[dpdk-dev,1/5] app/test: unit test for rx and tx cycles/packet
diff mbox

Message ID 1408947174-11323-2-git-send-email-cunming.liang@intel.com
State Superseded, archived
Headers show

Commit Message

Cunming Liang Aug. 25, 2014, 6:12 a.m. UTC
Signed-off-by: Cunming Liang <cunming.liang@intel.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
---
 app/test/Makefile                   |   1 +
 app/test/commands.c                 |  41 +++
 app/test/packet_burst_generator.c   |   4 +-
 app/test/test.h                     |   6 +
 app/test/test_pmd_perf.c            | 620 ++++++++++++++++++++++++++++++++++++
 lib/librte_pmd_ixgbe/ixgbe_ethdev.c |   6 +
 6 files changed, 676 insertions(+), 2 deletions(-)
 create mode 100644 app/test/test_pmd_perf.c

Patch
diff mbox

diff --git a/app/test/Makefile b/app/test/Makefile
index 0024737..8418b2e 100644
--- a/app/test/Makefile
+++ b/app/test/Makefile
@@ -104,6 +104,7 @@  SRCS-$(CONFIG_RTE_APP_TEST) += test_distributor_perf.c
 SRCS-$(CONFIG_RTE_APP_TEST) += test_devargs.c
 SRCS-$(CONFIG_RTE_APP_TEST) += virtual_pmd.c
 SRCS-$(CONFIG_RTE_APP_TEST) += packet_burst_generator.c
+SRCS-$(CONFIG_RTE_APP_TEST) += test_pmd_perf.c
 ifeq ($(CONFIG_RTE_APP_TEST),y)
 SRCS-$(CONFIG_RTE_LIBRTE_ACL) += test_acl.c
 SRCS-$(CONFIG_RTE_LIBRTE_PMD_BOND) += test_link_bonding.c
diff --git a/app/test/commands.c b/app/test/commands.c
index 5f23420..d0e583e 100644
--- a/app/test/commands.c
+++ b/app/test/commands.c
@@ -204,6 +204,8 @@  static void cmd_autotest_parsed(void *parsed_result,
 	if (!strcmp(res->autotest, "kvargs_autotest"))
 		ret |= test_kvargs();
 #endif /* RTE_LIBRTE_KVARGS */
+	if (!strcmp(res->autotest, "pmd_perf_autotest"))
+		ret = test_pmd_perf();
 
 	if (ret == 0)
 		printf("Test OK\n");
@@ -251,6 +253,7 @@  cmdline_parse_token_string_t cmd_autotest_autotest =
 #ifdef RTE_LIBRTE_KVARGS
 			"kvargs_autotest#"
 #endif
+			"pmd_perf_autotest#"
 			"common_autotest#"
 			"distributor_autotest#distributor_perf_autotest");
 
@@ -451,12 +454,50 @@  cmdline_parse_inst_t cmd_quit = {
 
 /****************/
 
+/****************/
+
+struct cmd_set_rxtx_result {
+	cmdline_fixed_string_t set;
+	cmdline_fixed_string_t mode;
+};
+
+static void cmd_set_rxtx_parsed(void *parsed_result, struct cmdline *cl,
+				__attribute__((unused)) void *data)
+{
+	struct cmd_set_rxtx_result *res = parsed_result;
+	if (test_set_rxtx_conf(res->mode) < 0)
+		cmdline_printf(cl, "Cannot find such mode\n");
+}
+
+cmdline_parse_token_string_t cmd_set_rxtx_set =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxtx_result, set,
+				 "set_rxtx_mode");
+
+cmdline_parse_token_string_t cmd_set_rxtx_mode =
+	TOKEN_STRING_INITIALIZER(struct cmd_set_rxtx_result, mode, NULL);
+
+cmdline_parse_inst_t cmd_set_rxtx = {
+	.f = cmd_set_rxtx_parsed,  /* function to call */
+	.data = NULL,      /* 2nd arg of func */
+	.help_str = "set rxtx routine: "
+			"set_rxtx <mode>",
+	.tokens = {        /* token list, NULL terminated */
+		(void *)&cmd_set_rxtx_set,
+		(void *)&cmd_set_rxtx_mode,
+		NULL,
+	},
+};
+
+/****************/
+
+
 cmdline_parse_ctx_t main_ctx[] = {
 	(cmdline_parse_inst_t *)&cmd_autotest,
 	(cmdline_parse_inst_t *)&cmd_dump,
 	(cmdline_parse_inst_t *)&cmd_dump_one,
 	(cmdline_parse_inst_t *)&cmd_set_ring,
 	(cmdline_parse_inst_t *)&cmd_quit,
+	(cmdline_parse_inst_t *)&cmd_set_rxtx,
 	NULL,
 };
 
