net/pcap: fix possible mbuf double freeing
[dpdk.git] / drivers / net / pcap / rte_eth_pcap.c
index a015a9d..d1dca2a 100644 (file)
@@ -7,6 +7,14 @@
 #include <time.h>
 
 #include <net/if.h>
+#include <sys/socket.h>
+#include <sys/ioctl.h>
+#include <unistd.h>
+
+#if defined(RTE_EXEC_ENV_FREEBSD)
+#include <sys/sysctl.h>
+#include <net/if_dl.h>
+#endif
 
 #include <pcap.h>
 
 #include <rte_malloc.h>
 #include <rte_mbuf.h>
 #include <rte_bus_vdev.h>
+#include <rte_string_fns.h>
 
 #define RTE_ETH_PCAP_SNAPSHOT_LEN 65535
-#define RTE_ETH_PCAP_SNAPLEN ETHER_MAX_JUMBO_FRAME_LEN
+#define RTE_ETH_PCAP_SNAPLEN RTE_ETHER_MAX_JUMBO_FRAME_LEN
 #define RTE_ETH_PCAP_PROMISC 1
 #define RTE_ETH_PCAP_TIMEOUT -1
 
@@ -29,6 +38,8 @@
 #define ETH_PCAP_RX_IFACE_IN_ARG "rx_iface_in"
 #define ETH_PCAP_TX_IFACE_ARG "tx_iface"
 #define ETH_PCAP_IFACE_ARG    "iface"
+#define ETH_PCAP_PHY_MAC_ARG  "phy_mac"
+#define ETH_PCAP_INFINITE_RX_ARG  "infinite_rx"
 
 #define ETH_PCAP_ARG_MAXLEN    64
 
@@ -39,6 +50,7 @@ static unsigned char tx_pcap_data[RTE_ETH_PCAP_SNAPLEN];
 static struct timeval start_time;
 static uint64_t start_cycles;
 static uint64_t hz;
+static uint8_t iface_idx;
 
 struct queue_stat {
        volatile unsigned long pkts;
@@ -47,17 +59,20 @@ struct queue_stat {
 };
 
 struct pcap_rx_queue {
-       pcap_t *pcap;
-       uint16_t in_port;
+       uint16_t port_id;
+       uint16_t queue_id;
        struct rte_mempool *mb_pool;
        struct queue_stat rx_stat;
        char name[PATH_MAX];
        char type[ETH_PCAP_ARG_MAXLEN];
+
+       /* Contains pre-generated packets to be looped through */
+       struct rte_ring *pkts;
 };
 
 struct pcap_tx_queue {
-       pcap_dumper_t *dumper;
-       pcap_t *pcap;
+       uint16_t port_id;
+       uint16_t queue_id;
        struct queue_stat tx_stat;
        char name[PATH_MAX];
        char type[ETH_PCAP_ARG_MAXLEN];
@@ -66,8 +81,18 @@ struct pcap_tx_queue {
 struct pmd_internals {
        struct pcap_rx_queue rx_queue[RTE_PMD_PCAP_MAX_QUEUES];
        struct pcap_tx_queue tx_queue[RTE_PMD_PCAP_MAX_QUEUES];
+       char devargs[ETH_PCAP_ARG_MAXLEN];
+       struct rte_ether_addr eth_addr;
        int if_index;
        int single_iface;
+       int phy_mac;
+       unsigned int infinite_rx;
+};
+
+struct pmd_process_private {
+       pcap_t *rx_pcap[RTE_PMD_PCAP_MAX_QUEUES];
+       pcap_t *tx_pcap[RTE_PMD_PCAP_MAX_QUEUES];
+       pcap_dumper_t *tx_dumper[RTE_PMD_PCAP_MAX_QUEUES];
 };
 
 struct pmd_devargs {
@@ -78,6 +103,16 @@ struct pmd_devargs {
                const char *name;
                const char *type;
        } queue[RTE_PMD_PCAP_MAX_QUEUES];
+       int phy_mac;
+};
+
+struct pmd_devargs_all {
+       struct pmd_devargs rx_queues;
+       struct pmd_devargs tx_queues;
+       int single_iface;
+       unsigned int is_tx_pcap;
+       unsigned int is_tx_iface;
+       unsigned int infinite_rx;
 };
 
 static const char *valid_arguments[] = {
@@ -87,13 +122,11 @@ static const char *valid_arguments[] = {
        ETH_PCAP_RX_IFACE_IN_ARG,
        ETH_PCAP_TX_IFACE_ARG,
        ETH_PCAP_IFACE_ARG,
+       ETH_PCAP_PHY_MAC_ARG,
+       ETH_PCAP_INFINITE_RX_ARG,
        NULL
 };
 
-static struct ether_addr eth_addr = {
-       .addr_bytes = { 0, 0, 0, 0x1, 0x2, 0x3 }
-};
-
 static struct rte_eth_link pmd_link = {
                .link_speed = ETH_SPEED_NUM_10G,
                .link_duplex = ETH_LINK_FULL_DUPLEX,
@@ -160,19 +193,61 @@ eth_pcap_gather_data(unsigned char *data, struct rte_mbuf *mbuf)
        }
 }
 
