xref: /freebsd/sys/netinet/ip_var.h (revision 2e3f49888ec8851bafb22011533217487764fdb0)
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 };
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 #if defined(_NETINET_IN_VAR_H_) && defined(_KERNEL)
86 /*
87  * Structure attached to inpcb.ip_moptions and
88  * passed to ip_output when IP multicast options are in use.
89  * This structure is lazy-allocated.
90  */
91 struct ip_moptions {
92 	struct	ifnet *imo_multicast_ifp; /* ifp for outgoing multicasts */
93 	struct in_addr imo_multicast_addr; /* ifindex/addr on MULTICAST_IF */
94 	u_long	imo_multicast_vif;	/* vif num outgoing multicasts */
95 	u_char	imo_multicast_ttl;	/* TTL for outgoing multicasts */
96 	u_char	imo_multicast_loop;	/* 1 => hear sends if a member */
97 	struct ip_mfilter_head imo_head; /* group membership list */
98 };
99 #else
100 struct ip_moptions;
101 #endif
102 
103 struct	ipstat {
104 	uint64_t ips_total;		/* total packets received */
105 	uint64_t ips_badsum;		/* checksum bad */
106 	uint64_t ips_tooshort;		/* packet too short */
107 	uint64_t ips_toosmall;		/* not enough data */
108 	uint64_t ips_badhlen;		/* ip header length < data size */
109 	uint64_t ips_badlen;		/* ip length < ip header length */
110 	uint64_t ips_fragments;		/* fragments received */
111 	uint64_t ips_fragdropped;	/* frags dropped (dups, out of space) */
112 	uint64_t ips_fragtimeout;	/* fragments timed out */
113 	uint64_t ips_forward;		/* packets forwarded */
114 	uint64_t ips_fastforward;	/* packets fast forwarded */
115 	uint64_t ips_cantforward;	/* packets rcvd for unreachable dest */
116 	uint64_t ips_redirectsent;	/* packets forwarded on same net */
117 	uint64_t ips_noproto;		/* unknown or unsupported protocol */
118 	uint64_t ips_delivered;		/* datagrams delivered to upper level*/
119 	uint64_t ips_localout;		/* total ip packets generated here */
120 	uint64_t ips_odropped;		/* lost packets due to nobufs, etc. */
121 	uint64_t ips_reassembled;	/* total packets reassembled ok */
122 	uint64_t ips_fragmented;	/* datagrams successfully fragmented */
123 	uint64_t ips_ofragments;	/* output fragments created */
124 	uint64_t ips_cantfrag;		/* don't fragment flag was set, etc. */
125 	uint64_t ips_badoptions;		/* error in option processing */
126 	uint64_t ips_noroute;		/* packets discarded due to no route */
127 	uint64_t ips_badvers;		/* ip version != 4 */
128 	uint64_t ips_rawout;		/* total raw ip packets generated */
129 	uint64_t ips_toolong;		/* ip length > max ip packet size */
130 	uint64_t ips_notmember;		/* multicasts for unregistered grps */
131 	uint64_t ips_nogif;		/* no match gif found */
132 	uint64_t ips_badaddr;		/* invalid address on header */
133 };
134 
135 #ifdef _KERNEL
136 
137 #include <sys/counter.h>
138 #include <net/vnet.h>
139 
140 VNET_PCPUSTAT_DECLARE(struct ipstat, ipstat);
141 /*
142  * In-kernel consumers can use these accessor macros directly to update
143  * stats.
144  */
145 #define	IPSTAT_ADD(name, val)	\
146     VNET_PCPUSTAT_ADD(struct ipstat, ipstat, name, (val))
147 #define	IPSTAT_SUB(name, val)	IPSTAT_ADD(name, -(val))
148 #define	IPSTAT_INC(name)	IPSTAT_ADD(name, 1)
149 #define	IPSTAT_DEC(name)	IPSTAT_SUB(name, 1)
150 
151 /*
152  * Kernel module consumers must use this accessor macro.
