sysctl_net_ipv4.c 36.8 KB
Newer Older
1
// SPDX-License-Identifier: GPL-2.0
L
Linus Torvalds 已提交
2 3 4 5 6 7 8 9
/*
 * sysctl_net_ipv4.c: sysctl interface to net IPV4 subsystem.
 *
 * Begun April 1, 1996, Mike Shaver.
 * Added /proc/sys/net/ipv4 directory entry (empty =) ). [MS]
 */

#include <linux/sysctl.h>
10
#include <linux/seqlock.h>
11
#include <linux/init.h>
12
#include <linux/slab.h>
13
#include <net/icmp.h>
L
Linus Torvalds 已提交
14
#include <net/ip.h>
15
#include <net/ip_fib.h>
L
Linus Torvalds 已提交
16
#include <net/tcp.h>
H
Hideo Aoki 已提交
17
#include <net/udp.h>
P
Paul Moore 已提交
18
#include <net/cipso_ipv4.h>
19
#include <net/ping.h>
20
#include <net/protocol.h>
21
#include <net/netevent.h>
L
Linus Torvalds 已提交
22

23
static int two = 2;
24
static int three __maybe_unused = 3;
N
Nandita Dukkipati 已提交
25
static int four = 4;
26
static int thousand = 1000;
27
static int tcp_retr1_max = 255;
L
Linus Torvalds 已提交
28 29
static int ip_local_port_range_min[] = { 1, 1 };
static int ip_local_port_range_max[] = { 65535, 65535 };
30 31
static int tcp_adv_win_scale_min = -31;
static int tcp_adv_win_scale_max = 31;
E
Eric Dumazet 已提交
32 33
static int tcp_min_snd_mss_min = TCP_MIN_SND_MSS;
static int tcp_min_snd_mss_max = 65535;
34 35
static int ip_privileged_port_min;
static int ip_privileged_port_max = 65535;
E
Eric Dumazet 已提交
36 37
static int ip_ttl_min = 1;
static int ip_ttl_max = 255;
38 39
static int tcp_syn_retries_min = 1;
static int tcp_syn_retries_max = MAX_TCP_SYNCNT;
40 41
static int ip_ping_group_range_min[] = { 0, 0 };
static int ip_ping_group_range_max[] = { GID_T_MAX, GID_T_MAX };
42
static u32 u32_max_div_HZ = UINT_MAX / HZ;
43
static int one_day_secs = 24 * 3600;
44 45
static u32 fib_multipath_hash_fields_all_mask __maybe_unused =
	FIB_MULTIPATH_HASH_FIELD_ALL_MASK;
L
Linus Torvalds 已提交
46

47 48 49
/* obsolete */
static int sysctl_tcp_low_latency __read_mostly;

50
/* Update system visible IP port range */
51
static void set_local_port_range(struct net *net, int range[2])
52
{
53 54
	bool same_parity = !((range[0] ^ range[1]) & 1);

55
	write_seqlock_bh(&net->ipv4.ip_local_ports.lock);
56 57 58 59
	if (same_parity && !net->ipv4.ip_local_ports.warned) {
		net->ipv4.ip_local_ports.warned = true;
		pr_err_ratelimited("ip_local_port_range: prefer different parity for start/end values.\n");
	}
60 61
	net->ipv4.ip_local_ports.range[0] = range[0];
	net->ipv4.ip_local_ports.range[1] = range[1];
62
	write_sequnlock_bh(&net->ipv4.ip_local_ports.lock);
63 64 65
}

/* Validate changes from /proc interface. */
66
static int ipv4_local_port_range(struct ctl_table *table, int write,
67
				 void *buffer, size_t *lenp, loff_t *ppos)
68
{
69
	struct net *net =
70
		container_of(table->data, struct net, ipv4.ip_local_ports.range);
71
	int ret;
E
Eric Dumazet 已提交
72
	int range[2];
73
	struct ctl_table tmp = {
74 75 76 77 78 79 80
		.data = &range,
		.maxlen = sizeof(range),
		.mode = table->mode,
		.extra1 = &ip_local_port_range_min,
		.extra2 = &ip_local_port_range_max,
	};

81 82
	inet_get_local_port_range(net, &range[0], &range[1]);

83
	ret = proc_dointvec_minmax(&tmp, write, buffer, lenp, ppos);
84 85

	if (write && ret == 0) {
86 87 88 89 90 91
		/* Ensure that the upper limit is not smaller than the lower,
		 * and that the lower does not encroach upon the privileged
		 * port limit.
		 */
		if ((range[1] < range[0]) ||
		    (range[0] < net->ipv4.sysctl_ip_prot_sock))
92 93
			ret = -EINVAL;
		else
94
			set_local_port_range(net, range);
95 96 97 98 99
	}

	return ret;
}

100 101
/* Validate changes from /proc interface. */
static int ipv4_privileged_ports(struct ctl_table *table, int write,
102
				void *buffer, size_t *lenp, loff_t *ppos)
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
{
	struct net *net = container_of(table->data, struct net,
	    ipv4.sysctl_ip_prot_sock);
	int ret;
	int pports;
	int range[2];
	struct ctl_table tmp = {
		.data = &pports,
		.maxlen = sizeof(pports),
		.mode = table->mode,
		.extra1 = &ip_privileged_port_min,
		.extra2 = &ip_privileged_port_max,
	};

	pports = net->ipv4.sysctl_ip_prot_sock;

	ret = proc_dointvec_minmax(&tmp, write, buffer, lenp, ppos);

	if (write && ret == 0) {
		inet_get_local_port_range(net, &range[0], &range[1]);
		/* Ensure that the local port range doesn't overlap with the
		 * privileged port range.
		 */
		if (range[0] < pports)
			ret = -EINVAL;
		else
			net->ipv4.sysctl_ip_prot_sock = pports;
	}

	return ret;
}
134

135
static void inet_get_ping_group_range_table(struct ctl_table *table, kgid_t *low, kgid_t *high)
136
{
137
	kgid_t *data = table->data;
138
	struct net *net =
139
		container_of(table->data, struct net, ipv4.ping_group_range.range);
140
	unsigned int seq;
141
	do {
142
		seq = read_seqbegin(&net->ipv4.ping_group_range.lock);
143 144 145

		*low = data[0];
		*high = data[1];
146
	} while (read_seqretry(&net->ipv4.ping_group_range.lock, seq));
147 148 149
}

/* Update system visible IP port range */
150
static void set_ping_group_range(struct ctl_table *table, kgid_t low, kgid_t high)
151
{
152
	kgid_t *data = table->data;
153
	struct net *net =
154
		container_of(table->data, struct net, ipv4.ping_group_range.range);
155
	write_seqlock(&net->ipv4.ping_group_range.lock);
156 157
	data[0] = low;
	data[1] = high;
158
	write_sequnlock(&net->ipv4.ping_group_range.lock);
159 160 161
}

/* Validate changes from /proc interface. */
162
static int ipv4_ping_group_range(struct ctl_table *table, int write,
163
				 void *buffer, size_t *lenp, loff_t *ppos)
164
{
165
	struct user_namespace *user_ns = current_user_ns();
166
	int ret;
167 168
	gid_t urange[2];
	kgid_t low, high;
169
	struct ctl_table tmp = {
170 171
		.data = &urange,
		.maxlen = sizeof(urange),
172 173 174 175 176
		.mode = table->mode,
		.extra1 = &ip_ping_group_range_min,
		.extra2 = &ip_ping_group_range_max,
	};