+static uint16_t
+eth_pcap_rx_infinite(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
+{
+       int i;
+       struct pcap_rx_queue *pcap_q = queue;
+       uint32_t rx_bytes = 0;
+
+       if (unlikely(nb_pkts == 0))
+               return 0;
+
+       if (rte_pktmbuf_alloc_bulk(pcap_q->mb_pool, bufs, nb_pkts) != 0)
+               return 0;
+
+       for (i = 0; i < nb_pkts; i++) {
+               struct rte_mbuf *pcap_buf;
+               int err = rte_ring_dequeue(pcap_q->pkts, (void **)&pcap_buf);
+               if (err)
+                       return i;
+
+               rte_memcpy(rte_pktmbuf_mtod(bufs[i], void *),
+                               rte_pktmbuf_mtod(pcap_buf, void *),
+                               pcap_buf->data_len);
+               bufs[i]->data_len = pcap_buf->data_len;
+               bufs[i]->pkt_len = pcap_buf->pkt_len;
+               bufs[i]->port = pcap_q->port_id;
+               rx_bytes += pcap_buf->data_len;
+
+               /* Enqueue packet back on ring to allow infinite rx. */
+               rte_ring_enqueue(pcap_q->pkts, pcap_buf);
+       }
+
+       pcap_q->rx_stat.pkts += i;
+       pcap_q->rx_stat.bytes += rx_bytes;
+
+       return i;
+}
+
 static uint16_t
 eth_pcap_rx(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
 {
        unsigned int i;
        struct pcap_pkthdr header;
+       struct pmd_process_private *pp;
        const u_char *packet;
        struct rte_mbuf *mbuf;
        struct pcap_rx_queue *pcap_q = queue;
        uint16_t num_rx = 0;
        uint16_t buf_size;
        uint32_t rx_bytes = 0;
+       pcap_t *pcap;
+
+       pp = rte_eth_devices[pcap_q->port_id].process_private;
+       pcap = pp->rx_pcap[pcap_q->queue_id];
 
-       if (unlikely(pcap_q->pcap == NULL || nb_pkts == 0))
+       if (unlikely(pcap == NULL || nb_pkts == 0))
                return 0;
 
        /* Reads the given number of packets from the pcap file one by one
@@ -180,7 +255,7 @@ eth_pcap_rx(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
         */
        for (i = 0; i < nb_pkts; i++) {
                /* Get the next PCAP packet */
-               packet = pcap_next(pcap_q->pcap, &header);
+               packet = pcap_next(pcap, &header);
                if (unlikely(packet == NULL))
                        break;
 
@@ -209,7 +284,7 @@ eth_pcap_rx(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
                }
 
                mbuf->pkt_len = (uint16_t)header.caplen;
-               mbuf->port = pcap_q->in_port;
+               mbuf->port = pcap_q->port_id;
                bufs[num_rx] = mbuf;
                num_rx++;
                rx_bytes += header.caplen;
@@ -239,12 +314,17 @@ eth_pcap_tx_dumper(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
 {
        unsigned int i;
        struct rte_mbuf *mbuf;
+       struct pmd_process_private *pp;
        struct pcap_tx_queue *dumper_q = queue;
        uint16_t num_tx = 0;
        uint32_t tx_bytes = 0;
        struct pcap_pkthdr header;
+       pcap_dumper_t *dumper;
+
+       pp = rte_eth_devices[dumper_q->port_id].process_private;
+       dumper = pp->tx_dumper[dumper_q->queue_id];
 
-       if (dumper_q->dumper == NULL || nb_pkts == 0)
+       if (dumper == NULL || nb_pkts == 0)
                return 0;
 
        /* writes the nb_pkts packets to the previously opened pcap file
@@ -256,20 +336,19 @@ eth_pcap_tx_dumper(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
                header.caplen = header.len;
 
                if (likely(mbuf->nb_segs == 1)) {
-                       pcap_dump((u_char *)dumper_q->dumper, &header,
+                       pcap_dump((u_char *)dumper, &header,
                                  rte_pktmbuf_mtod(mbuf, void*));
                } else {
-                       if (mbuf->pkt_len <= ETHER_MAX_JUMBO_FRAME_LEN) {
+                       if (mbuf->pkt_len <= RTE_ETHER_MAX_JUMBO_FRAME_LEN) {
                                eth_pcap_gather_data(tx_pcap_data, mbuf);
-                               pcap_dump((u_char *)dumper_q->dumper, &header,
+                               pcap_dump((u_char *)dumper, &header,
                                          tx_pcap_data);
                        } else {
                                PMD_LOG(ERR,
                                        "Dropping PCAP packet. Size (%d) > max jumbo size (%d).",
                                        mbuf->pkt_len,
-                                       ETHER_MAX_JUMBO_FRAME_LEN);
+                                       RTE_ETHER_MAX_JUMBO_FRAME_LEN);
 
-                               rte_pktmbuf_free(mbuf);
                                break;
                        }
                }
@@ -284,7 +363,7 @@ eth_pcap_tx_dumper(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
         * process stops and to make sure the pcap file is actually written,
         * we flush the pcap dumper within each burst.
         */
-       pcap_dump_flush(dumper_q->dumper);
+       pcap_dump_flush(dumper);
        dumper_q->tx_stat.pkts += num_tx;
        dumper_q->tx_stat.bytes += tx_bytes;
        dumper_q->tx_stat.err_pkts += nb_pkts - num_tx;
@@ -292,6 +371,30 @@ eth_pcap_tx_dumper(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
        return num_tx;
 }
 
+/*
+ * Callback to handle dropping packets in the infinite rx case.
+ */
+static uint16_t
+eth_tx_drop(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
+{
+       unsigned int i;
+       uint32_t tx_bytes = 0;
+       struct pcap_tx_queue *tx_queue = queue;
+
+       if (unlikely(nb_pkts == 0))
+               return 0;
+
+       for (i = 0; i < nb_pkts; i++) {
+               tx_bytes += bufs[i]->data_len;
+               rte_pktmbuf_free(bufs[i]);
+       }
+
+       tx_queue->tx_stat.pkts += nb_pkts;
+       tx_queue->tx_stat.bytes += tx_bytes;
+
+       return i;
+}
+
 /*
  * Callback to handle sending packets through a real NIC.
  */
@@ -301,32 +404,36 @@ eth_pcap_tx(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
        unsigned int i;
        int ret;
        struct rte_mbuf *mbuf;
+       struct pmd_process_private *pp;
        struct pcap_tx_queue *tx_queue = queue;
        uint16_t num_tx = 0;
        uint32_t tx_bytes = 0;
+       pcap_t *pcap;
 
