_TCP | ETH_RSS_IPV6_UDP,
+                       }
+       },
+};
+
+struct output_buffer {
+       unsigned count;
+       struct rte_mbuf *mbufs[BURST_SIZE];
+};
+
+/*
+ * Initialises a given port using global settings and with the rx buffers
+ * coming from the mbuf_pool passed as parameter
+ */
+static inline int
+port_init(uint8_t port, struct rte_mempool *mbuf_pool)
+{
+       struct rte_eth_conf port_conf = port_conf_default;
+       const uint16_t rxRings = 1, txRings = rte_lcore_count() - 1;
+       int retval;
+       uint16_t q;
+
+       if (port >= rte_eth_dev_count())
+               return -1;
+
+       retval = rte_eth_dev_configure(port, rxRings, txRings, &port_conf);
+       if (retval != 0)
+               return retval;
+
+       for (q = 0; q < rxRings; q++) {
+               retval = rte_eth_rx_queue_setup(port, q, RX_RING_SIZE,
+                                               rte_eth_dev_socket_id(port),
+                                               NULL, mbuf_pool);
+               if (retval < 0)
+                       return retval;
+       }
+
+       for (q = 0; q < txRings; q++) {
+               retval = rte_eth_tx_queue_setup(port, q, TX_RING_SIZE,
+                                               rte_eth_dev_socket_id(port),
+                                               NULL);
+               if (retval < 0)
+                       return retval;
+       }
+
+       retval = rte_eth_dev_start(port);
+       if (retval < 0)
+               return retval;
+
+       struct rte_eth_link link;
+       rte_eth_link_get_nowait(port, &link);
+       if (!link.link_status) {
+               sleep(1);
+               rte_eth_link_get_nowait(port, &link);
+       }
+
+       if (!link.link_status) {
+               printf("Link down on port %"PRIu8"\n", port);
+               return 0;
+       }
+
+       struct ether_addr addr;
+       rte_eth_macaddr_get(port, &addr);
+       printf("Port %u MAC: %02"PRIx8" %02"PRIx8" %02"PRIx8
+                       " %02"PRIx8" %02"PRIx8" %02"PRIx8"\n",
+                       (unsigned)port,
+                       addr.addr_bytes[0], addr.addr_bytes[1],
+                       addr.addr_bytes[2], addr.addr_bytes[3],
+                       addr.addr_bytes[4], addr.addr_bytes[5]);
+
+       rte_eth_promiscuous_enable(port);
+
+       return 0;
+}
+
+struct lcore_params {
+       unsigned worker_id;
+       struct rte_distributor *d;
+       struct rte_ring *r;
+       struct rte_mempool *mem_pool;
+};
+
+static void
+quit_workers(struct rte_distributor *d, struct rte_mempool *p)
+{
+       const unsigned num_workers = rte_lcore_count() - 2;
+       unsigned i;
+       struct rte_mbuf *bufs[num_workers];
+       rte_mempool_get_bulk(p, (void *)bufs, num_workers);
+
+       for (i = 0; i < num_workers; i++)
+               bufs[i]->hash.rss = i << 1;
+
+       rte_distributor_process(d, bufs, num_workers);
+       rte_mempool_put_bulk(p, (void *)bufs, num_workers);
+}
+
+static int
+lcore_rx(struct lcore_params *p)
+{
+       struct rte_distributor *d = p->d;
+       struct rte_mempool *mem_pool = p->mem_pool;
+       struct rte_ring *r = p->r;
+       const uint8_t nb_ports = rte_eth_dev_count();
+       const int socket_id = rte_socket_id();
+       uint8_t port;
+
+       for (port = 0; port < nb_ports; port++) {
+               /* skip ports that are not enabled */
+               if ((enabled_port_mask & (1 << port)) == 0)
+                       continue;
+
+               if (rte_eth_dev_socket_id(port) > 0 &&
+                               rte_eth_dev_socket_id(port) != socket_id)
+                       printf("WARNING, port %u is on remote NUMA node to "
+                                       "RX thread.\n\tPerformance will not "
+                                       "be optimal.\n", port);
+       }
+