177 178 179
	inet_get_ping_group_range_table(table, &low, &high);
	urange[0] = from_kgid_munged(user_ns, low);
	urange[1] = from_kgid_munged(user_ns, high);
180 181
	ret = proc_dointvec_minmax(&tmp, write, buffer, lenp, ppos);

182 183 184
	if (write && ret == 0) {
		low = make_kgid(user_ns, urange[0]);
		high = make_kgid(user_ns, urange[1]);
185 186 187
		if (!gid_valid(low) || !gid_valid(high))
			return -EINVAL;
		if (urange[1] < urange[0] || gid_lt(high, low)) {
188 189 190 191 192
			low = make_kgid(&init_user_ns, 1);
			high = make_kgid(&init_user_ns, 0);
		}
		set_ping_group_range(table, low, high);
	}
193 194 195 196

	return ret;
}

197
static int ipv4_fwd_update_priority(struct ctl_table *table, int write,
198
				    void *buffer, size_t *lenp, loff_t *ppos)
199 200 201 202 203 204
{
	struct net *net;
	int ret;

	net = container_of(table->data, struct net,
			   ipv4.sysctl_ip_fwd_update_priority);
205
	ret = proc_dou8vec_minmax(table, write, buffer, lenp, ppos);
206 207 208 209 210 211 212
	if (write && ret == 0)
		call_netevent_notifiers(NETEVENT_IPV4_FWD_UPDATE_PRIORITY_UPDATE,
					net);

	return ret;
}

213
static int proc_tcp_congestion_control(struct ctl_table *ctl, int write,
214
				       void *buffer, size_t *lenp, loff_t *ppos)
215
{
216 217
	struct net *net = container_of(ctl->data, struct net,
				       ipv4.tcp_congestion_control);
218
	char val[TCP_CA_NAME_MAX];
219
	struct ctl_table tbl = {
220 221 222 223 224
		.data = val,
		.maxlen = TCP_CA_NAME_MAX,
	};
	int ret;

225
	tcp_get_default_congestion_control(net, val);
226

227
	ret = proc_dostring(&tbl, write, buffer, lenp, ppos);
228
	if (write && ret == 0)
229
		ret = tcp_set_default_congestion_control(net, val);
230 231 232
	return ret;
}

233
static int proc_tcp_available_congestion_control(struct ctl_table *ctl,
234 235
						 int write, void *buffer,
						 size_t *lenp, loff_t *ppos)
236
{
237
	struct ctl_table tbl = { .maxlen = TCP_CA_BUF_MAX, };
238 239 240 241 242 243
	int ret;

	tbl.data = kmalloc(tbl.maxlen, GFP_USER);
	if (!tbl.data)
		return -ENOMEM;
	tcp_get_available_congestion_control(tbl.data, TCP_CA_BUF_MAX);
244
	ret = proc_dostring(&tbl, write, buffer, lenp, ppos);
245 246 247 248
	kfree(tbl.data);
	return ret;
}

249
static int proc_allowed_congestion_control(struct ctl_table *ctl,
250 251
					   int write, void *buffer,
					   size_t *lenp, loff_t *ppos)
252
{
253
	struct ctl_table tbl = { .maxlen = TCP_CA_BUF_MAX };
254 255 256 257 258 259 260
	int ret;

	tbl.data = kmalloc(tbl.maxlen, GFP_USER);
	if (!tbl.data)
		return -ENOMEM;

	tcp_get_allowed_congestion_control(tbl.data, tbl.maxlen);
261
	ret = proc_dostring(&tbl, write, buffer, lenp, ppos);
262 263 264 265 266 267
	if (write && ret == 0)
		ret = tcp_set_allowed_congestion_control(tbl.data);
	kfree(tbl.data);
	return ret;
}

268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285
static int sscanf_key(char *buf, __le32 *key)
{
	u32 user_key[4];
	int i, ret = 0;

	if (sscanf(buf, "%x-%x-%x-%x", user_key, user_key + 1,
		   user_key + 2, user_key + 3) != 4) {
		ret = -EINVAL;
	} else {
		for (i = 0; i < ARRAY_SIZE(user_key); i++)
			key[i] = cpu_to_le32(user_key[i]);
	}
	pr_debug("proc TFO key set 0x%x-%x-%x-%x <- 0x%s: %u\n",
		 user_key[0], user_key[1], user_key[2], user_key[3], buf, ret);

	return ret;
}

286
static int proc_tcp_fastopen_key(struct ctl_table *table, int write,
287
				 void *buffer, size_t *lenp, loff_t *ppos)
288
{
289 290
	struct net *net = container_of(table->data, struct net,
	    ipv4.sysctl_tcp_fastopen);
291 292 293 294 295 296
	/* maxlen to print the list of keys in hex (*2), with dashes
	 * separating doublewords and a comma in between keys.
	 */
	struct ctl_table tbl = { .maxlen = ((TCP_FASTOPEN_KEY_LENGTH *
					    2 * TCP_FASTOPEN_KEY_MAX) +
					    (TCP_FASTOPEN_KEY_MAX * 5)) };
297 298
	u32 user_key[TCP_FASTOPEN_KEY_BUF_LENGTH / sizeof(u32)];
	__le32 key[TCP_FASTOPEN_KEY_BUF_LENGTH / sizeof(__le32)];
299
	char *backup_data;
300
	int ret, i = 0, off = 0, n_keys;
301 302 303 304 305

	tbl.data = kmalloc(tbl.maxlen, GFP_KERNEL);
	if (!tbl.data)
		return -ENOMEM;

306
	n_keys = tcp_fastopen_get_cipher(net, NULL, (u64 *)key);
307 308 309 310 311 312
	if (!n_keys) {
		memset(&key[0], 0, TCP_FASTOPEN_KEY_LENGTH);
		n_keys = 1;
	}

	for (i = 0; i < n_keys * 4; i++)
313 314
		user_key[i] = le32_to_cpu(key[i]);

315 316 317 318 319 320 321
	for (i = 0; i < n_keys; i++) {
		off += snprintf(tbl.data + off, tbl.maxlen - off,
				"%08x-%08x-%08x-%08x",
				user_key[i * 4],
				user_key[i * 4 + 1],
				user_key[i * 4 + 2],
				user_key[i * 4 + 3]);
322 323 324 325

		if (WARN_ON_ONCE(off >= tbl.maxlen - 1))
			break;

326 327 328 329
		if (i + 1 < n_keys)
			off += snprintf(tbl.data + off, tbl.maxlen - off, ",");
	}

330 331 332
	ret = proc_dostring(&tbl, write, buffer, lenp, ppos);

	if (write && ret == 0) {
333 334 335 336 337 338
		backup_data = strchr(tbl.data, ',');
		if (backup_data) {
			*backup_data = '\0';
			backup_data++;
		}
		if (sscanf_key(tbl.data, key)) {
339 340 341
			ret = -EINVAL;
			goto bad_key;
		}
342 343 344 345 346 347 348
		if (backup_data) {
			if (sscanf_key(backup_data, key + 4)) {
				ret = -EINVAL;
				goto bad_key;
			}
		}
		tcp_fastopen_reset_cipher(net, NULL, key,
349
					  backup_data ? key + 4 : NULL);
350 351 352 353 354 355 356
	}

bad_key:
	kfree(tbl.data);
	return ret;
}

357 358 359 360 361 362 363
static void proc_configure_early_demux(int enabled, int protocol)
{
	struct net_protocol *ipprot;
#if IS_ENABLED(CONFIG_IPV6)
	struct inet6_protocol *ip6prot;
#endif

364 365
	rcu_read_lock();