153  */
154 void	kmod_ipstat_inc(int statnum);
155 #define	KMOD_IPSTAT_INC(name)	\
156     kmod_ipstat_inc(offsetof(struct ipstat, name) / sizeof(uint64_t))
157 void	kmod_ipstat_dec(int statnum);
158 #define	KMOD_IPSTAT_DEC(name)	\
159     kmod_ipstat_dec(offsetof(struct ipstat, name) / sizeof(uint64_t))
160 
161 /* flags passed to ip_output as last parameter */
162 #define	IP_FORWARDING		0x1		/* most of ip header exists */
163 #define	IP_RAWOUTPUT		0x2		/* raw ip header exists */
164 #define	IP_SENDONES		0x4		/* send all-ones broadcast */
165 #define	IP_SENDTOIF		0x8		/* send on specific ifnet */
166 #define IP_ROUTETOIF		SO_DONTROUTE	/* 0x10 bypass routing tables */
167 #define IP_ALLOWBROADCAST	SO_BROADCAST	/* 0x20 can send broadcast packets */
168 #define	IP_NODEFAULTFLOWID	0x40		/* Don't set the flowid from inp */
169 #define IP_NO_SND_TAG_RL	0x80		/* Don't send down the ratelimit tag */
170 
171 #ifdef __NO_STRICT_ALIGNMENT
172 #define IP_HDR_ALIGNED_P(ip)	1
173 #else
174 #define IP_HDR_ALIGNED_P(ip)	((((intptr_t) (ip)) & 3) == 0)
175 #endif
176 
177 struct ip;
178 struct inpcb;
179 struct route;
180 struct sockopt;
181 struct inpcbinfo;
182 
183 VNET_DECLARE(int, ip_defttl);			/* default IP ttl */
184 VNET_DECLARE(int, ipforwarding);		/* ip forwarding */
185 VNET_DECLARE(int, ipsendredirects);
186 #ifdef IPSTEALTH
187 VNET_DECLARE(int, ipstealth);			/* stealth forwarding */
188 #endif
189 VNET_DECLARE(struct socket *, ip_rsvpd);	/* reservation protocol daemon*/
190 VNET_DECLARE(struct socket *, ip_mrouter);	/* multicast routing daemon */
191 extern int	(*legal_vif_num)(int);
192 extern u_long	(*ip_mcast_src)(int);
193 VNET_DECLARE(int, rsvp_on);
194 VNET_DECLARE(int, drop_redirect);
195 
196 #define	V_ip_id			VNET(ip_id)
197 #define	V_ip_defttl		VNET(ip_defttl)
198 #define	V_ipforwarding		VNET(ipforwarding)
199 #define	V_ipsendredirects	VNET(ipsendredirects)
200 #ifdef IPSTEALTH
201 #define	V_ipstealth		VNET(ipstealth)
202 #endif
203 #define	V_ip_rsvpd		VNET(ip_rsvpd)
204 #define	V_ip_mrouter		VNET(ip_mrouter)
205 #define	V_rsvp_on		VNET(rsvp_on)
206 #define	V_drop_redirect		VNET(drop_redirect)
207 
208 void	inp_freemoptions(struct ip_moptions *);
209 int	inp_getmoptions(struct inpcb *, struct sockopt *);
210 int	inp_setmoptions(struct inpcb *, struct sockopt *);
211 
212 int	ip_ctloutput(struct socket *, struct sockopt *sopt);
213 int	ip_fragment(struct ip *ip, struct mbuf **m_frag, int mtu,
214 	    u_long if_hwassist_flags);
215 void	ip_forward(struct mbuf *m, int srcrt);
216 extern int
217 	(*ip_mforward)(struct ip *, struct ifnet *, struct mbuf *,
218 	    struct ip_moptions *);
219 int	ip_output(struct mbuf *,
220 	    struct mbuf *, struct route *, int, struct ip_moptions *,
221 	    struct inpcb *);
222 struct mbuf *
223 	ip_reass(struct mbuf *);
224 void	ip_savecontrol(struct inpcb *, struct mbuf **, struct ip *,
225 	    struct mbuf *);
226 void	ip_fillid(struct ip *);
227 int	rip_ctloutput(struct socket *, struct sockopt *);
228 int	ipip_input(struct mbuf **, int *, int);
229 int	rsvp_input(struct mbuf **, int *, int);
230 
231 int	ip_rsvp_init(struct socket *);
232 int	ip_rsvp_done(void);
233 extern int	(*ip_rsvp_vif)(struct socket *, struct sockopt *);
234 extern void	(*ip_rsvp_force_done)(struct socket *);
235 extern int	(*rsvp_input_p)(struct mbuf **, int *, int);
236 
237 typedef int	ipproto_input_t(struct mbuf **, int *, int);
238 struct icmp;
239 typedef void	ipproto_ctlinput_t(struct icmp *);
240 int	ipproto_register(uint8_t, ipproto_input_t, ipproto_ctlinput_t);
241 int	ipproto_unregister(uint8_t);
242 #define	IPPROTO_REGISTER(prot, input, ctl)	do {			\
243 	int error __diagused;						\
244 	error = ipproto_register(prot, input, ctl);			\
245 	MPASS(error == 0);						\
246 } while (0)
247 
248 ipproto_input_t		rip_input;
249 ipproto_ctlinput_t	rip_ctlinput;
250 
251 VNET_DECLARE(struct pfil_head *, inet_pfil_head);
252 #define	V_inet_pfil_head	VNET(inet_pfil_head)
253 #define	PFIL_INET_NAME		"inet"
254 
255 VNET_DECLARE(struct pfil_head *, inet_local_pfil_head);
256 #define	V_inet_local_pfil_head	VNET(inet_local_pfil_head)
257 #define	PFIL_INET_LOCAL_NAME	"inet-local"
258 
259 void	in_delayed_cksum(struct mbuf *m);
260 
261 /* Hooks for ipfw, dummynet, divert etc. Most are declared in raw_ip.c */
262 /*
263  * Reference to an ipfw or packet filter rule that can be carried
264  * outside critical sections.