+       printf("\nCore %u doing packet RX.\n", rte_lcore_id());
+       port = 0;
+       while (!quit_signal_rx) {
+
+               /* skip ports that are not enabled */
+               if ((enabled_port_mask & (1 << port)) == 0) {
+                       if (++port == nb_ports)
+                               port = 0;
+                       continue;
+               }
+               struct rte_mbuf *bufs[BURST_SIZE*2];
+               const uint16_t nb_rx = rte_eth_rx_burst(port, 0, bufs,
+                               BURST_SIZE);
+               app_stats.rx.rx_pkts += nb_rx;
+
+               rte_distributor_process(d, bufs, nb_rx);
+               const uint16_t nb_ret = rte_distributor_returned_pkts(d,
+                               bufs, BURST_SIZE*2);
+               app_stats.rx.returned_pkts += nb_ret;
+               if (unlikely(nb_ret == 0))
+                       continue;
+
+               uint16_t sent = rte_ring_enqueue_burst(r, (void *)bufs, nb_ret);
+               app_stats.rx.enqueued_pkts += sent;
+               if (unlikely(sent < nb_ret)) {
+                       LOG_DEBUG(DISTRAPP, "%s:Packet loss due to full ring\n", __func__);
+                       while (sent < nb_ret)
+                               rte_pktmbuf_free(bufs[sent++]);
+               }
+               if (++port == nb_ports)
+                       port = 0;
+       }
+       rte_distributor_process(d, NULL, 0);
+       /* flush distributor to bring to known state */
+       rte_distributor_flush(d);
+       /* set worker & tx threads quit flag */
+       quit_signal = 1;
+       /*
+        * worker threads may hang in get packet as
+        * distributor process is not running, just make sure workers
+        * get packets till quit_signal is actually been
+        * received and they gracefully shutdown
+        */
+       quit_workers(d, mem_pool);
+       /* rx thread should quit at last */
+       return 0;
+}
+
+static inline void
+flush_one_port(struct output_buffer *outbuf, uint8_t outp)
+{
+       unsigned nb_tx = rte_eth_tx_burst(outp, 0, outbuf->mbufs,
+                       outbuf->count);
+       app_stats.tx.tx_pkts += nb_tx;
+
+       if (unlikely(nb_tx < outbuf->count)) {
+               LOG_DEBUG(DISTRAPP, "%s:Packet loss with tx_burst\n", __func__);
+               do {
+                       rte_pktmbuf_free(outbuf->mbufs[nb_tx]);
+               } while (++nb_tx < outbuf->count);
+       }
+       outbuf->count = 0;
+}
+
+static inline void
+flush_all_ports(struct output_buffer *tx_buffers, uint8_t nb_ports)
+{
+       uint8_t outp;
+       for (outp = 0; outp < nb_ports; outp++) {
+               /* skip ports that are not enabled */
+               if ((enabled_port_mask & (1 << outp)) == 0)
+                       continue;
+
+               if (tx_buffers[outp].count == 0)
+                       continue;
+
+               flush_one_port(&tx_buffers[outp], outp);
+       }
+}
+
+static int
+lcore_tx(struct rte_ring *in_r)
+{
+       static struct output_buffer tx_buffers[RTE_MAX_ETHPORTS];
+       const uint8_t nb_ports = rte_eth_dev_count();
+       const int socket_id = rte_socket_id();
+       uint8_t port;
+
+       for (port = 0; port < nb_ports; port++) {
+               /* skip ports that are not enabled */
+               if ((enabled_port_mask & (1 << port)) == 0)
+                       continue;
+
+               if (rte_eth_dev_socket_id(port) > 0 &&
+                               rte_eth_dev_socket_id(port) != socket_id)
+                       printf("WARNING, port %u is on remote NUMA node to "
+                                       "TX thread.\n\tPerformance will not "
+                                       "be optimal.\n", port);
+       }
+
+       printf("\nCore %u doing packet TX.\n", rte_lcore_id());
+       while (!quit_signal) {
+
+               for (port = 0; port < nb_ports; port++) {
