[thirdparty/linux.git] / tools / testing / selftests / net / psock_fanout.c

// SPDX-License-Identifier: GPL-2.0-only
/*
 * Copyright 2013 Google Inc.
 * Author: Willem de Bruijn (willemb@google.com)
 *
 * A basic test of packet socket fanout behavior.
 *
 * Control:
 * - create fanout fails as expected with illegal flag combinations
 * - join   fanout fails as expected with diverging types or flags
 *
 * Datapath:
 *   Open a pair of packet sockets and a pair of INET sockets, send a known
 *   number of packets across the two INET sockets and count the number of
 *   packets enqueued onto the two packet sockets.
 *
 *   The test currently runs for
 *   - PACKET_FANOUT_HASH
 *   - PACKET_FANOUT_HASH with PACKET_FANOUT_FLAG_ROLLOVER
 *   - PACKET_FANOUT_LB
 *   - PACKET_FANOUT_CPU
 *   - PACKET_FANOUT_ROLLOVER
 *   - PACKET_FANOUT_CBPF
 *   - PACKET_FANOUT_EBPF
 *
 * Todo:
 * - functionality: PACKET_FANOUT_FLAG_DEFRAG
 */

#define _GNU_SOURCE		/* for sched_setaffinity */

#include <arpa/inet.h>
#include <errno.h>
#include <fcntl.h>
#include <linux/unistd.h>	/* for __NR_bpf */
#include <linux/filter.h>
#include <linux/bpf.h>
#include <linux/if_packet.h>
#include <net/if.h>
#include <net/ethernet.h>
#include <netinet/ip.h>
#include <netinet/udp.h>
#include <poll.h>
#include <sched.h>
#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/mman.h>
#include <sys/socket.h>
#include <sys/stat.h>
#include <sys/types.h>
#include <unistd.h>

#include "psock_lib.h"

#define RING_NUM_FRAMES			20

/* Open a socket in a given fanout mode.
 * @return -1 if mode is bad, a valid socket otherwise */
static int sock_fanout_open(uint16_t typeflags, uint16_t group_id)
{
	struct sockaddr_ll addr = {0};
	int fd, val;

	fd = socket(PF_PACKET, SOCK_RAW, 0);
	if (fd < 0) {
		perror("socket packet");
		exit(1);
	}

	pair_udp_setfilter(fd);

	addr.sll_family = AF_PACKET;
	addr.sll_protocol = htons(ETH_P_IP);
	addr.sll_ifindex = if_nametoindex("lo");
	if (addr.sll_ifindex == 0) {
		perror("if_nametoindex");
		exit(1);
	}
	if (bind(fd, (void *) &addr, sizeof(addr))) {
		perror("bind packet");
		exit(1);
	}

	val = (((int) typeflags) << 16) | group_id;
	if (setsockopt(fd, SOL_PACKET, PACKET_FANOUT, &val, sizeof(val))) {
		if (close(fd)) {
			perror("close packet");
			exit(1);
		}
		return -1;
	}

	return fd;
}

static void sock_fanout_set_cbpf(int fd)
{
	struct sock_filter bpf_filter[] = {
		BPF_STMT(BPF_LD+BPF_B+BPF_ABS, 80),	      /* ldb [80] */
		BPF_STMT(BPF_RET+BPF_A, 0),		      /* ret A */
	};
	struct sock_fprog bpf_prog;

	bpf_prog.filter = bpf_filter;
	bpf_prog.len = sizeof(bpf_filter) / sizeof(struct sock_filter);

	if (setsockopt(fd, SOL_PACKET, PACKET_FANOUT_DATA, &bpf_prog,
		       sizeof(bpf_prog))) {
		perror("fanout data cbpf");
		exit(1);
	}
}

static void sock_fanout_getopts(int fd, uint16_t *typeflags, uint16_t *group_id)
{
	int sockopt;
	socklen_t sockopt_len = sizeof(sockopt);

	if (getsockopt(fd, SOL_PACKET, PACKET_FANOUT,
		       &sockopt, &sockopt_len)) {
		perror("failed to getsockopt");
		exit(1);
	}
	*typeflags = sockopt >> 16;
	*group_id = sockopt & 0xfffff;
}

static void sock_fanout_set_ebpf(int fd)
{
	static char log_buf[65536];

	const int len_off = __builtin_offsetof(struct __sk_buff, len);
	struct bpf_insn prog[] = {
		{ BPF_ALU64 | BPF_MOV | BPF_X,   6, 1, 0, 0 },
		{ BPF_LDX   | BPF_W   | BPF_MEM, 0, 6, len_off, 0 },
		{ BPF_JMP   | BPF_JGE | BPF_K,   0, 0, 1, DATA_LEN },
		{ BPF_JMP   | BPF_JA  | BPF_K,   0, 0, 4, 0 },
		{ BPF_LD    | BPF_B   | BPF_ABS, 0, 0, 0, 0x50 },
		{ BPF_JMP   | BPF_JEQ | BPF_K,   0, 0, 2, DATA_CHAR },
		{ BPF_JMP   | BPF_JEQ | BPF_K,   0, 0, 1, DATA_CHAR_1 },
		{ BPF_ALU   | BPF_MOV | BPF_K,   0, 0, 0, 0 },
		{ BPF_JMP   | BPF_EXIT,          0, 0, 0, 0 }
	};
	union bpf_attr attr;
	int pfd;

