DPDK L2fwd 源码阅读

代码部分

/* SPDX-License-Identifier: BSD-3-Clause
 * Copyright(c) 2010-2016 Intel Corporation
 */

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <stdint.h>
#include <inttypes.h>
#include <sys/types.h>
#include <sys/queue.h>
#include <netinet/in.h>
#include <setjmp.h>
#include <stdarg.h>
#include <ctype.h>
#include <errno.h>
#include <getopt.h>
#include <signal.h>
#include <stdbool.h>

#include <rte_common.h>
#include <rte_log.h>
#include <rte_malloc.h>
#include <rte_memory.h>
#include <rte_memcpy.h>
#include <rte_eal.h>
#include <rte_launch.h>
#include <rte_atomic.h>
#include <rte_cycles.h>
#include <rte_prefetch.h>
#include <rte_lcore.h>
#include <rte_per_lcore.h>
#include <rte_branch_prediction.h>
#include <rte_interrupts.h>
#include <rte_random.h>
#include <rte_debug.h>
#include <rte_ether.h>
#include <rte_ethdev.h>
#include <rte_mempool.h>
#include <rte_mbuf.h>

static volatile bool force_quit;

/* MAC updating enabled by default */
static int mac_updating = 1;
/* MAC updating，默认开启。若不开启，则是和basicfw一样的模式。开启后，会有如下影响：

The source MAC address is replaced by the TX_PORT MAC address
源MAC地址会改写成发送端口的MAC地址

The destination MAC address is replaced by 02:00:00:00:00:TX_PORT_ID
改写目的MAC地址，改写为 02:00:00:00:00:<发送端口的port id>
*/

#define RTE_LOGTYPE_L2FWD RTE_LOGTYPE_USER1

#define MAX_PKT_BURST 32
#define BURST_TX_DRAIN_US 100 /* TX drain every ~100us */
#define MEMPOOL_CACHE_SIZE 256

/*
 * Configurable number of RX/TX ring descriptors
 */
#define RTE_TEST_RX_DESC_DEFAULT 1024
#define RTE_TEST_TX_DESC_DEFAULT 1024
static uint16_t nb_rxd = RTE_TEST_RX_DESC_DEFAULT;
static uint16_t nb_txd = RTE_TEST_TX_DESC_DEFAULT;

/* ethernet addresses of ports */
static struct ether_addr l2fwd_ports_eth_addr[RTE_MAX_ETHPORTS];

/* mask of enabled ports */
static uint32_t l2fwd_enabled_port_mask = 0;

/* list of enabled ports */
static uint32_t l2fwd_dst_ports[RTE_MAX_ETHPORTS];

static unsigned int l2fwd_rx_queue_per_lcore = 1; // 每个逻辑核最多可以用来处理几个端口/队列（L2fwd 一个端口分配各一个收发队列）

#define MAX_RX_QUEUE_PER_LCORE 16
#define MAX_TX_QUEUE_PER_PORT 16
struct lcore_queue_conf { // 逻辑核上的队列配置
	unsigned n_rx_port; // 该 lcore 上绑定多少个端口，也作为下一个数组的下标（0 ~ n-1）。
	unsigned rx_port_list[MAX_RX_QUEUE_PER_LCORE]; // 存放一系列端口号，绑定哪些端口。
} __rte_cache_aligned;

// 这也就是 poll module driver 思想。绑定 lcore 和 port，特定的 lcore 轮询对应的一个或多个 port

struct lcore_queue_conf lcore_queue_conf[RTE_MAX_LCORE]; // 数组下标是 lcore id

static struct rte_eth_dev_tx_buffer *tx_buffer[RTE_MAX_ETHPORTS]; // 用于缓冲未来要发送的数据包的结构，API rte_eth_tx_buffer 和 rte_eth_tx_buffer_flush使用的结构

static struct rte_eth_conf port_conf = {
	.rxmode = { // RX feature 见 flow_filtering
		.split_hdr_size = 0,
		.ignore_offload_bitfield = 1,
		.offloads = DEV_RX_OFFLOAD_CRC_STRIP,
	},
	.txmode = { // TX feature
		.mq_mode = ETH_MQ_TX_NONE, // mq_多队列选项，有一些宏来定义用多队列发包的方法
	},
};

struct rte_mempool * l2fwd_pktmbuf_pool = NULL;