+                       /* skip ports that are not enabled */
+                       if ((enabled_port_mask & (1 << port)) == 0)
+                               continue;
+
+                       struct rte_mbuf *bufs[BURST_SIZE];
+                       const uint16_t nb_rx = rte_ring_dequeue_burst(in_r,
+                                       (void *)bufs, BURST_SIZE);
+                       app_stats.tx.dequeue_pkts += nb_rx;
+
+                       /* if we get no traffic, flush anything we have */
+                       if (unlikely(nb_rx == 0)) {
+                               flush_all_ports(tx_buffers, nb_ports);
+                               continue;
+                       }
+
+                       /* for traffic we receive, queue it up for transmit */
+                       uint16_t i;
+                       _mm_prefetch(bufs[0], 0);
+                       _mm_prefetch(bufs[1], 0);
+                       _mm_prefetch(bufs[2], 0);
+                       for (i = 0; i < nb_rx; i++) {
+                               struct output_buffer *outbuf;
+                               uint8_t outp;
+                               _mm_prefetch(bufs[i + 3], 0);
+                               /*
+                                * workers should update in_port to hold the
+                                * output port value
+                                */
+                               outp = bufs[i]->port;
+                               /* skip ports that are not enabled */
+                               if ((enabled_port_mask & (1 << outp)) == 0)
+                                       continue;
+
+                               outbuf = &tx_buffers[outp];
+                               outbuf->mbufs[outbuf->count++] = bufs[i];
+                               if (outbuf->count == BURST_SIZE)
+                                       flush_one_port(outbuf, outp);
+                       }
+               }
+       }
+       return 0;
+}
+
+static void
+int_handler(int sig_num)
+{
+       printf("Exiting on signal %d\n", sig_num);
+       /* set quit flag for rx thread to exit */
+       quit_signal_rx = 1;
+}
+
+static void
+print_stats(void)
+{
+       struct rte_eth_stats eth_stats;
+       unsigned i;
+
+       printf("\nRX thread stats:\n");
+       printf(" - Received:    %"PRIu64"\n", app_stats.rx.rx_pkts);
+       printf(" - Processed:   %"PRIu64"\n", app_stats.rx.returned_pkts);
+       printf(" - Enqueued:    %"PRIu64"\n", app_stats.rx.enqueued_pkts);
+
+       printf("\nTX thread stats:\n");
+       printf(" - Dequeued:    %"PRIu64"\n", app_stats.tx.dequeue_pkts);
+       printf(" - Transmitted: %"PRIu64"\n", app_stats.tx.tx_pkts);
+
+       for (i = 0; i < rte_eth_dev_count(); i++) {
+               rte_eth_stats_get(i, &eth_stats);
+               printf("\nPort %u stats:\n", i);
+               printf(" - Pkts in:   %"PRIu64"\n", eth_stats.ipackets);
+               printf(" - Pkts out:  %"PRIu64"\n", eth_stats.opackets);
+               printf(" - In Errs:   %"PRIu64"\n", eth_stats.ierrors);
+               printf(" - Out Errs:  %"PRIu64"\n", eth_stats.oerrors);
+               printf(" - Mbuf Errs: %"PRIu64"\n", eth_stats.rx_nombuf);
+       }
+}
+
+static int
+lcore_worker(struct lcore_params *p)
+{
+       struct rte_distributor *d = p->d;
+       const unsigned id = p->worker_id;
+       /*
+        * for single port, xor_val will be zero so we won't modify the output
+        * port, otherwise we send traffic from 0 to 1, 2 to 3, and vice versa
+        */
+       const unsigned xor_val = (rte_eth_dev_count() > 1);
+       struct rte_mbuf *buf = NULL;
+
+       printf("\nCore %u acting as worker core.\n", rte_lcore_id());
+       while (!quit_signal) {
+               buf = rte_distributor_get_pkt(d, id, buf);
+               buf->port ^= xor_val;
