xref: /freebsd/sys/netinet/ip_var.h (revision e32fecd0c2c3ee37c47ee100f169e7eb0282a873)
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  *	@(#)ip_var.h	8.2 (Berkeley) 1/9/95
32  * $FreeBSD$
33  */
34 
35 #ifndef _NETINET_IP_VAR_H_
36 #define	_NETINET_IP_VAR_H_
37 
38 #include <sys/epoch.h>
39 #include <sys/queue.h>
40 #include <sys/types.h>
41 
42 #include <netinet/in.h>
43 
44 /*
45  * Overlay for ip header used by other protocols (tcp, udp).
46  */
47 struct ipovly {
48 	u_char	ih_x1[9];		/* (unused) */
49 	u_char	ih_pr;			/* protocol */
50 	u_short	ih_len;			/* protocol length */
51 	struct	in_addr ih_src;		/* source internet address */
52 	struct	in_addr ih_dst;		/* destination internet address */
53 };
54 
55 #ifdef _KERNEL
56 /*
57  * Ip reassembly queue structure.  Each fragment
58  * being reassembled is attached to one of these structures.
59  * They are timed out after net.inet.ip.fragttl seconds, and may also be
60  * reclaimed if memory becomes tight.
61  */
62 struct ipq {
63 	TAILQ_ENTRY(ipq) ipq_list;	/* to other reass headers */
64 	time_t	ipq_expire;		/* time_uptime when ipq expires */
65 	u_char	ipq_nfrags;		/* # frags in this packet */
66 	u_char	ipq_p;			/* protocol of this fragment */
67 	u_short	ipq_id;			/* sequence id for reassembly */
68 	int	ipq_maxoff;		/* total length of packet */
69 	struct mbuf *ipq_frags;		/* to ip headers of fragments */
70 	struct	in_addr ipq_src,ipq_dst;
71 	struct label *ipq_label;	/* MAC label */
72 };
73 #endif /* _KERNEL */
74 
75 /*
76  * Structure stored in mbuf in inpcb.ip_options
77  * and passed to ip_output when ip options are in use.
78  * The actual length of the options (including ipopt_dst)
79  * is in m_len.
80  */
81 #define MAX_IPOPTLEN	40
82 
83 struct ipoption {
84 	struct	in_addr ipopt_dst;	/* first-hop dst if source routed */
85 	char	ipopt_list[MAX_IPOPTLEN];	/* options proper */
86 };
87 
88 #if defined(_NETINET_IN_VAR_H_) && defined(_KERNEL)
89 /*
90  * Structure attached to inpcb.ip_moptions and
91  * passed to ip_output when IP multicast options are in use.
92  * This structure is lazy-allocated.
93  */
94 struct ip_moptions {
95 	struct	ifnet *imo_multicast_ifp; /* ifp for outgoing multicasts */
96 	struct in_addr imo_multicast_addr; /* ifindex/addr on MULTICAST_IF */
97 	u_long	imo_multicast_vif;	/* vif num outgoing multicasts */
98 	u_char	imo_multicast_ttl;	/* TTL for outgoing multicasts */
99 	u_char	imo_multicast_loop;	/* 1 => hear sends if a member */
100 	struct ip_mfilter_head imo_head; /* group membership list */
101 };
102 #else
103 struct ip_moptions;
104 #endif
105 
106 struct	ipstat {
107 	uint64_t ips_total;		/* total packets received */
108 	uint64_t ips_badsum;		/* checksum bad */
109 	uint64_t ips_tooshort;		/* packet too short */
110 	uint64_t ips_toosmall;		/* not enough data */
111 	uint64_t ips_badhlen;		/* ip header length < data size */
112 	uint64_t ips_badlen;		/* ip length < ip header length */
113 	uint64_t ips_fragments;		/* fragments received */
114 	uint64_t ips_fragdropped;	/* frags dropped (dups, out of space) */
