summaryrefslogtreecommitdiffstats
path: root/examples/server_node_efd/server/main.c
blob: 597b4c250d5c6d2c2ae9b6b604cb36b01a3d7a04 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
/*-
 *   BSD LICENSE
 *
 *   Copyright(c) 2016-2017 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 <stdlib.h>
#include <string.h>
#include <unistd.h>
#include <stdint.h>
#include <stdarg.h>
#include <inttypes.h>
#include <inttypes.h>
#include <sys/queue.h>
#include <errno.h>
#include <netinet/ip.h>

#include <rte_common.h>
#include <rte_memory.h>
#include <rte_memzone.h>
#include <rte_eal.h>
#include <rte_byteorder.h>
#include <rte_launch.h>
#include <rte_per_lcore.h>
#include <rte_lcore.h>
#include <rte_branch_prediction.h>
#include <rte_atomic.h>
#include <rte_ring.h>
#include <rte_log.h>
#include <rte_debug.h>
#include <rte_mempool.h>
#include <rte_memcpy.h>
#include <rte_mbuf.h>
#include <rte_ether.h>
#include <rte_interrupts.h>
#include <rte_pci.h>
#include <rte_ethdev.h>
#include <rte_byteorder.h>
#include <rte_malloc.h>
#include <rte_string_fns.h>
#include <rte_efd.h>
#include <rte_ip.h>

#include "common.h"
#include "args.h"
#include "init.h"

/*
 * When doing reads from the NIC or the node queues,
 * use this batch size
 */
#define PACKET_READ_SIZE 32

/*
 * Local buffers to put packets in, used to send packets in bursts to the
 * nodes
 */
struct node_rx_buf {
	struct rte_mbuf *buffer[PACKET_READ_SIZE];
	uint16_t count;
};

struct efd_stats {
	uint64_t distributed;
	uint64_t drop;
} flow_dist_stats;

/* One buffer per node rx queue - dynamically allocate array */
static struct node_rx_buf *cl_rx_buf;

static const char *
get_printable_mac_addr(uint8_t port)
{
	static const char err_address[] = "00:00:00:00:00:00";
	static char addresses[RTE_MAX_ETHPORTS][sizeof(err_address)];
	struct ether_addr mac;

	if (unlikely(port >= RTE_MAX_ETHPORTS))
		return err_address;
	if (unlikely(addresses[port][0] == '\0')) {
		rte_eth_macaddr_get(port, &mac);
		snprintf(addresses[port], sizeof(addresses[port]),
				"%02x:%02x:%02x:%02x:%02x:%02x\n",
				mac.addr_bytes[0], mac.addr_bytes[1],
				mac.addr_bytes[2], mac.addr_bytes[3],
				mac.addr_bytes[4], mac.addr_bytes[5]);
	}
	return addresses[port];
}

/*
 * This function displays the recorded statistics for each port
 * and for each node. It uses ANSI terminal codes to clear
 * screen when called. It is called from a single non-master
 * thread in the server process, when the process is run with more
 * than one lcore enabled.
 */
static void
do_stats_display(void)
{
	unsigned int i, j;
	const char clr[] = {27, '[', '2', 'J', '\0'};
	const char topLeft[] = {27, '[', '1', ';', '1', 'H', '\0'};
	uint64_t port_tx[RTE_MAX_ETHPORTS], port_tx_drop[RTE_MAX_ETHPORTS];
	uint64_t node_tx[MAX_NODES], node_tx_drop[MAX_NODES];

	/* to get TX stats, we need to do some summing calculations */
	memset(port_tx, 0, sizeof(port_tx));
	memset(port_tx_drop, 0, sizeof(port_tx_drop));
	memset(node_tx, 0, sizeof(node_tx));
	memset(node_tx_drop, 0, sizeof(node_tx_drop));

	for (i = 0; i < num_nodes; i++) {
		const struct tx_stats *tx = &info->tx_stats[i];

		for (j = 0; j < info->num_ports; j++) {
			const uint64_t tx_val = tx->tx[info->id[j]];
			const uint64_t drop_val = tx->tx_drop[info->id[j]];

			port_tx[j] += tx_val;
			port_tx_drop[j] += drop_val;
			node_tx[i] += tx_val;
			node_tx_drop[i] += drop_val;
		}
	}

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

	printf("PORTS\n");
	printf("-----\n");
	for (i = 0; i < info->num_ports; i++)
		printf("Port %u: '%s'\t", (unsigned int)info->id[i],
				get_printable_mac_addr(info->id[i]));
	printf("\n\n");
	for (i = 0; i < info->num_ports; i++) {
		printf("Port %u - rx: %9"PRIu64"\t"
				"tx: %9"PRIu64"\n",
				(unsigned int)info->id[i], info->rx_stats.rx[i],
				port_tx[i]);
	}

	printf("\nSERVER\n");
	printf("-----\n");
	printf("distributed: %9"PRIu64", drop: %9"PRIu64"\n",
			flow_dist_stats.distributed, flow_dist_stats.drop);

	printf("\nNODES\n");
	printf("-------\n");
	for (i = 0; i < num_nodes; i++) {
		const unsigned long long rx = nodes[i].stats.rx;
		const unsigned long long rx_drop = nodes[i].stats.rx_drop;
		const struct filter_stats *filter = &info->filter_stats[i];

		printf("Node %2u - rx: %9llu, rx_drop: %9llu\n"
				"            tx: %9"PRIu64", tx_drop: %9"PRIu64"\n"
				"            filter_passed: %9"PRIu64", "
				"filter_drop: %9"PRIu64"\n",
				i, rx, rx_drop, node_tx[i], node_tx_drop[i],
				filter->passed, filter->drop);
	}

	printf("\n");
}

/*
 * The function called from each non-master lcore used by the process.
 * The test_and_set function is used to randomly pick a single lcore on which
 * the code to display the statistics will run. Otherwise, the code just
 * repeatedly sleeps.
 */
static int
sleep_lcore(__attribute__((unused)) void *dummy)
{
	/* Used to pick a display thread - static, so zero-initialised */
	static rte_atomic32_t display_stats;