diff --git a/app/test/packet_burst_generator.c b/app/test/packet_burst_generator.c
index 5d539f1..943ea98 100644
--- a/app/test/packet_burst_generator.c
+++ b/app/test/packet_burst_generator.c
@@ -190,12 +190,12 @@  initialize_ipv4_header(struct ipv4_hdr *ip_hdr, uint32_t src_addr,
  */
 #define RTE_MAX_SEGS_PER_PKT 255 /**< pkt.nb_segs is a 8-bit unsigned char. */
 
-#define TXONLY_DEF_PACKET_LEN 64
+#define TXONLY_DEF_PACKET_LEN 60
 #define TXONLY_DEF_PACKET_LEN_128 128
 
 uint16_t tx_pkt_length = TXONLY_DEF_PACKET_LEN;
 uint16_t tx_pkt_seg_lengths[RTE_MAX_SEGS_PER_PKT] = {
-		TXONLY_DEF_PACKET_LEN_128,
+		TXONLY_DEF_PACKET_LEN,
 };
 
 uint8_t  tx_pkt_nb_segs = 1;
diff --git a/app/test/test.h b/app/test/test.h
index 181c38e..ad01eec 100644
--- a/app/test/test.h
+++ b/app/test/test.h
@@ -123,6 +123,9 @@  int unit_test_suite_runner(struct unit_test_suite *suite);
 
 #define RECURSIVE_ENV_VAR "RTE_TEST_RECURSIVE"
 
+#include <cmdline_parse.h>
+#include <cmdline_parse_string.h>
+
 extern const char *prgname;
 
 int main(int argc, char **argv);
@@ -180,6 +183,9 @@  int test_distributor_perf(void);
 int test_kvargs(void);
 int test_devargs(void);
 int test_link_bonding(void);
+int test_pmd_perf(void);
+int test_set_rxtx_conf(cmdline_fixed_string_t mode);
+
 
 int test_pci_run;
 
