aboutsummaryrefslogtreecommitdiffstats
path: root/sys/netinet/ip_state.c
blob: a6bda8a170ede3c06b36286ddf7396a700fa5776 (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
/*
 * (C)opyright 1995 by Darren Reed.
 *
 * Redistribution and use in source and binary forms are permitted
 * provided that this notice is preserved and due credit is given
 * to the original author and the contributors.
 */
#if !defined(lint) && defined(LIBC_SCCS)
static	char	sccsid[] = "@(#)ip_state.c	1.8 6/5/96 (C) 1993-1995 Darren Reed";
static	char	rcsid[] = "$Id: ip_state.c,v 2.0.2.12 1997/05/24 07:34:10 darrenr Exp $";
#endif

#if !defined(_KERNEL) && !defined(KERNEL)
# include <stdlib.h>
# include <string.h>
#endif
#include <sys/errno.h>
#include <sys/types.h>
#include <sys/param.h>
#include <sys/time.h>
#include <sys/file.h>
#if defined(KERNEL) && (__FreeBSD_version >= 220000)
# include <sys/filio.h>
# include <sys/fcntl.h>
#else
# include <sys/ioctl.h>
#endif
#include <sys/uio.h>
#include <sys/protosw.h>
#include <sys/socket.h>
#ifdef _KERNEL
# include <sys/systm.h>
#endif
#if !defined(__SVR4) && !defined(__svr4__)
# include <sys/mbuf.h>
#else
# include <sys/filio.h>
# include <sys/byteorder.h>
# include <sys/dditypes.h>
# include <sys/stream.h>
# include <sys/kmem.h>
#endif

#include <net/if.h>
#ifdef sun
#include <net/af.h>
#endif
#include <net/route.h>
#include <netinet/in.h>
#include <netinet/in_systm.h>
#include <netinet/ip.h>
#include <netinet/ip_var.h>
#include <netinet/tcp.h>
#include <netinet/tcp_fsm.h>
#include <netinet/udp.h>
#include <netinet/tcpip.h>
#include <netinet/ip_icmp.h>
#include "netinet/ip_compat.h"
#include "netinet/ip_fil.h"
#include "netinet/ip_nat.h"
#include "netinet/ip_state.h"
#ifndef	MIN
#define	MIN(a,b)	(((a)<(b))?(a):(b))
#endif

#define	TCP_CLOSE	(TH_FIN|TH_RST)

ipstate_t *ips_table[IPSTATE_SIZE];
int	ips_num = 0;
ips_stat_t ips_stats;
#if	SOLARIS && defined(_KERNEL)
extern	kmutex_t	ipf_state;
#endif


#define	FIVE_DAYS	(2 * 5 * 86400)	/* 5 days: half closed session */

u_long	fr_tcpidletimeout = FIVE_DAYS,
	fr_tcpclosewait = 60,
	fr_tcplastack = 20,
	fr_tcptimeout = 120,
	fr_tcpclosed = 1,
	fr_udptimeout = 120,
	fr_icmptimeout = 120;


ips_stat_t *fr_statetstats()
{
	ips_stats.iss_active = ips_num;
	ips_stats.iss_table = ips_table;
	return &ips_stats;
}


int fr_state_ioctl(data, cmd, mode)
caddr_t data;
int cmd;
int mode;
{
	switch (cmd)
	{
	case SIOCGIPST :
		IWCOPY((caddr_t)fr_statetstats(), data, sizeof(ips_stat_t));
		break;
	case FIONREAD :
#ifdef	IPFILTER_LOG
		*(int *)data = iplused[IPL_LOGSTATE];
#endif
		break;
	default :
		return -1;
	}
	return 0;
}