/* Per-port statistics struct */
struct l2fwd_port_statistics {
	uint64_t tx; // 发包的数量
	uint64_t rx; // 收包的数量
	uint64_t dropped; // 丢包的数量
} __rte_cache_aligned;
struct l2fwd_port_statistics port_statistics[RTE_MAX_ETHPORTS];

#define MAX_TIMER_PERIOD 86400 /* 1 day max */
/* A tsc-based timer responsible for triggering statistics printout */
// timer 负责每隔一段时间触发打印数据
static uint64_t timer_period = 10; /* default period is 10 seconds */

/* Print out statistics on packets dropped */
static void
print_stats(void)
{
	uint64_t total_packets_dropped, total_packets_tx, total_packets_rx;
	unsigned portid;

	total_packets_dropped = 0;
	total_packets_tx = 0;
	total_packets_rx = 0;

	const char clr[] = { 27, '[', '2', 'J', '\0' };
	const char topLeft[] = { 27, '[', '1', ';', '1', 'H','\0' };

		/* Clear screen and move to top left */
	printf("%s%s", clr, topLeft);

	printf("\nPort statistics ====================================");

	for (portid = 0; portid < RTE_MAX_ETHPORTS; portid++) {
		/* skip disabled ports */
		if ((l2fwd_enabled_port_mask & (1 << portid)) == 0)
			continue;
		printf("\nStatistics for port %u ------------------------------"
			   "\nPackets sent: %24"PRIu64
			   "\nPackets received: %20"PRIu64
			   "\nPackets dropped: %21"PRIu64,
			   portid,
			   port_statistics[portid].tx,
			   port_statistics[portid].rx,
			   port_statistics[portid].dropped);

		total_packets_dropped += port_statistics[portid].dropped;
		total_packets_tx += port_statistics[portid].tx;
		total_packets_rx += port_statistics[portid].rx;
	}
	printf("\nAggregate statistics ==============================="
		   "\nTotal packets sent: %18"PRIu64
		   "\nTotal packets received: %14"PRIu64
		   "\nTotal packets dropped: %15"PRIu64,
		   total_packets_tx,
		   total_packets_rx,
		   total_packets_dropped);
	printf("\n====================================================\n");
}

static void
l2fwd_mac_updating(struct rte_mbuf *m, unsigned dest_portid) // 改写包的 MAC 层信息
{
	struct ether_hdr *eth;
	void *tmp;

	eth = rte_pktmbuf_mtod(m, struct ether_hdr *);

	/* 02:00:00:00:00:xx */
	tmp = &eth->d_addr.addr_bytes[0]; // 改写目的 MAC 地址为 02:00:00:00:00:<发送端口的port id>
	*((uint64_t *)tmp) = 0x000000000002 + ((uint64_t)dest_portid << 40);

	/* src addr */
	// 改写 源 MAC 地址 改写成发送端口的MAC地址
	ether_addr_copy(&l2fwd_ports_eth_addr[dest_portid], &eth->s_addr);
}

static void
l2fwd_simple_forward(struct rte_mbuf *m, unsigned portid)
{
	unsigned dst_port;
	int sent;
	struct rte_eth_dev_tx_buffer *buffer;

	dst_port = l2fwd_dst_ports[portid]; // 与之配对的端口

	if (mac_updating) // 如果开启了 mac updating 模式
		l2fwd_mac_updating(m, dst_port); // 调整 MAC 地址

	buffer = tx_buffer[dst_port]; // 该端口的 tx_buffer
	sent = rte_eth_tx_buffer(dst_port, 0, buffer, m); // 将收到的包缓存在 tx_buffer 里，用于未来的发送。
	// 返回值 如果为0，表示 pkt 已经被缓存
	// 返回值 N>0，表示由于缓冲区被flush导致N个pkt被发送。

	if (sent)
		port_statistics[dst_port].tx += sent;
}

/* main processing loop */
static void
l2fwd_main_loop(void)
{
	struct rte_mbuf *pkts_burst[MAX_PKT_BURST];
	struct rte_mbuf *m;
	int sent;
	unsigned lcore_id;
	uint64_t prev_tsc, diff_tsc, cur_tsc, timer_tsc;
	unsigned i, j, portid, nb_rx;
	struct lcore_queue_conf *qconf;
	const uint64_t drain_tsc = (rte_get_tsc_hz() + US_PER_S - 1) / US_PER_S *
			BURST_TX_DRAIN_US;
	struct rte_eth_dev_tx_buffer *buffer;

	prev_tsc = 0;
	timer_tsc = 0;

	lcore_id = rte_lcore_id(); // 获取自己的 lcore id
	qconf = &lcore_queue_conf[lcore_id];

	if (qconf->n_rx_port == 0) { // 因为对每一个 lcore 都执行 main 线程，如果该 lcore 上没有绑定端口，就无事可做。
		RTE_LOG(INFO, L2FWD, "lcore %u has nothing to do\n", lcore_id);
		return;
	}