366 367 368 369 370 371 372 373 374 375 376
	ipprot = rcu_dereference(inet_protos[protocol]);
	if (ipprot)
		ipprot->early_demux = enabled ? ipprot->early_demux_handler :
						NULL;

#if IS_ENABLED(CONFIG_IPV6)
	ip6prot = rcu_dereference(inet6_protos[protocol]);
	if (ip6prot)
		ip6prot->early_demux = enabled ? ip6prot->early_demux_handler :
						 NULL;
#endif
377
	rcu_read_unlock();
378 379 380
}

static int proc_tcp_early_demux(struct ctl_table *table, int write,
381
				void *buffer, size_t *lenp, loff_t *ppos)
382 383 384
{
	int ret = 0;

385
	ret = proc_dou8vec_minmax(table, write, buffer, lenp, ppos);
386 387 388 389 390 391 392 393 394 395 396

	if (write && !ret) {
		int enabled = init_net.ipv4.sysctl_tcp_early_demux;

		proc_configure_early_demux(enabled, IPPROTO_TCP);
	}

	return ret;
}

static int proc_udp_early_demux(struct ctl_table *table, int write,
397
				void *buffer, size_t *lenp, loff_t *ppos)
398 399 400
{
	int ret = 0;

401
	ret = proc_dou8vec_minmax(table, write, buffer, lenp, ppos);
402 403 404 405 406 407 408 409 410 411

	if (write && !ret) {
		int enabled = init_net.ipv4.sysctl_udp_early_demux;

		proc_configure_early_demux(enabled, IPPROTO_UDP);
	}

	return ret;
}

412
static int proc_tfo_blackhole_detect_timeout(struct ctl_table *table,
413
					     int write, void *buffer,
414 415
					     size_t *lenp, loff_t *ppos)
{
416 417
	struct net *net = container_of(table->data, struct net,
	    ipv4.sysctl_tcp_fastopen_blackhole_timeout);
418 419 420 421
	int ret;

	ret = proc_dointvec_minmax(table, write, buffer, lenp, ppos);
	if (write && ret == 0)
422
		atomic_set(&net->ipv4.tfo_active_disable_times, 0);
D
Dave Watson 已提交
423 424 425 426 427

	return ret;
}

static int proc_tcp_available_ulp(struct ctl_table *ctl,
428
				  int write, void *buffer, size_t *lenp,
D
Dave Watson 已提交
429 430 431 432 433 434 435 436 437 438 439 440
				  loff_t *ppos)
{
	struct ctl_table tbl = { .maxlen = TCP_ULP_BUF_MAX, };
	int ret;

	tbl.data = kmalloc(tbl.maxlen, GFP_USER);
	if (!tbl.data)
		return -ENOMEM;
	tcp_get_available_ulp(tbl.data, TCP_ULP_BUF_MAX);
	ret = proc_dostring(&tbl, write, buffer, lenp, ppos);
	kfree(tbl.data);

441 442 443
	return ret;
}

444 445
#ifdef CONFIG_IP_ROUTE_MULTIPATH
static int proc_fib_multipath_hash_policy(struct ctl_table *table, int write,
446
					  void *buffer, size_t *lenp,
447 448 449 450 451 452
					  loff_t *ppos)
{
	struct net *net = container_of(table->data, struct net,
	    ipv4.sysctl_fib_multipath_hash_policy);
	int ret;

453
	ret = proc_dou8vec_minmax(table, write, buffer, lenp, ppos);
454
	if (write && ret == 0)
455
		call_netevent_notifiers(NETEVENT_IPV4_MPATH_HASH_UPDATE, net);
456 457 458

	return ret;
}
459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474

static int proc_fib_multipath_hash_fields(struct ctl_table *table, int write,
					  void *buffer, size_t *lenp,
					  loff_t *ppos)
{
	struct net *net;
	int ret;

	net = container_of(table->data, struct net,
			   ipv4.sysctl_fib_multipath_hash_fields);
	ret = proc_douintvec_minmax(table, write, buffer, lenp, ppos);
	if (write && ret == 0)
		call_netevent_notifiers(NETEVENT_IPV4_MPATH_HASH_UPDATE, net);

	return ret;
}
475 476
#endif

477
static struct ctl_table ipv4_table[] = {
L
Linus Torvalds 已提交
478 479 480 481 482
	{
		.procname	= "tcp_max_orphans",
		.data		= &sysctl_tcp_max_orphans,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
483
		.proc_handler	= proc_dointvec
L
Linus Torvalds 已提交
484 485 486 487 488 489
	},
	{
		.procname	= "inet_peer_threshold",
		.data		= &inet_peer_threshold,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
490
		.proc_handler	= proc_dointvec
L
Linus Torvalds 已提交
491 492 493 494 495 496
	},
	{
		.procname	= "inet_peer_minttl",
		.data		= &inet_peer_minttl,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
497
		.proc_handler	= proc_dointvec_jiffies,
L
Linus Torvalds 已提交
498 499 500 501 502 503
	},
	{
		.procname	= "inet_peer_maxttl",
		.data		= &inet_peer_maxttl,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
504
		.proc_handler	= proc_dointvec_jiffies,
L
Linus Torvalds 已提交
505
	},
506 507 508 509 510 511 512
	{
		.procname	= "tcp_mem",
		.maxlen		= sizeof(sysctl_tcp_mem),
		.data		= &sysctl_tcp_mem,
		.mode		= 0644,
		.proc_handler	= proc_doulongvec_minmax,
	},
L
Linus Torvalds 已提交
513 514 515 516 517
	{
		.procname	= "tcp_low_latency",
		.data		= &sysctl_tcp_low_latency,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
518
		.proc_handler	= proc_dointvec
L
Linus Torvalds 已提交
519
	},
P
Paul Moore 已提交
520 521 522 523 524 525
#ifdef CONFIG_NETLABEL
	{
		.procname	= "cipso_cache_enable",
		.data		= &cipso_v4_cache_enabled,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
526
		.proc_handler	= proc_dointvec,
P
Paul Moore 已提交
527 528 529 530 531 532
	},
	{
		.procname	= "cipso_cache_bucket_size",
		.data		= &cipso_v4_cache_bucketsize,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
533
		.proc_handler	= proc_dointvec,
P
Paul Moore 已提交
534 535 536 537 538 539
	},
	{
		.procname	= "cipso_rbm_optfmt",
		.data		= &cipso_v4_rbm_optfmt,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
540
		.proc_handler	= proc_dointvec,
P
Paul Moore 已提交
541 542 543 544 545 546
	},
	{
		.procname	= "cipso_rbm_strictvalid",
		.data		= &cipso_v4_rbm_strictvalid,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
547
		.proc_handler	= proc_dointvec,
P
Paul Moore 已提交
548 549
	},
#endif /* CONFIG_NETLABEL */
D
Dave Watson 已提交
550 551 552 553 554 555
	{
		.procname	= "tcp_available_ulp",
		.maxlen		= TCP_ULP_BUF_MAX,
		.mode		= 0444,
		.proc_handler   = proc_tcp_available_ulp,
	},
556 557 558 559 560 561
	{
		.procname	= "icmp_msgs_per_sec",
		.data		= &sysctl_icmp_msgs_per_sec,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
562
		.extra1		= SYSCTL_ZERO,
563 564 565 566 567 568 569
	},
	{
		.procname	= "icmp_msgs_burst",
		.data		= &sysctl_icmp_msgs_burst,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
570
		.extra1		= SYSCTL_ZERO,
571
	},
H
Hideo Aoki 已提交
572 573 574 575 576
	{
		.procname	= "udp_mem",
		.data		= &sysctl_udp_mem,
		.maxlen		= sizeof(sysctl_udp_mem),
		.mode		= 0644,
E
Eric Dumazet 已提交
577
		.proc_handler	= proc_doulongvec_minmax,
H
Hideo Aoki 已提交
578
	},
579 580 581 582 583 584 585 586 587
	{
		.procname	= "fib_sync_mem",
		.data		= &sysctl_fib_sync_mem,
		.maxlen		= sizeof(sysctl_fib_sync_mem),
		.mode		= 0644,
		.proc_handler	= proc_douintvec_minmax,
		.extra1		= &sysctl_fib_sync_mem_min,
		.extra2		= &sysctl_fib_sync_mem_max,
	},
588
	{ }
L
Linus Torvalds 已提交
589
};
590