/*
 * Create a new ipstate structure and hang it off the hash table.
 */
int fr_addstate(ip, fin, pass)
ip_t *ip;
fr_info_t *fin;
u_int pass;
{
	ipstate_t ips;
	register ipstate_t *is = &ips;
	register u_int hv;

	if ((ip->ip_off & 0x1fff) || (fin->fin_fi.fi_fl & FI_SHORT))
		return -1;
	if (ips_num == IPSTATE_MAX) {
		ips_stats.iss_max++;
		return -1;
	}
	ips.is_age = 1;
	ips.is_state[0] = 0;
	ips.is_state[1] = 0;
	/*
	 * Copy and calculate...
	 */
	hv = (is->is_p = ip->ip_p);
	hv += (is->is_src.s_addr = ip->ip_src.s_addr);
	hv += (is->is_dst.s_addr = ip->ip_dst.s_addr);

	switch (ip->ip_p)
	{
	case IPPROTO_ICMP :
	    {
		struct icmp *ic = (struct icmp *)fin->fin_dp;

		switch (ic->icmp_type)
		{
		case ICMP_ECHO :
			is->is_icmp.ics_type = 0;
			hv += (is->is_icmp.ics_id = ic->icmp_id);
			hv += (is->is_icmp.ics_seq = ic->icmp_seq);
			break;
		case ICMP_TSTAMP :
		case ICMP_IREQ :
		case ICMP_MASKREQ :
			is->is_icmp.ics_type = ic->icmp_type + 1;
			break;
		default :
			return -1;
		}
		ips_stats.iss_icmp++;
		is->is_age = fr_icmptimeout;
		break;
	    }
	case IPPROTO_TCP :
	    {
		register tcphdr_t *tcp = (tcphdr_t *)fin->fin_dp;

		/*
		 * The endian of the ports doesn't matter, but the ack and
		 * sequence numbers do as we do mathematics on them later.
		 */
		hv += (is->is_dport = tcp->th_dport);
		hv += (is->is_sport = tcp->th_sport);
		is->is_seq = ntohl(tcp->th_seq);
		is->is_ack = ntohl(tcp->th_ack);
		is->is_swin = ntohs(tcp->th_win);
		is->is_dwin = is->is_swin;	/* start them the same */
		ips_stats.iss_tcp++;
		/*
		 * If we're creating state for a starting connectoin, start the
		 * timer on it as we'll never see an error if it fails to
		 * connect.
		 */
		if ((tcp->th_flags & (TH_SYN|TH_ACK)) == TH_SYN)
			is->is_ack = 0;	/* Trumpet WinSock 'ism */
		fr_tcp_age(&is->is_age, is->is_state, ip, fin,
			   tcp->th_sport == is->is_sport);
		break;
	    }
	case IPPROTO_UDP :
	    {
		register tcphdr_t *tcp = (tcphdr_t *)fin->fin_dp;

		hv += (is->is_dport = tcp->th_dport);
		hv += (is->is_sport = tcp->th_sport);
		ips_stats.iss_udp++;
		is->is_age = fr_udptimeout;
		break;
	    }
	default :
		return -1;
	}

	KMALLOC(is, ipstate_t *, sizeof(*is));
	if (is == NULL) {
		ips_stats.iss_nomem++;
		return -1;
	}
	bcopy((char *)&ips, (char *)is, sizeof(*is));
	hv %= IPSTATE_SIZE;
	MUTEX_ENTER(&ipf_state);
	is->is_next = ips_table[hv];
	ips_table[hv] = is;
	is->is_pass = pass;
	is->is_pkts = 1;
	is->is_bytes = ip->ip_len;
	if (pass & FR_LOGFIRST)
		is->is_pass &= ~(FR_LOGFIRST|FR_LOG);
	ips_num++;
#ifdef	IPFILTER_LOG
	ipstate_log(is, ISL_NEW);
#endif
	MUTEX_EXIT(&ipf_state);
	if (fin->fin_fi.fi_fl & FI_FRAG)
		ipfr_newfrag(ip, fin, pass ^ FR_KEEPSTATE);
	return 0;
}


/*
 * check to see if a packet with TCP headers fits within the TCP window.
 * change timeout depending on whether new packet is a SYN-ACK returning for a
 * SYN or a RST or FIN which indicate time to close up shop.
 */
int fr_tcpstate(is, fin, ip, tcp, sport)
register ipstate_t *is;
fr_info_t *fin;
ip_t *ip;
tcphdr_t *tcp;
u_short sport;
{
	register int seqskew, ackskew;
	register u_short swin, dwin;
	register tcp_seq seq, ack;
	int source;