	RTE_LOG(INFO, L2FWD, "entering main loop on lcore %u\n", lcore_id);

	for (i = 0; i < qconf->n_rx_port; i++) {

		portid = qconf->rx_port_list[i];
		RTE_LOG(INFO, L2FWD, " -- lcoreid=%u portid=%u\n", lcore_id,
			portid); // 显示一下 lcore 和 port 的对应关系

	}

	while (!force_quit) {

		cur_tsc = rte_rdtsc(); // 获取从开机起至当前的时间戳

		/*
		 * TX burst queue drain
		 * 发送逻辑
		 */
		diff_tsc = cur_tsc - prev_tsc;
		if (unlikely(diff_tsc > drain_tsc)) { // 时间到了
			// 如果tx_buffer满，会发送一批 pkt 出去。如果没满，为了保证没有没被发出的 pkt，所以每个一小段时间，也会发送队列中的包
			for (i = 0; i < qconf->n_rx_port; i++) { // 对 lcore 负责的每个端口

				portid = l2fwd_dst_ports[qconf->rx_port_list[i]]; // 与之配对的端口
				buffer = tx_buffer[portid];

				sent = rte_eth_tx_buffer_flush(portid, 0, buffer); // 将 buffer 里的 pkt 全部从 port id 的 0号 Tx queue 发出去
				if (sent) // 返回值是成功发出的 pkt 数量
					port_statistics[portid].tx += sent;

			}

			/* if timer is enabled */
			if (timer_period > 0) {

				/* advance the timer */
				timer_tsc += diff_tsc;

				/* if timer has reached its timeout */
				if (unlikely(timer_tsc >= timer_period)) {

					/* do this only on master core */
					if (lcore_id == rte_get_master_lcore()) { // 如果计时器到了，就打印一下信息。只在主核心打印信息
						print_stats();
						/* reset the timer */
						timer_tsc = 0;
					}
				}
			}

			prev_tsc = cur_tsc;
		}

		/*
		 * Read packet from RX queues
		 * 接收逻辑
		 */
		for (i = 0; i < qconf->n_rx_port; i++) { // 对 lcore 负责的每个端口

			portid = qconf->rx_port_list[i]; // 获取端口号
			nb_rx = rte_eth_rx_burst(portid, 0,
						 pkts_burst, MAX_PKT_BURST); // 收包，收到该端口的 0 号 rx queue

			port_statistics[portid].rx += nb_rx; // 更新端口上的收包计数器

			for (j = 0; j < nb_rx; j++) { // 对每一个包
				m = pkts_burst[j]; // 包的 mbuf 指针
				// Prefetch: 预取一个 cache 行。参数是要取的地址，类型 void *
				// rte_pktmbuf_mtod：返回 mbuf 中 data 的起始地址
				rte_prefetch0(rte_pktmbuf_mtod(m, void *));
				l2fwd_simple_forward(m, portid); // 收包后进行 L2fwd !!
			}
		}
	}
}

static int
l2fwd_launch_one_lcore(__attribute__((unused)) void *dummy)
{
	l2fwd_main_loop();
	return 0;
}

/* display usage */
static void
l2fwd_usage(const char *prgname)
{
	printf("%s [EAL options] -- -p PORTMASK [-q NQ]\n"
	       "  -p PORTMASK: hexadecimal bitmask of ports to configure\n"
	       "  -q NQ: number of queue (=ports) per lcore (default is 1)\n"
		   "  -T PERIOD: statistics will be refreshed each PERIOD seconds (0 to disable, 10 default, 86400 maximum)\n"
		   "  --[no-]mac-updating: Enable or disable MAC addresses updating (enabled by default)\n"
		   "      When enabled:\n"
		   "       - The source MAC address is replaced by the TX port MAC address\n"
		   "       - The destination MAC address is replaced by 02:00:00:00:00:TX_PORT_ID\n",
	       prgname);
}

static int
l2fwd_parse_portmask(const char *portmask)
{
	char *end = NULL;
	unsigned long pm;