+       }
+       return 0;
+}
+
+/* display usage */
+static void
+print_usage(const char *prgname)
+{
+       printf("%s [EAL options] -- -p PORTMASK\n"
+                       "  -p PORTMASK: hexadecimal bitmask of ports to configure\n",
+                       prgname);
+}
+
+static int
+parse_portmask(const char *portmask)
+{
+       char *end = NULL;
+       unsigned long pm;
+
+       /* parse hexadecimal string */
+       pm = strtoul(portmask, &end, 16);
+       if ((portmask[0] == '\0') || (end == NULL) || (*end != '\0'))
+               return -1;
+
+       if (pm == 0)
+               return -1;
+
+       return pm;
+}
+
+/* Parse the argument given in the command line of the application */
+static int
+parse_args(int argc, char **argv)
+{
+       int opt;
+       char **argvopt;
+       int option_index;
+       char *prgname = argv[0];
+       static struct option lgopts[] = {
+               {NULL, 0, 0, 0}
+       };
+
+       argvopt = argv;
+
+       while ((opt = getopt_long(argc, argvopt, "p:",
+                       lgopts, &option_index)) != EOF) {
+
+               switch (opt) {
+               /* portmask */
+               case 'p':
+                       enabled_port_mask = parse_portmask(optarg);
+                       if (enabled_port_mask == 0) {
+                               printf("invalid portmask\n");
+                               print_usage(prgname);
+                               return -1;
+                       }
+                       break;
+
+               default:
+                       print_usage(prgname);
+                       return -1;
+               }
+       }
+
+       if (optind <= 1) {
+               print_usage(prgname);
+               return -1;
+       }
+
+       argv[optind-1] = prgname;
+
+       optind = 0; /* reset getopt lib */
+       return 0;
+}
+
+/* Main function, does initialization and calls the per-lcore functions */
+int
+MAIN(int argc, char *argv[])
+{
+       struct rte_mempool *mbuf_pool;
+       struct rte_distributor *d;
+       struct rte_ring *output_ring;
+       unsigned lcore_id, worker_id = 0;
+       unsigned nb_ports;
+       uint8_t portid;
+       uint8_t nb_ports_available;
+
+       /* catch ctrl-c so we can print on exit */
+       signal(SIGINT, int_handler);
+
+       /* init EAL */
+       int ret = rte_eal_init(argc, argv);
+       if (ret < 0)
+               rte_exit(EXIT_FAILURE, "Error with EAL initialization\n");
+       argc -= ret;
+       argv += ret;
+
+       /* parse application arguments (after the EAL ones) */
+       ret = parse_args(argc, argv);
+       if (ret < 0)
+               rte_exit(EXIT_FAILURE, "Invalid distributor parameters\n");
+
+       if (rte_lcore_count() < 3)
+               rte_exit(EXIT_FAILURE, "Error, This application needs at "
+                               "least 3 logical cores to run:\n"
+                               "1 lcore for packet RX and distribution\n"
+                               "1 lcore for packet TX\n"
+                               "and at least 1 lcore for worker threads\n");
+
+       nb_ports = rte_eth_dev_count();
+       if (nb_ports == 0)
+               rte_exit(EXIT_FAILURE, "Error: no ethernet ports detected\n");
+       if (nb_ports != 1 && (nb_ports & 1))
+               rte_exit(EXIT_FAILURE, "Error: number of ports must be even, except "
+                               "when using a single port\n");
+
+       mbuf_pool = rte_mempool_create("MBUF_POOL", NUM_MBUFS * nb_ports,
+                       MBUF_SIZE, MBUF_CACHE_SIZE,
+                       sizeof(struct rte_pktmbuf_pool_private),
+                       rte_pktmbuf_pool_init, NULL,
+                       rte_pktmbuf_init, NULL,
+                       rte_socket_id(), 0);
+       if (mbuf_pool == NULL)
+               rte_exit(EXIT_FAILURE, "Cannot create mbuf pool\n");
