sysctl_net_ipv4.c 26.1 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10
/*
 * 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/mm.h>
#include <linux/module.h>
#include <linux/sysctl.h>
11
#include <linux/igmp.h>
12
#include <linux/inetdevice.h>
13
#include <linux/seqlock.h>
14
#include <linux/init.h>
15
#include <linux/slab.h>
16
#include <linux/nsproxy.h>
G
Glauber Costa 已提交
17
#include <linux/swap.h>
L
Linus Torvalds 已提交
18
#include <net/snmp.h>
19
#include <net/icmp.h>
L
Linus Torvalds 已提交
20 21 22
#include <net/ip.h>
#include <net/route.h>
#include <net/tcp.h>
H
Hideo Aoki 已提交
23
#include <net/udp.h>
P
Paul Moore 已提交
24
#include <net/cipso_ipv4.h>
25
#include <net/inet_frag.h>
26
#include <net/ping.h>
L
Linus Torvalds 已提交
27

H
Herbert Xu 已提交
28
static int zero;
29
static int one = 1;
N
Nandita Dukkipati 已提交
30
static int four = 4;
31
static int thousand = 1000;
E
Eric Dumazet 已提交
32
static int gso_max_segs = GSO_MAX_SEGS;
33
static int tcp_retr1_max = 255;
L
Linus Torvalds 已提交
34 35
static int ip_local_port_range_min[] = { 1, 1 };
static int ip_local_port_range_max[] = { 65535, 65535 };
36 37
static int tcp_adv_win_scale_min = -31;
static int tcp_adv_win_scale_max = 31;
38 39
static int ip_privileged_port_min;
static int ip_privileged_port_max = 65535;
E
Eric Dumazet 已提交
40 41
static int ip_ttl_min = 1;
static int ip_ttl_max = 255;
42 43
static int tcp_syn_retries_min = 1;
static int tcp_syn_retries_max = MAX_TCP_SYNCNT;
44 45
static int ip_ping_group_range_min[] = { 0, 0 };
static int ip_ping_group_range_max[] = { GID_T_MAX, GID_T_MAX };
L
Linus Torvalds 已提交
46

47
/* Update system visible IP port range */
48
static void set_local_port_range(struct net *net, int range[2])
49
{
50 51
	bool same_parity = !((range[0] ^ range[1]) & 1);

52
	write_seqlock_bh(&net->ipv4.ip_local_ports.lock);
53 54 55 56
	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");
	}
57 58
	net->ipv4.ip_local_ports.range[0] = range[0];
	net->ipv4.ip_local_ports.range[1] = range[1];
59
	write_sequnlock_bh(&net->ipv4.ip_local_ports.lock);
60 61 62
}

/* Validate changes from /proc interface. */
63
static int ipv4_local_port_range(struct ctl_table *table, int write,
64 65 66
				 void __user *buffer,
				 size_t *lenp, loff_t *ppos)
{
67
	struct net *net =
68
		container_of(table->data, struct net, ipv4.ip_local_ports.range);
69
	int ret;
E
Eric Dumazet 已提交
70
	int range[2];
71
	struct ctl_table tmp = {
72 73 74 75 76 77 78
		.data = &range,
		.maxlen = sizeof(range),
		.mode = table->mode,
		.extra1 = &ip_local_port_range_min,
		.extra2 = &ip_local_port_range_max,
	};

79 80
	inet_get_local_port_range(net, &range[0], &range[1]);

81
	ret = proc_dointvec_minmax(&tmp, write, buffer, lenp, ppos);
82 83

	if (write && ret == 0) {
84 85 86 87 88 89
		/* 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))
90 91
			ret = -EINVAL;
		else
92
			set_local_port_range(net, range);
93 94 95 96 97
	}

	return ret;
}

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
/* Validate changes from /proc interface. */
static int ipv4_privileged_ports(struct ctl_table *table, int write,
				void __user *buffer, size_t *lenp, loff_t *ppos)
{
	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;
}
132

133
static void inet_get_ping_group_range_table(struct ctl_table *table, kgid_t *low, kgid_t *high)
134
{
135
	kgid_t *data = table->data;
136
	struct net *net =
137
		container_of(table->data, struct net, ipv4.ping_group_range.range);
138
	unsigned int seq;
139
	do {
140
		seq = read_seqbegin(&net->ipv4.ping_group_range.lock);
141 142 143

		*low = data[0];
		*high = data[1];
144
	} while (read_seqretry(&net->ipv4.ping_group_range.lock, seq));
145 146 147
}