	memset(&attr, 0, sizeof(attr));
	attr.prog_type = BPF_PROG_TYPE_SOCKET_FILTER;
	attr.insns = (unsigned long) prog;
	attr.insn_cnt = sizeof(prog) / sizeof(prog[0]);
	attr.license = (unsigned long) "GPL";
	attr.log_buf = (unsigned long) log_buf,
	attr.log_size = sizeof(log_buf),
	attr.log_level = 1,

	pfd = syscall(__NR_bpf, BPF_PROG_LOAD, &attr, sizeof(attr));
	if (pfd < 0) {
		perror("bpf");
		fprintf(stderr, "bpf verifier:\n%s\n", log_buf);
		exit(1);
	}

	if (setsockopt(fd, SOL_PACKET, PACKET_FANOUT_DATA, &pfd, sizeof(pfd))) {
		perror("fanout data ebpf");
		exit(1);
	}

	if (close(pfd)) {
		perror("close ebpf");
		exit(1);
	}
}

static char *sock_fanout_open_ring(int fd)
{
	struct tpacket_req req = {
		.tp_block_size = getpagesize(),
		.tp_frame_size = getpagesize(),
		.tp_block_nr   = RING_NUM_FRAMES,
		.tp_frame_nr   = RING_NUM_FRAMES,
	};
	char *ring;
	int val = TPACKET_V2;

	if (setsockopt(fd, SOL_PACKET, PACKET_VERSION, (void *) &val,
		       sizeof(val))) {
		perror("packetsock ring setsockopt version");
		exit(1);
	}
	if (setsockopt(fd, SOL_PACKET, PACKET_RX_RING, (void *) &req,
		       sizeof(req))) {
		perror("packetsock ring setsockopt");
		exit(1);
	}

	ring = mmap(0, req.tp_block_size * req.tp_block_nr,
		    PROT_READ | PROT_WRITE, MAP_SHARED, fd, 0);
	if (ring == MAP_FAILED) {
		perror("packetsock ring mmap");
		exit(1);
	}

	return ring;
}

static int sock_fanout_read_ring(int fd, void *ring)
{
	struct tpacket2_hdr *header = ring;
	int count = 0;

	while (count < RING_NUM_FRAMES && header->tp_status & TP_STATUS_USER) {
		count++;
		header = ring + (count * getpagesize());
	}

	return count;
}

static int sock_fanout_read(int fds[], char *rings[], const int expect[])
{
	int ret[2];

	ret[0] = sock_fanout_read_ring(fds[0], rings[0]);
	ret[1] = sock_fanout_read_ring(fds[1], rings[1]);

	fprintf(stderr, "info: count=%d,%d, expect=%d,%d\n",
			ret[0], ret[1], expect[0], expect[1]);

	if ((!(ret[0] == expect[0] && ret[1] == expect[1])) &&
	    (!(ret[0] == expect[1] && ret[1] == expect[0]))) {
		fprintf(stderr, "warning: incorrect queue lengths\n");
		return 1;
	}

	return 0;
}

/* Test illegal mode + flag combination */
static void test_control_single(void)
{
	fprintf(stderr, "test: control single socket\n");

	if (sock_fanout_open(PACKET_FANOUT_ROLLOVER |
			       PACKET_FANOUT_FLAG_ROLLOVER, 0) != -1) {
		fprintf(stderr, "ERROR: opened socket with dual rollover\n");
		exit(1);
	}
}

/* Test illegal group with different modes or flags */
static void test_control_group(void)
{
	int fds[2];

	fprintf(stderr, "test: control multiple sockets\n");

	fds[0] = sock_fanout_open(PACKET_FANOUT_HASH, 0);
	if (fds[0] == -1) {
		fprintf(stderr, "ERROR: failed to open HASH socket\n");
		exit(1);
	}
	if (sock_fanout_open(PACKET_FANOUT_HASH |
			       PACKET_FANOUT_FLAG_DEFRAG, 0) != -1) {
		fprintf(stderr, "ERROR: joined group with wrong flag defrag\n");
		exit(1);
	}
	if (sock_fanout_open(PACKET_FANOUT_HASH |
			       PACKET_FANOUT_FLAG_ROLLOVER, 0) != -1) {
		fprintf(stderr, "ERROR: joined group with wrong flag ro\n");
		exit(1);
	}
	if (sock_fanout_open(PACKET_FANOUT_CPU, 0) != -1) {
		fprintf(stderr, "ERROR: joined group with wrong mode\n");
		exit(1);
	}
	fds[1] = sock_fanout_open(PACKET_FANOUT_HASH, 0);
	if (fds[1] == -1) {
		fprintf(stderr, "ERROR: failed to join group\n");
		exit(1);
	}
	if (close(fds[1]) || close(fds[0])) {
		fprintf(stderr, "ERROR: closing sockets\n");
		exit(1);
	}
}