diff --git a/app/test/test_pmd_perf.c b/app/test/test_pmd_perf.c
new file mode 100644
index 0000000..ee527d3
--- /dev/null
+++ b/app/test/test_pmd_perf.c
@@ -0,0 +1,620 @@ 
+/*-
+ *   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 <stdio.h>
+#include <inttypes.h>
+#include <signal.h>
+#include <unistd.h>
+#include <rte_cycles.h>
+#include <rte_ethdev.h>
+#include <rte_byteorder.h>
+#include "packet_burst_generator.h"
+#include "test.h"
+
+#define NB_ETHPORTS_USED                (1)
+#define NB_SOCKETS                      (2)
+#define MEMPOOL_CACHE_SIZE 250
+#define MBUF_SIZE (2048 + sizeof(struct rte_mbuf) + RTE_PKTMBUF_HEADROOM)
+#define MAX_PKT_BURST                   (32)
+#define RTE_TEST_RX_DESC_DEFAULT        (128)
+#define RTE_TEST_TX_DESC_DEFAULT        (512)
+#define RTE_PORT_ALL            (~(uint8_t)0x0)
+
+/* how long test would take at full line rate */
+#define RTE_TEST_DURATION                (2)
+
+/*
+ * RX and TX Prefetch, Host, and Write-back threshold values should be
+ * carefully set for optimal performance. Consult the network
+ * controller's datasheet and supporting DPDK documentation for guidance
+ * on how these parameters should be set.
+ */
+#define RX_PTHRESH 8 /**< Default values of RX prefetch threshold reg. */
+#define RX_HTHRESH 8 /**< Default values of RX host threshold reg. */
+#define RX_WTHRESH 0 /**< Default values of RX write-back threshold reg. */
+
+/*
+ * These default values are optimized for use with the Intel(R) 82599 10 GbE
+ * Controller and the DPDK ixgbe PMD. Consider using other values for other
+ * network controllers and/or network drivers.
+ */
+#define TX_PTHRESH 32 /**< Default values of TX prefetch threshold reg. */
+#define TX_HTHRESH 0  /**< Default values of TX host threshold reg. */
+#define TX_WTHRESH 0  /**< Default values of TX write-back threshold reg. */
+
+
+#define NB_MBUF RTE_MAX(						\
+	  (unsigned)(nb_ports*nb_rx_queue*RTE_TEST_RX_DESC_DEFAULT +	\
+	   nb_ports*nb_lcores*MAX_PKT_BURST +				\
+	   nb_ports*nb_tx_queue*RTE_TEST_TX_DESC_DEFAULT +		\
+	   nb_lcores*MEMPOOL_CACHE_SIZE),				\
+	  (unsigned)8192)
+
+
+static struct rte_mempool *mbufpool[NB_SOCKETS];
+/* ethernet addresses of ports */
+static struct ether_addr ports_eth_addr[RTE_MAX_ETHPORTS];
+
+static struct rte_eth_conf port_conf = {
+	.rxmode = {
+		.mq_mode = ETH_MQ_RX_NONE,
+		.max_rx_pkt_len = ETHER_MAX_LEN,
+		.split_hdr_size = 0,
+		.header_split   = 0, /**< Header Split disabled */
+		.hw_ip_checksum = 0, /**< IP checksum offload enabled */
+		.hw_vlan_filter = 0, /**< VLAN filtering disabled */
+		.hw_vlan_strip  = 0, /**< VLAN strip enabled. */
+		.hw_vlan_extend = 0, /**< Extended VLAN disabled. */
+		.jumbo_frame    = 0, /**< Jumbo Frame Support disabled */
+		.hw_strip_crc   = 0, /**< CRC stripped by hardware */