-       if (unlikely(nb_pkts == 0 || tx_queue->pcap == NULL))
+       pp = rte_eth_devices[tx_queue->port_id].process_private;
+       pcap = pp->tx_pcap[tx_queue->queue_id];
+
+       if (unlikely(nb_pkts == 0 || pcap == NULL))
                return 0;
 
        for (i = 0; i < nb_pkts; i++) {
                mbuf = bufs[i];
 
                if (likely(mbuf->nb_segs == 1)) {
-                       ret = pcap_sendpacket(tx_queue->pcap,
+                       ret = pcap_sendpacket(pcap,
                                        rte_pktmbuf_mtod(mbuf, u_char *),
                                        mbuf->pkt_len);
                } else {
-                       if (mbuf->pkt_len <= ETHER_MAX_JUMBO_FRAME_LEN) {
+                       if (mbuf->pkt_len <= RTE_ETHER_MAX_JUMBO_FRAME_LEN) {
                                eth_pcap_gather_data(tx_pcap_data, mbuf);
-                               ret = pcap_sendpacket(tx_queue->pcap,
+                               ret = pcap_sendpacket(pcap,
                                                tx_pcap_data, mbuf->pkt_len);
                        } else {
                                PMD_LOG(ERR,
                                        "Dropping PCAP packet. Size (%d) > max jumbo size (%d).",
                                        mbuf->pkt_len,
-                                       ETHER_MAX_JUMBO_FRAME_LEN);
+                                       RTE_ETHER_MAX_JUMBO_FRAME_LEN);
 
-                               rte_pktmbuf_free(mbuf);
                                break;
                        }
                }
@@ -414,11 +521,30 @@ open_single_rx_pcap(const char *pcap_filename, pcap_t **pcap)
        return 0;
 }
 
+static uint64_t
+count_packets_in_pcap(pcap_t **pcap, struct pcap_rx_queue *pcap_q)
+{
+       const u_char *packet;
+       struct pcap_pkthdr header;
+       uint64_t pcap_pkt_count = 0;
+
+       while ((packet = pcap_next(*pcap, &header)))
+               pcap_pkt_count++;
+
+       /* The pcap is reopened so it can be used as normal later. */
+       pcap_close(*pcap);
+       *pcap = NULL;
+       open_single_rx_pcap(pcap_q->name, pcap);
+
+       return pcap_pkt_count;
+}
+
 static int
 eth_dev_start(struct rte_eth_dev *dev)
 {
        unsigned int i;
        struct pmd_internals *internals = dev->data->dev_private;
+       struct pmd_process_private *pp = dev->process_private;
        struct pcap_tx_queue *tx;
        struct pcap_rx_queue *rx;
 
@@ -427,10 +553,11 @@ eth_dev_start(struct rte_eth_dev *dev)
                tx = &internals->tx_queue[0];
                rx = &internals->rx_queue[0];
 
-               if (!tx->pcap && strcmp(tx->type, ETH_PCAP_IFACE_ARG) == 0) {
-                       if (open_single_iface(tx->name, &tx->pcap) < 0)
+               if (!pp->tx_pcap[0] &&
+                       strcmp(tx->type, ETH_PCAP_IFACE_ARG) == 0) {
+                       if (open_single_iface(tx->name, &pp->tx_pcap[0]) < 0)
                                return -1;
-                       rx->pcap = tx->pcap;
+                       pp->rx_pcap[0] = pp->tx_pcap[0];
                }
 
                goto status_up;
@@ -440,13 +567,14 @@ eth_dev_start(struct rte_eth_dev *dev)
        for (i = 0; i < dev->data->nb_tx_queues; i++) {
                tx = &internals->tx_queue[i];
 
-               if (!tx->dumper &&
+               if (!pp->tx_dumper[i] &&
                                strcmp(tx->type, ETH_PCAP_TX_PCAP_ARG) == 0) {
-                       if (open_single_tx_pcap(tx->name, &tx->dumper) < 0)
+                       if (open_single_tx_pcap(tx->name,
+                               &pp->tx_dumper[i]) < 0)
                                return -1;
-               } else if (!tx->pcap &&
+               } else if (!pp->tx_pcap[i] &&
                                strcmp(tx->type, ETH_PCAP_TX_IFACE_ARG) == 0) {
-                       if (open_single_iface(tx->name, &tx->pcap) < 0)
+                       if (open_single_iface(tx->name, &pp->tx_pcap[i]) < 0)
                                return -1;
                }
        }
@@ -455,14 +583,14 @@ eth_dev_start(struct rte_eth_dev *dev)
        for (i = 0; i < dev->data->nb_rx_queues; i++) {
                rx = &internals->rx_queue[i];
 
-               if (rx->pcap != NULL)
+               if (pp->rx_pcap[i] != NULL)
                        continue;
 
                if (strcmp(rx->type, ETH_PCAP_RX_PCAP_ARG) == 0) {
-                       if (open_single_rx_pcap(rx->name, &rx->pcap) < 0)
+                       if (open_single_rx_pcap(rx->name, &pp->rx_pcap[i]) < 0)
                                return -1;
                } else if (strcmp(rx->type, ETH_PCAP_RX_IFACE_ARG) == 0) {
-                       if (open_single_iface(rx->name, &rx->pcap) < 0)
+                       if (open_single_iface(rx->name, &pp->rx_pcap[i]) < 0)
                                return -1;
                }
        }
@@ -489,39 +617,32 @@ eth_dev_stop(struct rte_eth_dev *dev)
 {
        unsigned int i;
        struct pmd_internals *internals = dev->data->dev_private;
-       struct pcap_tx_queue *tx;
-       struct pcap_rx_queue *rx;
+       struct pmd_process_private *pp = dev->process_private;
 