	/* Only one core should display stats */
	if (rte_atomic32_test_and_set(&display_stats)) {
		const unsigned int sleeptime = 1;

		printf("Core %u displaying statistics\n", rte_lcore_id());

		/* Longer initial pause so above printf is seen */
		sleep(sleeptime * 3);

		/* Loop forever: sleep always returns 0 or <= param */
		while (sleep(sleeptime) <= sleeptime)
			do_stats_display();
	}
	return 0;
}

/*
 * Function to set all the node statistic values to zero.
 * Called at program startup.
 */
static void
clear_stats(void)
{
	unsigned int i;

	for (i = 0; i < num_nodes; i++)
		nodes[i].stats.rx = nodes[i].stats.rx_drop = 0;
}

/*
 * send a burst of traffic to a node, assuming there are packets
 * available to be sent to this node
 */
static void
flush_rx_queue(uint16_t node)
{
	uint16_t j;
	struct node *cl;

	if (cl_rx_buf[node].count == 0)
		return;

	cl = &nodes[node];
	if (rte_ring_enqueue_bulk(cl->rx_q, (void **)cl_rx_buf[node].buffer,
			cl_rx_buf[node].count, NULL) != cl_rx_buf[node].count){
		for (j = 0; j < cl_rx_buf[node].count; j++)
			rte_pktmbuf_free(cl_rx_buf[node].buffer[j]);
		cl->stats.rx_drop += cl_rx_buf[node].count;
	} else
		cl->stats.rx += cl_rx_buf[node].count;

	cl_rx_buf[node].count = 0;
}

/*
 * marks a packet down to be sent to a particular node process
 */
static inline void
enqueue_rx_packet(uint8_t node, struct rte_mbuf *buf)
{
	cl_rx_buf[node].buffer[cl_rx_buf[node].count++] = buf;
}

/*
 * This function takes a group of packets and routes them
 * individually to the node process. Very simply round-robins the packets
 * without checking any of the packet contents.
 */
static void
process_packets(uint32_t port_num __rte_unused, struct rte_mbuf *pkts[],
		uint16_t rx_count, unsigned int socket_id)
{
	uint16_t i;
	uint8_t node;
	efd_value_t data[RTE_EFD_BURST_MAX];
	const void *key_ptrs[RTE_EFD_BURST_MAX];

	struct ipv4_hdr *ipv4_hdr;
	uint32_t ipv4_dst_ip[RTE_EFD_BURST_MAX];

	for (i = 0; i < rx_count; i++) {
		/* Handle IPv4 header.*/
		ipv4_hdr = rte_pktmbuf_mtod_offset(pkts[i], struct ipv4_hdr *,
				sizeof(struct ether_hdr));
		ipv4_dst_ip[i] = ipv4_hdr->dst_addr;
		key_ptrs[i] = (void *)&ipv4_dst_ip[i];
	}

	rte_efd_lookup_bulk(efd_table, socket_id, rx_count,
				(const void **) key_ptrs, data);
	for (i = 0; i < rx_count; i++) {
		node = (uint8_t) ((uintptr_t)data[i]);

		if (node >= num_nodes) {
			/*
			 * Node is out of range, which means that
			 * flow has not been inserted
			 */
			flow_dist_stats.drop++;
			rte_pktmbuf_free(pkts[i]);
		} else {
			flow_dist_stats.distributed++;
			enqueue_rx_packet(node, pkts[i]);
		}
	}

	for (i = 0; i < num_nodes; i++)
		flush_rx_queue(i);
}

/*
 * Function called by the master lcore of the DPDK process.
 */
static void
do_packet_forwarding(void)
{
	unsigned int port_num = 0; /* indexes the port[] array */
	unsigned int socket_id = rte_socket_id();

	for (;;) {
		struct rte_mbuf *buf[PACKET_READ_SIZE];
		uint16_t rx_count;

		/* read a port */
		rx_count = rte_eth_rx_burst(info->id[port_num], 0,
				buf, PACKET_READ_SIZE);
		info->rx_stats.rx[port_num] += rx_count;

		/* Now process the NIC packets read */
		if (likely(rx_count > 0))
			process_packets(port_num, buf, rx_count, socket_id);

		/* move to next port */
		if (++port_num == info->num_ports)
			port_num = 0;
	}
}

int
main(int argc, char *argv[])
{
	/* initialise the system */
	if (init(argc, argv) < 0)
		return -1;
	RTE_LOG(INFO, APP, "Finished Process Init.\n");

	cl_rx_buf = calloc(num_nodes, sizeof(cl_rx_buf[0]));

	/* clear statistics */
	clear_stats();

	/* put all other cores to sleep bar master */
	rte_eal_mp_remote_launch(sleep_lcore, NULL, SKIP_MASTER);

	do_packet_forwarding();
	return 0;
}