591
static struct ctl_table ipv4_net_table[] = {
592 593 594 595 596 597 598 599
	/* tcp_max_tw_buckets must be first in this table. */
	{
		.procname	= "tcp_max_tw_buckets",
/*		.data		= &init_net.ipv4.tcp_death_row.sysctl_max_tw_buckets, */
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
600 601 602
	{
		.procname	= "icmp_echo_ignore_all",
		.data		= &init_net.ipv4.sysctl_icmp_echo_ignore_all,
603
		.maxlen		= sizeof(u8),
604
		.mode		= 0644,
605
		.proc_handler	= proc_dou8vec_minmax,
606
	},
607 608 609
	{
		.procname	= "icmp_echo_enable_probe",
		.data		= &init_net.ipv4.sysctl_icmp_echo_enable_probe,
610
		.maxlen		= sizeof(u8),
611
		.mode		= 0644,
612
		.proc_handler	= proc_dou8vec_minmax,
613 614 615
		.extra1		= SYSCTL_ZERO,
		.extra2		= SYSCTL_ONE
	},
616 617 618
	{
		.procname	= "icmp_echo_ignore_broadcasts",
		.data		= &init_net.ipv4.sysctl_icmp_echo_ignore_broadcasts,
619
		.maxlen		= sizeof(u8),
620
		.mode		= 0644,
621
		.proc_handler	= proc_dou8vec_minmax,
622 623 624 625
	},
	{
		.procname	= "icmp_ignore_bogus_error_responses",
		.data		= &init_net.ipv4.sysctl_icmp_ignore_bogus_error_responses,
626
		.maxlen		= sizeof(u8),
627
		.mode		= 0644,
628
		.proc_handler	= proc_dou8vec_minmax,
629 630 631 632
	},
	{
		.procname	= "icmp_errors_use_inbound_ifaddr",
		.data		= &init_net.ipv4.sysctl_icmp_errors_use_inbound_ifaddr,
633
		.maxlen		= sizeof(u8),
634
		.mode		= 0644,
635
		.proc_handler	= proc_dou8vec_minmax,
636 637 638 639 640 641
	},
	{
		.procname	= "icmp_ratelimit",
		.data		= &init_net.ipv4.sysctl_icmp_ratelimit,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
642
		.proc_handler	= proc_dointvec_ms_jiffies,
643 644 645 646 647 648
	},
	{
		.procname	= "icmp_ratemask",
		.data		= &init_net.ipv4.sysctl_icmp_ratemask,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
649
		.proc_handler	= proc_dointvec
650
	},
651 652
	{
		.procname	= "ping_group_range",
653
		.data		= &init_net.ipv4.ping_group_range.range,
654
		.maxlen		= sizeof(gid_t)*2,
655 656 657
		.mode		= 0644,
		.proc_handler	= ipv4_ping_group_range,
	},
658 659 660 661
#ifdef CONFIG_NET_L3_MASTER_DEV
	{
		.procname	= "raw_l3mdev_accept",
		.data		= &init_net.ipv4.sysctl_raw_l3mdev_accept,
662
		.maxlen		= sizeof(u8),
663
		.mode		= 0644,
664
		.proc_handler	= proc_dou8vec_minmax,
665 666
		.extra1		= SYSCTL_ZERO,
		.extra2		= SYSCTL_ONE,
667 668
	},
#endif
669 670 671
	{
		.procname	= "tcp_ecn",
		.data		= &init_net.ipv4.sysctl_tcp_ecn,
672
		.maxlen		= sizeof(u8),
673
		.mode		= 0644,
674
		.proc_handler	= proc_dou8vec_minmax,
675
	},
676 677 678
	{
		.procname	= "tcp_ecn_fallback",
		.data		= &init_net.ipv4.sysctl_tcp_ecn_fallback,
679
		.maxlen		= sizeof(u8),
680
		.mode		= 0644,
681
		.proc_handler	= proc_dou8vec_minmax,
682
	},
683 684 685
	{
		.procname	= "ip_dynaddr",
		.data		= &init_net.ipv4.sysctl_ip_dynaddr,
686
		.maxlen		= sizeof(u8),
687
		.mode		= 0644,
688
		.proc_handler	= proc_dou8vec_minmax,
689
	},
690 691 692
	{
		.procname	= "ip_early_demux",
		.data		= &init_net.ipv4.sysctl_ip_early_demux,
693
		.maxlen		= sizeof(u8),
694
		.mode		= 0644,
695
		.proc_handler	= proc_dou8vec_minmax,
696
	},
697 698 699
	{
		.procname       = "udp_early_demux",
		.data           = &init_net.ipv4.sysctl_udp_early_demux,
700
		.maxlen         = sizeof(u8),
701 702 703 704 705 706
		.mode           = 0644,
		.proc_handler   = proc_udp_early_demux
	},
	{
		.procname       = "tcp_early_demux",
		.data           = &init_net.ipv4.sysctl_tcp_early_demux,
707
		.maxlen         = sizeof(u8),
708 709 710
		.mode           = 0644,
		.proc_handler   = proc_tcp_early_demux
	},
711 712 713
	{
		.procname       = "nexthop_compat_mode",
		.data           = &init_net.ipv4.sysctl_nexthop_compat_mode,
714
		.maxlen         = sizeof(u8),
715
		.mode           = 0644,
716
		.proc_handler   = proc_dou8vec_minmax,
717 718 719
		.extra1		= SYSCTL_ZERO,
		.extra2		= SYSCTL_ONE,
	},
720 721 722
	{
		.procname	= "ip_default_ttl",
		.data		= &init_net.ipv4.sysctl_ip_default_ttl,
723
		.maxlen		= sizeof(u8),
724
		.mode		= 0644,
725
		.proc_handler	= proc_dou8vec_minmax,
726 727 728
		.extra1		= &ip_ttl_min,
		.extra2		= &ip_ttl_max,
	},
729 730
	{
		.procname	= "ip_local_port_range",
731 732
		.maxlen		= sizeof(init_net.ipv4.ip_local_ports.range),
		.data		= &init_net.ipv4.ip_local_ports.range,
733 734 735
		.mode		= 0644,
		.proc_handler	= ipv4_local_port_range,
	},
736 737 738 739 740 741 742
	{
		.procname	= "ip_local_reserved_ports",
		.data		= &init_net.ipv4.sysctl_local_reserved_ports,
		.maxlen		= 65536,
		.mode		= 0644,
		.proc_handler	= proc_do_large_bitmap,
	},
743 744 745
	{
		.procname	= "ip_no_pmtu_disc",
		.data		= &init_net.ipv4.sysctl_ip_no_pmtu_disc,
746
		.maxlen		= sizeof(u8),
747
		.mode		= 0644,
748
		.proc_handler	= proc_dou8vec_minmax,
749
	},
750 751 752
	{
		.procname	= "ip_forward_use_pmtu",
		.data		= &init_net.ipv4.sysctl_ip_fwd_use_pmtu,
753
		.maxlen		= sizeof(u8),
754
		.mode		= 0644,
755
		.proc_handler	= proc_dou8vec_minmax,
756
	},
757 758 759
	{
		.procname	= "ip_forward_update_priority",
		.data		= &init_net.ipv4.sysctl_ip_fwd_update_priority,
760
		.maxlen		= sizeof(u8),
761
		.mode		= 0644,
762
		.proc_handler   = ipv4_fwd_update_priority,
763 764
		.extra1		= SYSCTL_ZERO,
		.extra2		= SYSCTL_ONE,
765
	},
766 767 768
	{
		.procname	= "ip_nonlocal_bind",
		.data		= &init_net.ipv4.sysctl_ip_nonlocal_bind,
769
		.maxlen		= sizeof(u8),
770
		.mode		= 0644,
771
		.proc_handler	= proc_dou8vec_minmax,
772
	},
773 774 775
	{
		.procname	= "ip_autobind_reuse",
		.data		= &init_net.ipv4.sysctl_ip_autobind_reuse,
776
		.maxlen		= sizeof(u8),
777
		.mode		= 0644,
778
		.proc_handler	= proc_dou8vec_minmax,
779 780 781
		.extra1         = SYSCTL_ZERO,
		.extra2         = SYSCTL_ONE,
	},
782 783 784
	{
		.procname	= "fwmark_reflect",
		.data		= &init_net.ipv4.sysctl_fwmark_reflect,
785
		.maxlen		= sizeof(u8),
786
		.mode		= 0644,
787
		.proc_handler	= proc_dou8vec_minmax,
788
	},
