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 */
 33
 34#ifndef _NETINET_IP_VAR_H_
 35#define	_NETINET_IP_VAR_H_
 36
 37#include <sys/epoch.h>
 38#include <sys/queue.h>
 39#include <sys/types.h>
 40
 41#include <netinet/in.h>
 42
 43/*
 44 * Overlay for ip header used by other protocols (tcp, udp).
 45 */
 46struct ipovly {
 47	u_char	ih_x1[9];		/* (unused) */
 48	u_char	ih_pr;			/* protocol */
 49	u_short	ih_len;			/* protocol length */
 50	struct	in_addr ih_src;		/* source internet address */
 51	struct	in_addr ih_dst;		/* destination internet address */
 52};
 53
 54#ifdef _KERNEL
 55/*
 56 * Ip reassembly queue structure.  Each fragment
 57 * being reassembled is attached to one of these structures.
 58 * They are timed out after net.inet.ip.fragttl seconds, and may also be
 59 * reclaimed if memory becomes tight.
 60 */
 61struct ipq {
 62	TAILQ_ENTRY(ipq) ipq_list;	/* to other reass headers */
 63	time_t	ipq_expire;		/* time_uptime when ipq expires */
 64	u_char	ipq_nfrags;		/* # frags in this packet */
 65	u_char	ipq_p;			/* protocol of this fragment */
 66	u_short	ipq_id;			/* sequence id for reassembly */
 67	int	ipq_maxoff;		/* total length of packet */
 68	struct mbuf *ipq_frags;		/* to ip headers of fragments */
 69	struct	in_addr ipq_src,ipq_dst;
 70	struct label *ipq_label;	/* MAC label */
 71};
 72#endif /* _KERNEL */
 73
 74/*
 75 * Structure stored in mbuf in inpcb.ip_options
 76 * and passed to ip_output when ip options are in use.
 77 * The actual length of the options (including ipopt_dst)
 78 * is in m_len.
 79 */
 80#define MAX_IPOPTLEN	40
 81
 82struct ipoption {
 83	struct	in_addr ipopt_dst;	/* first-hop dst if source routed */
 84	char	ipopt_list[MAX_IPOPTLEN];	/* options proper */
 85};
 86
 87#if defined(_NETINET_IN_VAR_H_) && defined(_KERNEL)
 88/*
 89 * Structure attached to inpcb.ip_moptions and
 90 * passed to ip_output when IP multicast options are in use.
 91 * This structure is lazy-allocated.
 92 */
 93struct ip_moptions {
 94	struct	ifnet *imo_multicast_ifp; /* ifp for outgoing multicasts */
 95	struct in_addr imo_multicast_addr; /* ifindex/addr on MULTICAST_IF */
 96	u_long	imo_multicast_vif;	/* vif num outgoing multicasts */
 97	u_char	imo_multicast_ttl;	/* TTL for outgoing multicasts */
 98	u_char	imo_multicast_loop;	/* 1 => hear sends if a member */
 99	struct ip_mfilter_head imo_head; /* group membership list */
100};
101#else
102struct ip_moptions;
103#endif
104
105struct	ipstat {
106	uint64_t ips_total;		/* total packets received */
107	uint64_t ips_badsum;		/* checksum bad */
108	uint64_t ips_tooshort;		/* packet too short */
109	uint64_t ips_toosmall;		/* not enough data */
110	uint64_t ips_badhlen;		/* ip header length < data size */
111	uint64_t ips_badlen;		/* ip length < ip header length */
112	uint64_t ips_fragments;		/* fragments received */
113	uint64_t ips_fragdropped;	/* frags dropped (dups, out of space) */
114	uint64_t ips_fragtimeout;	/* fragments timed out */
115	uint64_t ips_forward;		/* packets forwarded */
116	uint64_t ips_fastforward;	/* packets fast forwarded */
117	uint64_t ips_cantforward;	/* packets rcvd for unreachable dest */
118	uint64_t ips_redirectsent;	/* packets forwarded on same net */