	/* parse hexadecimal string */
	pm = strtoul(portmask, &end, 16); // 将字符串 portmask 转成 16 进制无符号长整形
	if ((portmask[0] == '\0') || (end == NULL) || (*end != '\0'))
		return -1;

	if (pm == 0)
		return -1;

	return pm;
}

static unsigned int
l2fwd_parse_nqueue(const char *q_arg)
{
	char *end = NULL;
	unsigned long n;

	/* parse hexadecimal string */
	n = strtoul(q_arg, &end, 10);
	if ((q_arg[0] == '\0') || (end == NULL) || (*end != '\0'))
		return 0;
	if (n == 0)
		return 0;
	if (n >= MAX_RX_QUEUE_PER_LCORE)
		return 0;

	return n;
}

static int
l2fwd_parse_timer_period(const char *q_arg)
{
	char *end = NULL;
	int n;

	/* parse number string */
	n = strtol(q_arg, &end, 10);
	if ((q_arg[0] == '\0') || (end == NULL) || (*end != '\0'))
		return -1;
	if (n >= MAX_TIMER_PERIOD)
		return -1;

	return n;
}

static const char short_options[] =
	"p:"  /* portmask */
	"q:"  /* number of queues */
	"T:"  /* timer period */
	;

#define CMD_LINE_OPT_MAC_UPDATING "mac-updating"
#define CMD_LINE_OPT_NO_MAC_UPDATING "no-mac-updating"

enum {
	/* long options mapped to a short option */

	/* first long only option value must be >= 256, so that we won't
	 * conflict with short options */
	CMD_LINE_OPT_MIN_NUM = 256,
};

static const struct option lgopts[] = {
	{ CMD_LINE_OPT_MAC_UPDATING, no_argument, &mac_updating, 1},
	{ CMD_LINE_OPT_NO_MAC_UPDATING, no_argument, &mac_updating, 0},
	{NULL, 0, 0, 0}
};

/* Parse the argument given in the command line of the application */
static int
l2fwd_parse_args(int argc, char **argv)
{
	int opt, ret, timer_secs;
	char **argvopt;
	int option_index;
	char *prgname = argv[0]; // l2fwd

	argvopt = argv;

	while ((opt = getopt_long(argc, argvopt, short_options,
				  lgopts, &option_index)) != EOF) { // linux 下解析命令行参数的函数。支持由两个横杠开头的长选项。
		// 关于这个函数可以 man getopt_long
		switch (opt) { // 解析成功时返回字符
		/* portmask */
		case 'p': // 端口掩码
			l2fwd_enabled_port_mask = l2fwd_parse_portmask(optarg); // 解析成功时，将字符后面的参数放到 optarg 里
			if (l2fwd_enabled_port_mask == 0) {
				printf("invalid portmask\n");
				l2fwd_usage(prgname);
				return -1;
			}
			break;

		/* nqueue */
		case 'q': // A number of queues (=ports) per lcore (default is 1)
				  // q 后面跟着的数字是每个逻辑核心上要绑定多少个队列（端口）
				  // 例如 -q 4 意味着该应用使用一个 lcore 轮询 4个端口。如果共有16个端口，则只需要4个lcore
			l2fwd_rx_queue_per_lcore = l2fwd_parse_nqueue(optarg);
			if (l2fwd_rx_queue_per_lcore == 0) {
				printf("invalid queue number\n");
				l2fwd_usage(prgname);
				return -1;
			}
			break;

		/* timer period */
		case 'T':
			timer_secs = l2fwd_parse_timer_period(optarg);
			if (timer_secs < 0) {
				printf("invalid timer period\n");
				l2fwd_usage(prgname);
				return -1;
			}
			timer_period = timer_secs;
			break;

		/* long options */
		case 0: // 解析到了长选项 会返回0，长选项形如 --arg=param or --arg param.
			break;

		default:
			l2fwd_usage(prgname);
			return -1;
		}
	}

	if (optind >= 0) // optind 是 argv 中下一个要被处理的参数的 index
		argv[optind-1] = prgname;

	ret = optind-1;
	optind = 1; /* reset getopt lib */ // 解析完所有的参数要让 optind 重新指向 1
	return ret;
}