+		.enable_scatter = 0, /**< scatter rx disabled */
+	},
+	.txmode = {
+		.mq_mode = ETH_MQ_TX_NONE,
+	},
+	.lpbk_mode = 1,  /* enable loopback */
+};
+
+static struct rte_eth_rxconf rx_conf = {
+	.rx_thresh = {
+		.pthresh = RX_PTHRESH,
+		.hthresh = RX_HTHRESH,
+		.wthresh = RX_WTHRESH,
+	},
+	.rx_free_thresh = 32,
+};
+
+static struct rte_eth_txconf tx_conf = {
+	.tx_thresh = {
+		.pthresh = TX_PTHRESH,
+		.hthresh = TX_HTHRESH,
+		.wthresh = TX_WTHRESH,
+	},
+	.tx_free_thresh = 32, /* Use PMD default values */
+	.tx_rs_thresh = 32, /* Use PMD default values */
+	.txq_flags = (ETH_TXQ_FLAGS_NOMULTSEGS |
+		      ETH_TXQ_FLAGS_NOVLANOFFL |
+		      ETH_TXQ_FLAGS_NOXSUMSCTP |
+		      ETH_TXQ_FLAGS_NOXSUMUDP |
+		      ETH_TXQ_FLAGS_NOXSUMTCP)
+};
+
+enum {
+	LCORE_INVALID = 0,
+	LCORE_AVAIL,
+	LCORE_USED,
+};
+
+struct lcore_conf {
+	uint8_t status;
+	uint8_t socketid;
+	uint16_t nb_ports;
+	uint8_t portlist[RTE_MAX_ETHPORTS];
+} __rte_cache_aligned;
+
+struct lcore_conf lcore_conf[RTE_MAX_LCORE];
+
+static uint64_t link_mbps;
+
+/* Check the link status of all ports in up to 3s, and print them finally */
+static void
+check_all_ports_link_status(uint8_t port_num, uint32_t port_mask)
+{
+#define CHECK_INTERVAL 100 /* 100ms */
+#define MAX_CHECK_TIME 30 /* 3s (30 * 100ms) in total */
+	uint8_t portid, count, all_ports_up, print_flag = 0;
+	struct rte_eth_link link;
+
+	printf("Checking link statuses...\n");
+	fflush(stdout);
+	for (count = 0; count <= MAX_CHECK_TIME; count++) {
+		all_ports_up = 1;
+		for (portid = 0; portid < port_num; portid++) {
+			if ((port_mask & (1 << portid)) == 0)
+				continue;
+			memset(&link, 0, sizeof(link));
+			rte_eth_link_get_nowait(portid, &link);
+			/* print link status if flag set */
+			if (print_flag == 1) {
+				if (link.link_status) {
+					printf("Port %d Link Up - speed %u "
+						"Mbps - %s\n", (uint8_t)portid,
+						(unsigned)link.link_speed,
+				(link.link_duplex == ETH_LINK_FULL_DUPLEX) ?
+					("full-duplex") : ("half-duplex\n"));
+					if (link_mbps == 0)
+						link_mbps = link.link_speed;
+				} else
+					printf("Port %d Link Down\n",
+						(uint8_t)portid);
+				continue;
+			}
+			/* clear all_ports_up flag if any link down */
+			if (link.link_status == 0) {
+				all_ports_up = 0;
+				break;
+			}
+		}
+		/* after finally printing all link status, get out */
+		if (print_flag == 1)
+			break;
+
+		if (all_ports_up == 0) {
+			fflush(stdout);
+			rte_delay_ms(CHECK_INTERVAL);
+		}
+
+		/* set the print_flag if all ports up or timeout */
+		if (all_ports_up == 1 || count == (MAX_CHECK_TIME - 1))
+			print_flag = 1;
+	}
+}
+
+static void
+print_ethaddr(const char *name, const struct ether_addr *eth_addr)
+{
+	printf("%s%02X:%02X:%02X:%02X:%02X:%02X", name,
+		eth_addr->addr_bytes[0],
+		eth_addr->addr_bytes[1],
+		eth_addr->addr_bytes[2],
