1 /*-
2 * SPDX-License-Identifier: BSD-3-Clause
3 *
4 * Copyright (c) 1982, 1986, 1993
5 * The Regents of the University of California. All rights reserved.
6 *
7 * Redistribution and use in source and binary forms, with or without
8 * modification, are permitted provided that the following conditions
9 * are met:
10 * 1. Redistributions of source code must retain the above copyright
11 * notice, this list of conditions and the following disclaimer.
12 * 2. Redistributions in binary form must reproduce the above copyright
13 * notice, this list of conditions and the following disclaimer in the
14 * documentation and/or other materials provided with the distribution.
15 * 3. Neither the name of the University nor the names of its contributors
16 * may be used to endorse or promote products derived from this software
17 * without specific prior written permission.
18 *
19 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
20 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
21 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
22 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
23 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
24 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
25 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
26 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
27 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
28 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
29 * SUCH DAMAGE.
30 */
31
32 #ifndef _NETINET_IP_VAR_H_
33 #define _NETINET_IP_VAR_H_
34
35 #include <sys/epoch.h>
36 #include <sys/queue.h>
37 #include <sys/types.h>
38
39 #include <netinet/in.h>
40
41 /*
42 * Overlay for ip header used by other protocols (tcp, udp).
43 */
44 struct ipovly {
45 u_char ih_x1[9]; /* (unused) */
46 u_char ih_pr; /* protocol */
47 u_short ih_len; /* protocol length */
48 struct in_addr ih_src; /* source internet address */
49 struct in_addr ih_dst; /* destination internet address */
50 } __packed;
51
52 #ifdef _KERNEL
53 /*
54 * Ip reassembly queue structure. Each fragment
55 * being reassembled is attached to one of these structures.
56 * They are timed out after net.inet.ip.fragttl seconds, and may also be
57 * reclaimed if memory becomes tight.
58 */
59 struct ipq {
60 TAILQ_ENTRY(ipq) ipq_list; /* to other reass headers */
61 time_t ipq_expire; /* time_uptime when ipq expires */
62 u_char ipq_nfrags; /* # frags in this packet */
63 u_char ipq_p; /* protocol of this fragment */
64 u_short ipq_id; /* sequence id for reassembly */
65 int ipq_maxoff; /* total length of packet */
66 struct mbuf *ipq_frags; /* to ip headers of fragments */
67 struct in_addr ipq_src,ipq_dst;
68 struct label *ipq_label; /* MAC label */
69 };
70 #endif /* _KERNEL */
71
72 /*
73 * Structure stored in mbuf in inpcb.ip_options
74 * and passed to ip_output when ip options are in use.
75 * The actual length of the options (including ipopt_dst)
76 * is in m_len.
77 */
78 #define MAX_IPOPTLEN 40
79
80 struct ipoption {
81 struct in_addr ipopt_dst; /* first-hop dst if source routed */
82 char ipopt_list[MAX_IPOPTLEN]; /* options proper */
83 };
84
85 struct ipstat {
86 uint64_t ips_total; /* total packets received */
87 uint64_t ips_badsum; /* checksum bad */
88 uint64_t ips_tooshort; /* packet too short */
89 uint64_t ips_toosmall; /* not enough data */
90 uint64_t ips_badhlen; /* ip header length < data size */
