xref: /freebsd/share/man/man9/pfil.9 (revision 74ca7bf1d4c7173d5575ba168bc4b5f6d181ff5a)
1.\"	$NetBSD: pfil.9,v 1.22 2003/07/01 13:04:06 wiz Exp $
2.\"
3.\" Copyright (c) 1996 Matthew R. Green
4.\" All rights reserved.
5.\"
6.\" Redistribution and use in source and binary forms, with or without
7.\" modification, are permitted provided that the following conditions
8.\" are met:
9.\" 1. Redistributions of source code must retain the above copyright
10.\"    notice, this list of conditions and the following disclaimer.
11.\" 2. Redistributions in binary form must reproduce the above copyright
12.\"    notice, this list of conditions and the following disclaimer in the
13.\"    documentation and/or other materials provided with the distribution.
14.\" 3. The name of the author may not be used to endorse or promote products
15.\"    derived from this software without specific prior written permission.
16.\"
17.\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
18.\" IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
19.\" OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
20.\" IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
21.\" INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
22.\" BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
23.\" LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
24.\" AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
25.\" OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
27.\" SUCH DAMAGE.
28.\"
29.\" $FreeBSD$
30.\"
31.Dd March 10, 2018
32.Dt PFIL 9
33.Os
34.Sh NAME
35.Nm pfil ,
36.Nm pfil_head_register ,
37.Nm pfil_head_unregister ,
38.Nm pfil_head_get ,
39.Nm pfil_add_hook ,
40.Nm pfil_add_hook_flags ,
41.Nm pfil_remove_hook ,
42.Nm pfil_remove_hook_flags ,
43.Nm pfil_run_hooks ,
44.Nm pfil_rlock ,
45.Nm pfil_runlock ,
46.Nm pfil_wlock ,
47.Nm pfil_wunlock
48.Nd packet filter interface
49.Sh SYNOPSIS
50.In sys/param.h
51.In sys/mbuf.h
52.In net/if.h
53.In net/pfil.h
54.Bd -literal
55typedef int (*pfil_func_t)(void *arg, struct mbuf **mp, struct ifnet *, int dir, struct inpcb);
56.Bd -literal
57typedef int (*pfil_func_flags_t)(void *arg, struct mbuf **mp, struct ifnet *, int dir, int flags, struct inpcb);
58.Ft int
59.Fn pfil_head_register "struct pfil_head *head"
60.Ft int
61.Fn pfil_head_unregister "struct pfil_head *head"
62.Ft "struct pfil_head *"
63.Fn pfil_head_get "int af" "u_long dlt"
64.Ft int
65.Fn pfil_add_hook "pfil_func_t" "void *arg" "struct pfil_head *"
66.Ft int
67.Fn pfil_add_hook_flags "pfil_func_flags_t" "void *arg" "int flags" "struct pfil_head *"
68.Ft int
69.Fn pfil_remove_hook "pfil_func_t" "void *arg" "struct pfil_head *"
70.Ft int
71.Fn pfil_remove_hook_flags "pfil_func_flags_t" "void *arg" "int flags" "struct pfil_head *"
72.Ft int
73.Fn pfil_run_hooks "struct pfil_head *head" "struct mbuf **mp" "struct ifnet *" "int dir" "int flags" "struct inpcb *"
74.Ft void
75.Fn pfil_rlock "struct pfil_head *" "struct rm_priotracker *"
76.Ft void
77.Fn pfil_runlock "struct pfil_head *" "struct rm_priotracker *"
78.Ft void
79.Fn pfil_wlock "struct pfil_head *"
80.Ft void
81.Fn pfil_wunlock "struct pfil_head *"
82.Ed
83.Sh DESCRIPTION
84The
85.Nm
86framework allows for a specified function to be invoked for every
87incoming or outgoing packet for a particular network I/O stream.
88These hooks may be used to implement a firewall or perform packet
89transformations.
90.Pp
91Packet filtering points are registered with
92.Fn pfil_head_register .
93Filtering points are identified by a key
94.Pq Vt "void *"
95and a data link type
96.Pq Vt int
97in the
98.Vt pfil_head
99structure.
100Packet filters use the key and data link type to look up the filtering
101point with which they register themselves.
102The key is unique to the filtering point.
103The data link type is a
104.Xr bpf 4
105DLT constant indicating what kind of header is present on the packet
106at the filtering point.
107Each filtering point uses common per-VNET rmlock by default.