        /* Special iface case. Single pcap is open and shared between tx/rx. */
        if (internals->single_iface) {
-               tx = &internals->tx_queue[0];
-               rx = &internals->rx_queue[0];
-               pcap_close(tx->pcap);
-               tx->pcap = NULL;
-               rx->pcap = NULL;
+               pcap_close(pp->tx_pcap[0]);
+               pp->tx_pcap[0] = NULL;
+               pp->rx_pcap[0] = NULL;
                goto status_down;
        }
 
        for (i = 0; i < dev->data->nb_tx_queues; i++) {
-               tx = &internals->tx_queue[i];
-
-               if (tx->dumper != NULL) {
-                       pcap_dump_close(tx->dumper);
-                       tx->dumper = NULL;
+               if (pp->tx_dumper[i] != NULL) {
+                       pcap_dump_close(pp->tx_dumper[i]);
+                       pp->tx_dumper[i] = NULL;
                }
 
-               if (tx->pcap != NULL) {
-                       pcap_close(tx->pcap);
-                       tx->pcap = NULL;
+               if (pp->tx_pcap[i] != NULL) {
+                       pcap_close(pp->tx_pcap[i]);
+                       pp->tx_pcap[i] = NULL;
                }
        }
 
        for (i = 0; i < dev->data->nb_rx_queues; i++) {
-               rx = &internals->rx_queue[i];
-
-               if (rx->pcap != NULL) {
-                       pcap_close(rx->pcap);
-                       rx->pcap = NULL;
+               if (pp->rx_pcap[i] != NULL) {
+                       pcap_close(pp->rx_pcap[i]);
+                       pp->rx_pcap[i] = NULL;
                }
        }
 
@@ -576,10 +697,9 @@ eth_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *stats)
                        i < dev->data->nb_tx_queues; i++) {
                stats->q_opackets[i] = internal->tx_queue[i].tx_stat.pkts;
                stats->q_obytes[i] = internal->tx_queue[i].tx_stat.bytes;
-               stats->q_errors[i] = internal->tx_queue[i].tx_stat.err_pkts;
                tx_packets_total += stats->q_opackets[i];
                tx_bytes_total += stats->q_obytes[i];
-               tx_packets_err_total += stats->q_errors[i];
+               tx_packets_err_total += internal->tx_queue[i].tx_stat.err_pkts;
        }
 
        stats->ipackets = rx_packets_total;
@@ -610,8 +730,25 @@ eth_stats_reset(struct rte_eth_dev *dev)
 }
 
 static void
-eth_dev_close(struct rte_eth_dev *dev __rte_unused)
+eth_dev_close(struct rte_eth_dev *dev)
 {
+       unsigned int i;
+       struct pmd_internals *internals = dev->data->dev_private;
+
+       /* Device wide flag, but cleanup must be performed per queue. */
+       if (internals->infinite_rx) {
+               for (i = 0; i < dev->data->nb_rx_queues; i++) {
+                       struct pcap_rx_queue *pcap_q = &internals->rx_queue[i];
+                       struct rte_mbuf *pcap_buf;
+
+                       while (!rte_ring_dequeue(pcap_q->pkts,
+                                       (void **)&pcap_buf))
+                               rte_pktmbuf_free(pcap_buf);
+
+                       rte_ring_free(pcap_q->pkts);
+               }
+       }
+
 }
 
 static void
@@ -638,8 +775,62 @@ eth_rx_queue_setup(struct rte_eth_dev *dev,
        struct pcap_rx_queue *pcap_q = &internals->rx_queue[rx_queue_id];
 
        pcap_q->mb_pool = mb_pool;
+       pcap_q->port_id = dev->data->port_id;
+       pcap_q->queue_id = rx_queue_id;
        dev->data->rx_queues[rx_queue_id] = pcap_q;
-       pcap_q->in_port = dev->data->port_id;
+
+       if (internals->infinite_rx) {
+               struct pmd_process_private *pp;
+               char ring_name[NAME_MAX];
+               static uint32_t ring_number;
+               uint64_t pcap_pkt_count = 0;
+               struct rte_mbuf *bufs[1];
+               pcap_t **pcap;
+
+               pp = rte_eth_devices[pcap_q->port_id].process_private;
+               pcap = &pp->rx_pcap[pcap_q->queue_id];
+
+               if (unlikely(*pcap == NULL))
+                       return -ENOENT;
+
+               pcap_pkt_count = count_packets_in_pcap(pcap, pcap_q);
+
+               snprintf(ring_name, sizeof(ring_name), "PCAP_RING%" PRIu16,
+                               ring_number);
+
+               pcap_q->pkts = rte_ring_create(ring_name,
+                               rte_align64pow2(pcap_pkt_count + 1), 0,
+                               RING_F_SP_ENQ | RING_F_SC_DEQ);
+               ring_number++;
+               if (!pcap_q->pkts)
+                       return -ENOENT;
+
+               /* Fill ring with packets from PCAP file one by one. */
+               while (eth_pcap_rx(pcap_q, bufs, 1)) {
+                       /* Check for multiseg mbufs. */
+                       if (bufs[0]->nb_segs != 1) {
+                               rte_pktmbuf_free(*bufs);
+
+                               while (!rte_ring_dequeue(pcap_q->pkts,
+                                               (void **)bufs))
+                                       rte_pktmbuf_free(*bufs);
+
+                               rte_ring_free(pcap_q->pkts);
+                               PMD_LOG(ERR, "Multiseg mbufs are not supported in infinite_rx "
+                                               "mode.");
+                               return -EINVAL;
+                       }
+
+                       rte_ring_enqueue_bulk(pcap_q->pkts,
+                                       (void * const *)bufs, 1, NULL);
+               }
+               /*
+                * Reset the stats for this queue since eth_pcap_rx calls above
+                * didn't result in the application receiving packets.
+                */
+               pcap_q->rx_stat.pkts = 0;
+               pcap_q->rx_stat.bytes = 0;
+       }
 
        return 0;
 }
@@ -652,8 +843,11 @@ eth_tx_queue_setup(struct rte_eth_dev *dev,
                const struct rte_eth_txconf *tx_conf __rte_unused)
 {
        struct pmd_internals *internals = dev->data->dev_private;
+       struct pcap_tx_queue *pcap_q = &internals->tx_queue[tx_queue_id];
 