119	uint64_t ips_noproto;		/* unknown or unsupported protocol */
120	uint64_t ips_delivered;		/* datagrams delivered to upper level*/
121	uint64_t ips_localout;		/* total ip packets generated here */
122	uint64_t ips_odropped;		/* lost packets due to nobufs, etc. */
123	uint64_t ips_reassembled;	/* total packets reassembled ok */
124	uint64_t ips_fragmented;	/* datagrams successfully fragmented */
125	uint64_t ips_ofragments;	/* output fragments created */
126	uint64_t ips_cantfrag;		/* don't fragment flag was set, etc. */
127	uint64_t ips_badoptions;		/* error in option processing */
128	uint64_t ips_noroute;		/* packets discarded due to no route */
129	uint64_t ips_badvers;		/* ip version != 4 */
130	uint64_t ips_rawout;		/* total raw ip packets generated */
131	uint64_t ips_toolong;		/* ip length > max ip packet size */
132	uint64_t ips_notmember;		/* multicasts for unregistered grps */
133	uint64_t ips_nogif;		/* no match gif found */
134	uint64_t ips_badaddr;		/* invalid address on header */
135};
136
137#ifdef _KERNEL
138
139#include <sys/counter.h>
140#include <net/vnet.h>
141
142VNET_PCPUSTAT_DECLARE(struct ipstat, ipstat);
143/*
144 * In-kernel consumers can use these accessor macros directly to update
145 * stats.
146 */
147#define	IPSTAT_ADD(name, val)	\
148    VNET_PCPUSTAT_ADD(struct ipstat, ipstat, name, (val))
149#define	IPSTAT_SUB(name, val)	IPSTAT_ADD(name, -(val))
150#define	IPSTAT_INC(name)	IPSTAT_ADD(name, 1)
151#define	IPSTAT_DEC(name)	IPSTAT_SUB(name, 1)
152
153/*
154 * Kernel module consumers must use this accessor macro.
155 */
156void	kmod_ipstat_inc(int statnum);
157#define	KMOD_IPSTAT_INC(name)	\
158    kmod_ipstat_inc(offsetof(struct ipstat, name) / sizeof(uint64_t))
159void	kmod_ipstat_dec(int statnum);
160#define	KMOD_IPSTAT_DEC(name)	\
161    kmod_ipstat_dec(offsetof(struct ipstat, name) / sizeof(uint64_t))
162
163/* flags passed to ip_output as last parameter */
164#define	IP_FORWARDING		0x1		/* most of ip header exists */
165#define	IP_RAWOUTPUT		0x2		/* raw ip header exists */
166#define	IP_SENDONES		0x4		/* send all-ones broadcast */
167#define	IP_SENDTOIF		0x8		/* send on specific ifnet */
168#define IP_ROUTETOIF		SO_DONTROUTE	/* 0x10 bypass routing tables */
169#define IP_ALLOWBROADCAST	SO_BROADCAST	/* 0x20 can send broadcast packets */
170#define	IP_NODEFAULTFLOWID	0x40		/* Don't set the flowid from inp */
171#define IP_NO_SND_TAG_RL	0x80		/* Don't send down the ratelimit tag */
172
173#ifdef __NO_STRICT_ALIGNMENT
174#define IP_HDR_ALIGNED_P(ip)	1
175#else
176#define IP_HDR_ALIGNED_P(ip)	((((intptr_t) (ip)) & 3) == 0)
177#endif
178
179struct ip;
180struct inpcb;
181struct route;
182struct sockopt;
183struct inpcbinfo;
184
185VNET_DECLARE(int, ip_defttl);			/* default IP ttl */
186VNET_DECLARE(int, ipforwarding);		/* ip forwarding */
187VNET_DECLARE(int, ipsendredirects);
188#ifdef IPSTEALTH
189VNET_DECLARE(int, ipstealth);			/* stealth forwarding */
190#endif
191VNET_DECLARE(struct socket *, ip_rsvpd);	/* reservation protocol daemon*/
192VNET_DECLARE(struct socket *, ip_mrouter);	/* multicast routing daemon */
193extern int	(*legal_vif_num)(int);
194extern u_long	(*ip_mcast_src)(int);
