dpdk-pingpong/pingpong.c

320 lines
8.5 KiB
C

/* SPDX-License-Identifier: BSD-3-Clause
* Copyright(c) 2021+ Vitaliy Filippov
*/
// DPDK-based ping-pong
// USAGE:
// Host 1: RECV=1 ./pingpong
// Host 2: ./pingpong
#include <stdint.h>
#include <inttypes.h>
#include <rte_eal.h>
#include <rte_ethdev.h>
#include <rte_cycles.h>
#include <rte_lcore.h>
#include <rte_mbuf.h>
#include <rte_ether.h>
#include <rte_ip.h>
#include <rte_udp.h>
#define RX_RING_SIZE 1024
#define TX_RING_SIZE 1024
#define NUM_MBUFS 8191
#define MBUF_CACHE_SIZE 250
#define BURST_SIZE 32
static const struct rte_eth_conf port_conf_default = {
.rxmode = {
.max_rx_pkt_len = RTE_ETHER_MAX_LEN,
},
};
/*
* Initializes a given port using global settings and with the RX buffers
* coming from the mbuf_pool passed as a parameter.
*/
static inline int
port_init(uint16_t port, struct rte_mempool *mbuf_pool)
{
struct rte_eth_conf port_conf = port_conf_default;
const uint16_t rx_rings = 1, tx_rings = 1;
uint16_t nb_rxd = RX_RING_SIZE;
uint16_t nb_txd = TX_RING_SIZE;
int retval;
uint16_t q;
struct rte_eth_dev_info dev_info;
struct rte_eth_txconf txconf;
if (!rte_eth_dev_is_valid_port(port))
return -1;
retval = rte_eth_dev_info_get(port, &dev_info);
if (retval != 0) {
printf("Error during getting device (port %u) info: %s\n",
port, strerror(-retval));
return retval;
}
if (dev_info.tx_offload_capa & DEV_TX_OFFLOAD_MBUF_FAST_FREE)
port_conf.txmode.offloads |=
DEV_TX_OFFLOAD_MBUF_FAST_FREE;
/* Configure the Ethernet device. */
retval = rte_eth_dev_configure(port, rx_rings, tx_rings, &port_conf);
if (retval != 0)
return retval;
retval = rte_eth_dev_adjust_nb_rx_tx_desc(port, &nb_rxd, &nb_txd);
if (retval != 0)
return retval;
/* Allocate and set up 1 RX queue per Ethernet port. */
for (q = 0; q < rx_rings; q++) {
retval = rte_eth_rx_queue_setup(port, q, nb_rxd,
rte_eth_dev_socket_id(port), NULL, mbuf_pool);
if (retval < 0)
return retval;
}
txconf = dev_info.default_txconf;
txconf.offloads = port_conf.txmode.offloads;
/* Allocate and set up 1 TX queue per Ethernet port. */
for (q = 0; q < tx_rings; q++) {
retval = rte_eth_tx_queue_setup(port, q, nb_txd,
rte_eth_dev_socket_id(port), &txconf);
if (retval < 0)
return retval;
}
if (9000 > dev_info.max_mtu || 9000 < dev_info.min_mtu) {
printf("Set MTU failed. MTU:%u is not in valid range, min:%u - max:%u\n",
9000, dev_info.min_mtu, dev_info.max_mtu);
return -1;
}
retval = rte_eth_dev_set_mtu(port, 9000);
if (retval < 0) {
printf("Set MTU failed. retval=%d\n", retval);
return retval;
}
/* Start the Ethernet port. */
retval = rte_eth_dev_start(port);
if (retval < 0)
return retval;
/* Display the port MAC address. */
struct rte_ether_addr addr;
retval = rte_eth_macaddr_get(port, &addr);
if (retval != 0)
return retval;
printf("Port %u MAC: %02" PRIx8 " %02" PRIx8 " %02" PRIx8
" %02" PRIx8 " %02" PRIx8 " %02" PRIx8 "\n",
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]);
/* Enable RX in promiscuous mode for the Ethernet device. */
retval = rte_eth_promiscuous_enable(port);
if (retval != 0)
return retval;
return 0;
}
/*
* The lcore main. This is the main thread that does the work, reading from
* an input port and writing to an output port.
*/
static __attribute__((noreturn)) void
lcore_recv_main(void)
{
uint16_t port;
/*
* Check that the port is on the same NUMA node as the polling thread
* for best performance.
*/
RTE_ETH_FOREACH_DEV(port)
if (rte_eth_dev_socket_id(port) > 0 &&
rte_eth_dev_socket_id(port) !=
(int)rte_socket_id())
printf("WARNING, port %u is on remote NUMA node to "
"polling thread.\n\tPerformance will "
"not be optimal.\n", port);
printf("\nCore %u echoing packets. [Ctrl+C to quit]\n",
rte_lcore_id());
uint16_t buf;
struct rte_ether_addr s_addr;
/* Run until the application is quit or killed. */
for (;;) {
/*
* Receive packets on a port and forward them on the paired
* port. The mapping is 0 -> 1, 1 -> 0, 2 -> 3, 3 -> 2, etc.