/* Test creating a unique fanout group ids */
static void test_unique_fanout_group_ids(void)
{
	int fds[3];
	uint16_t typeflags, first_group_id, second_group_id;

	fprintf(stderr, "test: unique ids\n");

	fds[0] = sock_fanout_open(PACKET_FANOUT_HASH |
				  PACKET_FANOUT_FLAG_UNIQUEID, 0);
	if (fds[0] == -1) {
		fprintf(stderr, "ERROR: failed to create a unique id group.\n");
		exit(1);
	}

	sock_fanout_getopts(fds[0], &typeflags, &first_group_id);
	if (typeflags != PACKET_FANOUT_HASH) {
		fprintf(stderr, "ERROR: unexpected typeflags %x\n", typeflags);
		exit(1);
	}

	if (sock_fanout_open(PACKET_FANOUT_CPU, first_group_id) != -1) {
		fprintf(stderr, "ERROR: joined group with wrong type.\n");
		exit(1);
	}

	fds[1] = sock_fanout_open(PACKET_FANOUT_HASH, first_group_id);
	if (fds[1] == -1) {
		fprintf(stderr,
			"ERROR: failed to join previously created group.\n");
		exit(1);
	}

	fds[2] = sock_fanout_open(PACKET_FANOUT_HASH |
				  PACKET_FANOUT_FLAG_UNIQUEID, 0);
	if (fds[2] == -1) {
		fprintf(stderr,
			"ERROR: failed to create a second unique id group.\n");
		exit(1);
	}

	sock_fanout_getopts(fds[2], &typeflags, &second_group_id);
	if (sock_fanout_open(PACKET_FANOUT_HASH | PACKET_FANOUT_FLAG_UNIQUEID,
			     second_group_id) != -1) {
		fprintf(stderr,
			"ERROR: specified a group id when requesting unique id\n");
		exit(1);
	}

	if (close(fds[0]) || close(fds[1]) || close(fds[2])) {
		fprintf(stderr, "ERROR: closing sockets\n");
		exit(1);
	}
}

static int test_datapath(uint16_t typeflags, int port_off,
			 const int expect1[], const int expect2[])
{
	const int expect0[] = { 0, 0 };
	char *rings[2];
	uint8_t type = typeflags & 0xFF;
	int fds[2], fds_udp[2][2], ret;

	fprintf(stderr, "\ntest: datapath 0x%hx ports %hu,%hu\n",
		typeflags, PORT_BASE, PORT_BASE + port_off);

	fds[0] = sock_fanout_open(typeflags, 0);
	fds[1] = sock_fanout_open(typeflags, 0);
	if (fds[0] == -1 || fds[1] == -1) {
		fprintf(stderr, "ERROR: failed open\n");
		exit(1);
	}
	if (type == PACKET_FANOUT_CBPF)
		sock_fanout_set_cbpf(fds[0]);
	else if (type == PACKET_FANOUT_EBPF)
		sock_fanout_set_ebpf(fds[0]);

	rings[0] = sock_fanout_open_ring(fds[0]);
	rings[1] = sock_fanout_open_ring(fds[1]);
	pair_udp_open(fds_udp[0], PORT_BASE);
	pair_udp_open(fds_udp[1], PORT_BASE + port_off);
	sock_fanout_read(fds, rings, expect0);

	/* Send data, but not enough to overflow a queue */
	pair_udp_send(fds_udp[0], 15);
	pair_udp_send_char(fds_udp[1], 5, DATA_CHAR_1);
	ret = sock_fanout_read(fds, rings, expect1);

	/* Send more data, overflow the queue */
	pair_udp_send_char(fds_udp[0], 15, DATA_CHAR_1);
	/* TODO: ensure consistent order between expect1 and expect2 */
	ret |= sock_fanout_read(fds, rings, expect2);

	if (munmap(rings[1], RING_NUM_FRAMES * getpagesize()) ||
	    munmap(rings[0], RING_NUM_FRAMES * getpagesize())) {
		fprintf(stderr, "close rings\n");
		exit(1);
	}
	if (close(fds_udp[1][1]) || close(fds_udp[1][0]) ||
	    close(fds_udp[0][1]) || close(fds_udp[0][0]) ||
	    close(fds[1]) || close(fds[0])) {
		fprintf(stderr, "close datapath\n");
		exit(1);
	}

	return ret;
}

static int set_cpuaffinity(int cpuid)
{
	cpu_set_t mask;

	CPU_ZERO(&mask);
	CPU_SET(cpuid, &mask);
	if (sched_setaffinity(0, sizeof(mask), &mask)) {
		if (errno != EINVAL) {
			fprintf(stderr, "setaffinity %d\n", cpuid);
			exit(1);
		}
		return 1;
	}

	return 0;
}

int main(int argc, char **argv)
{
	const int expect_hash[2][2]	= { { 15, 5 },  { 20, 5 } };
	const int expect_hash_rb[2][2]	= { { 15, 5 },  { 20, 15 } };
	const int expect_lb[2][2]	= { { 10, 10 }, { 18, 17 } };
	const int expect_rb[2][2]	= { { 15, 5 },  { 20, 15 } };
	const int expect_cpu0[2][2]	= { { 20, 0 },  { 20, 0 } };
	const int expect_cpu1[2][2]	= { { 0, 20 },  { 0, 20 } };
	const int expect_bpf[2][2]	= { { 15, 5 },  { 15, 20 } };
	const int expect_uniqueid[2][2] = { { 20, 20},  { 20, 20 } };
	int port_off = 2, tries = 20, ret;

	test_control_single();
	test_control_group();
	test_unique_fanout_group_ids();