/* Update system visible IP port range */
148
static void set_ping_group_range(struct ctl_table *table, kgid_t low, kgid_t high)
149
{
150
	kgid_t *data = table->data;
151
	struct net *net =
152
		container_of(table->data, struct net, ipv4.ping_group_range.range);
153
	write_seqlock(&net->ipv4.ping_group_range.lock);
154 155
	data[0] = low;
	data[1] = high;
156
	write_sequnlock(&net->ipv4.ping_group_range.lock);
157 158 159
}

/* Validate changes from /proc interface. */
160
static int ipv4_ping_group_range(struct ctl_table *table, int write,
161 162 163
				 void __user *buffer,
				 size_t *lenp, loff_t *ppos)
{
164
	struct user_namespace *user_ns = current_user_ns();
165
	int ret;
166 167
	gid_t urange[2];
	kgid_t low, high;
168
	struct ctl_table tmp = {
169 170
		.data = &urange,
		.maxlen = sizeof(urange),
171 172 173 174 175
		.mode = table->mode,
		.extra1 = &ip_ping_group_range_min,
		.extra2 = &ip_ping_group_range_max,
	};

176 177 178
	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);
179 180
	ret = proc_dointvec_minmax(&tmp, write, buffer, lenp, ppos);

181 182 183 184 185 186 187 188 189 190
	if (write && ret == 0) {
		low = make_kgid(user_ns, urange[0]);
		high = make_kgid(user_ns, urange[1]);
		if (!gid_valid(low) || !gid_valid(high) ||
		    (urange[1] < urange[0]) || gid_lt(high, low)) {
			low = make_kgid(&init_user_ns, 1);
			high = make_kgid(&init_user_ns, 0);
		}
		set_ping_group_range(table, low, high);
	}
191 192 193 194

	return ret;
}

195
static int proc_tcp_congestion_control(struct ctl_table *ctl, int write,
196 197 198
				       void __user *buffer, size_t *lenp, loff_t *ppos)
{
	char val[TCP_CA_NAME_MAX];
199
	struct ctl_table tbl = {
200 201 202 203 204 205 206
		.data = val,
		.maxlen = TCP_CA_NAME_MAX,
	};
	int ret;

	tcp_get_default_congestion_control(val);

207
	ret = proc_dostring(&tbl, write, buffer, lenp, ppos);
208 209 210 211 212
	if (write && ret == 0)
		ret = tcp_set_default_congestion_control(val);
	return ret;
}

213
static int proc_tcp_available_congestion_control(struct ctl_table *ctl,
214
						 int write,
215 216 217
						 void __user *buffer, size_t *lenp,
						 loff_t *ppos)
{
218
	struct ctl_table tbl = { .maxlen = TCP_CA_BUF_MAX, };
219 220 221 222 223 224
	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);
225
	ret = proc_dostring(&tbl, write, buffer, lenp, ppos);
226 227 228 229
	kfree(tbl.data);
	return ret;
}

230
static int proc_allowed_congestion_control(struct ctl_table *ctl,
231
					   int write,
232 233 234
					   void __user *buffer, size_t *lenp,
					   loff_t *ppos)
{
235
	struct ctl_table tbl = { .maxlen = TCP_CA_BUF_MAX };
236 237 238 239 240 241 242
	int ret;

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

	tcp_get_allowed_congestion_control(tbl.data, tbl.maxlen);
243
	ret = proc_dostring(&tbl, write, buffer, lenp, ppos);
244 245 246 247 248 249
	if (write && ret == 0)
		ret = tcp_set_allowed_congestion_control(tbl.data);
	kfree(tbl.data);
	return ret;
}

250 251 252
static int proc_tcp_fastopen_key(struct ctl_table *ctl, int write,
				 void __user *buffer, size_t *lenp,
				 loff_t *ppos)
253
{
254
	struct ctl_table tbl = { .maxlen = (TCP_FASTOPEN_KEY_LENGTH * 2 + 10) };
255 256 257 258 259 260 261 262 263 264 265 266
	struct tcp_fastopen_context *ctxt;
	int ret;
	u32  user_key[4]; /* 16 bytes, matching TCP_FASTOPEN_KEY_LENGTH */