-       dev->data->tx_queues[tx_queue_id] = &internals->tx_queue[tx_queue_id];
+       pcap_q->port_id = dev->data->port_id;
+       pcap_q->queue_id = tx_queue_id;
+       dev->data->tx_queues[tx_queue_id] = pcap_q;
 
        return 0;
 }
@@ -861,7 +1055,33 @@ open_tx_iface(const char *key, const char *value, void *extra_args)
        return open_iface(key, value, extra_args);
 }
 
-static struct rte_vdev_driver pmd_pcap_drv;
+static int
+select_phy_mac(const char *key __rte_unused, const char *value,
+               void *extra_args)
+{
+       if (extra_args) {
+               const int phy_mac = atoi(value);
+               int *enable_phy_mac = extra_args;
+
+               if (phy_mac)
+                       *enable_phy_mac = 1;
+       }
+       return 0;
+}
+
+static int
+get_infinite_rx_arg(const char *key __rte_unused,
+               const char *value, void *extra_args)
+{
+       if (extra_args) {
+               const int infinite_rx = atoi(value);
+               int *enable_infinite_rx = extra_args;
+
+               if (infinite_rx > 0)
+                       *enable_infinite_rx = 1;
+       }
+       return 0;
+}
 
 static int
 pmd_init_internals(struct rte_vdev_device *vdev,
@@ -871,16 +1091,29 @@ pmd_init_internals(struct rte_vdev_device *vdev,
                struct rte_eth_dev **eth_dev)
 {
        struct rte_eth_dev_data *data;
+       struct pmd_process_private *pp;
        unsigned int numa_node = vdev->device.numa_node;
 
        PMD_LOG(INFO, "Creating pcap-backed ethdev on numa socket %d",
                numa_node);
 
+       pp = (struct pmd_process_private *)
+               rte_zmalloc(NULL, sizeof(struct pmd_process_private),
+                               RTE_CACHE_LINE_SIZE);
+
+       if (pp == NULL) {
+               PMD_LOG(ERR,
+                       "Failed to allocate memory for process private");
+               return -1;
+       }
+
        /* reserve an ethdev entry */
        *eth_dev = rte_eth_vdev_allocate(vdev, sizeof(**internals));
-       if (!(*eth_dev))
+       if (!(*eth_dev)) {
+               rte_free(pp);
                return -1;
-
+       }
+       (*eth_dev)->process_private = pp;
        /* now put it all together
         * - store queue data in internals,
         * - store numa_node info in eth_dev
@@ -888,11 +1121,20 @@ pmd_init_internals(struct rte_vdev_device *vdev,
         * - and point eth_dev structure to new eth_dev_data structure
         */
        *internals = (*eth_dev)->data->dev_private;
+       /*
+        * Interface MAC = 02:70:63:61:70:<iface_idx>
+        * derived from: 'locally administered':'p':'c':'a':'p':'iface_idx'
+        * where the middle 4 characters are converted to hex.
+        */
+       (*internals)->eth_addr = (struct rte_ether_addr) {
+               .addr_bytes = { 0x02, 0x70, 0x63, 0x61, 0x70, iface_idx++ }
+       };
+       (*internals)->phy_mac = 0;
        data = (*eth_dev)->data;
        data->nb_rx_queues = (uint16_t)nb_rx_queues;
        data->nb_tx_queues = (uint16_t)nb_tx_queues;
        data->dev_link = pmd_link;
-       data->mac_addrs = &eth_addr;
+       data->mac_addrs = &(*internals)->eth_addr;
 
        /*
         * NOTE: we'll replace the data element, of originally allocated
@@ -900,18 +1142,106 @@ pmd_init_internals(struct rte_vdev_device *vdev,
         */
        (*eth_dev)->dev_ops = &ops;
 