+		eth_addr->addr_bytes[3],
+		eth_addr->addr_bytes[4],
+		eth_addr->addr_bytes[5]);
+}
+
+static int
+init_traffic(struct rte_mempool *mp,
+	     struct rte_mbuf **pkts_burst, uint32_t burst_size)
+{
+	struct ether_hdr pkt_eth_hdr;
+	struct ipv4_hdr pkt_ipv4_hdr;
+	struct udp_hdr pkt_udp_hdr;
+	uint32_t pktlen;
+	static uint8_t src_mac[] = { 0x00, 0xFF, 0xAA, 0xFF, 0xAA, 0xFF };
+	static uint8_t dst_mac[] = { 0x00, 0xAA, 0xFF, 0xAA, 0xFF, 0xAA };
+
+
+	initialize_eth_header(&pkt_eth_hdr,
+		(struct ether_addr *)src_mac,
+		(struct ether_addr *)dst_mac, 0, 0);
+	pkt_eth_hdr.ether_type = rte_cpu_to_be_16(ETHER_TYPE_IPv4);
+
+	pktlen = initialize_ipv4_header(&pkt_ipv4_hdr,
+					IPV4_ADDR(10, 0, 0, 1),
+					IPV4_ADDR(10, 0, 0, 2), 26);
+	printf("IPv4 pktlen %u\n", pktlen);
+
+	pktlen = initialize_udp_header(&pkt_udp_hdr, 0, 0, 18);
+
+	printf("UDP pktlen %u\n", pktlen);
+
+	return generate_packet_burst(mp, pkts_burst, &pkt_eth_hdr,
+				     0, &pkt_ipv4_hdr, 1,
+				     &pkt_udp_hdr, burst_size);
+}
+
+static int
+init_lcores(void)
+{
+	unsigned lcore_id;
+
+	for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) {
+		lcore_conf[lcore_id].socketid =
+			rte_lcore_to_socket_id(lcore_id);
+		if (rte_lcore_is_enabled(lcore_id) == 0) {
+			lcore_conf[lcore_id].status = LCORE_INVALID;
+			continue;
+		} else
+			lcore_conf[lcore_id].status = LCORE_AVAIL;
+	}
+	return 0;
+}
+
+static int
+init_mbufpool(unsigned nb_mbuf)
+{
+	int socketid;
+	unsigned lcore_id;
+	char s[64];
+
+	for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) {
+		if (rte_lcore_is_enabled(lcore_id) == 0)
+			continue;
+
+		socketid = rte_lcore_to_socket_id(lcore_id);
+		if (socketid >= NB_SOCKETS) {
+			rte_exit(EXIT_FAILURE,
+				"Socket %d of lcore %u is out of range %d\n",
+				socketid, lcore_id, NB_SOCKETS);
+		}
+		if (mbufpool[socketid] == NULL) {
+			snprintf(s, sizeof(s), "mbuf_pool_%d", socketid);
+			mbufpool[socketid] =
+				rte_mempool_create(s, nb_mbuf, MBUF_SIZE,
+					MEMPOOL_CACHE_SIZE,
+					sizeof(struct rte_pktmbuf_pool_private),
+					rte_pktmbuf_pool_init, NULL,
+					rte_pktmbuf_init, NULL,
+					socketid, 0);
+			if (mbufpool[socketid] == NULL)
+				rte_exit(EXIT_FAILURE,
+					"Cannot init mbuf pool on socket %d\n",
+					socketid);
+			else
+				printf("Allocated mbuf pool on socket %d\n",
+					socketid);
+		}
+	}
+	return 0;
+}
+
+static uint16_t
+alloc_lcore(uint16_t socketid)
+{
+	unsigned lcore_id;
+
+	for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) {
+		if (LCORE_AVAIL != lcore_conf[lcore_id].status ||
+		    lcore_conf[lcore_id].socketid != socketid ||
+		    lcore_id == rte_get_master_lcore())
+			continue;
+		lcore_conf[lcore_id].status = LCORE_USED;
+		lcore_conf[lcore_id].nb_ports = 0;
+		return lcore_id;
+	}
+
+	return (uint16_t)-1;
+}
+