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

	rcu_read_lock();
	ctxt = rcu_dereference(tcp_fastopen_ctx);
	if (ctxt)
		memcpy(user_key, ctxt->key, TCP_FASTOPEN_KEY_LENGTH);
267 268
	else
		memset(user_key, 0, sizeof(user_key));
269 270 271 272 273 274 275 276 277 278 279 280
	rcu_read_unlock();

	snprintf(tbl.data, tbl.maxlen, "%08x-%08x-%08x-%08x",
		user_key[0], user_key[1], user_key[2], user_key[3]);
	ret = proc_dostring(&tbl, write, buffer, lenp, ppos);

	if (write && ret == 0) {
		if (sscanf(tbl.data, "%x-%x-%x-%x", user_key, user_key + 1,
			   user_key + 2, user_key + 3) != 4) {
			ret = -EINVAL;
			goto bad_key;
		}
281 282 283 284 285
		/* Generate a dummy secret but don't publish it. This
		 * is needed so we don't regenerate a new key on the
		 * first invocation of tcp_fastopen_cookie_gen
		 */
		tcp_fastopen_init_key_once(false);
286 287 288 289 290 291 292 293 294 295 296
		tcp_fastopen_reset_cipher(user_key, TCP_FASTOPEN_KEY_LENGTH);
	}

bad_key:
	pr_debug("proc FO key set 0x%x-%x-%x-%x <- 0x%s: %u\n",
	       user_key[0], user_key[1], user_key[2], user_key[3],
	       (char *)tbl.data, ret);
	kfree(tbl.data);
	return ret;
}