91 uint64_t ips_badlen; /* ip length < ip header length */
92 uint64_t ips_fragments; /* fragments received */
93 uint64_t ips_fragdropped; /* frags dropped (dups, out of space) */
94 uint64_t ips_fragtimeout; /* fragments timed out */
95 uint64_t ips_forward; /* packets forwarded */
96 uint64_t ips_fastforward; /* packets fast forwarded */
97 uint64_t ips_cantforward; /* packets rcvd for unreachable dest */
98 uint64_t ips_redirectsent; /* packets forwarded on same net */
99 uint64_t ips_noproto; /* unknown or unsupported protocol */
100 uint64_t ips_delivered; /* datagrams delivered to upper level*/
101 uint64_t ips_localout; /* total ip packets generated here */
102 uint64_t ips_odropped; /* lost packets due to nobufs, etc. */
103 uint64_t ips_reassembled; /* total packets reassembled ok */
104 uint64_t ips_fragmented; /* datagrams successfully fragmented */
105 uint64_t ips_ofragments; /* output fragments created */
106 uint64_t ips_cantfrag; /* don't fragment flag was set, etc. */
107 uint64_t ips_badoptions; /* error in option processing */
108 uint64_t ips_noroute; /* packets discarded due to no route */
109 uint64_t ips_badvers; /* ip version != 4 */
110 uint64_t ips_rawout; /* total raw ip packets generated */
111 uint64_t ips_toolong; /* ip length > max ip packet size */
112 uint64_t ips_notmember; /* multicasts for unregistered grps */
113 uint64_t ips_nogif; /* no match gif found */
114 uint64_t ips_badaddr; /* invalid address on header */
115 };
116
117 #ifdef _KERNEL
118
119 #include <sys/counter.h>
120 #include <net/vnet.h>
121 #include <netinet/in_kdtrace.h>
122
123 VNET_PCPUSTAT_DECLARE(struct ipstat, ipstat);
124 /*
125 * In-kernel consumers can use these accessor macros directly to update
126 * stats.
127 */
128 #define IPSTAT_ADD(name, val) \
129 do { \
130 MIB_SDT_PROBE1(ip, count, name, (val)); \
131 VNET_PCPUSTAT_ADD(struct ipstat, ipstat, name, (val)); \
132 } while (0)
133 #define IPSTAT_SUB(name, val) IPSTAT_ADD(name, -(val))
134 #define IPSTAT_INC(name) IPSTAT_ADD(name, 1)
135 #define IPSTAT_DEC(name) IPSTAT_SUB(name, 1)
136
137 /*
138 * Kernel module consumers must use this accessor macro.
139 */
140 void kmod_ipstat_inc(int statnum);
141 #define KMOD_IPSTAT_INC(name) \
142 do { \
143 MIB_SDT_PROBE1(ip, count, name, 1); \
144 kmod_ipstat_inc( \
145 offsetof(struct ipstat, name) / sizeof(uint64_t)); \
146 } while (0)
147 void kmod_ipstat_dec(int statnum);
148 #define KMOD_IPSTAT_DEC(name) \
149 do { \
150 MIB_SDT_PROBE1(ip, count, name, -1); \
151 kmod_ipstat_dec( \
152 offsetof(struct ipstat, name) / sizeof(uint64_t)); \
153 } while (0)
154
155 /* flags passed to ip_output as last parameter */
156 #define IP_FORWARDING 0x1 /* most of ip header exists */
157 #define IP_RAWOUTPUT 0x2 /* raw ip header exists */
158 #define IP_SENDONES 0x4 /* send all-ones broadcast */
159 #define IP_SENDTOIF 0x8 /* send on specific ifnet */
160 #define IP_ROUTETOIF SO_DONTROUTE /* 0x10 bypass routing tables */
161 #define IP_ALLOWBROADCAST SO_BROADCAST /* 0x20 can send broadcast packets */
162 #define IP_NODEFAULTFLOWID 0x40 /* Don't set the flowid from inp */
163 #define IP_NO_SND_TAG_RL 0x80 /* Don't send down the ratelimit tag */
164
165 #ifdef __NO_STRICT_ALIGNMENT
166 #define IP_HDR_ALIGNED_P(ip) 1
167 #else
168 #define IP_HDR_ALIGNED_P(ip) ((((intptr_t) (ip)) & 3) == 0)
169 #endif
170
171 struct ip;
172 struct inpcb;