+       strlcpy((*internals)->devargs, rte_vdev_device_args(vdev),
+                       ETH_PCAP_ARG_MAXLEN);
+
+       return 0;
+}
+
+static int
+eth_pcap_update_mac(const char *if_name, struct rte_eth_dev *eth_dev,
+               const unsigned int numa_node)
+{
+#if defined(RTE_EXEC_ENV_LINUX)
+       void *mac_addrs;
+       struct ifreq ifr;
+       int if_fd = socket(AF_INET, SOCK_DGRAM, 0);
+
+       if (if_fd == -1)
+               return -1;
+
+       rte_strscpy(ifr.ifr_name, if_name, sizeof(ifr.ifr_name));
+       if (ioctl(if_fd, SIOCGIFHWADDR, &ifr)) {
+               close(if_fd);
+               return -1;
+       }
+
+       mac_addrs = rte_zmalloc_socket(NULL, RTE_ETHER_ADDR_LEN, 0, numa_node);
+       if (!mac_addrs) {
+               close(if_fd);
+               return -1;
+       }
+
+       PMD_LOG(INFO, "Setting phy MAC for %s", if_name);
+       eth_dev->data->mac_addrs = mac_addrs;
+       rte_memcpy(eth_dev->data->mac_addrs[0].addr_bytes,
+                       ifr.ifr_hwaddr.sa_data, RTE_ETHER_ADDR_LEN);
+
+       close(if_fd);
+
        return 0;
+
+#elif defined(RTE_EXEC_ENV_FREEBSD)
+       void *mac_addrs;
+       struct if_msghdr *ifm;
+       struct sockaddr_dl *sdl;
+       int mib[6];
+       size_t len = 0;
+       char *buf;
+
+       mib[0] = CTL_NET;
+       mib[1] = AF_ROUTE;
+       mib[2] = 0;
+       mib[3] = AF_LINK;
+       mib[4] = NET_RT_IFLIST;
+       mib[5] = if_nametoindex(if_name);
+
+       if (sysctl(mib, 6, NULL, &len, NULL, 0) < 0)
+               return -1;
+
+       if (len == 0)
+               return -1;
+
+       buf = rte_malloc(NULL, len, 0);
+       if (!buf)
+               return -1;
+
+       if (sysctl(mib, 6, buf, &len, NULL, 0) < 0) {
+               rte_free(buf);
+               return -1;
+       }
+       ifm = (struct if_msghdr *)buf;
+       sdl = (struct sockaddr_dl *)(ifm + 1);
+
+       mac_addrs = rte_zmalloc_socket(NULL, RTE_ETHER_ADDR_LEN, 0, numa_node);
+       if (!mac_addrs) {
+               rte_free(buf);
+               return -1;
+       }
+
+       PMD_LOG(INFO, "Setting phy MAC for %s", if_name);
+       eth_dev->data->mac_addrs = mac_addrs;
+       rte_memcpy(eth_dev->data->mac_addrs[0].addr_bytes,
+                       LLADDR(sdl), RTE_ETHER_ADDR_LEN);
+
+       rte_free(buf);
+
+       return 0;
+#else
+       return -1;
+#endif
 }
 
 static int
 eth_from_pcaps_common(struct rte_vdev_device *vdev,
-               struct pmd_devargs *rx_queues, const unsigned int nb_rx_queues,
-               struct pmd_devargs *tx_queues, const unsigned int nb_tx_queues,
-               struct rte_kvargs *kvlist, struct pmd_internals **internals,
-               struct rte_eth_dev **eth_dev)
+               struct pmd_devargs_all *devargs_all,
+               struct pmd_internals **internals, struct rte_eth_dev **eth_dev)
 {
-       struct rte_kvargs_pair *pair = NULL;
-       unsigned int k_idx;
+       struct pmd_process_private *pp;
+       struct pmd_devargs *rx_queues = &devargs_all->rx_queues;
+       struct pmd_devargs *tx_queues = &devargs_all->tx_queues;
+       const unsigned int nb_rx_queues = rx_queues->num_of_queue;
+       const unsigned int nb_tx_queues = tx_queues->num_of_queue;
        unsigned int i;
 
        /* do some parameter checking */
@@ -924,52 +1254,41 @@ eth_from_pcaps_common(struct rte_vdev_device *vdev,
                        eth_dev) < 0)
                return -1;
 
+       pp = (*eth_dev)->process_private;
        for (i = 0; i < nb_rx_queues; i++) {
                struct pcap_rx_queue *rx = &(*internals)->rx_queue[i];
                struct devargs_queue *queue = &rx_queues->queue[i];
 
-               rx->pcap = queue->pcap;
-               snprintf(rx->name, sizeof(rx->name), "%s", queue->name);
-               snprintf(rx->type, sizeof(rx->type), "%s", queue->type);
+               pp->rx_pcap[i] = queue->pcap;
+               strlcpy(rx->name, queue->name, sizeof(rx->name));
+               strlcpy(rx->type, queue->type, sizeof(rx->type));
        }
 
        for (i = 0; i < nb_tx_queues; i++) {
                struct pcap_tx_queue *tx = &(*internals)->tx_queue[i];
                struct devargs_queue *queue = &tx_queues->queue[i];
 
-               tx->dumper = queue->dumper;
-               tx->pcap = queue->pcap;
-               snprintf(tx->name, sizeof(tx->name), "%s", queue->name);
-               snprintf(tx->type, sizeof(tx->type), "%s", queue->type);
-       }
-
-       for (k_idx = 0; k_idx < kvlist->count; k_idx++) {
-               pair = &kvlist->pairs[k_idx];
-               if (strstr(pair->key, ETH_PCAP_IFACE_ARG) != NULL)
-                       break;
+               pp->tx_dumper[i] = queue->dumper;
+               pp->tx_pcap[i] = queue->pcap;
+               strlcpy(tx->name, queue->name, sizeof(tx->name));
+               strlcpy(tx->type, queue->type, sizeof(tx->type));
        }
 
-       if (pair == NULL)
-               (*internals)->if_index = 0;
-       else
-               (*internals)->if_index = if_nametoindex(pair->value);
-
        return 0;
 }
 
 static int
 eth_from_pcaps(struct rte_vdev_device *vdev,
-               struct pmd_devargs *rx_queues, const unsigned int nb_rx_queues,
-               struct pmd_devargs *tx_queues, const unsigned int nb_tx_queues,
-               struct rte_kvargs *kvlist, int single_iface,
-               unsigned int using_dumpers)
+               struct pmd_devargs_all *devargs_all)
 {
        struct pmd_internals *internals = NULL;
        struct rte_eth_dev *eth_dev = NULL;
+       struct pmd_devargs *rx_queues = &devargs_all->rx_queues;
+       int single_iface = devargs_all->single_iface;
+       unsigned int infinite_rx = devargs_all->infinite_rx;
        int ret;
 
-       ret = eth_from_pcaps_common(vdev, rx_queues, nb_rx_queues,
-               tx_queues, nb_tx_queues, kvlist, &internals, &eth_dev);
+       ret = eth_from_pcaps_common(vdev, devargs_all, &internals, &eth_dev);
 
        if (ret < 0)
                return ret;
@@ -977,12 +1296,32 @@ eth_from_pcaps(struct rte_vdev_device *vdev,
        /* store weather we are using a single interface for rx/tx or not */
        internals->single_iface = single_iface;
 