195VNET_DECLARE(int, rsvp_on);
196VNET_DECLARE(int, drop_redirect);
197
198#define	V_ip_id			VNET(ip_id)
199#define	V_ip_defttl		VNET(ip_defttl)
200#define	V_ipforwarding		VNET(ipforwarding)
201#define	V_ipsendredirects	VNET(ipsendredirects)
202#ifdef IPSTEALTH
203#define	V_ipstealth		VNET(ipstealth)
204#endif
205#define	V_ip_rsvpd		VNET(ip_rsvpd)
206#define	V_ip_mrouter		VNET(ip_mrouter)
207#define	V_rsvp_on		VNET(rsvp_on)
208#define	V_drop_redirect		VNET(drop_redirect)
209
210void	inp_freemoptions(struct ip_moptions *);
211int	inp_getmoptions(struct inpcb *, struct sockopt *);
212int	inp_setmoptions(struct inpcb *, struct sockopt *);
213
214int	ip_ctloutput(struct socket *, struct sockopt *sopt);
215int	ip_fragment(struct ip *ip, struct mbuf **m_frag, int mtu,
216	    u_long if_hwassist_flags);
217void	ip_forward(struct mbuf *m, int srcrt);
218extern int
219	(*ip_mforward)(struct ip *, struct ifnet *, struct mbuf *,
220	    struct ip_moptions *);
221int	ip_output(struct mbuf *,
222	    struct mbuf *, struct route *, int, struct ip_moptions *,
223	    struct inpcb *);
224struct mbuf *
225	ip_reass(struct mbuf *);
226void	ip_savecontrol(struct inpcb *, struct mbuf **, struct ip *,
227	    struct mbuf *);
228void	ip_fillid(struct ip *);
229int	rip_ctloutput(struct socket *, struct sockopt *);
230int	ipip_input(struct mbuf **, int *, int);
231int	rsvp_input(struct mbuf **, int *, int);
232
233int	ip_rsvp_init(struct socket *);
234int	ip_rsvp_done(void);
235extern int	(*ip_rsvp_vif)(struct socket *, struct sockopt *);
236extern void	(*ip_rsvp_force_done)(struct socket *);
237extern int	(*rsvp_input_p)(struct mbuf **, int *, int);
238
239typedef int	ipproto_input_t(struct mbuf **, int *, int);
240struct icmp;
241typedef void	ipproto_ctlinput_t(struct icmp *);
242int	ipproto_register(uint8_t, ipproto_input_t, ipproto_ctlinput_t);
243int	ipproto_unregister(uint8_t);
244#define	IPPROTO_REGISTER(prot, input, ctl)	do {			\
245	int error __diagused;						\
246	error = ipproto_register(prot, input, ctl);			\
247	MPASS(error == 0);						\
248} while (0)
249
250ipproto_input_t		rip_input;
251ipproto_ctlinput_t	rip_ctlinput;
252
253VNET_DECLARE(struct pfil_head *, inet_pfil_head);
254#define	V_inet_pfil_head	VNET(inet_pfil_head)
255#define	PFIL_INET_NAME		"inet"
256
257VNET_DECLARE(struct pfil_head *, inet_local_pfil_head);
258#define	V_inet_local_pfil_head	VNET(inet_local_pfil_head)
259#define	PFIL_INET_LOCAL_NAME	"inet-local"
260
261void	in_delayed_cksum(struct mbuf *m);
262
263/* Hooks for ipfw, dummynet, divert etc. Most are declared in raw_ip.c */
264/*
265 * Reference to an ipfw or packet filter rule that can be carried
266 * outside critical sections.
267 * A rule is identified by rulenum:rule_id which is ordered.
268 * In version chain_id the rule can be found in slot 'slot', so
269 * we don't need a lookup if chain_id == chain->id.
270 *
271 * On exit from the firewall this structure refers to the rule after
272 * the matching one (slot points to the new rule; rulenum:rule_id-1
273 * is the matching rule), and additional info (e.g. info often contains
274 * the insn argument or tablearg in the low 16 bits, in host format).