173 struct route;
174 struct sockopt;
175 struct inpcbinfo;
176 struct ip_moptions;
177
178 VNET_DECLARE(int, ip_defttl); /* default IP ttl */
179 VNET_DECLARE(int, ipforwarding); /* ip forwarding */
180 VNET_DECLARE(int, ipsendredirects);
181 #ifdef IPSTEALTH
182 VNET_DECLARE(int, ipstealth); /* stealth forwarding */
183 #endif
184 VNET_DECLARE(struct socket *, ip_rsvpd); /* reservation protocol daemon*/
185 VNET_DECLARE(int, rsvp_on);
186 VNET_DECLARE(int, drop_redirect);
187 VNET_DECLARE(int, ip_random_id);
188
189 #define V_ip_id VNET(ip_id)
190 #define V_ip_defttl VNET(ip_defttl)
191 #define V_ipforwarding VNET(ipforwarding)
192 #define V_ipsendredirects VNET(ipsendredirects)
193 #ifdef IPSTEALTH
194 #define V_ipstealth VNET(ipstealth)
195 #endif
196 #define V_ip_rsvpd VNET(ip_rsvpd)
197 #define V_rsvp_on VNET(rsvp_on)
198 #define V_drop_redirect VNET(drop_redirect)
199 #define V_ip_random_id VNET(ip_random_id)
200
201 void inp_freemoptions(struct ip_moptions *);
202 int inp_getmoptions(struct inpcb *, struct sockopt *);
203 int inp_setmoptions(struct inpcb *, struct sockopt *);
204
205 int ip_ctloutput(struct socket *, struct sockopt *sopt);
206 int ip_fragment(struct ip *ip, struct mbuf **m_frag, int mtu,
207 u_long if_hwassist_flags);
208 void ip_forward(struct mbuf *m, int srcrt);
209 int ip_output(struct mbuf *,
210 struct mbuf *, struct route *, int, struct ip_moptions *,
211 struct inpcb *);
212 struct mbuf *
213 ip_reass(struct mbuf *);
214 void ip_savecontrol(struct inpcb *, struct mbuf **, struct ip *,
215 struct mbuf *);
216 void ip_fillid(struct ip *, bool);
217 int rip_ctloutput(struct socket *, struct sockopt *);
218 int ipip_input(struct mbuf **, int *, int);
219 int rsvp_input(struct mbuf **, int *, int);
220
221 int ip_rsvp_init(struct socket *);
222 int ip_rsvp_done(void);
223
224 typedef int ipproto_input_t(struct mbuf **, int *, int);
225 struct icmp;
226 typedef void ipproto_ctlinput_t(struct icmp *);
227 int ipproto_register(uint8_t, ipproto_input_t, ipproto_ctlinput_t);
228 int ipproto_unregister(uint8_t);
229 #define IPPROTO_REGISTER(prot, input, ctl) do { \
230 int error __diagused; \
231 error = ipproto_register(prot, input, ctl); \
232 MPASS(error == 0); \
233 } while (0)
234
235 ipproto_input_t rip_input;
236 ipproto_ctlinput_t rip_ctlinput;
237
238 VNET_DECLARE(struct pfil_head *, inet_pfil_head);
239 #define V_inet_pfil_head VNET(inet_pfil_head)
240 #define PFIL_INET_NAME "inet"
241
242 VNET_DECLARE(struct pfil_head *, inet_local_pfil_head);
243 #define V_inet_local_pfil_head VNET(inet_local_pfil_head)
244 #define PFIL_INET_LOCAL_NAME "inet-local"
245
246 void in_delayed_cksum(struct mbuf *m);
247 void in_delayed_cksum_o(struct mbuf *m, uint16_t o);
248
249 /* Hooks for ipfw, dummynet, divert etc. Most are declared in raw_ip.c */
250 /*
251 * Reference to an ipfw or packet filter rule that can be carried
252 * outside critical sections.
253 * A rule is identified by rulenum:rule_id which is ordered.
254 * In version chain_id the rule can be found in slot 'slot', so
255 * we don't need a lookup if chain_id == chain->id.
256 *
257 * On exit from the firewall this structure refers to the rule after
258 * the matching one (slot points to the new rule; rulenum:rule_id-1
259 * is the matching rule), and additional info (e.g. info often contains
260 * the insn argument or tablearg in the low 16 bits, in host format).