*/
RTE_ETH_FOREACH_DEV(port) {
/* Get burst of RX packets, from first port of pair. */
struct rte_mbuf *bufs[BURST_SIZE];
const uint16_t nb_rx = rte_eth_rx_burst(port, 0,
bufs, BURST_SIZE);
if (unlikely(nb_rx == 0))
continue;
for (buf = 0; buf < nb_rx; buf++) {
struct rte_ether_hdr *eth_hdr = rte_pktmbuf_mtod(bufs[buf], struct rte_ether_hdr*);
s_addr = eth_hdr->s_addr;
eth_hdr->s_addr = eth_hdr->d_addr;
eth_hdr->d_addr = s_addr;
}
/* Send burst of TX packets, to second port of pair. */
const uint16_t nb_tx = rte_eth_tx_burst(port, 0,
bufs, nb_rx);
/* Free any unsent packets. */
if (unlikely(nb_tx < nb_rx)) {
for (buf = nb_tx; buf < nb_rx; buf++)
rte_pktmbuf_free(bufs[buf]);
}
}
}
}
static struct rte_mempool *mbuf_pool;
static uint64_t lat_count = 0, lat_us = 0, lat_min = 0, lat_max = 0;
static double lat_mean = 0, lat_m2 = 0;
static __attribute__((noreturn)) void
lcore_send_main(void)
{
uint16_t port;
/*
* Check that the port is on the same NUMA node as the polling thread
* for best performance.
*/
RTE_ETH_FOREACH_DEV(port)
if (rte_eth_dev_socket_id(port) > 0 &&
rte_eth_dev_socket_id(port) !=
(int)rte_socket_id())
printf("WARNING, port %u is on remote NUMA node to "
"polling thread.\n\tPerformance will "
"not be optimal.\n", port);
printf("\nCore %u sending echo packets. [Ctrl+C to quit]\n",
rte_lcore_id());
struct rte_ether_hdr *eth_hdr;
struct rte_ether_addr s_addr = {{0xc6,0x47,0x2c,0x07,0x8a,0x24}};
struct rte_ether_addr d_addr = {{0xbc,0x01,0xa4,0x9b,0x03,0x98}};
uint16_t ether_type = 0x0a00;
uint64_t lat_this;
/* Run until the application is quit or killed. */
struct timespec ckpt, start, end;
clock_gettime(CLOCK_REALTIME, &ckpt);
for (;;) {
RTE_ETH_FOREACH_DEV(port) {
clock_gettime(CLOCK_REALTIME, &start);
struct rte_mbuf *pkt[BURST_SIZE];
int i;
for (i = 0; i < 1; i++) {
pkt[i] = rte_pktmbuf_alloc(mbuf_pool);
eth_hdr = (struct rte_ether_hdr*)rte_pktmbuf_append(pkt[i], sizeof(struct rte_ether_hdr));
eth_hdr->d_addr = d_addr;
eth_hdr->s_addr = s_addr;
eth_hdr->ether_type = ether_type;
char *data = rte_pktmbuf_append(pkt[i], 4096);
memset(data, 0xaa, 4096);
}
uint16_t nb_tx = 0;
while (nb_tx == 0) {
nb_tx = rte_eth_tx_burst(port, 0, pkt, 1);
}
// Receive reply
pkt[0] = 0;
uint16_t nb_rx = 0;
while (nb_rx == 0) {
nb_rx = rte_eth_rx_burst(port, 0, pkt, BURST_SIZE);
}
for (uint16_t buf = 0; buf < nb_rx; buf++)
rte_pktmbuf_free(pkt[buf]);
clock_gettime(CLOCK_REALTIME, &end);
lat_this = (end.tv_sec-start.tv_sec)*1000000 + (end.tv_nsec-start.tv_nsec)/1000;
lat_count++;
lat_us += lat_this;
if (lat_min == 0 || lat_min > lat_this)
lat_min = lat_this;
if (lat_max < lat_this)
lat_max = lat_this;
double delta = lat_this-lat_mean;
lat_mean += delta/lat_count;
lat_m2 += delta*(lat_this-lat_mean);
if (end.tv_sec > ckpt.tv_sec+5) {
printf("latency: avg %lu us, min %lu us, max %lu us, stddev %lu us (%lu packets)\n",
lat_us/lat_count, lat_min, lat_max, (uint64_t)(lat_m2/(lat_count-1)), lat_count);
lat_count = lat_us = lat_min = lat_max = 0;
lat_mean = lat_m2 = 0;
ckpt = end;
}
}
}
}
/*
* The main function, which does initialization and calls the per-lcore
* functions.
*/
int
main(int argc, char *argv[])
{
unsigned nb_ports;
uint16_t portid;
/* Initialize the Environment Abstraction Layer (EAL). */
int ret = rte_eal_init(argc, argv);
if (ret < 0)
rte_exit(EXIT_FAILURE, "Error with EAL initialization\n");
argc -= ret;
argv += ret;
/* Check that there is a port to send/receive on. */
nb_ports = rte_eth_dev_count_avail();
if (!nb_ports)
rte_exit(EXIT_FAILURE, "Error: no ports\n");
/* Creates a new mempool in memory to hold the mbufs. */
mbuf_pool = rte_pktmbuf_pool_create("MBUF_POOL", NUM_MBUFS * nb_ports,
MBUF_CACHE_SIZE, 0, 9000, rte_socket_id());
if (mbuf_pool == NULL)
rte_exit(EXIT_FAILURE, "Cannot create mbuf pool\n");
// FIXME: Use only one device
/* Initialize all ports. */
RTE_ETH_FOREACH_DEV(portid)
if (port_init(portid, mbuf_pool) != 0)
rte_exit(EXIT_FAILURE, "Cannot init port %"PRIu16 "\n",
portid);
if (rte_lcore_count() > 1)
printf("\nWARNING: Too many lcores enabled. Only 1 used.\n");
/* Call lcore_main on the master core only. */
if (getenv("RECV"))
lcore_recv_main();
else
lcore_send_main();
return 0;
}