-       eth_dev->rx_pkt_burst = eth_pcap_rx;
+       if (single_iface) {
+               internals->if_index = if_nametoindex(rx_queues->queue[0].name);
 
-       if (using_dumpers)
-               eth_dev->tx_pkt_burst = eth_pcap_tx_dumper;
+               /* phy_mac arg is applied only only if "iface" devarg is provided */
+               if (rx_queues->phy_mac) {
+                       int ret = eth_pcap_update_mac(rx_queues->queue[0].name,
+                                       eth_dev, vdev->device.numa_node);
+                       if (ret == 0)
+                               internals->phy_mac = 1;
+               }
+       }
+
+       internals->infinite_rx = infinite_rx;
+       /* Assign rx ops. */
+       if (infinite_rx)
+               eth_dev->rx_pkt_burst = eth_pcap_rx_infinite;
        else
+               eth_dev->rx_pkt_burst = eth_pcap_rx;
+
+       /* Assign tx ops. */
+       if (devargs_all->is_tx_pcap)
+               eth_dev->tx_pkt_burst = eth_pcap_tx_dumper;
+       else if (devargs_all->is_tx_iface)
                eth_dev->tx_pkt_burst = eth_pcap_tx;
+       else
+               eth_dev->tx_pkt_burst = eth_tx_drop;
 
        rte_eth_dev_probing_finish(eth_dev);
        return 0;
@@ -992,14 +1331,21 @@ static int
 pmd_pcap_probe(struct rte_vdev_device *dev)
 {
        const char *name;
-       unsigned int is_rx_pcap = 0, is_tx_pcap = 0;
+       unsigned int is_rx_pcap = 0;
        struct rte_kvargs *kvlist;
        struct pmd_devargs pcaps = {0};
        struct pmd_devargs dumpers = {0};
-       struct rte_eth_dev *eth_dev;
-       int single_iface = 0;
+       struct rte_eth_dev *eth_dev =  NULL;
+       struct pmd_internals *internal;
        int ret;
 
+       struct pmd_devargs_all devargs_all = {
+               .single_iface = 0,
+               .is_tx_pcap = 0,
+               .is_tx_iface = 0,
+               .infinite_rx = 0,
+       };
+
        name = rte_vdev_device_name(dev);
        PMD_LOG(INFO, "Initializing pmd_pcap for %s", name);
 
@@ -1007,23 +1353,24 @@ pmd_pcap_probe(struct rte_vdev_device *dev)
        start_cycles = rte_get_timer_cycles();
        hz = rte_get_timer_hz();
 
-       if (rte_eal_process_type() == RTE_PROC_SECONDARY &&
-           strlen(rte_vdev_device_args(dev)) == 0) {
+       if (rte_eal_process_type() == RTE_PROC_SECONDARY) {
                eth_dev = rte_eth_dev_attach_secondary(name);
                if (!eth_dev) {
                        PMD_LOG(ERR, "Failed to probe %s", name);
                        return -1;
                }
-               /* TODO: request info from primary to set up Rx and Tx */
-               eth_dev->dev_ops = &ops;
-               eth_dev->device = &dev->device;
-               rte_eth_dev_probing_finish(eth_dev);
-               return 0;
-       }
 
-       kvlist = rte_kvargs_parse(rte_vdev_device_args(dev), valid_arguments);
-       if (kvlist == NULL)
-               return -1;
+               internal = eth_dev->data->dev_private;
+
+               kvlist = rte_kvargs_parse(internal->devargs, valid_arguments);
+               if (kvlist == NULL)
+                       return -1;
+       } else {
+               kvlist = rte_kvargs_parse(rte_vdev_device_args(dev),
+                               valid_arguments);
+               if (kvlist == NULL)
+                       return -1;
+       }
 
        /*
         * If iface argument is passed we open the NICs and use them for
@@ -1033,13 +1380,19 @@ pmd_pcap_probe(struct rte_vdev_device *dev)
 
                ret = rte_kvargs_process(kvlist, ETH_PCAP_IFACE_ARG,
                                &open_rx_tx_iface, &pcaps);
-
                if (ret < 0)
                        goto free_kvlist;
 
                dumpers.queue[0] = pcaps.queue[0];
 
-               single_iface = 1;
+               ret = rte_kvargs_process(kvlist, ETH_PCAP_PHY_MAC_ARG,
+                               &select_phy_mac, &pcaps.phy_mac);
+               if (ret < 0)
+                       goto free_kvlist;
+
+               dumpers.phy_mac = pcaps.phy_mac;
+
+               devargs_all.single_iface = 1;
                pcaps.num_of_queue = 1;
                dumpers.num_of_queue = 1;
 
@@ -1054,6 +1407,29 @@ pmd_pcap_probe(struct rte_vdev_device *dev)
        pcaps.num_of_queue = 0;
 
        if (is_rx_pcap) {
+               /*
+                * We check whether we want to infinitely rx the pcap file.
+                */
+               unsigned int infinite_rx_arg_cnt = rte_kvargs_count(kvlist,
+                               ETH_PCAP_INFINITE_RX_ARG);
+
+               if (infinite_rx_arg_cnt == 1) {
+                       ret = rte_kvargs_process(kvlist,
+                                       ETH_PCAP_INFINITE_RX_ARG,
+                                       &get_infinite_rx_arg,
+                                       &devargs_all.infinite_rx);
+                       if (ret < 0)
+                               goto free_kvlist;
+                       PMD_LOG(INFO, "infinite_rx has been %s for %s",
+                                       devargs_all.infinite_rx ? "enabled" : "disabled",
+                                       name);
+
+               } else if (infinite_rx_arg_cnt > 1) {
+                       PMD_LOG(WARNING, "infinite_rx has not been enabled since the "
+                                       "argument has been provided more than once "
+                                       "for %s", name);
+               }
+
                ret = rte_kvargs_process(kvlist, ETH_PCAP_RX_PCAP_ARG,
                                &open_rx_pcap, &pcaps);
        } else {
@@ -1065,25 +1441,80 @@ pmd_pcap_probe(struct rte_vdev_device *dev)
                goto free_kvlist;
 