115 	uint64_t ips_fragtimeout;	/* fragments timed out */
116 	uint64_t ips_forward;		/* packets forwarded */
117 	uint64_t ips_fastforward;	/* packets fast forwarded */
118 	uint64_t ips_cantforward;	/* packets rcvd for unreachable dest */
119 	uint64_t ips_redirectsent;	/* packets forwarded on same net */
120 	uint64_t ips_noproto;		/* unknown or unsupported protocol */
121 	uint64_t ips_delivered;		/* datagrams delivered to upper level*/
122 	uint64_t ips_localout;		/* total ip packets generated here */
123 	uint64_t ips_odropped;		/* lost packets due to nobufs, etc. */
124 	uint64_t ips_reassembled;	/* total packets reassembled ok */
125 	uint64_t ips_fragmented;	/* datagrams successfully fragmented */
126 	uint64_t ips_ofragments;	/* output fragments created */
127 	uint64_t ips_cantfrag;		/* don't fragment flag was set, etc. */
128 	uint64_t ips_badoptions;		/* error in option processing */
129 	uint64_t ips_noroute;		/* packets discarded due to no route */
130 	uint64_t ips_badvers;		/* ip version != 4 */
131 	uint64_t ips_rawout;		/* total raw ip packets generated */
132 	uint64_t ips_toolong;		/* ip length > max ip packet size */
133 	uint64_t ips_notmember;		/* multicasts for unregistered grps */
134 	uint64_t ips_nogif;		/* no match gif found */
135 	uint64_t ips_badaddr;		/* invalid address on header */
136 };
137 
138 #ifdef _KERNEL
139 
140 #include <sys/counter.h>
141 #include <net/vnet.h>
142 
143 VNET_PCPUSTAT_DECLARE(struct ipstat, ipstat);
144 /*
145  * In-kernel consumers can use these accessor macros directly to update
146  * stats.
147  */
148 #define	IPSTAT_ADD(name, val)	\
149     VNET_PCPUSTAT_ADD(struct ipstat, ipstat, name, (val))
150 #define	IPSTAT_SUB(name, val)	IPSTAT_ADD(name, -(val))
151 #define	IPSTAT_INC(name)	IPSTAT_ADD(name, 1)
152 #define	IPSTAT_DEC(name)	IPSTAT_SUB(name, 1)
153 
154 /*
155  * Kernel module consumers must use this accessor macro.
156  */
157 void	kmod_ipstat_inc(int statnum);
158 #define	KMOD_IPSTAT_INC(name)	\
159     kmod_ipstat_inc(offsetof(struct ipstat, name) / sizeof(uint64_t))
160 void	kmod_ipstat_dec(int statnum);
161 #define	KMOD_IPSTAT_DEC(name)	\
162     kmod_ipstat_dec(offsetof(struct ipstat, name) / sizeof(uint64_t))
163 
164 /* flags passed to ip_output as last parameter */
165 #define	IP_FORWARDING		0x1		/* most of ip header exists */
166 #define	IP_RAWOUTPUT		0x2		/* raw ip header exists */
167 #define	IP_SENDONES		0x4		/* send all-ones broadcast */
168 #define	IP_SENDTOIF		0x8		/* send on specific ifnet */
169 #define IP_ROUTETOIF		SO_DONTROUTE	/* 0x10 bypass routing tables */
170 #define IP_ALLOWBROADCAST	SO_BROADCAST	/* 0x20 can send broadcast packets */
171 #define	IP_NODEFAULTFLOWID	0x40		/* Don't set the flowid from inp */
172 #define IP_NO_SND_TAG_RL	0x80		/* Don't send down the ratelimit tag */
173 
174 #ifdef __NO_STRICT_ALIGNMENT
175 #define IP_HDR_ALIGNED_P(ip)	1
176 #else
177 #define IP_HDR_ALIGNED_P(ip)	((((intptr_t) (ip)) & 3) == 0)
178 #endif
179 
180 struct ip;
181 struct inpcb;
182 struct route;
183 struct sockopt;
184 struct inpcbinfo;