	/* find a set of ports that do not collide onto the same socket */
	ret = test_datapath(PACKET_FANOUT_HASH, port_off,
			    expect_hash[0], expect_hash[1]);
	while (ret) {
		fprintf(stderr, "info: trying alternate ports (%d)\n", tries);
		ret = test_datapath(PACKET_FANOUT_HASH, ++port_off,
				    expect_hash[0], expect_hash[1]);
		if (!--tries) {
			fprintf(stderr, "too many collisions\n");
			return 1;
		}
	}

	ret |= test_datapath(PACKET_FANOUT_HASH | PACKET_FANOUT_FLAG_ROLLOVER,
			     port_off, expect_hash_rb[0], expect_hash_rb[1]);
	ret |= test_datapath(PACKET_FANOUT_LB,
			     port_off, expect_lb[0], expect_lb[1]);
	ret |= test_datapath(PACKET_FANOUT_ROLLOVER,
			     port_off, expect_rb[0], expect_rb[1]);

	ret |= test_datapath(PACKET_FANOUT_CBPF,
			     port_off, expect_bpf[0], expect_bpf[1]);
	ret |= test_datapath(PACKET_FANOUT_EBPF,
			     port_off, expect_bpf[0], expect_bpf[1]);

	set_cpuaffinity(0);
	ret |= test_datapath(PACKET_FANOUT_CPU, port_off,
			     expect_cpu0[0], expect_cpu0[1]);
	if (!set_cpuaffinity(1))
		/* TODO: test that choice alternates with previous */
		ret |= test_datapath(PACKET_FANOUT_CPU, port_off,
				     expect_cpu1[0], expect_cpu1[1]);

	ret |= test_datapath(PACKET_FANOUT_FLAG_UNIQUEID, port_off,
			     expect_uniqueid[0], expect_uniqueid[1]);

	if (ret)
		return 1;