	/*
	 * Find difference between last checked packet and this packet.
	 */
	seq = ntohl(tcp->th_seq);
	ack = ntohl(tcp->th_ack);
	if (sport == is->is_sport) {
		seqskew = seq - is->is_seq;
		ackskew = ack - is->is_ack;
	} else {
		seqskew = ack - is->is_seq;
		if (!is->is_ack)
			/*
			 * Must be a SYN-ACK in reply to a SYN.
			 */
			is->is_ack = seq;
		ackskew = seq - is->is_ack;
	}

	/*
	 * Make skew values absolute
	 */
	if (seqskew < 0)
		seqskew = -seqskew;
	if (ackskew < 0)
		ackskew = -ackskew;

	/*
	 * If the difference in sequence and ack numbers is within the
	 * window size of the connection, store these values and match
	 * the packet.
	 */
	if ((source = (sport == is->is_sport))) {
		swin = is->is_swin;
		dwin = is->is_dwin;
	} else {
		dwin = is->is_swin;
		swin = is->is_dwin;
	}

	if ((seqskew <= swin) && (ackskew <= dwin)) {
		if (source) {
			is->is_seq = seq;
			is->is_ack = ack;
			is->is_swin = ntohs(tcp->th_win);
		} else {
			is->is_seq = ack;
			is->is_ack = seq;
			is->is_dwin = ntohs(tcp->th_win);
		}
		ips_stats.iss_hits++;
		is->is_pkts++;
		is->is_bytes += ip->ip_len;
		/*
		 * Nearing end of connection, start timeout.
		 */
		fr_tcp_age(&is->is_age, is->is_state, ip, fin,
			   tcp->th_sport == is->is_sport);
		return 1;
	}
	return 0;
}


/*
 * Check if a packet has a registered state.
 */
int fr_checkstate(ip, fin)
ip_t *ip;
fr_info_t *fin;
{
	register struct in_addr dst, src;
	register ipstate_t *is, **isp;
	register u_char pr;
	struct icmp *ic;
	tcphdr_t *tcp;
	u_int hv, hlen, pass;

	if ((ip->ip_off & 0x1fff) || (fin->fin_fi.fi_fl & FI_SHORT))
		return 0;

	hlen = fin->fin_hlen;
	tcp = (tcphdr_t *)((char *)ip + hlen);
	ic = (struct icmp *)tcp;
	hv = (pr = ip->ip_p);
	hv += (src.s_addr = ip->ip_src.s_addr);
	hv += (dst.s_addr = ip->ip_dst.s_addr);

	/*
	 * Search the hash table for matching packet header info.
	 */
	switch (ip->ip_p)
	{
	case IPPROTO_ICMP :
		hv += ic->icmp_id;
		hv += ic->icmp_seq;
		hv %= IPSTATE_SIZE;
		MUTEX_ENTER(&ipf_state);
		for (isp = &ips_table[hv]; (is = *isp); isp = &is->is_next)
			if ((is->is_p == pr) &&
			    (ic->icmp_id == is->is_icmp.ics_id) &&
			    (ic->icmp_seq == is->is_icmp.ics_seq) &&
			    IPPAIR(src, dst, is->is_src, is->is_dst)) {
				/*
				 * If we have type 0 stored, allow any icmp
				 * replies through.
				 */
				if (is->is_icmp.ics_type &&
				    is->is_icmp.ics_type != ic->icmp_type)
					continue;
				is->is_age = fr_icmptimeout;
				is->is_pkts++;
				is->is_bytes += ip->ip_len;
				ips_stats.iss_hits++;
				pass = is->is_pass;
				MUTEX_EXIT(&ipf_state);
				return pass;
			}
		MUTEX_EXIT(&ipf_state);
		break;
	case IPPROTO_TCP :
	    {
		register u_short dport = tcp->th_dport, sport = tcp->th_sport;