+volatile uint64_t stop;
+uint64_t count;
+uint64_t drop;
+uint64_t idle;
+
+static void
+reset_count(void)
+{
+	count = 0;
+	drop = 0;
+	idle = 0;
+}
+
+static void
+stats_display(uint8_t port_id)
+{
+	struct rte_eth_stats stats;
+	rte_eth_stats_get(port_id, &stats);
+
+	printf("  RX-packets: %-10"PRIu64" RX-missed: %-10"PRIu64" RX-bytes:  "
+	       "%-"PRIu64"\n",
+	       stats.ipackets, stats.imissed, stats.ibytes);
+	printf("  RX-badcrc:  %-10"PRIu64" RX-badlen: %-10"PRIu64" RX-errors: "
+	       "%-"PRIu64"\n",
+	       stats.ibadcrc, stats.ibadlen, stats.ierrors);
+	printf("  RX-nombuf:  %-10"PRIu64"\n",
+	       stats.rx_nombuf);
+	printf("  TX-packets: %-10"PRIu64" TX-errors: %-10"PRIu64" TX-bytes:  "
+	       "%-"PRIu64"\n",
+	       stats.opackets, stats.oerrors, stats.obytes);
+}
+
+static void
+signal_handler(int signum)
+{
+	/* When we receive a USR1 signal, print stats */
+	if (signum == SIGUSR1) {
+		printf("Force Stop!\n");
+		stop = 1;
+	}
+	if (signum == SIGUSR2)
+		stats_display(0);
+}
+
+#define MAX_TRAFIC_BURST                (4096)
+struct rte_mbuf *tx_burst[MAX_TRAFIC_BURST];
+
+/* main processing loop */
+static int
+main_loop(__rte_unused void *args)
+{
+#define PACKET_SIZE 64
+#define FRAME_GAP 12
+#define MAC_PREAMBLE 8
+	struct rte_mbuf *pkts_burst[MAX_PKT_BURST];
+	unsigned lcore_id;
+	unsigned i, portid, nb_rx = 0, nb_tx = 0;
+	struct lcore_conf *conf;
+	uint64_t prev_tsc, cur_tsc;
+	int pkt_per_port;
+	uint64_t packets_per_second, total_packets;
+
+	lcore_id = rte_lcore_id();
+	conf = &lcore_conf[lcore_id];
+	if (conf->status != LCORE_USED)
+		return 0;
+
+	pkt_per_port =  MAX_TRAFIC_BURST / conf->nb_ports;
+
+	int idx = 0;
+	for (i = 0; i < conf->nb_ports; i++) {
+		int num = pkt_per_port;
+		portid = conf->portlist[i];
+		printf("inject %d packet to port %d\n", num, portid);
+		while (num) {
+			nb_tx = RTE_MIN(MAX_PKT_BURST, num);
+			nb_tx = rte_eth_tx_burst(portid, 0,
+						&tx_burst[idx], nb_tx);
+			num -= nb_tx;
+			idx += nb_tx;
+		}
+	}
+	printf("Total packets inject to prime ports = %u\n", idx);
+
+	packets_per_second = (link_mbps * 1000 * 1000) /
+		+((PACKET_SIZE + FRAME_GAP + MAC_PREAMBLE) * CHAR_BIT);
+	printf("Each port will do %"PRIu64" packets per second\n",
+		+packets_per_second);
+
+	total_packets = RTE_TEST_DURATION * conf->nb_ports * packets_per_second;
+	printf("Test will stop after at least %"PRIu64" packets received\n",
+		+ total_packets);
+
+	prev_tsc = rte_rdtsc();
+
+	while (likely(!stop)) {
+		for (i = 0; i < conf->nb_ports; i++) {
+			portid = conf->portlist[i];
+			nb_rx = rte_eth_rx_burst((uint8_t) portid, 0,
+						 pkts_burst, MAX_PKT_BURST);
+			if (unlikely(nb_rx == 0)) {
+				idle++;
+				continue;
+			}
+
+			count += nb_rx;
+			nb_tx = rte_eth_tx_burst(portid, 0, pkts_burst, nb_rx);
+			if (unlikely(nb_tx < nb_rx)) {