789 790 791
	{
		.procname	= "tcp_fwmark_accept",
		.data		= &init_net.ipv4.sysctl_tcp_fwmark_accept,
792
		.maxlen		= sizeof(u8),
793
		.mode		= 0644,
794
		.proc_handler	= proc_dou8vec_minmax,
795
	},
796 797 798 799
#ifdef CONFIG_NET_L3_MASTER_DEV
	{
		.procname	= "tcp_l3mdev_accept",
		.data		= &init_net.ipv4.sysctl_tcp_l3mdev_accept,
800
		.maxlen		= sizeof(u8),
801
		.mode		= 0644,
802
		.proc_handler	= proc_dou8vec_minmax,
803 804
		.extra1		= SYSCTL_ZERO,
		.extra2		= SYSCTL_ONE,
805 806
	},
#endif
F
Fan Du 已提交
807 808 809
	{
		.procname	= "tcp_mtu_probing",
		.data		= &init_net.ipv4.sysctl_tcp_mtu_probing,
810
		.maxlen		= sizeof(u8),
F
Fan Du 已提交
811
		.mode		= 0644,
812
		.proc_handler	= proc_dou8vec_minmax,
F
Fan Du 已提交
813 814 815 816 817 818 819 820
	},
	{
		.procname	= "tcp_base_mss",
		.data		= &init_net.ipv4.sysctl_tcp_base_mss,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
E
Eric Dumazet 已提交
821 822 823 824 825 826 827 828 829
	{
		.procname	= "tcp_min_snd_mss",
		.data		= &init_net.ipv4.sysctl_tcp_min_snd_mss,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &tcp_min_snd_mss_min,
		.extra2		= &tcp_min_snd_mss_max,
	},
830 831 832 833 834 835 836 837 838
	{
		.procname	= "tcp_mtu_probe_floor",
		.data		= &init_net.ipv4.sysctl_tcp_mtu_probe_floor,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &tcp_min_snd_mss_min,
		.extra2		= &tcp_min_snd_mss_max,
	},
839 840 841 842 843 844 845
	{
		.procname	= "tcp_probe_threshold",
		.data		= &init_net.ipv4.sysctl_tcp_probe_threshold,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
846 847 848
	{
		.procname	= "tcp_probe_interval",
		.data		= &init_net.ipv4.sysctl_tcp_probe_interval,
849
		.maxlen		= sizeof(u32),
850
		.mode		= 0644,
851 852
		.proc_handler	= proc_douintvec_minmax,
		.extra2		= &u32_max_div_HZ,
853
	},
854 855
	{
		.procname	= "igmp_link_local_mcast_reports",
856
		.data		= &init_net.ipv4.sysctl_igmp_llm_reports,
857
		.maxlen		= sizeof(u8),
858
		.mode		= 0644,
859
		.proc_handler	= proc_dou8vec_minmax,
860
	},
861 862 863 864 865 866 867
	{
		.procname	= "igmp_max_memberships",
		.data		= &init_net.ipv4.sysctl_igmp_max_memberships,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
868 869 870 871 872 873 874
	{
		.procname	= "igmp_max_msf",
		.data		= &init_net.ipv4.sysctl_igmp_max_msf,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
875 876 877 878 879 880 881
#ifdef CONFIG_IP_MULTICAST
	{
		.procname	= "igmp_qrv",
		.data		= &init_net.ipv4.sysctl_igmp_qrv,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
882
		.extra1		= SYSCTL_ONE
883 884
	},
#endif
885 886 887 888 889 890 891
	{
		.procname	= "tcp_congestion_control",
		.data		= &init_net.ipv4.tcp_congestion_control,
		.mode		= 0644,
		.maxlen		= TCP_CA_NAME_MAX,
		.proc_handler	= proc_tcp_congestion_control,
	},
892 893 894 895 896 897 898 899 900 901 902 903
	{
		.procname	= "tcp_available_congestion_control",
		.maxlen		= TCP_CA_BUF_MAX,
		.mode		= 0444,
		.proc_handler   = proc_tcp_available_congestion_control,
	},
	{
		.procname	= "tcp_allowed_congestion_control",
		.maxlen		= TCP_CA_BUF_MAX,
		.mode		= 0644,
		.proc_handler   = proc_allowed_congestion_control,
	},
904 905 906 907 908 909 910
	{
		.procname	= "tcp_keepalive_time",
		.data		= &init_net.ipv4.sysctl_tcp_keepalive_time,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_jiffies,
	},
911 912 913
	{
		.procname	= "tcp_keepalive_probes",
		.data		= &init_net.ipv4.sysctl_tcp_keepalive_probes,
914
		.maxlen		= sizeof(u8),
915
		.mode		= 0644,
916
		.proc_handler	= proc_dou8vec_minmax,
917
	},
918 919 920 921 922 923 924
	{
		.procname	= "tcp_keepalive_intvl",
		.data		= &init_net.ipv4.sysctl_tcp_keepalive_intvl,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_jiffies,
	},
925 926 927
	{
		.procname	= "tcp_syn_retries",
		.data		= &init_net.ipv4.sysctl_tcp_syn_retries,
928
		.maxlen		= sizeof(u8),
929
		.mode		= 0644,
930
		.proc_handler	= proc_dou8vec_minmax,
931 932 933
		.extra1		= &tcp_syn_retries_min,
		.extra2		= &tcp_syn_retries_max
	},
934 935 936
	{
		.procname	= "tcp_synack_retries",
		.data		= &init_net.ipv4.sysctl_tcp_synack_retries,
937
		.maxlen		= sizeof(u8),
938
		.mode		= 0644,
939
		.proc_handler	= proc_dou8vec_minmax,
940
	},
941 942 943 944
#ifdef CONFIG_SYN_COOKIES
	{
		.procname	= "tcp_syncookies",
		.data		= &init_net.ipv4.sysctl_tcp_syncookies,
945
		.maxlen		= sizeof(u8),
946
		.mode		= 0644,
947
		.proc_handler	= proc_dou8vec_minmax,
948 949
	},
#endif
950 951 952 953 954 955 956 957 958
	{
		.procname	= "tcp_migrate_req",
		.data		= &init_net.ipv4.sysctl_tcp_migrate_req,
		.maxlen		= sizeof(u8),
		.mode		= 0644,
		.proc_handler	= proc_dou8vec_minmax,
		.extra1		= SYSCTL_ZERO,
		.extra2		= SYSCTL_ONE
	},
959 960 961 962 963 964 965
	{
		.procname	= "tcp_reordering",
		.data		= &init_net.ipv4.sysctl_tcp_reordering,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
966 967 968
	{
		.procname	= "tcp_retries1",
		.data		= &init_net.ipv4.sysctl_tcp_retries1,
969
		.maxlen		= sizeof(u8),
970
		.mode		= 0644,
971
		.proc_handler	= proc_dou8vec_minmax,
972 973
		.extra2		= &tcp_retr1_max
	},
974 975 976
	{
		.procname	= "tcp_retries2",
		.data		= &init_net.ipv4.sysctl_tcp_retries2,
977
		.maxlen		= sizeof(u8),
978
		.mode		= 0644,
979
		.proc_handler	= proc_dou8vec_minmax,
980
	},
981 982 983
	{
		.procname	= "tcp_orphan_retries",
		.data		= &init_net.ipv4.sysctl_tcp_orphan_retries,
984
		.maxlen		= sizeof(u8),
985
		.mode		= 0644,
986
		.proc_handler	= proc_dou8vec_minmax,
987
	},
988 989 990 991 992 993 994
	{
		.procname	= "tcp_fin_timeout",
		.data		= &init_net.ipv4.sysctl_tcp_fin_timeout,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_jiffies,
	},
995 996 997 998 999
	{
		.procname	= "tcp_notsent_lowat",
		.data		= &init_net.ipv4.sysctl_tcp_notsent_lowat,
		.maxlen		= sizeof(unsigned int),
		.mode		= 0644,
1000
		.proc_handler	= proc_douintvec,
1001
	},
1002 1003 1004
	{
		.procname	= "tcp_tw_reuse",