        /*
-        * We check whether we want to open a TX stream to a real NIC or a
-        * pcap file
+        * We check whether we want to open a TX stream to a real NIC,
+        * a pcap file, or drop packets on tx
         */
-       is_tx_pcap = rte_kvargs_count(kvlist, ETH_PCAP_TX_PCAP_ARG) ? 1 : 0;
+       devargs_all.is_tx_pcap =
+               rte_kvargs_count(kvlist, ETH_PCAP_TX_PCAP_ARG) ? 1 : 0;
+       devargs_all.is_tx_iface =
+               rte_kvargs_count(kvlist, ETH_PCAP_TX_IFACE_ARG) ? 1 : 0;
        dumpers.num_of_queue = 0;
 
-       if (is_tx_pcap)
+       if (devargs_all.is_tx_pcap) {
                ret = rte_kvargs_process(kvlist, ETH_PCAP_TX_PCAP_ARG,
                                &open_tx_pcap, &dumpers);
-       else
+       } else if (devargs_all.is_tx_iface) {
                ret = rte_kvargs_process(kvlist, ETH_PCAP_TX_IFACE_ARG,
                                &open_tx_iface, &dumpers);
+       } else {
+               unsigned int i;
+
+               PMD_LOG(INFO, "Dropping packets on tx since no tx queues were provided.");
+
+               /* Add 1 dummy queue per rxq which counts and drops packets. */
+               for (i = 0; i < pcaps.num_of_queue; i++)
+                       ret = add_queue(&dumpers, "dummy", "tx_drop", NULL,
+                                       NULL);
+       }
 
        if (ret < 0)
                goto free_kvlist;
 
 create_eth:
-       ret = eth_from_pcaps(dev, &pcaps, pcaps.num_of_queue, &dumpers,
-               dumpers.num_of_queue, kvlist, single_iface, is_tx_pcap);
+       if (rte_eal_process_type() == RTE_PROC_SECONDARY) {
+               struct pmd_process_private *pp;
+               unsigned int i;
+
+               internal = eth_dev->data->dev_private;
+                       pp = (struct pmd_process_private *)
+                               rte_zmalloc(NULL,
+                                       sizeof(struct pmd_process_private),
+                                       RTE_CACHE_LINE_SIZE);
+
+               if (pp == NULL) {
+                       PMD_LOG(ERR,
+                               "Failed to allocate memory for process private");
+                       ret = -1;
+                       goto free_kvlist;
+               }
+
+               eth_dev->dev_ops = &ops;
+               eth_dev->device = &dev->device;
+
+               /* setup process private */
+               for (i = 0; i < pcaps.num_of_queue; i++)
+                       pp->rx_pcap[i] = pcaps.queue[i].pcap;
+
+               for (i = 0; i < dumpers.num_of_queue; i++) {
+                       pp->tx_dumper[i] = dumpers.queue[i].dumper;
+                       pp->tx_pcap[i] = dumpers.queue[i].pcap;
+               }
+
+               eth_dev->process_private = pp;
+               eth_dev->rx_pkt_burst = eth_pcap_rx;
+               if (devargs_all.is_tx_pcap)
+                       eth_dev->tx_pkt_burst = eth_pcap_tx_dumper;
+               else
+                       eth_dev->tx_pkt_burst = eth_pcap_tx;
+
+               rte_eth_dev_probing_finish(eth_dev);
+               goto free_kvlist;
+       }
+
+       devargs_all.rx_queues = pcaps;
+       devargs_all.tx_queues = dumpers;
+
+       ret = eth_from_pcaps(dev, &devargs_all);
 
 free_kvlist:
        rte_kvargs_free(kvlist);
@@ -1094,6 +1525,7 @@ free_kvlist:
 static int
 pmd_pcap_remove(struct rte_vdev_device *dev)
 {
+       struct pmd_internals *internals = NULL;
        struct rte_eth_dev *eth_dev = NULL;
 
        PMD_LOG(INFO, "Closing pcap ethdev on numa socket %d",
@@ -1107,8 +1539,16 @@ pmd_pcap_remove(struct rte_vdev_device *dev)
        if (eth_dev == NULL)
                return -1;
 
-       rte_free(eth_dev->data->dev_private);
+       if (rte_eal_process_type() == RTE_PROC_PRIMARY) {
+               internals = eth_dev->data->dev_private;
+               if (internals != NULL && internals->phy_mac == 0)
+                       /* not dynamically allocated, must not be freed */
+                       eth_dev->data->mac_addrs = NULL;
+       }
+
+       eth_dev_close(eth_dev);
 
+       rte_free(eth_dev->process_private);
        rte_eth_dev_release_port(eth_dev);
 
        return 0;
@@ -1127,7 +1567,9 @@ RTE_PMD_REGISTER_PARAM_STRING(net_pcap,
        ETH_PCAP_RX_IFACE_ARG "=<ifc> "
        ETH_PCAP_RX_IFACE_IN_ARG "=<ifc> "
        ETH_PCAP_TX_IFACE_ARG "=<ifc> "
-       ETH_PCAP_IFACE_ARG "=<ifc>");
+       ETH_PCAP_IFACE_ARG "=<ifc> "
+       ETH_PCAP_PHY_MAC_ARG "=<int>"
+       ETH_PCAP_INFINITE_RX_ARG "=<0|1>");
 
 RTE_INIT(eth_pcap_init_log)
 {