+				drop += (nb_rx - nb_tx);
+				do {
+					rte_pktmbuf_free(pkts_burst[nb_tx]);
+				} while (++nb_tx < nb_rx);
+			}
+		}
+		if (unlikely(count >= total_packets))
+			break;
+	}
+
+	cur_tsc = rte_rdtsc();
+
+	for (i = 0; i < conf->nb_ports; i++) {
+		portid = conf->portlist[i];
+		int nb_free = pkt_per_port;
+		do { /* dry out */
+			nb_rx = rte_eth_rx_burst((uint8_t) portid, 0,
+						 pkts_burst, MAX_PKT_BURST);
+			nb_tx = 0;
+			while (nb_tx < nb_rx)
+				rte_pktmbuf_free(pkts_burst[nb_tx++]);
+			nb_free -= nb_rx;
+		} while (nb_free != 0);
+		printf("free %d mbuf left in port %u\n", pkt_per_port, portid);
+	}
+
+	if (count == 0)
+		return -1;
+
+	printf("%lu packet, %lu drop, %lu idle\n", count, drop, idle);
+	printf("Result: %ld cycles per packet\n", (cur_tsc - prev_tsc) / count);
+
+	return 0;
+}
+
+int
+test_pmd_perf(void)
+{
+	uint16_t nb_ports, num, nb_lcores, slave_id = (uint16_t)-1;
+	uint16_t nb_rxd = RTE_TEST_RX_DESC_DEFAULT;
+	uint16_t nb_txd = RTE_TEST_TX_DESC_DEFAULT;
+	uint16_t portid;
+	uint16_t nb_rx_queue = 1, nb_tx_queue = 1;
+	int socketid = -1;
+	int ret;
+
+	printf("Start PMD RXTX cycles cost test.\n");
+
+	signal(SIGUSR1, signal_handler);
+	signal(SIGUSR2, signal_handler);
+
+	nb_ports = rte_eth_dev_count();
+	if (nb_ports < NB_ETHPORTS_USED) {
+		printf("At least %u port(s) used for perf. test\n",
+		       NB_ETHPORTS_USED);
+		return -1;
+	}
+
+	if (nb_ports > RTE_MAX_ETHPORTS)
+		nb_ports = RTE_MAX_ETHPORTS;
+
+	nb_lcores = rte_lcore_count();
+
+	memset(lcore_conf, 0, sizeof(lcore_conf));
+	init_lcores();
+
+	init_mbufpool(NB_MBUF);
+
+	reset_count();
+	num = 0;
+	for (portid = 0; portid < nb_ports; portid++) {
+		if (socketid == -1) {
+			socketid = rte_eth_dev_socket_id(portid);
+			slave_id = alloc_lcore(socketid);
+			if (slave_id == (uint16_t)-1) {
+				printf("No avail lcore to run test\n");
+				return -1;
+			}
+			printf("Performance test runs on lcore %u socket %u\n",
+			       slave_id, socketid);
+		}
+
+		if (socketid != rte_eth_dev_socket_id(portid)) {
+			printf("Skip port %d\n", portid);
+			continue;
+		}
+
+		/* port configure */
+		ret = rte_eth_dev_configure(portid, nb_rx_queue,
+					    nb_tx_queue, &port_conf);
+		if (ret < 0)
+			rte_exit(EXIT_FAILURE,
+				"Cannot configure device: err=%d, port=%d\n",
+				 ret, portid);
+
+		rte_eth_macaddr_get(portid, &ports_eth_addr[portid]);
+		printf("Port %u ", portid);
+		print_ethaddr("Address:", &ports_eth_addr[portid]);
+		printf("\n");
+
+		/* tx queue setup */
+		ret = rte_eth_tx_queue_setup(portid, 0, nb_txd,
+					     socketid, &tx_conf);
+		if (ret < 0)
+			rte_exit(EXIT_FAILURE,
+				"rte_eth_tx_queue_setup: err=%d, "
+				"port=%d\n", ret, portid);
+
+		/* rx queue steup */
+		ret = rte_eth_rx_queue_setup(portid, 0, nb_rxd,
+						socketid, &rx_conf,