261 * On entry, the structure is valid if slot>0, and refers to the starting
262 * rules. 'info' contains the reason for reinject, e.g. divert port,
263 * divert direction, and so on.
264 *
265 * Packet Mark is an analogue to ipfw tags with O(1) lookup from mbuf while
266 * regular tags require a single-linked list traversal. Mark is a 32-bit
267 * number that can be looked up in a table [with 'number' table-type], matched
268 * or compared with a number with optional mask applied before comparison.
269 * Having generic nature, Mark can be used in a variety of needs.
270 * For example, it could be used as a security group: mark will hold a
271 * security group id and represent a group of packet flows that shares same
272 * access control policy.
273 * O_MASK opcode can match mark value bitwise so one can build a hierarchical
274 * model designating different meanings for a bit range(s).
275 */
276 struct ipfw_rule_ref {
277 /* struct m_tag spans 24 bytes above this point, see mbuf_tags(9) */
278 /* spare space just to be save in case struct m_tag grows */
279 /* -- 32 bytes -- */
280 uint32_t slot; /* slot for matching rule */
281 uint32_t rulenum; /* matching rule number */
282 uint32_t rule_id; /* matching rule id */
283 uint32_t chain_id; /* ruleset id */
284 uint32_t info; /* see below */
285 uint32_t pkt_mark; /* packet mark */
286 uint32_t spare[2];
287 /* -- 64 bytes -- */
288 };
289
290 enum {
291 IPFW_INFO_MASK = 0x0000ffff,
292 IPFW_INFO_OUT = 0x00000000, /* outgoing, just for convenience */
293 IPFW_INFO_IN = 0x80000000, /* incoming, overloads dir */
294 IPFW_ONEPASS = 0x40000000, /* One-pass, do not reinject */
295 IPFW_IS_MASK = 0x30000000, /* which source ? */
296 IPFW_IS_DIVERT = 0x20000000,
297 IPFW_IS_DUMMYNET =0x10000000,
298 IPFW_IS_PIPE = 0x08000000, /* pipe=1, queue = 0 */
299 };
300 #define MTAG_IPFW 1148380143 /* IPFW-tagged cookie */
301 #define MTAG_IPFW_RULE 1262273568 /* rule reference */
302 #define MTAG_IPFW_CALL 1308397630 /* call stack */
303
304 struct ip_fw_args;
305 typedef int (*ip_fw_ctl_ptr_t)(struct sockopt *);
306 VNET_DECLARE(ip_fw_ctl_ptr_t, ip_fw_ctl_ptr);
307 #define V_ip_fw_ctl_ptr VNET(ip_fw_ctl_ptr)
308
309 /* Divert hooks. */
310 extern void (*ip_divert_ptr)(struct mbuf *m, uint64_t id, bool incoming);
311 /* ng_ipfw hooks -- XXX make it the same as divert and dummynet */
312 extern int (*ng_ipfw_input_p)(struct mbuf **, struct ip_fw_args *, bool);
313 extern int (*ip_dn_ctl_ptr)(struct sockopt *);
314 extern int (*ip_dn_io_ptr)(struct mbuf **, struct ip_fw_args *);
315
316 /* pf specific mtag for divert(4) support */
__enum_uint8_decl(pf_mtag_dir)317 __enum_uint8_decl(pf_mtag_dir) {
318 PF_DIVERT_MTAG_DIR_IN = 1,
319 PF_DIVERT_MTAG_DIR_OUT = 2
320 };
321 struct pf_divert_mtag {
322 __enum_uint8(pf_mtag_dir) idir; /* initial pkt direction */
323 union {
324 __enum_uint8(pf_mtag_dir) ndir; /* new dir after re-enter */
325 uint16_t port; /* initial divert(4) port */
326 };
327 };
328 #define MTAG_PF_DIVERT 1262273569
329
330 #endif /* _KERNEL */
331
332 #endif /* !_NETINET_IP_VAR_H_ */
333