275 * On entry, the structure is valid if slot>0, and refers to the starting
276 * rules. 'info' contains the reason for reinject, e.g. divert port,
277 * divert direction, and so on.
278 *
279 * Packet Mark is an analogue to ipfw tags with O(1) lookup from mbuf while
280 * regular tags require a single-linked list traversal. Mark is a 32-bit
281 * number that can be looked up in a table [with 'number' table-type], matched
282 * or compared with a number with optional mask applied before comparison.
283 * Having generic nature, Mark can be used in a variety of needs.
284 * For example, it could be used as a security group: mark will hold a
285 * security group id and represent a group of packet flows that shares same
286 * access control policy.
287 * O_MASK opcode can match mark value bitwise so one can build a hierarchical
288 * model designating different meanings for a bit range(s).
289 */
290struct ipfw_rule_ref {
291/* struct m_tag spans 24 bytes above this point, see mbuf_tags(9) */
292	/* spare space just to be save in case struct m_tag grows */
293/* -- 32 bytes -- */
294	uint32_t	slot;		/* slot for matching rule	*/
295	uint32_t	rulenum;	/* matching rule number		*/
296	uint32_t	rule_id;	/* matching rule id		*/
297	uint32_t	chain_id;	/* ruleset id			*/
298	uint32_t	info;		/* see below			*/
299	uint32_t	pkt_mark;	/* packet mark			*/
300	uint32_t	spare[2];
301/* -- 64 bytes -- */
302};
303
304enum {
305	IPFW_INFO_MASK	= 0x0000ffff,
306	IPFW_INFO_OUT	= 0x00000000,	/* outgoing, just for convenience */
307	IPFW_INFO_IN	= 0x80000000,	/* incoming, overloads dir */
308	IPFW_ONEPASS	= 0x40000000,	/* One-pass, do not reinject */
309	IPFW_IS_MASK	= 0x30000000,	/* which source ? */
310	IPFW_IS_DIVERT	= 0x20000000,
311	IPFW_IS_DUMMYNET =0x10000000,
312	IPFW_IS_PIPE	= 0x08000000,	/* pipe=1, queue = 0 */
313};
314#define MTAG_IPFW	1148380143	/* IPFW-tagged cookie */
315#define MTAG_IPFW_RULE	1262273568	/* rule reference */
316#define	MTAG_IPFW_CALL	1308397630	/* call stack */
317
318struct ip_fw_args;
319typedef int	(*ip_fw_ctl_ptr_t)(struct sockopt *);
320VNET_DECLARE(ip_fw_ctl_ptr_t, ip_fw_ctl_ptr);
321#define	V_ip_fw_ctl_ptr		VNET(ip_fw_ctl_ptr)
322
323/* Divert hooks. */
324extern void	(*ip_divert_ptr)(struct mbuf *m, bool incoming);
325/* ng_ipfw hooks -- XXX make it the same as divert and dummynet */
326extern int	(*ng_ipfw_input_p)(struct mbuf **, struct ip_fw_args *, bool);
327extern int	(*ip_dn_ctl_ptr)(struct sockopt *);
328extern int	(*ip_dn_io_ptr)(struct mbuf **, struct ip_fw_args *);
329
330/* pf specific mtag for divert(4) support */
331__enum_uint8_decl(pf_mtag_dir) {
332	PF_DIVERT_MTAG_DIR_IN = 1,
333	PF_DIVERT_MTAG_DIR_OUT = 2
334};
335struct pf_divert_mtag {
336	__enum_uint8(pf_mtag_dir) idir;		/* initial pkt direction */
337	union {
338		__enum_uint8(pf_mtag_dir) ndir;	/* new dir after re-enter */
339		uint16_t port;			/* initial divert(4) port */
340	};
341};
342#define MTAG_PF_DIVERT	1262273569
343
344#endif /* _KERNEL */
345
346#endif /* !_NETINET_IP_VAR_H_ */