	printf("OK. All tests passed\n");
	return 0;
}
Commit	Line	Data
fc44ef5a	1	// SPDX-License-Identifier: GPL-2.0-only
77f65ebd WB	2	/*
	3	* Copyright 2013 Google Inc.
	4	* Author: Willem de Bruijn (willemb@google.com)
	5	*
	6	* A basic test of packet socket fanout behavior.
	7	*
	8	* Control:
	9	* - create fanout fails as expected with illegal flag combinations
	10	* - join fanout fails as expected with diverging types or flags
	11	*
	12	* Datapath:
	13	* Open a pair of packet sockets and a pair of INET sockets, send a known
	14	* number of packets across the two INET sockets and count the number of
	15	* packets enqueued onto the two packet sockets.
	16	*
	17	* The test currently runs for
	18	* - PACKET_FANOUT_HASH
	19	* - PACKET_FANOUT_HASH with PACKET_FANOUT_FLAG_ROLLOVER
23a9072e WB	20	* - PACKET_FANOUT_LB
23a9072e WB	21	* - PACKET_FANOUT_CPU
77f65ebd	22	* - PACKET_FANOUT_ROLLOVER
95e22792	23	* - PACKET_FANOUT_CBPF
30da679e	24	* - PACKET_FANOUT_EBPF
77f65ebd WB	25	*
77f65ebd WB	26	* Todo:
77f65ebd	27	* - functionality: PACKET_FANOUT_FLAG_DEFRAG
77f65ebd WB	28	*/
77f65ebd WB	29
23a9072e WB	30	#define _GNU_SOURCE /* for sched_setaffinity */
23a9072e WB	31
77f65ebd WB	32	#include <arpa/inet.h>
77f65ebd WB	33	#include <errno.h>
23a9072e	34	#include <fcntl.h>
30da679e	35	#include <linux/unistd.h> /* for __NR_bpf */
77f65ebd	36	#include <linux/filter.h>
30da679e	37	#include <linux/bpf.h>
77f65ebd	38	#include <linux/if_packet.h>
cc30c93f	39	#include <net/if.h>
77f65ebd WB	40	#include <net/ethernet.h>
	41	#include <netinet/ip.h>
	42	#include <netinet/udp.h>
23a9072e WB	43	#include <poll.h>
23a9072e WB	44	#include <sched.h>
77f65ebd WB	45	#include <stdint.h>
	46	#include <stdio.h>
	47	#include <stdlib.h>
	48	#include <string.h>
23a9072e	49	#include <sys/mman.h>
77f65ebd WB	50	#include <sys/socket.h>
	51	#include <sys/stat.h>
	52	#include <sys/types.h>
	53	#include <unistd.h>
	54
23a95442	55	#include "psock_lib.h"
77f65ebd	56
23a95442	57	#define RING_NUM_FRAMES 20
77f65ebd WB	58
	59	/* Open a socket in a given fanout mode.
	60	* @return -1 if mode is bad, a valid socket otherwise */
28be04f5	61	static int sock_fanout_open(uint16_t typeflags, uint16_t group_id)
77f65ebd	62	{
cc30c93f	63	struct sockaddr_ll addr = {0};
77f65ebd WB	64	int fd, val;
77f65ebd WB	65
cc30c93f	66	fd = socket(PF_PACKET, SOCK_RAW, 0);
77f65ebd WB	67	if (fd < 0) {
	68	perror("socket packet");
	69	exit(1);
	70	}
	71
cc30c93f WB	72	pair_udp_setfilter(fd);
	73
	74	addr.sll_family = AF_PACKET;
	75	addr.sll_protocol = htons(ETH_P_IP);
	76	addr.sll_ifindex = if_nametoindex("lo");
	77	if (addr.sll_ifindex == 0) {
	78	perror("if_nametoindex");
	79	exit(1);
	80	}
	81	if (bind(fd, (void *) &addr, sizeof(addr))) {
	82	perror("bind packet");
	83	exit(1);
	84	}
	85
28be04f5	86	val = (((int) typeflags) << 16) \| group_id;
77f65ebd WB	87	if (setsockopt(fd, SOL_PACKET, PACKET_FANOUT, &val, sizeof(val))) {
	88	if (close(fd)) {
	89	perror("close packet");
	90	exit(1);
	91	}
	92	return -1;
	93	}
	94
77f65ebd WB	95	return fd;
	96	}
	97
c1f8d0f9 MM	98	static void sock_fanout_set_cbpf(int fd)
	99	{
	100	struct sock_filter bpf_filter[] = {
	101	BPF_STMT(BPF_LD+BPF_B+BPF_ABS, 80), /* ldb [80] */
	102	BPF_STMT(BPF_RET+BPF_A, 0), /* ret A */
	103	};
	104	struct sock_fprog bpf_prog;
	105
	106	bpf_prog.filter = bpf_filter;
	107	bpf_prog.len = sizeof(bpf_filter) / sizeof(struct sock_filter);
	108
	109	if (setsockopt(fd, SOL_PACKET, PACKET_FANOUT_DATA, &bpf_prog,
	110	sizeof(bpf_prog))) {
	111	perror("fanout data cbpf");
	112	exit(1);
	113	}
	114	}
	115
28be04f5 MM	116	static void sock_fanout_getopts(int fd, uint16_t typeflags, uint16_t group_id)
	117	{
	118	int sockopt;
	119	socklen_t sockopt_len = sizeof(sockopt);
	120
	121	if (getsockopt(fd, SOL_PACKET, PACKET_FANOUT,
	122	&sockopt, &sockopt_len)) {
	123	perror("failed to getsockopt");
	124	exit(1);
	125	}
	126	*typeflags = sockopt >> 16;
	127	*group_id = sockopt & 0xfffff;