		hv += dport;
		hv += sport;
		hv %= IPSTATE_SIZE;
		MUTEX_ENTER(&ipf_state);
		for (isp = &ips_table[hv]; (is = *isp); isp = &is->is_next) {
			if ((is->is_p == pr) &&
			    PAIRS(sport, dport, is->is_sport, is->is_dport) &&
			    IPPAIR(src, dst, is->is_src, is->is_dst))
				if (fr_tcpstate(is, fin, ip, tcp, sport)) {
					pass = is->is_pass;
#ifdef	_KERNEL
					MUTEX_EXIT(&ipf_state);
#else

					if (tcp->th_flags & TCP_CLOSE) {
						*isp = is->is_next;
						isp = &ips_table[hv];
						KFREE(is);
					}
#endif
					return pass;
				}
		}
		MUTEX_EXIT(&ipf_state);
		break;
	    }
	case IPPROTO_UDP :
	    {
		register u_short dport = tcp->th_dport, sport = tcp->th_sport;

		hv += dport;
		hv += sport;
		hv %= IPSTATE_SIZE;
		/*
		 * Nothing else to match on but ports. and IP#'s
		 */
		MUTEX_ENTER(&ipf_state);
		for (is = ips_table[hv]; is; is = is->is_next)
			if ((is->is_p == pr) &&
			    PAIRS(sport, dport, is->is_sport, is->is_dport) &&
			    IPPAIR(src, dst, is->is_src, is->is_dst)) {
				ips_stats.iss_hits++;
				is->is_pkts++;
				is->is_bytes += ip->ip_len;
				is->is_age = fr_udptimeout;
				pass = is->is_pass;
				MUTEX_EXIT(&ipf_state);
				return pass;
			}
		MUTEX_EXIT(&ipf_state);
		break;
	    }
	default :
		break;
	}
	ips_stats.iss_miss++;
	return 0;
}


/*
 * Free memory in use by all state info. kept.
 */
void fr_stateunload()
{
	register int i;
	register ipstate_t *is, **isp;
	int s;

	MUTEX_ENTER(&ipf_state);
	SPLNET(s);
	for (i = 0; i < IPSTATE_SIZE; i++)
		for (isp = &ips_table[i]; (is = *isp); ) {
			*isp = is->is_next;
			KFREE(is);
		}
	SPLX(s);
	MUTEX_EXIT(&ipf_state);
}


/*
 * Slowly expire held state for thingslike UDP and ICMP.  Timeouts are set
 * in expectation of this being called twice per second.
 */
void fr_timeoutstate()
{
	register int i;
	register ipstate_t *is, **isp;
	int s;

	MUTEX_ENTER(&ipf_state);
	SPLNET(s);
	for (i = 0; i < IPSTATE_SIZE; i++)
		for (isp = &ips_table[i]; (is = *isp); )
			if (is->is_age && !--is->is_age) {
				*isp = is->is_next;
				if (is->is_p == IPPROTO_TCP)
					ips_stats.iss_fin++;
				else
					ips_stats.iss_expire++;
#ifdef	IPFILTER_LOG
				ipstate_log(is, ISL_EXPIRE);
#endif
				KFREE(is);
				ips_num--;
			} else
				isp = &is->is_next;
	SPLX(s);
	MUTEX_EXIT(&ipf_state);
}


/*
 * Original idea freom Pradeep Krishnan for use primarily with NAT code.
 * (pkrishna@netcom.com)
 */
void fr_tcp_age(age, state, ip, fin, dir)
u_long *age;
u_char *state;
ip_t *ip;
fr_info_t *fin;
int dir;
{
	tcphdr_t *tcp = (tcphdr_t *)fin->fin_dp;
	u_char flags = tcp->th_flags;
	int dlen, ostate;

	ostate = state[1 - dir];

	dlen = ip->ip_len - fin->fin_hlen - (tcp->th_off << 2);

	if (flags & TH_RST) {
		if (!(tcp->th_flags & TH_PUSH) && !dlen) {
			*age = fr_tcpclosed;
			state[dir] = TCPS_CLOSED;
		} else {
			*age = fr_tcpclosewait;
			state[dir] = TCPS_CLOSE_WAIT;
		}
		return;
	}