+						mbufpool[socketid]);
+		if (ret < 0)
+			rte_exit(EXIT_FAILURE, "rte_eth_rx_queue_setup: err=%d,"
+				 "port=%d\n", ret, portid);
+
+		/* Start device */
+		stop = 0;
+		ret = rte_eth_dev_start(portid);
+		if (ret < 0)
+			rte_exit(EXIT_FAILURE,
+				"rte_eth_dev_start: err=%d, port=%d\n",
+				ret, portid);
+
+		/* always eanble promiscuous */
+		rte_eth_promiscuous_enable(portid);
+
+		lcore_conf[slave_id].portlist[num++] = portid;
+		lcore_conf[slave_id].nb_ports++;
+	}
+	check_all_ports_link_status(nb_ports, RTE_PORT_ALL);
+
+	init_traffic(mbufpool[socketid], tx_burst, MAX_TRAFIC_BURST);
+
+	rte_eal_remote_launch(main_loop, NULL, slave_id);
+	if (rte_eal_wait_lcore(slave_id) < 0)
+		return -1;
+
+	/* port tear down */
+	for (portid = 0; portid < nb_ports; portid++) {
+		if (socketid != rte_eth_dev_socket_id(portid))
+			continue;
+
+		rte_eth_dev_stop(portid);
+	}
+
+	return 0;
+}
+
+int
+test_set_rxtx_conf(cmdline_fixed_string_t mode)
+{
+	printf("mode is %s\n", mode);
+
+	if (!strcmp(mode, "vector")) {
+		/* vector rx, tx */
+		tx_conf.txq_flags = 0xf01;
+		tx_conf.tx_rs_thresh = 32;
+		tx_conf.tx_free_thresh = 32;
+		port_conf.rxmode.hw_ip_checksum = 0;
+		port_conf.rxmode.enable_scatter = 0;
+		return 0;
+	} else if (!strcmp(mode, "scalar")) {
+		/* bulk alloc rx, simple tx */
+		tx_conf.txq_flags = 0xf01;
+		tx_conf.tx_rs_thresh = 128;
+		tx_conf.tx_free_thresh = 128;
+		port_conf.rxmode.hw_ip_checksum = 1;
+		port_conf.rxmode.enable_scatter = 0;
+		return 0;
+	} else if (!strcmp(mode, "hybrid")) {
+		/* bulk alloc rx, vector tx
+		 * when vec macro not define,
+		 * using the same rx/tx as scalar
+		 */
+		tx_conf.txq_flags = 0xf01;
+		tx_conf.tx_rs_thresh = 32;
+		tx_conf.tx_free_thresh = 32;
+		port_conf.rxmode.hw_ip_checksum = 1;
+		port_conf.rxmode.enable_scatter = 0;
+		return 0;
+	} else if (!strcmp(mode, "full")) {
+		/* full feature rx,tx pair */
+		tx_conf.txq_flags = 0x0;   /* must condition */
+		tx_conf.tx_rs_thresh = 32;
+		tx_conf.tx_free_thresh = 32;
+		port_conf.rxmode.hw_ip_checksum = 0;
+		port_conf.rxmode.enable_scatter = 1; /* must condition */
+		return 0;
+	}
+
+	return -1;
+}
diff --git a/lib/librte_pmd_ixgbe/ixgbe_ethdev.c b/lib/librte_pmd_ixgbe/ixgbe_ethdev.c
index 59122a1..c581dbf 100644
--- a/lib/librte_pmd_ixgbe/ixgbe_ethdev.c
+++ b/lib/librte_pmd_ixgbe/ixgbe_ethdev.c
@@ -1582,6 +1582,9 @@  ixgbe_dev_stop(struct rte_eth_dev *dev)
 
 	ixgbe_dev_clear_queues(dev);
 
+	/* Clear stored conf */
+	dev->data->scattered_rx = 0;
+
 	/* Clear recorded link status */
 	memset(&link, 0, sizeof(link));
 	rte_ixgbe_dev_atomic_write_link_status(dev, &link);
@@ -2853,6 +2856,9 @@  ixgbevf_dev_stop(struct rte_eth_dev *dev)
 	  */
 	ixgbevf_set_vfta_all(dev,0);
 
+	/* Clear stored conf */
+	dev->data->scattered_rx = 0;
+
 	ixgbe_dev_clear_queues(dev);
 }