/* Check the link status of all ports in up to 9s, and print them finally */
static void
check_all_ports_link_status(uint32_t port_mask)
{
#define CHECK_INTERVAL 100 /* 100ms */
#define MAX_CHECK_TIME 90 /* 9s (90 * 100ms) in total */
	uint16_t portid;
	uint8_t count, all_ports_up, print_flag = 0;
	struct rte_eth_link link;

	printf("\nChecking link status");
	fflush(stdout);
	for (count = 0; count <= MAX_CHECK_TIME; count++) {
		if (force_quit)
			return;
		all_ports_up = 1;
		RTE_ETH_FOREACH_DEV(portid) {
			if (force_quit)
				return;
			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",
						portid, link.link_speed,
				(link.link_duplex == ETH_LINK_FULL_DUPLEX) ?
					("full-duplex") : ("half-duplex\n"));
				else
					printf("Port %d Link Down\n", portid);
				continue;
			}
			/* clear all_ports_up flag if any link down */
			if (link.link_status == ETH_LINK_DOWN) {
				all_ports_up = 0;
				break;
			}
		}
		/* after finally printing all link status, get out */
		if (print_flag == 1)
			break;

		if (all_ports_up == 0) {
			printf(".");
			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;
			printf("done\n");
		}
	}
}

static void
signal_handler(int signum)
{
	if (signum == SIGINT || signum == SIGTERM) {
		printf("\n\nSignal %d received, preparing to exit...\n",
				signum);
		force_quit = true; //当我们退出是ctrl+c不是直接将进程杀死，而是会将force_quit置为true，让程序自然退出，这样程序就来得及完成最后退出之前的操作。
	}
}

int
main(int argc, char **argv)
{
	struct lcore_queue_conf *qconf;
	int ret;
	uint16_t nb_ports;
	uint16_t nb_ports_available = 0;
	uint16_t portid, last_port;
	unsigned lcore_id, rx_lcore_id;
	unsigned nb_ports_in_mask = 0;
	unsigned int nb_lcores = 0;
	unsigned int nb_mbufs;

	/* init EAL */
	// 解析 EAL 的参数
	ret = rte_eal_init(argc, argv);
	if (ret < 0)
		rte_exit(EXIT_FAILURE, "Invalid EAL arguments\n");
	argc -= ret;
	argv += ret;

	force_quit = false;
	signal(SIGINT, signal_handler);
	signal(SIGTERM, signal_handler);

	/* parse application arguments (after the EAL ones) */
	// 解析 l2fwd 的运行参数
	ret = l2fwd_parse_args(argc, argv);
	if (ret < 0)
		rte_exit(EXIT_FAILURE, "Invalid L2FWD arguments\n");

	printf("MAC updating %s\n", mac_updating ? "enabled" : "disabled"); // 默认开启 mac updating 这一功能。

	/* convert to number of cycles */
	timer_period *= rte_get_timer_hz(); // 获得CPU主频，单位hz (1s多少个cycle)，位于rte_cycles.h

	nb_ports = rte_eth_dev_count(); // 网口数量
	if (nb_ports == 0)
		rte_exit(EXIT_FAILURE, "No Ethernet ports - bye\n");

	/* check port mask to possible port mask */
	// 检查掩码和可用网口数量是否有冲突
	if (l2fwd_enabled_port_mask & ~((1 << nb_ports) - 1))
		rte_exit(EXIT_FAILURE, "Invalid portmask; possible (0x%x)\n",
			(1 << nb_ports) - 1);

	/* reset l2fwd_dst_ports */

	for (portid = 0; portid < RTE_MAX_ETHPORTS; portid++)
		l2fwd_dst_ports[portid] = 0; // 先重置这个数组
	last_port = 0; 

	/*
	 * Each logical core is assigned a dedicated TX queue on each port.
	 */
	RTE_ETH_FOREACH_DEV(portid) { // 使用RTE_ETH_FOREACH_DEV()宏来访问所有的 ethdev
		/* skip ports that are not enabled */
		if ((l2fwd_enabled_port_mask & (1 << portid)) == 0)
			continue;

		if (nb_ports_in_mask % 2) {
			l2fwd_dst_ports[portid] = last_port;
			l2fwd_dst_ports[last_port] = portid;
		}
		else
			last_port = portid;

		nb_ports_in_mask++; // 这些逻辑可以实现 basicfwd 那样的一对对端口互相转发。
	}
	if (nb_ports_in_mask % 2) {
		printf("Notice: odd number of ports in portmask.\n");
		l2fwd_dst_ports[last_port] = last_port; // 如果是奇数个端口，会有最后一个端口的 dst_port 是自己
	}

	rx_lcore_id = 0; // 从逻辑核心id 0开始
	qconf = NULL;