297
static struct ctl_table ipv4_table[] = {
298
	{
L
Linus Torvalds 已提交
299 300 301 302
		.procname	= "tcp_timestamps",
		.data		= &sysctl_tcp_timestamps,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
303
		.proc_handler	= proc_dointvec
L
Linus Torvalds 已提交
304
	},
305
	{
L
Linus Torvalds 已提交
306 307 308 309
		.procname	= "tcp_window_scaling",
		.data		= &sysctl_tcp_window_scaling,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
310
		.proc_handler	= proc_dointvec
L
Linus Torvalds 已提交
311
	},
312
	{
L
Linus Torvalds 已提交
313 314 315 316
		.procname	= "tcp_sack",
		.data		= &sysctl_tcp_sack,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
317
		.proc_handler	= proc_dointvec
L
Linus Torvalds 已提交
318
	},
319
	{
L
Linus Torvalds 已提交
320 321 322 323
		.procname	= "tcp_retrans_collapse",
		.data		= &sysctl_tcp_retrans_collapse,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
324
		.proc_handler	= proc_dointvec
L
Linus Torvalds 已提交
325 326 327 328 329 330
	},
	{
		.procname	= "tcp_max_orphans",
		.data		= &sysctl_tcp_max_orphans,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
331
		.proc_handler	= proc_dointvec
L
Linus Torvalds 已提交
332
	},
Y
Yuchung Cheng 已提交
333 334 335 336 337 338 339
	{
		.procname	= "tcp_fastopen",
		.data		= &sysctl_tcp_fastopen,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
340 341 342 343 344 345
	{
		.procname	= "tcp_fastopen_key",
		.mode		= 0600,
		.maxlen		= ((TCP_FASTOPEN_KEY_LENGTH * 2) + 10),
		.proc_handler	= proc_tcp_fastopen_key,
	},
L
Linus Torvalds 已提交
346 347 348 349 350
	{
		.procname	= "tcp_abort_on_overflow",
		.data		= &sysctl_tcp_abort_on_overflow,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
351
		.proc_handler	= proc_dointvec
L
Linus Torvalds 已提交
352 353 354 355 356 357
	},
	{
		.procname	= "tcp_stdurg",
		.data		= &sysctl_tcp_stdurg,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
358
		.proc_handler	= proc_dointvec
L
Linus Torvalds 已提交
359 360 361 362 363 364
	},
	{
		.procname	= "tcp_rfc1337",
		.data		= &sysctl_tcp_rfc1337,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
365
		.proc_handler	= proc_dointvec
L
Linus Torvalds 已提交
366 367 368 369 370 371
	},
	{
		.procname	= "inet_peer_threshold",
		.data		= &inet_peer_threshold,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
372
		.proc_handler	= proc_dointvec
L
Linus Torvalds 已提交
373 374 375 376 377 378
	},
	{
		.procname	= "inet_peer_minttl",
		.data		= &inet_peer_minttl,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
379
		.proc_handler	= proc_dointvec_jiffies,
L
Linus Torvalds 已提交
380 381 382 383 384 385
	},
	{
		.procname	= "inet_peer_maxttl",
		.data		= &inet_peer_maxttl,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
386
		.proc_handler	= proc_dointvec_jiffies,
L
Linus Torvalds 已提交
387 388 389 390 391 392
	},
	{
		.procname	= "tcp_fack",
		.data		= &sysctl_tcp_fack,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
393
		.proc_handler	= proc_dointvec
L
Linus Torvalds 已提交
394
	},
Y
Yuchung Cheng 已提交
395 396 397 398 399 400 401
	{
		.procname	= "tcp_recovery",
		.data		= &sysctl_tcp_recovery,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
402 403 404 405 406 407 408
	{
		.procname	= "tcp_max_reordering",
		.data		= &sysctl_tcp_max_reordering,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
L
Linus Torvalds 已提交
409 410 411 412 413
	{
		.procname	= "tcp_dsack",
		.data		= &sysctl_tcp_dsack,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
414
		.proc_handler	= proc_dointvec
L
Linus Torvalds 已提交
415
	},
416 417 418 419 420 421 422
	{
		.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
423 424 425 426 427
	{
		.procname	= "tcp_wmem",
		.data		= &sysctl_tcp_wmem,
		.maxlen		= sizeof(sysctl_tcp_wmem),
		.mode		= 0644,
428
		.proc_handler	= proc_dointvec_minmax,
429
		.extra1		= &one,
L
Linus Torvalds 已提交
430 431 432 433 434 435
	},
	{
		.procname	= "tcp_rmem",
		.data		= &sysctl_tcp_rmem,
		.maxlen		= sizeof(sysctl_tcp_rmem),
		.mode		= 0644,
436
		.proc_handler	= proc_dointvec_minmax,
437
		.extra1		= &one,
L
Linus Torvalds 已提交
438 439 440 441 442 443
	},
	{
		.procname	= "tcp_app_win",
		.data		= &sysctl_tcp_app_win,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
444
		.proc_handler	= proc_dointvec
L
Linus Torvalds 已提交
445 446 447 448 449 450
	},
	{
		.procname	= "tcp_adv_win_scale",
		.data		= &sysctl_tcp_adv_win_scale,
		.maxlen		= sizeof(int),
		.mode		= 0644,
451 452 453
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &tcp_adv_win_scale_min,
		.extra2		= &tcp_adv_win_scale_max,
L
Linus Torvalds 已提交
454 455 456 457 458
	},
	{
		.procname	= "tcp_frto",
		.data		= &sysctl_tcp_frto,
		.maxlen		= sizeof(int),
459 460 461 462 463 464 465
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
	{
		.procname	= "tcp_min_rtt_wlen",
		.data		= &sysctl_tcp_min_rtt_wlen,
		.maxlen		= sizeof(int),
L
Linus Torvalds 已提交
466
		.mode		= 0644,
A
Alexey Dobriyan 已提交
467
		.proc_handler	= proc_dointvec
L
Linus Torvalds 已提交
468 469 470 471 472 473
	},
	{
		.procname	= "tcp_low_latency",
		.data		= &sysctl_tcp_low_latency,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
474
		.proc_handler	= proc_dointvec
L
Linus Torvalds 已提交
475 476 477 478 479 480
	},
	{
		.procname	= "tcp_no_metrics_save",
		.data		= &sysctl_tcp_nometrics_save,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
481
		.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
482 483 484 485 486 487
	},
	{