185 
186 VNET_DECLARE(int, ip_defttl);			/* default IP ttl */
187 VNET_DECLARE(int, ipforwarding);		/* ip forwarding */
188 VNET_DECLARE(int, ipsendredirects);
189 #ifdef IPSTEALTH
190 VNET_DECLARE(int, ipstealth);			/* stealth forwarding */
191 #endif
192 VNET_DECLARE(struct socket *, ip_rsvpd);	/* reservation protocol daemon*/
193 VNET_DECLARE(struct socket *, ip_mrouter);	/* multicast routing daemon */
194 extern int	(*legal_vif_num)(int);
195 extern u_long	(*ip_mcast_src)(int);
196 VNET_DECLARE(int, rsvp_on);
197 VNET_DECLARE(int, drop_redirect);
198 
199 #define	V_ip_id			VNET(ip_id)
200 #define	V_ip_defttl		VNET(ip_defttl)
201 #define	V_ipforwarding		VNET(ipforwarding)
202 #define	V_ipsendredirects	VNET(ipsendredirects)
203 #ifdef IPSTEALTH
204 #define	V_ipstealth		VNET(ipstealth)
205 #endif
206 #define	V_ip_rsvpd		VNET(ip_rsvpd)
207 #define	V_ip_mrouter		VNET(ip_mrouter)
208 #define	V_rsvp_on		VNET(rsvp_on)
209 #define	V_drop_redirect		VNET(drop_redirect)
210 
211 void	inp_freemoptions(struct ip_moptions *);
212 int	inp_getmoptions(struct inpcb *, struct sockopt *);
213 int	inp_setmoptions(struct inpcb *, struct sockopt *);
214 
215 int	ip_ctloutput(struct socket *, struct sockopt *sopt);
216 int	ip_fragment(struct ip *ip, struct mbuf **m_frag, int mtu,
217 	    u_long if_hwassist_flags);
218 void	ip_forward(struct mbuf *m, int srcrt);
219 extern int
220 	(*ip_mforward)(struct ip *, struct ifnet *, struct mbuf *,
221 	    struct ip_moptions *);
222 int	ip_output(struct mbuf *,
223 	    struct mbuf *, struct route *, int, struct ip_moptions *,
224 	    struct inpcb *);
225 struct mbuf *
226 	ip_reass(struct mbuf *);
227 void	ip_savecontrol(struct inpcb *, struct mbuf **, struct ip *,
228 	    struct mbuf *);
229 void	ip_fillid(struct ip *);
230 int	rip_ctloutput(struct socket *, struct sockopt *);
231 int	ipip_input(struct mbuf **, int *, int);
232 int	rsvp_input(struct mbuf **, int *, int);
233 
234 int	ip_rsvp_init(struct socket *);
235 int	ip_rsvp_done(void);
236 extern int	(*ip_rsvp_vif)(struct socket *, struct sockopt *);
237 extern void	(*ip_rsvp_force_done)(struct socket *);
238 extern int	(*rsvp_input_p)(struct mbuf **, int *, int);
239 
240 typedef int	ipproto_input_t(struct mbuf **, int *, int);
241 struct icmp;
242 typedef void	ipproto_ctlinput_t(struct icmp *);
243 int	ipproto_register(uint8_t, ipproto_input_t, ipproto_ctlinput_t);
244 int	ipproto_unregister(uint8_t);
245 #define	IPPROTO_REGISTER(prot, input, ctl)	do {			\
246 	int error __diagused;						\
247 	error = ipproto_register(prot, input, ctl);			\
248 	MPASS(error == 0);						\
249 } while (0)
250 
251 ipproto_input_t		rip_input;
252 ipproto_ctlinput_t	rip_ctlinput;
253 
254 VNET_DECLARE(struct pfil_head *, inet_pfil_head);
255 #define	V_inet_pfil_head	VNET(inet_pfil_head)
256 #define	PFIL_INET_NAME		"inet"
257 
258 void	in_delayed_cksum(struct mbuf *m);
259 
260 /* Hooks for ipfw, dummynet, divert etc. Most are declared in raw_ip.c */
261 /*
262  * Reference to an ipfw or packet filter rule that can be carried
263  * outside critical sections.