	/* Initialize the port/queue configuration of each logical core */
	// 在每一个端口上，配置逻辑核、配置队列。
	RTE_ETH_FOREACH_DEV(portid) {
		/* skip ports that are not enabled */
		if ((l2fwd_enabled_port_mask & (1 << portid)) == 0)
			continue;

		/* get the lcore_id for this port */
		// 为该端口配置一个逻辑核。
		while (rte_lcore_is_enabled(rx_lcore_id) == 0 ||
		       lcore_queue_conf[rx_lcore_id].n_rx_port ==
		       l2fwd_rx_queue_per_lcore) {
			/*从lcore id = 0 开始循环：
			如果：如果该 lcore id 有效（已经被占用），则检查下一个逻辑核。
			如果该 lcore 是空闲的，要检查该 lcore 上绑定了多少个端口，如果到达了最大端口数量限制也会循环。*/
			rx_lcore_id++;

			if (rx_lcore_id >= RTE_MAX_LCORE) // RTE_MAX_LCORE 宏 64
				rte_exit(EXIT_FAILURE, "Not enough cores\n"); // 逻辑核心不足
		}
		// 跳出循环时，rx_lcore_id 变量存储了一个可用的 lcore id，绑定该端口到这个 lcore

		if (qconf != &lcore_queue_conf[rx_lcore_id]) {
			/* Assigned a new logical core in the loop above. */
			qconf = &lcore_queue_conf[rx_lcore_id];
			nb_lcores++;
			// qconf 是一个指针，指向当前进行配置的 lcore 的，用于存放配置信息的结构体
		}
		qconf->rx_port_list[qconf->n_rx_port] = portid;
		qconf->n_rx_port++;
		// 绑定就是在这个核处理的端口列表中加上当前这个端口，然后该核绑定的端口数加 1。
		printf("Lcore %u: RX port %u\n", rx_lcore_id, portid);
	}

	nb_mbufs = RTE_MAX(nb_ports * (nb_rxd + nb_txd + MAX_PKT_BURST +
		nb_lcores * MEMPOOL_CACHE_SIZE), 8192U);
		// mbuf中的元素个数，取 8192 和 （端口数 * （队列长度 * 2 + 一个 Burst 的 pkt 数量 + 逻辑核数 * cache size）） 两者中较大的一个。

	/* create the mbuf pool */
	// 初始化内存池，用于 rx 队列接收 pkt 用
	l2fwd_pktmbuf_pool = rte_pktmbuf_pool_create("mbuf_pool", nb_mbufs,
		MEMPOOL_CACHE_SIZE, 0, RTE_MBUF_DEFAULT_BUF_SIZE,
		rte_socket_id());
	if (l2fwd_pktmbuf_pool == NULL)
		rte_exit(EXIT_FAILURE, "Cannot init mbuf pool\n");

	/* Initialise each port */
	// 端口初始化
	RTE_ETH_FOREACH_DEV(portid) {
		struct rte_eth_rxconf rxq_conf; // rx queue 的配置信息
		struct rte_eth_txconf txq_conf; // tx queue 的配置信息
		struct rte_eth_conf local_port_conf = port_conf; // 配置端口时使用的配置信息
		struct rte_eth_dev_info dev_info; // 以太网设备的信息

		/* skip ports that are not enabled */
		if ((l2fwd_enabled_port_mask & (1 << portid)) == 0) {
			printf("Skipping disabled port %u\n", portid);
			continue;
		}
		nb_ports_available++;

		/* init port */
		printf("Initializing port %u... ", portid);
		fflush(stdout); // 清除写缓冲区，强迫未写入磁盘的内容立即写入
		rte_eth_dev_info_get(portid, &dev_info); // 获取以太网设备信息
		if (dev_info.tx_offload_capa & DEV_TX_OFFLOAD_MBUF_FAST_FREE)
			local_port_conf.txmode.offloads |=
				DEV_TX_OFFLOAD_MBUF_FAST_FREE; // mbuf fast free，支持快速发包
		ret = rte_eth_dev_configure(portid, 1, 1, &local_port_conf); // 配置收发队列各 1 条
		/*本程序中，Rx队列只能有一条，确保一个 lcore 负责轮询一个 port
		 Tx 队列则可以根据可用的 lcore 数目更改。*/

		if (ret < 0)
			rte_exit(EXIT_FAILURE, "Cannot configure device: err=%d, port=%u\n",
				  ret, portid);

		ret = rte_eth_dev_adjust_nb_rx_tx_desc(portid, &nb_rxd,
						       &nb_txd);
		if (ret < 0)
			rte_exit(EXIT_FAILURE,
				 "Cannot adjust number of descriptors: err=%d, port=%u\n",
				 ret, portid);

		rte_eth_macaddr_get(portid,&l2fwd_ports_eth_addr[portid]); // 获取设备的MAC地址，写在后一个结构体里