	128	}
	129
30da679e WB	130	static void sock_fanout_set_ebpf(int fd)
30da679e WB	131	{
ddd00103 PB	132	static char log_buf[65536];
ddd00103 PB	133
30da679e WB	134	const int len_off = __builtin_offsetof(struct __sk_buff, len);
	135	struct bpf_insn prog[] = {
	136	{ BPF_ALU64 \| BPF_MOV \| BPF_X, 6, 1, 0, 0 },
	137	{ BPF_LDX \| BPF_W \| BPF_MEM, 0, 6, len_off, 0 },
	138	{ BPF_JMP \| BPF_JGE \| BPF_K, 0, 0, 1, DATA_LEN },
	139	{ BPF_JMP \| BPF_JA \| BPF_K, 0, 0, 4, 0 },
	140	{ BPF_LD \| BPF_B \| BPF_ABS, 0, 0, 0, 0x50 },
	141	{ BPF_JMP \| BPF_JEQ \| BPF_K, 0, 0, 2, DATA_CHAR },
	142	{ BPF_JMP \| BPF_JEQ \| BPF_K, 0, 0, 1, DATA_CHAR_1 },
	143	{ BPF_ALU \| BPF_MOV \| BPF_K, 0, 0, 0, 0 },
	144	{ BPF_JMP \| BPF_EXIT, 0, 0, 0, 0 }
	145	};
30da679e WB	146	union bpf_attr attr;
	147	int pfd;
	148
	149	memset(&attr, 0, sizeof(attr));
	150	attr.prog_type = BPF_PROG_TYPE_SOCKET_FILTER;
	151	attr.insns = (unsigned long) prog;
	152	attr.insn_cnt = sizeof(prog) / sizeof(prog[0]);
	153	attr.license = (unsigned long) "GPL";
	154	attr.log_buf = (unsigned long) log_buf,
	155	attr.log_size = sizeof(log_buf),
	156	attr.log_level = 1,
	157
	158	pfd = syscall(__NR_bpf, BPF_PROG_LOAD, &attr, sizeof(attr));
	159	if (pfd < 0) {
	160	perror("bpf");
	161	fprintf(stderr, "bpf verifier:\n%s\n", log_buf);
	162	exit(1);
	163	}
	164
	165	if (setsockopt(fd, SOL_PACKET, PACKET_FANOUT_DATA, &pfd, sizeof(pfd))) {
	166	perror("fanout data ebpf");
	167	exit(1);
	168	}
	169
	170	if (close(pfd)) {
	171	perror("close ebpf");
	172	exit(1);
	173	}
	174	}
	175
23a9072e	176	static char *sock_fanout_open_ring(int fd)
77f65ebd	177	{
23a9072e WB	178	struct tpacket_req req = {
	179	.tp_block_size = getpagesize(),
	180	.tp_frame_size = getpagesize(),
	181	.tp_block_nr = RING_NUM_FRAMES,
	182	.tp_frame_nr = RING_NUM_FRAMES,
	183	};
	184	char *ring;
98e821a2	185	int val = TPACKET_V2;
77f65ebd	186
98e821a2 WB	187	if (setsockopt(fd, SOL_PACKET, PACKET_VERSION, (void *) &val,
	188	sizeof(val))) {
	189	perror("packetsock ring setsockopt version");
	190	exit(1);
	191	}
23a9072e WB	192	if (setsockopt(fd, SOL_PACKET, PACKET_RX_RING, (void *) &req,
	193	sizeof(req))) {
	194	perror("packetsock ring setsockopt");
77f65ebd WB	195	exit(1);
77f65ebd WB	196	}
23a9072e WB	197
	198	ring = mmap(0, req.tp_block_size * req.tp_block_nr,
	199	PROT_READ \| PROT_WRITE, MAP_SHARED, fd, 0);
95e22792 WB	200	if (ring == MAP_FAILED) {
95e22792 WB	201	perror("packetsock ring mmap");
77f65ebd WB	202	exit(1);
77f65ebd WB	203	}
23a9072e WB	204
	205	return ring;
	206	}
	207
	208	static int sock_fanout_read_ring(int fd, void *ring)
	209	{
98e821a2	210	struct tpacket2_hdr *header = ring;
23a9072e WB	211	int count = 0;
23a9072e WB	212
fbf8e721	213	while (count < RING_NUM_FRAMES && header->tp_status & TP_STATUS_USER) {
23a9072e WB	214	count++;
	215	header = ring + (count * getpagesize());
	216	}
	217
	218	return count;
	219	}
	220
	221	static int sock_fanout_read(int fds[], char *rings[], const int expect[])
	222	{
	223	int ret[2];
	224
	225	ret[0] = sock_fanout_read_ring(fds[0], rings[0]);
	226	ret[1] = sock_fanout_read_ring(fds[1], rings[1]);
77f65ebd WB	227
	228	fprintf(stderr, "info: count=%d,%d, expect=%d,%d\n",
	229	ret[0], ret[1], expect[0], expect[1]);
	230
	231	if ((!(ret[0] == expect[0] && ret[1] == expect[1])) &&
	232	(!(ret[0] == expect[1] && ret[1] == expect[0]))) {
cc30c93f	233	fprintf(stderr, "warning: incorrect queue lengths\n");
23a9072e	234	return 1;
77f65ebd	235	}
23a9072e WB	236
23a9072e WB	237	return 0;
77f65ebd WB	238	}
	239
	240	/* Test illegal mode + flag combination */
	241	static void test_control_single(void)
	242	{
	243	fprintf(stderr, "test: control single socket\n");
	244
	245	if (sock_fanout_open(PACKET_FANOUT_ROLLOVER \|
28be04f5	246	PACKET_FANOUT_FLAG_ROLLOVER, 0) != -1) {
77f65ebd WB	247	fprintf(stderr, "ERROR: opened socket with dual rollover\n");
	248	exit(1);
	249	}
	250	}
	251