264  * A rule is identified by rulenum:rule_id which is ordered.
265  * In version chain_id the rule can be found in slot 'slot', so
266  * we don't need a lookup if chain_id == chain->id.
267  *
268  * On exit from the firewall this structure refers to the rule after
269  * the matching one (slot points to the new rule; rulenum:rule_id-1
270  * is the matching rule), and additional info (e.g. info often contains
271  * the insn argument or tablearg in the low 16 bits, in host format).
272  * On entry, the structure is valid if slot>0, and refers to the starting
273  * rules. 'info' contains the reason for reinject, e.g. divert port,
274  * divert direction, and so on.
275  *
276  * Packet Mark is an analogue to ipfw tags with O(1) lookup from mbuf while
277  * regular tags require a single-linked list traversal. Mark is a 32-bit
278  * number that can be looked up in a table [with 'number' table-type], matched
279  * or compared with a number with optional mask applied before comparison.
280  * Having generic nature, Mark can be used in a variety of needs.
281  * For example, it could be used as a security group: mark will hold a
282  * security group id and represent a group of packet flows that shares same
283  * access control policy.
284  * O_MASK opcode can match mark value bitwise so one can build a hierarchical
285  * model designating different meanings for a bit range(s).
286  */
287 struct ipfw_rule_ref {
288 /* struct m_tag spans 24 bytes above this point, see mbuf_tags(9) */
289 	/* spare space just to be save in case struct m_tag grows */
290 /* -- 32 bytes -- */
291 	uint32_t	slot;		/* slot for matching rule	*/
292 	uint32_t	rulenum;	/* matching rule number		*/
293 	uint32_t	rule_id;	/* matching rule id		*/
294 	uint32_t	chain_id;	/* ruleset id			*/
295 	uint32_t	info;		/* see below			*/
296 	uint32_t	pkt_mark;	/* packet mark			*/
297 	uint32_t	spare[2];
298 /* -- 64 bytes -- */
299 };
300 
301 enum {
302 	IPFW_INFO_MASK	= 0x0000ffff,
303 	IPFW_INFO_OUT	= 0x00000000,	/* outgoing, just for convenience */
304 	IPFW_INFO_IN	= 0x80000000,	/* incoming, overloads dir */
305 	IPFW_ONEPASS	= 0x40000000,	/* One-pass, do not reinject */
306 	IPFW_IS_MASK	= 0x30000000,	/* which source ? */
307 	IPFW_IS_DIVERT	= 0x20000000,
308 	IPFW_IS_DUMMYNET =0x10000000,
309 	IPFW_IS_PIPE	= 0x08000000,	/* pipe=1, queue = 0 */
310 };
311 #define MTAG_IPFW	1148380143	/* IPFW-tagged cookie */
312 #define MTAG_IPFW_RULE	1262273568	/* rule reference */
313 #define	MTAG_IPFW_CALL	1308397630	/* call stack */
314 
315 struct ip_fw_args;
316 typedef int	(*ip_fw_ctl_ptr_t)(struct sockopt *);
317 VNET_DECLARE(ip_fw_ctl_ptr_t, ip_fw_ctl_ptr);
318 #define	V_ip_fw_ctl_ptr		VNET(ip_fw_ctl_ptr)
319 
320 /* Divert hooks. */
321 extern void	(*ip_divert_ptr)(struct mbuf *m, bool incoming);
322 /* ng_ipfw hooks -- XXX make it the same as divert and dummynet */
323 extern int	(*ng_ipfw_input_p)(struct mbuf **, struct ip_fw_args *, bool);
324 extern int	(*ip_dn_ctl_ptr)(struct sockopt *);
325 extern int	(*ip_dn_io_ptr)(struct mbuf **, struct ip_fw_args *);
326 #endif /* _KERNEL */
327 
328 #endif /* !_NETINET_IP_VAR_H_ */
329