		/* init one RX queue */
		// 配置 rx 队列
		fflush(stdout);
		rxq_conf = dev_info.default_rxconf;
		rxq_conf.offloads = local_port_conf.rxmode.offloads;
		ret = rte_eth_rx_queue_setup(portid, 0, nb_rxd,
					     rte_eth_dev_socket_id(portid),
					     &rxq_conf,
					     l2fwd_pktmbuf_pool);
		if (ret < 0)
			rte_exit(EXIT_FAILURE, "rte_eth_rx_queue_setup:err=%d, port=%u\n",
				  ret, portid);

		/* init one TX queue on each port */
		// 每个 port 配置一条 tx 队列
		fflush(stdout);
		txq_conf = dev_info.default_txconf;
		txq_conf.txq_flags = ETH_TXQ_FLAGS_IGNORE;
		txq_conf.offloads = local_port_conf.txmode.offloads;
		ret = rte_eth_tx_queue_setup(portid, 0, nb_txd,
				rte_eth_dev_socket_id(portid),
				&txq_conf);
		if (ret < 0)
			rte_exit(EXIT_FAILURE, "rte_eth_tx_queue_setup:err=%d, port=%u\n",
				ret, portid);

		/* Initialize TX buffers */
		// 为每个端口的 Tx 分配发送缓冲区
		tx_buffer[portid] = rte_zmalloc_socket("tx_buffer",
				RTE_ETH_TX_BUFFER_SIZE(MAX_PKT_BURST), 0,
				rte_eth_dev_socket_id(portid)); // 为 tx buffer 分配空间。
												// 宏RTE_ETH_TX_BUFFER_SIZE(x) ：计算 tx buffer 的 size，参数x是包的个数

		if (tx_buffer[portid] == NULL)
			rte_exit(EXIT_FAILURE, "Cannot allocate buffer for tx on port %u\n",
					portid);

		rte_eth_tx_buffer_init(tx_buffer[portid], MAX_PKT_BURST); // 初始化 Tx buffer，参数是 buffer 指针和 buffer size。

		/* rte_eth_tx_buffer_set_err_callback() 对于不能被发送的 pkt 配置回调函数。
		在尝试发送一个 tx buffer 的所有 pkt，遇到问题不能全部成功发送，就会触发设置好的回调函数。
		默认行为是丢包。如果要其他的行为（例如重传，计数）则需要额外的代码。也有设置好的API例如rte_eth_count_unsent_packet_callback()等，和本函数中用的也是。
		参数 1. tx_buffer 指针，2.回调函数的指针。3. 回调函数的参数
		*/
		ret = rte_eth_tx_buffer_set_err_callback(tx_buffer[portid],
				rte_eth_tx_buffer_count_callback, // 丢包，并更新计数器
				&port_statistics[portid].dropped); // 计数器的指针放到第三个参数
		if (ret < 0)
			rte_exit(EXIT_FAILURE,
			"Cannot set error callback for tx buffer on port %u\n",
				 portid);

		/* Start device */
		// 启用设备
		ret = rte_eth_dev_start(portid);
		if (ret < 0)
			rte_exit(EXIT_FAILURE, "rte_eth_dev_start:err=%d, port=%u\n",
				  ret, portid);

		printf("done: \n");

		rte_eth_promiscuous_enable(portid); // 混杂模式

		printf("Port %u, MAC address: %02X:%02X:%02X:%02X:%02X:%02X\n\n",
				portid,
				l2fwd_ports_eth_addr[portid].addr_bytes[0],
				l2fwd_ports_eth_addr[portid].addr_bytes[1],
				l2fwd_ports_eth_addr[portid].addr_bytes[2],
				l2fwd_ports_eth_addr[portid].addr_bytes[3],
				l2fwd_ports_eth_addr[portid].addr_bytes[4],
				l2fwd_ports_eth_addr[portid].addr_bytes[5]);