265  * A rule is identified by rulenum:rule_id which is ordered.
266  * In version chain_id the rule can be found in slot 'slot', so
267  * we don't need a lookup if chain_id == chain->id.
268  *
269  * On exit from the firewall this structure refers to the rule after
270  * the matching one (slot points to the new rule; rulenum:rule_id-1
271  * is the matching rule), and additional info (e.g. info often contains
272  * the insn argument or tablearg in the low 16 bits, in host format).
273  * On entry, the structure is valid if slot>0, and refers to the starting
274  * rules. 'info' contains the reason for reinject, e.g. divert port,
275  * divert direction, and so on.
276  *
277  * Packet Mark is an analogue to ipfw tags with O(1) lookup from mbuf while
278  * regular tags require a single-linked list traversal. Mark is a 32-bit
279  * number that can be looked up in a table [with 'number' table-type], matched
280  * or compared with a number with optional mask applied before comparison.
281  * Having generic nature, Mark can be used in a variety of needs.
282  * For example, it could be used as a security group: mark will hold a
283  * security group id and represent a group of packet flows that shares same
284  * access control policy.
285  * O_MASK opcode can match mark value bitwise so one can build a hierarchical
286  * model designating different meanings for a bit range(s).
287  */
288 struct ipfw_rule_ref {
289 /* struct m_tag spans 24 bytes above this point, see mbuf_tags(9) */
290 	/* spare space just to be save in case struct m_tag grows */
291 /* -- 32 bytes -- */
292 	uint32_t	slot;		/* slot for matching rule	*/
293 	uint32_t	rulenum;	/* matching rule number		*/
294 	uint32_t	rule_id;	/* matching rule id		*/
295 	uint32_t	chain_id;	/* ruleset id			*/
296 	uint32_t	info;		/* see below			*/
297 	uint32_t	pkt_mark;	/* packet mark			*/
298 	uint32_t	spare[2];
299 /* -- 64 bytes -- */
300 };
301 
302 enum {
303 	IPFW_INFO_MASK	= 0x0000ffff,
304 	IPFW_INFO_OUT	= 0x00000000,	/* outgoing, just for convenience */
305 	IPFW_INFO_IN	= 0x80000000,	/* incoming, overloads dir */
306 	IPFW_ONEPASS	= 0x40000000,	/* One-pass, do not reinject */
307 	IPFW_IS_MASK	= 0x30000000,	/* which source ? */
308 	IPFW_IS_DIVERT	= 0x20000000,
309 	IPFW_IS_DUMMYNET =0x10000000,
310 	IPFW_IS_PIPE	= 0x08000000,	/* pipe=1, queue = 0 */
311 };
312 #define MTAG_IPFW	1148380143	/* IPFW-tagged cookie */
313 #define MTAG_IPFW_RULE	1262273568	/* rule reference */
314 #define	MTAG_IPFW_CALL	1308397630	/* call stack */
315 
316 struct ip_fw_args;
317 typedef int	(*ip_fw_ctl_ptr_t)(struct sockopt *);
318 VNET_DECLARE(ip_fw_ctl_ptr_t, ip_fw_ctl_ptr);
319 #define	V_ip_fw_ctl_ptr		VNET(ip_fw_ctl_ptr)
320 
321 /* Divert hooks. */
322 extern void	(*ip_divert_ptr)(struct mbuf *m, bool incoming);
323 /* ng_ipfw hooks -- XXX make it the same as divert and dummynet */
324 extern int	(*ng_ipfw_input_p)(struct mbuf **, struct ip_fw_args *, bool);
325 extern int	(*ip_dn_ctl_ptr)(struct sockopt *);
326 extern int	(*ip_dn_io_ptr)(struct mbuf **, struct ip_fw_args *);
327 
328 /* pf specific mtag for divert(4) support */
329 __enum_uint8_decl(pf_mtag_dir) {
330 	PF_DIVERT_MTAG_DIR_IN = 1,
331 	PF_DIVERT_MTAG_DIR_OUT = 2
332 };
333 struct pf_divert_mtag {
334 	__enum_uint8(pf_mtag_dir) idir;		/* initial pkt direction */
335 	union {
336 		__enum_uint8(pf_mtag_dir) ndir;	/* new dir after re-enter */
337 		uint16_t port;			/* initial divert(4) port */
338 	};
339 };
340 #define MTAG_PF_DIVERT	1262273569
341 
342 #endif /* _KERNEL */
343 
344 #endif /* !_NETINET_IP_VAR_H_ */
345