	*age = fr_tcptimeout; /* 1 min */

	switch(state[dir])
	{
	case TCPS_FIN_WAIT_2:
	case TCPS_CLOSED:
		if ((flags & TH_OPENING) == TH_OPENING)
			state[dir] = TCPS_SYN_RECEIVED;
		else if (flags & TH_SYN)
			state[dir] = TCPS_SYN_SENT;
		break;
	case TCPS_SYN_RECEIVED:
		if ((flags & (TH_FIN|TH_ACK)) == TH_ACK) {
			state[dir] = TCPS_ESTABLISHED;
			*age = fr_tcpidletimeout;
		}
		break;
	case TCPS_SYN_SENT:
		if ((flags & (TH_FIN|TH_ACK)) == TH_ACK) {
			state[dir] = TCPS_ESTABLISHED;
			*age = fr_tcpidletimeout;
		}
		break;
	case TCPS_ESTABLISHED:
		if (flags & TH_FIN) {
			state[dir] = TCPS_CLOSE_WAIT;
			if (!(flags & TH_PUSH) && !dlen &&
			    ostate > TCPS_ESTABLISHED)
				*age  = fr_tcplastack;
			else
				*age  = fr_tcpclosewait;
		} else
			*age = fr_tcpidletimeout;
		break;
	case TCPS_CLOSE_WAIT:
		if ((flags & TH_FIN) && !(flags & TH_PUSH) && !dlen &&
		    ostate > TCPS_ESTABLISHED) {
			*age  = fr_tcplastack;
			state[dir] = TCPS_LAST_ACK;
		} else
			*age  = fr_tcpclosewait;
		break;
	case TCPS_LAST_ACK:
		if (flags & TH_ACK) {
			state[dir] = TCPS_FIN_WAIT_2;
			if (!(flags & TH_PUSH) && !dlen &&
			    ostate > TCPS_ESTABLISHED)
				*age  = fr_tcplastack;
			else {
				*age  = fr_tcpclosewait;
				state[dir] = TCPS_CLOSE_WAIT;
			}
		}
		break;
	}
}


#ifdef	IPFILTER_LOG
void ipstate_log(is, type)
struct ipstate *is;
u_short type;
{
	struct	ipslog	ipsl;

	if (iplused[IPL_LOGSTATE] + sizeof(ipsl) > IPLLOGSIZE) {
		ips_stats.iss_logfail++;
		return;
	}

        if (iplh[IPL_LOGSTATE] == iplbuf[IPL_LOGSTATE] + IPLLOGSIZE)
                iplh[IPL_LOGSTATE] = iplbuf[IPL_LOGSTATE];

# ifdef	sun
	uniqtime(&ipsl);
# endif
# if BSD >= 199306 || defined(__FreeBSD__)
	microtime((struct timeval *)&ipsl);
# endif
	ipsl.isl_pkts = is->is_pkts;
	ipsl.isl_bytes = is->is_bytes;
	ipsl.isl_src = is->is_src;
	ipsl.isl_dst = is->is_dst;
	ipsl.isl_p = is->is_p;
	ipsl.isl_flags = is->is_flags;
	ipsl.isl_type = type;
	if (ipsl.isl_p == IPPROTO_TCP || ipsl.isl_p == IPPROTO_UDP) {
		ipsl.isl_sport = is->is_sport;
		ipsl.isl_dport = is->is_dport;
	} else if (ipsl.isl_p == IPPROTO_ICMP)
		ipsl.isl_itype = is->is_icmp.ics_type;
	else {
		ipsl.isl_ps.isl_filler[0] = 0;
		ipsl.isl_ps.isl_filler[1] = 0;
	}

	if (!fr_copytolog(IPL_LOGSTATE, (char *)&ipsl, sizeof(ipsl))) {
		iplused[IPL_LOGSTATE] += sizeof(ipsl);
		ips_stats.iss_logged++;
	} else
		ips_stats.iss_logfail++;
	wakeup(iplbuf[IPL_LOGSTATE]);
}
#endif