		.procname	= "tcp_moderate_rcvbuf",
		.data		= &sysctl_tcp_moderate_rcvbuf,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
488
		.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
489 490 491 492 493 494
	},
	{
		.procname	= "tcp_tso_win_divisor",
		.data		= &sysctl_tcp_tso_win_divisor,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
495
		.proc_handler	= proc_dointvec,
L
Linus Torvalds 已提交
496 497
	},
	{
498
		.procname	= "tcp_congestion_control",
L
Linus Torvalds 已提交
499
		.mode		= 0644,
500
		.maxlen		= TCP_CA_NAME_MAX,
A
Alexey Dobriyan 已提交
501
		.proc_handler	= proc_tcp_congestion_control,
L
Linus Torvalds 已提交
502
	},
503
	{
504 505 506 507
		.procname	= "tcp_workaround_signed_windows",
		.data		= &sysctl_tcp_workaround_signed_windows,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
508
		.proc_handler	= proc_dointvec
509
	},
E
Eric Dumazet 已提交
510 511 512 513 514 515 516
	{
		.procname	= "tcp_limit_output_bytes",
		.data		= &sysctl_tcp_limit_output_bytes,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
E
Eric Dumazet 已提交
517 518 519 520 521 522 523
	{
		.procname	= "tcp_challenge_ack_limit",
		.data		= &sysctl_tcp_challenge_ack_limit,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
524 525 526 527 528
	{
		.procname	= "tcp_slow_start_after_idle",
		.data		= &sysctl_tcp_slow_start_after_idle,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
529
		.proc_handler	= proc_dointvec
530
	},
P
Paul Moore 已提交
531 532 533 534 535 536
#ifdef CONFIG_NETLABEL
	{
		.procname	= "cipso_cache_enable",
		.data		= &cipso_v4_cache_enabled,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
537
		.proc_handler	= proc_dointvec,
P
Paul Moore 已提交
538 539 540 541 542 543
	},
	{
		.procname	= "cipso_cache_bucket_size",
		.data		= &cipso_v4_cache_bucketsize,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
544
		.proc_handler	= proc_dointvec,
P
Paul Moore 已提交
545 546 547 548 549 550
	},
	{
		.procname	= "cipso_rbm_optfmt",
		.data		= &cipso_v4_rbm_optfmt,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
551
		.proc_handler	= proc_dointvec,
P
Paul Moore 已提交
552 553 554 555 556 557
	},
	{
		.procname	= "cipso_rbm_strictvalid",
		.data		= &cipso_v4_rbm_strictvalid,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
558
		.proc_handler	= proc_dointvec,
P
Paul Moore 已提交
559 560
	},
#endif /* CONFIG_NETLABEL */
561 562 563 564
	{
		.procname	= "tcp_available_congestion_control",
		.maxlen		= TCP_CA_BUF_MAX,
		.mode		= 0444,
A
Alexey Dobriyan 已提交
565
		.proc_handler   = proc_tcp_available_congestion_control,
566
	},
567 568 569 570
	{
		.procname	= "tcp_allowed_congestion_control",
		.maxlen		= TCP_CA_BUF_MAX,
		.mode		= 0644,
A
Alexey Dobriyan 已提交
571
		.proc_handler   = proc_allowed_congestion_control,
572
	},
A
Andreas Petlund 已提交
573 574 575 576 577 578 579
	{
		.procname       = "tcp_thin_linear_timeouts",
		.data           = &sysctl_tcp_thin_linear_timeouts,
		.maxlen         = sizeof(int),
		.mode           = 0644,
		.proc_handler   = proc_dointvec
	},
Y
Yuchung Cheng 已提交
580 581 582 583 584 585 586
	{
		.procname	= "tcp_early_retrans",
		.data		= &sysctl_tcp_early_retrans,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &zero,
N
Nandita Dukkipati 已提交
587
		.extra2		= &four,
Y
Yuchung Cheng 已提交
588
	},
E
Eric Dumazet 已提交
589 590 591 592 593 594
	{
		.procname	= "tcp_min_tso_segs",
		.data		= &sysctl_tcp_min_tso_segs,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
595
		.extra1		= &one,
E
Eric Dumazet 已提交
596 597
		.extra2		= &gso_max_segs,
	},
598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615
	{
		.procname	= "tcp_pacing_ss_ratio",
		.data		= &sysctl_tcp_pacing_ss_ratio,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &zero,
		.extra2		= &thousand,
	},
	{
		.procname	= "tcp_pacing_ca_ratio",
		.data		= &sysctl_tcp_pacing_ca_ratio,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &zero,
		.extra2		= &thousand,
	},
E
Eric Dumazet 已提交
616 617 618 619 620 621 622 623 624
	{
		.procname	= "tcp_autocorking",
		.data		= &sysctl_tcp_autocorking,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &zero,
		.extra2		= &one,
	},
625 626 627 628 629 630 631
	{
		.procname	= "tcp_invalid_ratelimit",
		.data		= &sysctl_tcp_invalid_ratelimit,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_ms_jiffies,
	},
632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647
	{
		.procname	= "icmp_msgs_per_sec",
		.data		= &sysctl_icmp_msgs_per_sec,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &zero,
	},
	{
		.procname	= "icmp_msgs_burst",
		.data		= &sysctl_icmp_msgs_burst,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &zero,
	},
H
Hideo Aoki 已提交
648 649 650 651 652
	{
		.procname	= "udp_mem",
		.data		= &sysctl_udp_mem,
		.maxlen		= sizeof(sysctl_udp_mem),
		.mode		= 0644,
E
Eric Dumazet 已提交
653
		.proc_handler	= proc_doulongvec_minmax,
H
Hideo Aoki 已提交
654 655 656 657 658 659
	},
	{
		.procname	= "udp_rmem_min",
		.data		= &sysctl_udp_rmem_min,
		.maxlen		= sizeof(sysctl_udp_rmem_min),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
660
		.proc_handler	= proc_dointvec_minmax,
661
		.extra1		= &one
H
Hideo Aoki 已提交
662 663 664 665 666 667
	},
	{
		.procname	= "udp_wmem_min",
		.data		= &sysctl_udp_wmem_min,
		.maxlen		= sizeof(sysctl_udp_wmem_min),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
668
		.proc_handler	= proc_dointvec_minmax,
669
		.extra1		= &one
H
Hideo Aoki 已提交
670
	},
671
	{ }
L
Linus Torvalds 已提交
672
};
673