	252	/* Test illegal group with different modes or flags */
	253	static void test_control_group(void)
	254	{
	255	int fds[2];
	256
	257	fprintf(stderr, "test: control multiple sockets\n");
	258
28be04f5	259	fds[0] = sock_fanout_open(PACKET_FANOUT_HASH, 0);
77f65ebd WB	260	if (fds[0] == -1) {
	261	fprintf(stderr, "ERROR: failed to open HASH socket\n");
	262	exit(1);
	263	}
	264	if (sock_fanout_open(PACKET_FANOUT_HASH \|
28be04f5	265	PACKET_FANOUT_FLAG_DEFRAG, 0) != -1) {
77f65ebd WB	266	fprintf(stderr, "ERROR: joined group with wrong flag defrag\n");
	267	exit(1);
	268	}
	269	if (sock_fanout_open(PACKET_FANOUT_HASH \|
28be04f5	270	PACKET_FANOUT_FLAG_ROLLOVER, 0) != -1) {
77f65ebd WB	271	fprintf(stderr, "ERROR: joined group with wrong flag ro\n");
	272	exit(1);
	273	}
28be04f5	274	if (sock_fanout_open(PACKET_FANOUT_CPU, 0) != -1) {
77f65ebd WB	275	fprintf(stderr, "ERROR: joined group with wrong mode\n");
	276	exit(1);
	277	}
28be04f5	278	fds[1] = sock_fanout_open(PACKET_FANOUT_HASH, 0);
77f65ebd WB	279	if (fds[1] == -1) {
	280	fprintf(stderr, "ERROR: failed to join group\n");
	281	exit(1);
	282	}
	283	if (close(fds[1]) \|\| close(fds[0])) {
	284	fprintf(stderr, "ERROR: closing sockets\n");
	285	exit(1);
	286	}
	287	}
	288
28be04f5 MM	289	/* Test creating a unique fanout group ids */
	290	static void test_unique_fanout_group_ids(void)
	291	{
	292	int fds[3];
	293	uint16_t typeflags, first_group_id, second_group_id;
	294
	295	fprintf(stderr, "test: unique ids\n");
	296
	297	fds[0] = sock_fanout_open(PACKET_FANOUT_HASH \|
	298	PACKET_FANOUT_FLAG_UNIQUEID, 0);
	299	if (fds[0] == -1) {
	300	fprintf(stderr, "ERROR: failed to create a unique id group.\n");
	301	exit(1);
	302	}
	303
	304	sock_fanout_getopts(fds[0], &typeflags, &first_group_id);
	305	if (typeflags != PACKET_FANOUT_HASH) {
	306	fprintf(stderr, "ERROR: unexpected typeflags %x\n", typeflags);
	307	exit(1);
	308	}
	309
472ecf08	310	if (sock_fanout_open(PACKET_FANOUT_CPU, first_group_id) != -1) {
28be04f5 MM	311	fprintf(stderr, "ERROR: joined group with wrong type.\n");
	312	exit(1);
	313	}
	314
	315	fds[1] = sock_fanout_open(PACKET_FANOUT_HASH, first_group_id);
	316	if (fds[1] == -1) {
	317	fprintf(stderr,
	318	"ERROR: failed to join previously created group.\n");
	319	exit(1);
	320	}
	321
	322	fds[2] = sock_fanout_open(PACKET_FANOUT_HASH \|
	323	PACKET_FANOUT_FLAG_UNIQUEID, 0);
	324	if (fds[2] == -1) {
	325	fprintf(stderr,
	326	"ERROR: failed to create a second unique id group.\n");
	327	exit(1);
	328	}
	329
	330	sock_fanout_getopts(fds[2], &typeflags, &second_group_id);
	331	if (sock_fanout_open(PACKET_FANOUT_HASH \| PACKET_FANOUT_FLAG_UNIQUEID,
	332	second_group_id) != -1) {
	333	fprintf(stderr,
	334	"ERROR: specified a group id when requesting unique id\n");
	335	exit(1);
	336	}
	337
	338	if (close(fds[0]) \|\| close(fds[1]) \|\| close(fds[2])) {
	339	fprintf(stderr, "ERROR: closing sockets\n");
	340	exit(1);
	341	}
	342	}
	343
23a9072e WB	344	static int test_datapath(uint16_t typeflags, int port_off,
23a9072e WB	345	const int expect1[], const int expect2[])
77f65ebd WB	346	{
77f65ebd WB	347	const int expect0[] = { 0, 0 };
23a9072e	348	char *rings[2];
95e22792	349	uint8_t type = typeflags & 0xFF;
23a9072e	350	int fds[2], fds_udp[2][2], ret;
77f65ebd	351
cc30c93f WB	352	fprintf(stderr, "\ntest: datapath 0x%hx ports %hu,%hu\n",
cc30c93f WB	353	typeflags, PORT_BASE, PORT_BASE + port_off);
77f65ebd	354
28be04f5 MM	355	fds[0] = sock_fanout_open(typeflags, 0);
28be04f5 MM	356	fds[1] = sock_fanout_open(typeflags, 0);
77f65ebd WB	357	if (fds[0] == -1 \|\| fds[1] == -1) {
	358	fprintf(stderr, "ERROR: failed open\n");
	359	exit(1);
	360	}
95e22792	361	if (type == PACKET_FANOUT_CBPF)
c1f8d0f9	362	sock_fanout_set_cbpf(fds[0]);
30da679e WB	363	else if (type == PACKET_FANOUT_EBPF)
30da679e WB	364	sock_fanout_set_ebpf(fds[0]);
95e22792	365
23a9072e WB	366	rings[0] = sock_fanout_open_ring(fds[0]);