+       nb_ports_available = nb_ports;
+
+       /* initialize all ports */
+       for (portid = 0; portid < nb_ports; portid++) {
+               /* skip ports that are not enabled */
+               if ((enabled_port_mask & (1 << portid)) == 0) {
+                       printf("\nSkipping disabled port %d\n", portid);
+                       nb_ports_available--;
+                       continue;
+               }
+               /* init port */
+               printf("Initializing port %u... done\n", (unsigned) portid);
+
+               if (port_init(portid, mbuf_pool) != 0)
+                       rte_exit(EXIT_FAILURE, "Cannot initialize port %"PRIu8"\n",
+                                       portid);
+       }
+
+       if (!nb_ports_available) {
+               rte_exit(EXIT_FAILURE,
+                               "All available ports are disabled. Please set portmask.\n");
+       }
+
+       d = rte_distributor_create("PKT_DIST", rte_socket_id(),
+                       rte_lcore_count() - 2);
+       if (d == NULL)
+               rte_exit(EXIT_FAILURE, "Cannot create distributor\n");
+
+       /*
+        * scheduler ring is read only by the transmitter core, but written to
+        * by multiple threads
+        */
+       output_ring = rte_ring_create("Output_ring", RTE_RING_SZ,
+                       rte_socket_id(), RING_F_SC_DEQ);
+       if (output_ring == NULL)
+               rte_exit(EXIT_FAILURE, "Cannot create output ring\n");
+
+       RTE_LCORE_FOREACH_SLAVE(lcore_id) {
+               if (worker_id == rte_lcore_count() - 2)
+                       rte_eal_remote_launch((lcore_function_t *)lcore_tx,
+                                       output_ring, lcore_id);
+               else {
+                       struct lcore_params *p =
+                                       rte_malloc(NULL, sizeof(*p), 0);
+                       if (!p)
+                               rte_panic("malloc failure\n");
+                       *p = (struct lcore_params){worker_id, d, output_ring, mbuf_pool};
+
+                       rte_eal_remote_launch((lcore_function_t *)lcore_worker,
+                                       p, lcore_id);
+               }
+               worker_id++;
+       }
+       /* call lcore_main on master core only */
+       struct lcore_params p = { 0, d, output_ring, mbuf_pool};
+       lcore_rx(&p);
+
+       RTE_LCORE_FOREACH_SLAVE(lcore_id) {
+               if (rte_eal_wait_lcore(lcore_id) < 0)
+                       return -1;
+       }
+
+       print_stats();
+       return 0;
+}
diff --git a/examples/distributor/main.h b/examples/distributor/main.h
new file mode 100644 (file)
index 0000000..2682d15
--- /dev/null
@@ -0,0 +1,46 @@
+/*-
+ *   BSD LICENSE
+ *
+ *   Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
+ *   All rights reserved.
+ *
+ *   Redistribution and use in source and binary forms, with or without
+ *   modification, are permitted provided that the following conditions
+ *   are met:
+ *
+ *     * Redistributions of source code must retain the above copyright
+ *       notice, this list of conditions and the following disclaimer.
+ *     * Redistributions in binary form must reproduce the above copyright
+ *       notice, this list of conditions and the following disclaimer in
+ *       the documentation and/or other materials provided with the
+ *       distribution.
+ *     * Neither the name of Intel Corporation nor the names of its
+ *       contributors may be used to endorse or promote products derived
+ *       from this software without specific prior written permission.
+ *
+ *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#ifndef _MAIN_H_
+#define _MAIN_H_
+
+
+#ifdef RTE_EXEC_ENV_BAREMETAL
+#define MAIN _main
+#else
+#define MAIN main
+#endif
+
+int MAIN(int argc, char *argv[]);
+
+#endif /* ifndef _MAIN_H_ */