674 675 676 677 678 679
static struct ctl_table ipv4_net_table[] = {
	{
		.procname	= "icmp_echo_ignore_all",
		.data		= &init_net.ipv4.sysctl_icmp_echo_ignore_all,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
680
		.proc_handler	= proc_dointvec
681 682 683 684 685 686
	},
	{
		.procname	= "icmp_echo_ignore_broadcasts",
		.data		= &init_net.ipv4.sysctl_icmp_echo_ignore_broadcasts,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
687
		.proc_handler	= proc_dointvec
688 689 690 691 692 693
	},
	{
		.procname	= "icmp_ignore_bogus_error_responses",
		.data		= &init_net.ipv4.sysctl_icmp_ignore_bogus_error_responses,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
694
		.proc_handler	= proc_dointvec
695 696 697 698 699 700
	},
	{
		.procname	= "icmp_errors_use_inbound_ifaddr",
		.data		= &init_net.ipv4.sysctl_icmp_errors_use_inbound_ifaddr,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
701
		.proc_handler	= proc_dointvec
702 703 704 705 706 707
	},
	{
		.procname	= "icmp_ratelimit",
		.data		= &init_net.ipv4.sysctl_icmp_ratelimit,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
708
		.proc_handler	= proc_dointvec_ms_jiffies,
709 710 711 712 713 714
	},
	{
		.procname	= "icmp_ratemask",
		.data		= &init_net.ipv4.sysctl_icmp_ratemask,
		.maxlen		= sizeof(int),
		.mode		= 0644,
A
Alexey Dobriyan 已提交
715
		.proc_handler	= proc_dointvec
716
	},
717 718
	{
		.procname	= "ping_group_range",
719
		.data		= &init_net.ipv4.ping_group_range.range,
720
		.maxlen		= sizeof(gid_t)*2,
721 722 723
		.mode		= 0644,
		.proc_handler	= ipv4_ping_group_range,
	},
724 725 726 727 728 729 730
	{
		.procname	= "tcp_ecn",
		.data		= &init_net.ipv4.sysctl_tcp_ecn,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
731 732 733 734 735 736 737
	{
		.procname	= "tcp_ecn_fallback",
		.data		= &init_net.ipv4.sysctl_tcp_ecn_fallback,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
738 739 740 741 742 743 744
	{
		.procname	= "ip_dynaddr",
		.data		= &init_net.ipv4.sysctl_ip_dynaddr,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
745 746 747 748 749 750 751
	{
		.procname	= "ip_early_demux",
		.data		= &init_net.ipv4.sysctl_ip_early_demux,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
752 753 754 755 756 757 758 759 760
	{
		.procname	= "ip_default_ttl",
		.data		= &init_net.ipv4.sysctl_ip_default_ttl,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &ip_ttl_min,
		.extra2		= &ip_ttl_max,
	},
761 762
	{
		.procname	= "ip_local_port_range",
763 764
		.maxlen		= sizeof(init_net.ipv4.ip_local_ports.range),
		.data		= &init_net.ipv4.ip_local_ports.range,
765 766 767
		.mode		= 0644,
		.proc_handler	= ipv4_local_port_range,
	},
768 769 770 771 772 773 774
	{
		.procname	= "ip_local_reserved_ports",
		.data		= &init_net.ipv4.sysctl_local_reserved_ports,
		.maxlen		= 65536,
		.mode		= 0644,
		.proc_handler	= proc_do_large_bitmap,
	},
775 776 777 778 779 780 781
	{
		.procname	= "ip_no_pmtu_disc",
		.data		= &init_net.ipv4.sysctl_ip_no_pmtu_disc,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
782 783 784 785 786 787 788
	{
		.procname	= "ip_forward_use_pmtu",
		.data		= &init_net.ipv4.sysctl_ip_fwd_use_pmtu,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
789 790 791 792 793 794 795
	{
		.procname	= "ip_nonlocal_bind",
		.data		= &init_net.ipv4.sysctl_ip_nonlocal_bind,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
796 797 798 799 800 801 802
	{
		.procname	= "fwmark_reflect",
		.data		= &init_net.ipv4.sysctl_fwmark_reflect,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
803 804 805 806 807 808 809
	{
		.procname	= "tcp_fwmark_accept",
		.data		= &init_net.ipv4.sysctl_tcp_fwmark_accept,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
810 811 812 813 814 815 816 817 818 819 820
#ifdef CONFIG_NET_L3_MASTER_DEV
	{
		.procname	= "tcp_l3mdev_accept",
		.data		= &init_net.ipv4.sysctl_tcp_l3mdev_accept,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &zero,
		.extra2		= &one,
	},
#endif
F
Fan Du 已提交
821 822 823 824 825 826 827 828 829 830 831 832 833 834
	{
		.procname	= "tcp_mtu_probing",
		.data		= &init_net.ipv4.sysctl_tcp_mtu_probing,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "tcp_base_mss",
		.data		= &init_net.ipv4.sysctl_tcp_base_mss,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
835 836 837 838 839 840 841
	{
		.procname	= "tcp_probe_threshold",