	367	rings[1] = sock_fanout_open_ring(fds[1]);
	368	pair_udp_open(fds_udp[0], PORT_BASE);
	369	pair_udp_open(fds_udp[1], PORT_BASE + port_off);
	370	sock_fanout_read(fds, rings, expect0);
77f65ebd WB	371
	372	/* Send data, but not enough to overflow a queue */
	373	pair_udp_send(fds_udp[0], 15);
95e22792	374	pair_udp_send_char(fds_udp[1], 5, DATA_CHAR_1);
23a9072e	375	ret = sock_fanout_read(fds, rings, expect1);
77f65ebd WB	376
77f65ebd WB	377	/* Send more data, overflow the queue */
95e22792	378	pair_udp_send_char(fds_udp[0], 15, DATA_CHAR_1);
77f65ebd	379	/* TODO: ensure consistent order between expect1 and expect2 */
23a9072e	380	ret \|= sock_fanout_read(fds, rings, expect2);
77f65ebd	381
23a9072e WB	382	if (munmap(rings[1], RING_NUM_FRAMES * getpagesize()) \|\|
	383	munmap(rings[0], RING_NUM_FRAMES * getpagesize())) {
	384	fprintf(stderr, "close rings\n");
	385	exit(1);
	386	}
77f65ebd WB	387	if (close(fds_udp[1][1]) \|\| close(fds_udp[1][0]) \|\|
	388	close(fds_udp[0][1]) \|\| close(fds_udp[0][0]) \|\|
	389	close(fds[1]) \|\| close(fds[0])) {
	390	fprintf(stderr, "close datapath\n");
	391	exit(1);
	392	}
23a9072e WB	393
	394	return ret;
	395	}
	396
	397	static int set_cpuaffinity(int cpuid)
	398	{
	399	cpu_set_t mask;
	400
	401	CPU_ZERO(&mask);
	402	CPU_SET(cpuid, &mask);
	403	if (sched_setaffinity(0, sizeof(mask), &mask)) {
	404	if (errno != EINVAL) {
	405	fprintf(stderr, "setaffinity %d\n", cpuid);
	406	exit(1);
	407	}
	408	return 1;
	409	}
	410
	411	return 0;
77f65ebd WB	412	}
	413
	414	int main(int argc, char **argv)
	415	{
23a9072e WB	416	const int expect_hash[2][2] = { { 15, 5 }, { 20, 5 } };
	417	const int expect_hash_rb[2][2] = { { 15, 5 }, { 20, 15 } };
	418	const int expect_lb[2][2] = { { 10, 10 }, { 18, 17 } };
a2ad5d2a	419	const int expect_rb[2][2] = { { 15, 5 }, { 20, 15 } };
23a9072e WB	420	const int expect_cpu0[2][2] = { { 20, 0 }, { 20, 0 } };
23a9072e WB	421	const int expect_cpu1[2][2] = { { 0, 20 }, { 0, 20 } };
95e22792	422	const int expect_bpf[2][2] = { { 15, 5 }, { 15, 20 } };
28be04f5	423	const int expect_uniqueid[2][2] = { { 20, 20}, { 20, 20 } };
cc30c93f	424	int port_off = 2, tries = 20, ret;
77f65ebd WB	425
	426	test_control_single();
	427	test_control_group();
28be04f5	428	test_unique_fanout_group_ids();
77f65ebd	429
23a9072e WB	430	/* find a set of ports that do not collide onto the same socket */
	431	ret = test_datapath(PACKET_FANOUT_HASH, port_off,
	432	expect_hash[0], expect_hash[1]);
cc30c93f	433	while (ret) {
23a9072e WB	434	fprintf(stderr, "info: trying alternate ports (%d)\n", tries);
	435	ret = test_datapath(PACKET_FANOUT_HASH, ++port_off,
	436	expect_hash[0], expect_hash[1]);
cc30c93f WB	437	if (!--tries) {
	438	fprintf(stderr, "too many collisions\n");
	439	return 1;
	440	}
23a9072e WB	441	}
	442
	443	ret \|= test_datapath(PACKET_FANOUT_HASH \| PACKET_FANOUT_FLAG_ROLLOVER,
	444	port_off, expect_hash_rb[0], expect_hash_rb[1]);
	445	ret \|= test_datapath(PACKET_FANOUT_LB,
	446	port_off, expect_lb[0], expect_lb[1]);
	447	ret \|= test_datapath(PACKET_FANOUT_ROLLOVER,
	448	port_off, expect_rb[0], expect_rb[1]);
30da679e	449
95e22792 WB	450	ret \|= test_datapath(PACKET_FANOUT_CBPF,
95e22792 WB	451	port_off, expect_bpf[0], expect_bpf[1]);
30da679e WB	452	ret \|= test_datapath(PACKET_FANOUT_EBPF,
30da679e WB	453	port_off, expect_bpf[0], expect_bpf[1]);
23a9072e WB	454
	455	set_cpuaffinity(0);
	456	ret \|= test_datapath(PACKET_FANOUT_CPU, port_off,
	457	expect_cpu0[0], expect_cpu0[1]);
	458	if (!set_cpuaffinity(1))
	459	/* TODO: test that choice alternates with previous */
	460	ret \|= test_datapath(PACKET_FANOUT_CPU, port_off,
	461	expect_cpu1[0], expect_cpu1[1]);
	462
28be04f5 MM	463	ret \|= test_datapath(PACKET_FANOUT_FLAG_UNIQUEID, port_off,
	464	expect_uniqueid[0], expect_uniqueid[1]);
	465
23a9072e WB	466	if (ret)
23a9072e WB	467	return 1;
77f65ebd WB	468
	469	printf("OK. All tests passed\n");
	470	return 0;
	471	}