		.data		= &init_net.ipv4.sysctl_tcp_tw_reuse,
1005
		.maxlen		= sizeof(u8),
1006
		.mode		= 0644,
1007
		.proc_handler	= proc_dou8vec_minmax,
1008
		.extra1		= SYSCTL_ZERO,
1009
		.extra2		= &two,
1010
	},
1011 1012 1013 1014 1015 1016 1017
	{
		.procname	= "tcp_max_syn_backlog",
		.data		= &init_net.ipv4.sysctl_max_syn_backlog,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
1018 1019 1020 1021 1022 1023 1024
	{
		.procname	= "tcp_fastopen",
		.data		= &init_net.ipv4.sysctl_tcp_fastopen,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
1025 1026 1027 1028
	{
		.procname	= "tcp_fastopen_key",
		.mode		= 0600,
		.data		= &init_net.ipv4.sysctl_tcp_fastopen,
1029 1030 1031 1032 1033 1034
		/* maxlen to print the list of keys in hex (*2), with dashes
		 * separating doublewords and a comma in between keys.
		 */
		.maxlen		= ((TCP_FASTOPEN_KEY_LENGTH *
				   2 * TCP_FASTOPEN_KEY_MAX) +
				   (TCP_FASTOPEN_KEY_MAX * 5)),
1035 1036
		.proc_handler	= proc_tcp_fastopen_key,
	},
1037 1038 1039 1040 1041 1042
	{
		.procname	= "tcp_fastopen_blackhole_timeout_sec",
		.data		= &init_net.ipv4.sysctl_tcp_fastopen_blackhole_timeout,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_tfo_blackhole_detect_timeout,
1043
		.extra1		= SYSCTL_ZERO,
1044
	},
1045 1046 1047 1048
#ifdef CONFIG_IP_ROUTE_MULTIPATH
	{
		.procname	= "fib_multipath_use_neigh",
		.data		= &init_net.ipv4.sysctl_fib_multipath_use_neigh,
1049
		.maxlen		= sizeof(u8),
1050
		.mode		= 0644,
1051
		.proc_handler	= proc_dou8vec_minmax,
1052 1053
		.extra1		= SYSCTL_ZERO,
		.extra2		= SYSCTL_ONE,
1054 1055 1056 1057
	},
	{
		.procname	= "fib_multipath_hash_policy",
		.data		= &init_net.ipv4.sysctl_fib_multipath_hash_policy,
1058
		.maxlen		= sizeof(u8),
1059
		.mode		= 0644,
1060
		.proc_handler	= proc_fib_multipath_hash_policy,
1061
		.extra1		= SYSCTL_ZERO,
1062
		.extra2		= &three,
1063
	},
1064 1065 1066 1067 1068
	{
		.procname	= "fib_multipath_hash_fields",
		.data		= &init_net.ipv4.sysctl_fib_multipath_hash_fields,
		.maxlen		= sizeof(u32),
		.mode		= 0644,
1069
		.proc_handler	= proc_fib_multipath_hash_fields,
1070 1071 1072
		.extra1		= SYSCTL_ONE,
		.extra2		= &fib_multipath_hash_fields_all_mask,
	},
1073
#endif
1074 1075 1076 1077 1078 1079 1080
	{
		.procname	= "ip_unprivileged_port_start",
		.maxlen		= sizeof(int),
		.data		= &init_net.ipv4.sysctl_ip_prot_sock,
		.mode		= 0644,
		.proc_handler	= ipv4_privileged_ports,
	},
1081 1082 1083 1084
#ifdef CONFIG_NET_L3_MASTER_DEV
	{
		.procname	= "udp_l3mdev_accept",
		.data		= &init_net.ipv4.sysctl_udp_l3mdev_accept,
1085
		.maxlen		= sizeof(u8),
1086
		.mode		= 0644,
1087
		.proc_handler	= proc_dou8vec_minmax,
1088 1089
		.extra1		= SYSCTL_ZERO,
		.extra2		= SYSCTL_ONE,
1090 1091
	},
#endif
E
Eric Dumazet 已提交
1092 1093 1094
	{
		.procname	= "tcp_sack",
		.data		= &init_net.ipv4.sysctl_tcp_sack,
1095
		.maxlen		= sizeof(u8),
E
Eric Dumazet 已提交
1096
		.mode		= 0644,
1097
		.proc_handler	= proc_dou8vec_minmax,
E
Eric Dumazet 已提交
1098
	},
1099 1100 1101
	{
		.procname	= "tcp_window_scaling",
		.data		= &init_net.ipv4.sysctl_tcp_window_scaling,
1102
		.maxlen		= sizeof(u8),
1103
		.mode		= 0644,
1104
		.proc_handler	= proc_dou8vec_minmax,
1105
	},
1106 1107 1108
	{
		.procname	= "tcp_timestamps",
		.data		= &init_net.ipv4.sysctl_tcp_timestamps,
1109
		.maxlen		= sizeof(u8),
1110
		.mode		= 0644,
1111
		.proc_handler	= proc_dou8vec_minmax,
1112
	},
1113 1114 1115
	{
		.procname	= "tcp_early_retrans",
		.data		= &init_net.ipv4.sysctl_tcp_early_retrans,
1116
		.maxlen		= sizeof(u8),
1117
		.mode		= 0644,
1118
		.proc_handler	= proc_dou8vec_minmax,
1119
		.extra1		= SYSCTL_ZERO,
1120 1121
		.extra2		= &four,
	},
1122 1123 1124
	{
		.procname	= "tcp_recovery",
		.data		= &init_net.ipv4.sysctl_tcp_recovery,
1125
		.maxlen		= sizeof(u8),
1126
		.mode		= 0644,
1127
		.proc_handler	= proc_dou8vec_minmax,
1128
	},
1129 1130 1131
	{
		.procname       = "tcp_thin_linear_timeouts",
		.data           = &init_net.ipv4.sysctl_tcp_thin_linear_timeouts,
1132
		.maxlen         = sizeof(u8),
1133
		.mode           = 0644,
1134
		.proc_handler   = proc_dou8vec_minmax,
1135
	},
1136 1137 1138
	{
		.procname	= "tcp_slow_start_after_idle",
		.data		= &init_net.ipv4.sysctl_tcp_slow_start_after_idle,
1139
		.maxlen		= sizeof(u8),
1140
		.mode		= 0644,
1141
		.proc_handler	= proc_dou8vec_minmax,
1142
	},
1143 1144 1145
	{
		.procname	= "tcp_retrans_collapse",
		.data		= &init_net.ipv4.sysctl_tcp_retrans_collapse,
1146
		.maxlen		= sizeof(u8),
1147
		.mode		= 0644,
1148
		.proc_handler	= proc_dou8vec_minmax,
1149
	},
1150 1151 1152
	{
		.procname	= "tcp_stdurg",
		.data		= &init_net.ipv4.sysctl_tcp_stdurg,
1153
		.maxlen		= sizeof(u8),
1154
		.mode		= 0644,
1155
		.proc_handler	= proc_dou8vec_minmax,
1156
	},
1157 1158 1159
	{
		.procname	= "tcp_rfc1337",
		.data		= &init_net.ipv4.sysctl_tcp_rfc1337,
1160
		.maxlen		= sizeof(u8),
1161
		.mode		= 0644,
1162
		.proc_handler	= proc_dou8vec_minmax,
1163
	},
1164 1165 1166
	{
		.procname	= "tcp_abort_on_overflow",
		.data		= &init_net.ipv4.sysctl_tcp_abort_on_overflow,
1167
		.maxlen		= sizeof(u8),
1168
		.mode		= 0644,
1169
		.proc_handler	= proc_dou8vec_minmax,
1170
	},
E
Eric Dumazet 已提交
1171 1172 1173
	{
		.procname	= "tcp_fack",
		.data		= &init_net.ipv4.sysctl_tcp_fack,
1174
		.maxlen		= sizeof(u8),
E
Eric Dumazet 已提交
1175
		.mode		= 0644,
1176
		.proc_handler	= proc_dou8vec_minmax,
E
Eric Dumazet 已提交
1177
	},
1178 1179 1180 1181 1182 1183 1184
	{
		.procname	= "tcp_max_reordering",
		.data		= &init_net.ipv4.sysctl_tcp_max_reordering,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
E
Eric Dumazet 已提交
1185 1186 1187
	{
		.procname	= "tcp_dsack",
		.data		= &init_net.ipv4.sysctl_tcp_dsack,
1188
		.maxlen		= sizeof(u8),
E
Eric Dumazet 已提交
1189
		.mode		= 0644,
1190
		.proc_handler	= proc_dou8vec_minmax,
E
Eric Dumazet 已提交
1191
	},
1192 1193 1194
	{
		.procname	= "tcp_app_win",
		.data		= &init_net.ipv4.sysctl_tcp_app_win,
1195
		.maxlen		= sizeof(u8),