		.data		= &init_net.ipv4.sysctl_tcp_probe_threshold,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
842 843 844 845 846 847 848
	{
		.procname	= "tcp_probe_interval",
		.data		= &init_net.ipv4.sysctl_tcp_probe_interval,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
849 850
	{
		.procname	= "igmp_link_local_mcast_reports",
851
		.data		= &init_net.ipv4.sysctl_igmp_llm_reports,
852 853 854 855
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
856 857 858 859 860 861 862
	{
		.procname	= "igmp_max_memberships",
		.data		= &init_net.ipv4.sysctl_igmp_max_memberships,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
863 864 865 866 867 868 869
	{
		.procname	= "igmp_max_msf",
		.data		= &init_net.ipv4.sysctl_igmp_max_msf,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
870 871 872 873 874 875 876 877 878 879
#ifdef CONFIG_IP_MULTICAST
	{
		.procname	= "igmp_qrv",
		.data		= &init_net.ipv4.sysctl_igmp_qrv,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &one
	},
#endif
880 881 882 883 884 885 886
	{
		.procname	= "tcp_keepalive_time",
		.data		= &init_net.ipv4.sysctl_tcp_keepalive_time,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_jiffies,
	},
887 888 889 890 891 892 893
	{
		.procname	= "tcp_keepalive_probes",
		.data		= &init_net.ipv4.sysctl_tcp_keepalive_probes,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
894 895 896 897 898 899 900
	{
		.procname	= "tcp_keepalive_intvl",
		.data		= &init_net.ipv4.sysctl_tcp_keepalive_intvl,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_jiffies,
	},
901 902 903 904 905 906 907 908 909
	{
		.procname	= "tcp_syn_retries",
		.data		= &init_net.ipv4.sysctl_tcp_syn_retries,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &tcp_syn_retries_min,
		.extra2		= &tcp_syn_retries_max
	},
910 911 912 913 914 915 916
	{
		.procname	= "tcp_synack_retries",
		.data		= &init_net.ipv4.sysctl_tcp_synack_retries,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
917 918 919 920 921 922 923 924 925
#ifdef CONFIG_SYN_COOKIES
	{
		.procname	= "tcp_syncookies",
		.data		= &init_net.ipv4.sysctl_tcp_syncookies,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
#endif
926 927 928 929 930 931 932
	{
		.procname	= "tcp_reordering",
		.data		= &init_net.ipv4.sysctl_tcp_reordering,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
933 934 935 936 937 938 939 940
	{
		.procname	= "tcp_retries1",
		.data		= &init_net.ipv4.sysctl_tcp_retries1,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra2		= &tcp_retr1_max
	},
941 942 943 944 945 946 947
	{
		.procname	= "tcp_retries2",
		.data		= &init_net.ipv4.sysctl_tcp_retries2,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
948 949 950 951 952 953 954
	{
		.procname	= "tcp_orphan_retries",
		.data		= &init_net.ipv4.sysctl_tcp_orphan_retries,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
955 956 957 958 959 960 961
	{
		.procname	= "tcp_fin_timeout",
		.data		= &init_net.ipv4.sysctl_tcp_fin_timeout,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_jiffies,
	},
962 963 964 965 966
	{
		.procname	= "tcp_notsent_lowat",
		.data		= &init_net.ipv4.sysctl_tcp_notsent_lowat,
		.maxlen		= sizeof(unsigned int),
		.mode		= 0644,
967
		.proc_handler	= proc_douintvec,
968
	},
969 970 971 972 973 974 975
	{
		.procname	= "tcp_tw_reuse",
		.data		= &init_net.ipv4.sysctl_tcp_tw_reuse,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
976 977 978 979 980 981 982
	{
		.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
	},
983 984 985 986 987 988 989
	{
		.procname	= "tcp_max_syn_backlog",
		.data		= &init_net.ipv4.sysctl_max_syn_backlog,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
990 991 992 993 994 995 996 997 998 999 1000
#ifdef CONFIG_IP_ROUTE_MULTIPATH
	{
		.procname	= "fib_multipath_use_neigh",
		.data		= &init_net.ipv4.sysctl_fib_multipath_use_neigh,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &zero,
		.extra2		= &one,
	},
#endif
1001 1002 1003 1004 1005 1006 1007
	{
		.procname	= "ip_unprivileged_port_start",
		.maxlen		= sizeof(int),
		.data		= &init_net.ipv4.sysctl_ip_prot_sock,
		.mode		= 0644,
		.proc_handler	= ipv4_privileged_ports,
	},
1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018
#ifdef CONFIG_NET_L3_MASTER_DEV
	{
		.procname	= "udp_l3mdev_accept",
		.data		= &init_net.ipv4.sysctl_udp_l3mdev_accept,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &zero,
		.extra2		= &one,
	},
#endif
1019 1020 1021
	{ }
};