		/* initialize port stats */
		memset(&port_statistics, 0, sizeof(port_statistics));
	}

	if (!nb_ports_available) {
		rte_exit(EXIT_FAILURE,
			"All available ports are disabled. Please set portmask.\n");
	}

	check_all_ports_link_status(l2fwd_enabled_port_mask); // 检查所有链路的状态，可以参考 flow_filtering

	ret = 0;
	/* launch per-lcore init on every lcore */
	// 这里就是DPDK的典型执行方法，分配所有 lcore 执行函数
	rte_eal_mp_remote_launch(l2fwd_launch_one_lcore, NULL, CALL_MASTER);
	RTE_LCORE_FOREACH_SLAVE(lcore_id) {
		if (rte_eal_wait_lcore(lcore_id) < 0) {
			ret = -1;
			break;
		}
	}

	RTE_ETH_FOREACH_DEV(portid) {
		if ((l2fwd_enabled_port_mask & (1 << portid)) == 0)
			continue;
		printf("Closing port %d...", portid);
		rte_eth_dev_stop(portid);
		rte_eth_dev_close(portid);
		printf(" Done\n");
	}
	printf("Bye...\n");

	return ret;
}

二层转发和普通的端口转发有什么区别呢？简单来说有几个：

特点	L2fwd	basicfwd
端口数量	两者都用端口掩码来指定，L2fwd支持奇数个	只能是偶数个
lcore数量	多个，每个lcore负责一个port	一个lcore，执行类似repeater的程序
转发逻辑	转发时会改写MAC地址	只能是 0<-->1，2<-->3 这样的 pair 互相转发
Tx_buffer	有发包缓存队列，收的包会缓存到发包队列里，一段时间后或者队列满后才会转发	没有发包缓存，Rx收到包后直接Tx出去

运行情况

root@ubuntu:/home/chang/dpdk/examples/l2fwd/build# ./l2fwd -l 0-3 -n 4 -- -p 0x3EAL: Detected 8 lcore(s)
EAL: No free hugepages reported in hugepages-1048576kB
EAL: Multi-process socket /var/run/.rte_unix
EAL: Probing VFIO support...
EAL: PCI device 0000:02:01.0 on NUMA socket -1
EAL:   Invalid NUMA socket, default to 0
EAL:   probe driver: 8086:100f net_e1000_em
EAL: PCI device 0000:02:02.0 on NUMA socket -1
EAL:   Invalid NUMA socket, default to 0
EAL:   probe driver: 8086:100f net_e1000_em
EAL: PCI device 0000:02:03.0 on NUMA socket -1
EAL:   Invalid NUMA socket, default to 0
EAL:   probe driver: 8086:100f net_e1000_em
EAL: PCI device 0000:02:04.0 on NUMA socket -1
EAL:   Invalid NUMA socket, default to 0
EAL:   probe driver: 8086:100f net_e1000_em
MAC updating enabled
Lcore 0: RX port 0
Lcore 1: RX port 1
Initializing port 0... done:
Port 0, MAC address: 00:0C:29:F7:4D:25

Initializing port 1... done:
Port 1, MAC address: 00:0C:29:F7:4D:2F

Checking link statusdone
Port0 Link Up. Speed 1000 Mbps - full-duplex
Port1 Link Up. Speed 1000 Mbps - full-duplex
L2FWD: entering main loop on lcore 1
L2FWD:  -- lcoreid=1 portid=1
L2FWD: lcore 3 has nothing to do
L2FWD: entering main loop on lcore 0
L2FWD:  -- lcoreid=0 portid=0

Port statistics ====================================
Statistics for port 0 ------------------------------
Packets sent:                  2152346
Packets received:              2166674
Packets dropped:                     0
Statistics for port 1 ------------------------------
Packets sent:                  2166674
Packets received:              2152371
Packets dropped:                     0
Aggregate statistics ===============================
Total packets sent:            4319020
Total packets received:        4319045
Total packets dropped:               0
====================================================

用 wireshark 抓下包：

可以看到经过端口转发的包的目的MAC地址被程序改变了。

但是改变了目的MAC地址，自然无法通信。所以这个程序还是设置成测试速率用。如果把修改目的MAC地址的那一行代码注释掉，就可以正常通信。

Sample guide 里有一句话说：The L2 Forwarding application can also be used as a starting point for developing a new application based on the DPDK. 所以这个程序也是非常亲民的=。=

reference

作为典型应用，搜索能搜到很多相关代码阅读的博客。

参考了：https://blog.csdn.net/yangye2014/article/details/78064634?locationNum=6&fps=1