1196
		.mode		= 0644,
1197
		.proc_handler	= proc_dou8vec_minmax,
1198
	},
1199 1200 1201 1202 1203 1204 1205 1206 1207
	{
		.procname	= "tcp_adv_win_scale",
		.data		= &init_net.ipv4.sysctl_tcp_adv_win_scale,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &tcp_adv_win_scale_min,
		.extra2		= &tcp_adv_win_scale_max,
	},
E
Eric Dumazet 已提交
1208 1209 1210
	{
		.procname	= "tcp_frto",
		.data		= &init_net.ipv4.sysctl_tcp_frto,
1211
		.maxlen		= sizeof(u8),
E
Eric Dumazet 已提交
1212
		.mode		= 0644,
1213
		.proc_handler	= proc_dou8vec_minmax,
E
Eric Dumazet 已提交
1214
	},
1215 1216 1217
	{
		.procname	= "tcp_no_metrics_save",
		.data		= &init_net.ipv4.sysctl_tcp_nometrics_save,
1218
		.maxlen		= sizeof(u8),
1219
		.mode		= 0644,
1220
		.proc_handler	= proc_dou8vec_minmax,
1221
	},
1222 1223 1224
	{
		.procname	= "tcp_no_ssthresh_metrics_save",
		.data		= &init_net.ipv4.sysctl_tcp_no_ssthresh_metrics_save,
1225
		.maxlen		= sizeof(u8),
1226
		.mode		= 0644,
1227
		.proc_handler	= proc_dou8vec_minmax,
1228 1229 1230
		.extra1		= SYSCTL_ZERO,
		.extra2		= SYSCTL_ONE,
	},
1231 1232 1233
	{
		.procname	= "tcp_moderate_rcvbuf",
		.data		= &init_net.ipv4.sysctl_tcp_moderate_rcvbuf,
1234
		.maxlen		= sizeof(u8),
1235
		.mode		= 0644,
1236
		.proc_handler	= proc_dou8vec_minmax,
1237
	},
1238 1239 1240
	{
		.procname	= "tcp_tso_win_divisor",
		.data		= &init_net.ipv4.sysctl_tcp_tso_win_divisor,
1241
		.maxlen		= sizeof(u8),
1242
		.mode		= 0644,
1243
		.proc_handler	= proc_dou8vec_minmax,
1244
	},
1245 1246 1247
	{
		.procname	= "tcp_workaround_signed_windows",
		.data		= &init_net.ipv4.sysctl_tcp_workaround_signed_windows,
1248
		.maxlen		= sizeof(u8),
1249
		.mode		= 0644,
1250
		.proc_handler	= proc_dou8vec_minmax,
1251
	},
1252 1253 1254 1255 1256 1257 1258
	{
		.procname	= "tcp_limit_output_bytes",
		.data		= &init_net.ipv4.sysctl_tcp_limit_output_bytes,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
1259 1260 1261 1262 1263 1264 1265
	{
		.procname	= "tcp_challenge_ack_limit",
		.data		= &init_net.ipv4.sysctl_tcp_challenge_ack_limit,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
1266 1267 1268
	{
		.procname	= "tcp_min_tso_segs",
		.data		= &init_net.ipv4.sysctl_tcp_min_tso_segs,
1269
		.maxlen		= sizeof(u8),
1270
		.mode		= 0644,
1271
		.proc_handler	= proc_dou8vec_minmax,
1272
		.extra1		= SYSCTL_ONE,
1273
	},
1274 1275 1276 1277 1278
	{
		.procname	= "tcp_min_rtt_wlen",
		.data		= &init_net.ipv4.sysctl_tcp_min_rtt_wlen,
		.maxlen		= sizeof(int),
		.mode		= 0644,
1279
		.proc_handler	= proc_dointvec_minmax,
1280
		.extra1		= SYSCTL_ZERO,
1281
		.extra2		= &one_day_secs
1282
	},
1283 1284 1285
	{
		.procname	= "tcp_autocorking",
		.data		= &init_net.ipv4.sysctl_tcp_autocorking,
1286
		.maxlen		= sizeof(u8),
1287
		.mode		= 0644,
1288
		.proc_handler	= proc_dou8vec_minmax,
1289 1290
		.extra1		= SYSCTL_ZERO,
		.extra2		= SYSCTL_ONE,
1291
	},
1292 1293 1294 1295 1296 1297 1298
	{
		.procname	= "tcp_invalid_ratelimit",
		.data		= &init_net.ipv4.sysctl_tcp_invalid_ratelimit,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_ms_jiffies,
	},
1299 1300 1301 1302 1303 1304
	{
		.procname	= "tcp_pacing_ss_ratio",
		.data		= &init_net.ipv4.sysctl_tcp_pacing_ss_ratio,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
1305
		.extra1		= SYSCTL_ZERO,
1306 1307
		.extra2		= &thousand,
	},
1308 1309 1310 1311 1312 1313
	{
		.procname	= "tcp_pacing_ca_ratio",
		.data		= &init_net.ipv4.sysctl_tcp_pacing_ca_ratio,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
1314
		.extra1		= SYSCTL_ZERO,
1315 1316
		.extra2		= &thousand,
	},
1317 1318 1319 1320 1321 1322
	{
		.procname	= "tcp_wmem",
		.data		= &init_net.ipv4.sysctl_tcp_wmem,
		.maxlen		= sizeof(init_net.ipv4.sysctl_tcp_wmem),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
1323
		.extra1		= SYSCTL_ONE,
1324 1325 1326 1327 1328 1329 1330
	},
	{
		.procname	= "tcp_rmem",
		.data		= &init_net.ipv4.sysctl_tcp_rmem,
		.maxlen		= sizeof(init_net.ipv4.sysctl_tcp_rmem),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
1331
		.extra1		= SYSCTL_ONE,
1332
	},
1333 1334 1335 1336 1337 1338 1339
	{
		.procname	= "tcp_comp_sack_delay_ns",
		.data		= &init_net.ipv4.sysctl_tcp_comp_sack_delay_ns,
		.maxlen		= sizeof(unsigned long),
		.mode		= 0644,
		.proc_handler	= proc_doulongvec_minmax,
	},
1340 1341 1342 1343 1344 1345 1346
	{
		.procname	= "tcp_comp_sack_slack_ns",
		.data		= &init_net.ipv4.sysctl_tcp_comp_sack_slack_ns,
		.maxlen		= sizeof(unsigned long),
		.mode		= 0644,
		.proc_handler	= proc_doulongvec_minmax,
	},
E
Eric Dumazet 已提交
1347 1348 1349
	{
		.procname	= "tcp_comp_sack_nr",
		.data		= &init_net.ipv4.sysctl_tcp_comp_sack_nr,
1350
		.maxlen		= sizeof(u8),
E
Eric Dumazet 已提交
1351
		.mode		= 0644,
1352
		.proc_handler	= proc_dou8vec_minmax,
1353
		.extra1		= SYSCTL_ZERO,
E
Eric Dumazet 已提交
1354
	},
1355 1356 1357
	{
		.procname       = "tcp_reflect_tos",
		.data           = &init_net.ipv4.sysctl_tcp_reflect_tos,
1358
		.maxlen         = sizeof(u8),
1359
		.mode           = 0644,
1360
		.proc_handler   = proc_dou8vec_minmax,
1361 1362 1363
		.extra1         = SYSCTL_ZERO,
		.extra2         = SYSCTL_ONE,
	},
1364 1365 1366 1367 1368 1369
	{
		.procname	= "udp_rmem_min",
		.data		= &init_net.ipv4.sysctl_udp_rmem_min,
		.maxlen		= sizeof(init_net.ipv4.sysctl_udp_rmem_min),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
1370
		.extra1		= SYSCTL_ONE
1371 1372 1373 1374 1375 1376 1377
	},
	{
		.procname	= "udp_wmem_min",
		.data		= &init_net.ipv4.sysctl_udp_wmem_min,
		.maxlen		= sizeof(init_net.ipv4.sysctl_udp_wmem_min),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
1378
		.extra1		= SYSCTL_ONE
1379
	},
1380 1381 1382
	{
		.procname	= "fib_notify_on_flag_change",
		.data		= &init_net.ipv4.sysctl_fib_notify_on_flag_change,
1383
		.maxlen		= sizeof(u8),
1384
		.mode		= 0644,
1385
		.proc_handler	= proc_dou8vec_minmax,
1386
		.extra1		= SYSCTL_ZERO,
1387
		.extra2		= &two,
1388
	},
1389 1390 1391
	{ }
};