1022 1023
static __net_init int ipv4_sysctl_init_net(struct net *net)
{
1024 1025 1026
	struct ctl_table *table;

	table = ipv4_net_table;
O
Octavian Purdila 已提交
1027
	if (!net_eq(net, &init_net)) {
1028 1029
		int i;

1030
		table = kmemdup(table, sizeof(ipv4_net_table), GFP_KERNEL);
1031
		if (!table)
1032 1033
			goto err_alloc;

1034 1035 1036
		/* Update the variables to point into the current struct net */
		for (i = 0; i < ARRAY_SIZE(ipv4_net_table) - 1; i++)
			table[i].data += (void *)net - (void *)&init_net;
1037 1038
	}

1039
	net->ipv4.ipv4_hdr = register_net_sysctl(net, "net/ipv4", table);
1040
	if (!net->ipv4.ipv4_hdr)
1041 1042
		goto err_reg;

1043 1044 1045 1046
	net->ipv4.sysctl_local_reserved_ports = kzalloc(65536 / 8, GFP_KERNEL);
	if (!net->ipv4.sysctl_local_reserved_ports)
		goto err_ports;

1047
	return 0;
1048

1049 1050
err_ports:
	unregister_net_sysctl_table(net->ipv4.ipv4_hdr);
1051
err_reg:
O
Octavian Purdila 已提交
1052
	if (!net_eq(net, &init_net))
1053 1054 1055
		kfree(table);
err_alloc:
	return -ENOMEM;
1056 1057 1058 1059
}

static __net_exit void ipv4_sysctl_exit_net(struct net *net)
{
1060 1061
	struct ctl_table *table;

1062
	kfree(net->ipv4.sysctl_local_reserved_ports);
1063 1064 1065
	table = net->ipv4.ipv4_hdr->ctl_table_arg;
	unregister_net_sysctl_table(net->ipv4.ipv4_hdr);
	kfree(table);
1066 1067 1068 1069 1070 1071 1072
}

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

1073 1074 1075 1076
static __init int sysctl_ipv4_init(void)
{
	struct ctl_table_header *hdr;

1077
	hdr = register_net_sysctl(&init_net, "net/ipv4", ipv4_table);
1078
	if (!hdr)
1079 1080 1081
		return -ENOMEM;

	if (register_pernet_subsys(&ipv4_sysctl_ops)) {
1082
		unregister_net_sysctl_table(hdr);
1083 1084 1085 1086
		return -ENOMEM;
	}

	return 0;
1087 1088 1089
}

__initcall(sysctl_ipv4_init);