108This can be changed by specifying
109.Vt PFIL_FLAG_PRIVATE_LOCK
110as
111.Vt "flags"
112field in the
113.Vt pfil_head
114structure.
115Note that specifying private lock can break filters sharing the same
116ruleset and/or state between different data link types.
117Filtering points may be unregistered with the
118.Fn pfil_head_unregister
119function.
120.Pp
121Packet filters register/unregister themselves with a filtering point
122with the
123.Fn pfil_add_hook
124and
125.Fn pfil_remove_hook
126functions, respectively.
127.I
128The head is looked up using the
129.Fn pfil_head_get
130function, which takes the key and data link type that the packet filter
131expects.
132Filters may provide an argument to be passed to the filter when
133invoked on a packet.
134.Pp
135When a filter is invoked, the packet appears just as if it
136.Dq came off the wire .
137That is, all protocol fields are in network byte order.
138The filter is called with its specified argument, the pointer to the
139pointer to the
140.Vt mbuf
141containing the packet, the pointer to the network
142interface that the packet is traversing, and the direction
143.Dv ( PFIL_IN
144or
145.Dv PFIL_OUT )
146that the packet is traveling.
147The
148.Vt flags
149argument will indicate if an outgoing packet is simply being forwarded with the
150value PFIL_FWD.
151The filter may change which mbuf the
152.Vt "mbuf\ **"
153argument references.
154The filter returns an error (errno) if the packet processing is to stop, or 0
155if the processing is to continue.
156If the packet processing is to stop, it is the responsibility of the
157filter to free the packet.
158.Pp
159Every filter hook is called with
160.Nm
161read lock held.
162All heads uses the same lock within the same VNET instance.
163Packet filter can use this lock instead of own locking model to
164improve performance.
165Since
166.Nm
167uses
168.Xr rmlock 9
169.Fn pfil_rlock
170and
171.Fn pfil_runlock
172require
173.Va struct rm_priotracker
174to be passed as argument.
175Filter can acquire and release writer lock via
176.Fn pfil_wlock
177and
178.Fn pfil_wunlock
179functions.
180See
181.Xr rmlock 9
182for more details.
183.Sh FILTERING POINTS
184Currently, filtering points are implemented for the following link types:
185.Pp
186.Bl -tag -width "AF_INET6" -offset XXX -compact
187.It AF_INET
188IPv4 packets.
189.It AF_INET6
190IPv6 packets.
191.It AF_LINK
192Link-layer packets.
193.El
194.Sh RETURN VALUES
195If successful,
196.Fn pfil_head_get
197returns the
198.Vt pfil_head
199structure for the given key/dlt.
200The
201.Fn pfil_add_hook
202and
203.Fn pfil_remove_hook
204functions
205return 0 if successful.
206If called with flag
207.Dv PFIL_WAITOK ,
208.Fn pfil_remove_hook
209is expected to always succeed.
210.Pp
211The
212.Fn pfil_head_unregister
213function
214might sleep!
215.Sh SEE ALSO
216.Xr bpf 4 ,
217.Xr if_bridge 4 ,
218.Xr rmlock 9
219.Sh HISTORY
220The
221.Nm
222interface first appeared in
223.Nx 1.3 .
224The
225.Nm
226input and output lists were originally implemented as
227.In sys/queue.h
228.Dv LIST
229structures;
230however this was changed in
231.Nx 1.4
232to
233.Dv TAILQ
234structures.
235This change was to allow the input and output filters to be processed in
236reverse order, to allow the same path to be taken, in or out of the kernel.
237.Pp
238The
239.Nm
240interface was changed in 1.4T to accept a 3rd parameter to both
241.Fn pfil_add_hook
242and
243.Fn pfil_remove_hook ,
244introducing the capability of per-protocol filtering.
245This was done primarily in order to support filtering of IPv6.
246.Pp
247In 1.5K, the
248.Nm
249framework was changed to work with an arbitrary number of filtering points,
250as well as be less IP-centric.
251.Pp
252Fine-grained locking was added in
253.Fx 5.2 .
254.Nm
255lock export was added in
256.Fx 10.0 .
257.Sh BUGS
258When a
259.Vt pfil_head
260is being modified, no traffic is diverted
261(to avoid deadlock).
262This means that traffic may be dropped unconditionally for a short period
263of time.
264.Fn pfil_run_hooks
265will return
266.Er ENOBUFS
267to indicate this.
268