1392 1393
static __net_init int ipv4_sysctl_init_net(struct net *net)
{
1394 1395 1396
	struct ctl_table *table;

	table = ipv4_net_table;
O
Octavian Purdila 已提交
1397
	if (!net_eq(net, &init_net)) {
1398 1399
		int i;

1400
		table = kmemdup(table, sizeof(ipv4_net_table), GFP_KERNEL);
1401
		if (!table)
1402 1403
			goto err_alloc;

1404 1405
		/* skip first entry (sysctl_max_tw_buckets) */
		for (i = 1; i < ARRAY_SIZE(ipv4_net_table) - 1; i++) {
1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417
			if (table[i].data) {
				/* Update the variables to point into
				 * the current struct net
				 */
				table[i].data += (void *)net - (void *)&init_net;
			} else {
				/* Entries without data pointer are global;
				 * Make them read-only in non-init_net ns
				 */
				table[i].mode &= ~0222;
			}
		}
1418 1419
	}

1420 1421
	table[0].data = &net->ipv4.tcp_death_row->sysctl_max_tw_buckets;

1422
	net->ipv4.ipv4_hdr = register_net_sysctl(net, "net/ipv4", table);
1423
	if (!net->ipv4.ipv4_hdr)
1424 1425
		goto err_reg;

1426 1427 1428 1429
	net->ipv4.sysctl_local_reserved_ports = kzalloc(65536 / 8, GFP_KERNEL);
	if (!net->ipv4.sysctl_local_reserved_ports)
		goto err_ports;

1430
	return 0;
1431

1432 1433
err_ports:
	unregister_net_sysctl_table(net->ipv4.ipv4_hdr);
1434
err_reg:
O
Octavian Purdila 已提交
1435
	if (!net_eq(net, &init_net))
1436 1437 1438
		kfree(table);
err_alloc:
	return -ENOMEM;
1439 1440 1441 1442
}

static __net_exit void ipv4_sysctl_exit_net(struct net *net)
{
1443 1444
	struct ctl_table *table;

1445
	kfree(net->ipv4.sysctl_local_reserved_ports);
1446 1447 1448
	table = net->ipv4.ipv4_hdr->ctl_table_arg;
	unregister_net_sysctl_table(net->ipv4.ipv4_hdr);
	kfree(table);
1449 1450 1451 1452 1453 1454 1455
}

static __net_initdata struct pernet_operations ipv4_sysctl_ops = {
	.init = ipv4_sysctl_init_net,
	.exit = ipv4_sysctl_exit_net,
};

1456 1457 1458 1459
static __init int sysctl_ipv4_init(void)
{
	struct ctl_table_header *hdr;

1460
	hdr = register_net_sysctl(&init_net, "net/ipv4", ipv4_table);
1461
	if (!hdr)
1462 1463 1464
		return -ENOMEM;

	if (register_pernet_subsys(&ipv4_sysctl_ops)) {
1465
		unregister_net_sysctl_table(hdr);
1466 1467 1468 1469
		return -ENOMEM;
	}

	return 0;
1470 1471 1472
}

__initcall(sysctl_ipv4_init);