1 .\" $OpenBSD: pf.conf.5,v 1.292 2004/02/24 05:44:48 mcbride Exp $
3 .\" Copyright (c) 2002, Daniel Hartmeier
4 .\" All rights reserved.
6 .\" Redistribution and use in source and binary forms, with or without
7 .\" modification, are permitted provided that the following conditions
10 .\" - Redistributions of source code must retain the above copyright
11 .\" notice, this list of conditions and the following disclaimer.
12 .\" - Redistributions in binary form must reproduce the above
13 .\" copyright notice, this list of conditions and the following
14 .\" disclaimer in the documentation and/or other materials provided
15 .\" with the distribution.
17 .\" THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
18 .\" "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
19 .\" LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
20 .\" FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
21 .\" COPYRIGHT HOLDERS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
22 .\" INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
23 .\" BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
24 .\" LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
25 .\" CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
26 .\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
27 .\" ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
28 .\" POSSIBILITY OF SUCH DAMAGE.
35 .Nd packet filter configuration file
39 packet filter modifies, drops or passes packets according to rules or
40 definitions specified in
43 There are seven types of statements in
47 User-defined variables may be defined and used later, simplifying
48 the configuration file.
49 Macros must be defined before they are referenced in
52 Tables provide a mechanism for increasing the performance and flexibility of
53 rules with large numbers of source or destination addresses.
55 Options tune the behaviour of the packet filtering engine.
56 .It Cm Traffic Normalization Li (e.g. Em scrub )
57 Traffic normalization protects internal machines against inconsistencies
58 in Internet protocols and implementations.
60 Queueing provides rule-based bandwidth control.
61 .It Cm Translation Li (Various forms of NAT)
62 Translation rules specify how addresses are to be mapped or redirected to
64 .It Cm Packet Filtering
65 Stateful and stateless packet filtering provides rule-based blocking or
73 the types of statements should be grouped and appear in
75 in the order shown above, as this matches the operation of the underlying
76 packet filtering engine.
79 enforces this order (see
87 macros can be defined that will later be expanded in context.
88 Macro names must start with a letter, and may contain letters, digits
90 Macro names may not be reserved words (for example
94 Macros are not expanded inside quotes.
97 .Bd -literal -offset indent
99 all_ifs = \&"{\&" $ext_if lo0 \&"}\&"
100 pass out on $ext_if from any to any keep state
101 pass in on $ext_if proto tcp from any to any port 25 keep state
104 Tables are named structures which can hold a collection of addresses and
106 Lookups against tables in
108 are relatively fast, making a single rule with tables much more efficient,
110 processor usage and memory consumption, than a large number of rules which
111 differ only in IP address (either created explicitly or automatically by rule
114 Tables can be used as the source or destination of filter rules,
118 translation rules such as
122 (see below for details on the various rule types).
123 Tables can also be used for the redirect address of
127 rules and in the routing options of filter rules, but only for
131 Tables can be defined with any of the following
134 As with macros, reserved words may not be used as table names.
135 .Bl -tag -width "manually"
137 Persistent tables can be manually created with the
143 before or after the ruleset has been loaded.
145 Table definitions can be placed directly in this file, and loaded at the
146 same time as other rules are loaded, atomically.
147 Table definitions inside
151 statement, and are especially useful to define non-persistent tables.
152 The contents of a pre-existing table defined without a list of addresses
153 to initialize it is not altered when
156 A table initialized with the empty list,
158 will be cleared on load.
161 Tables may be defined with the following two attributes:
162 .Bl -tag -width persist
166 flag forces the kernel to keep the table even when no rules refer to it.
167 If the flag is not set, the kernel will automatically remove the table
168 when the last rule referring to it is flushed.
172 flag prevents the user from altering the contents of the table once it
176 can be used to add or remove addresses from the table at any time, even
183 .Bd -literal -offset indent
184 table <private> const { 10/8, 172.16/12, 192.168/16 }
185 table <badhosts> persist
186 block on fxp0 from { <private>, <badhosts> } to any
189 creates a table called private, to hold RFC 1918 private network
190 blocks, and a table called badhosts, which is initially empty.
191 A filter rule is set up to block all traffic coming from addresses listed in
193 The private table cannot have its contents changed and the badhosts table
194 will exist even when no active filter rules reference it.
195 Addresses may later be added to the badhosts table, so that traffic from
196 these hosts can be blocked by using
197 .Bd -literal -offset indent
198 # pfctl -t badhosts -Tadd 204.92.77.111
201 A table can also be initialized with an address list specified in one or more
202 external files, using the following syntax:
203 .Bd -literal -offset indent
204 table <spam> persist file \&"/etc/spammers\&" file \&"/etc/openrelays\&"
205 block on fxp0 from <spam> to any
212 list IP addresses, one per line.
213 Any lines beginning with a # are treated as comments and ignored.
214 In addition to being specified by IP address, hosts may also be
215 specified by their hostname.
216 When the resolver is called to add a hostname to a table,
218 resulting IPv4 and IPv6 addresses are placed into the table.
219 IP addresses can also be entered in a table by specifying a valid interface
222 keyword, in which case all addresses assigned to the interface(s) will be
226 may be tuned for various situations using the
232 .Bl -tag -width interval -compact
234 Interval between purging expired states and fragments.
236 Seconds before an unassembled fragment is expired.
238 Length of time to retain a source tracking entry after the last state
242 When a packet matches a stateful connection, the seconds to live for the
243 connection will be updated to that of the
245 which corresponds to the connection state.
246 Each packet which matches this state will reset the TTL.
247 Tuning these values may improve the performance of the
248 firewall at the risk of dropping valid idle connections.
250 .Bl -tag -width xxxx -compact
252 The state after the first packet.
254 The state before the destination host ever sends a packet.
255 .It Ar tcp.established
256 The fully established state.
258 The state after the first FIN has been sent.
260 The state after both FINs have been exchanged and the connection is closed.
261 Some hosts (notably web servers on Solaris) send TCP packets even after closing
267 can prevent blocking of such packets.
269 The state after one endpoint sends an RST.
272 ICMP and UDP are handled in a fashion similar to TCP, but with a much more
273 limited set of states:
275 .Bl -tag -width xxxx -compact
277 The state after the first packet.
279 The state if the source host sends more than one packet but the destination
280 host has never sent one back.
282 The state if both hosts have sent packets.
284 The state after the first packet.
286 The state after an ICMP error came back in response to an ICMP packet.
289 Other protocols are handled similarly to UDP:
291 .Bl -tag -width xxxx -compact
294 .It Ar other.multiple
297 Timeout values can be reduced adaptively as the number of state table
300 .Bl -tag -width xxxx -compact
301 .It Ar adaptive.start
302 When the number of state entries exceeds this value, adaptive scaling
304 All timeout values are scaled linearly with factor
305 (adaptive.end - number of states) / (adaptive.end - adaptive.start).
307 When reaching this number of state entries, all timeout values become
308 zero, effectively purging all state entries immediately.
309 This value is used to define the scale factor, it should not actually
310 be reached (set a lower state limit, see below).
313 These values can be defined both globally and for each rule.
314 When used on a per-rule basis, the values relate to the number of
315 states created by the rule, otherwise to the total number of
319 .Bd -literal -offset indent
320 set timeout tcp.first 120
321 set timeout tcp.established 86400
322 set timeout { adaptive.start 6000, adaptive.end 12000 }
323 set limit states 10000
326 With 9000 state table entries, the timeout values are scaled to 50%
327 (tcp.first 60, tcp.established 43200).
329 .It Ar set loginterface
330 Enable collection of packet and byte count statistics for the given interface.
331 These statistics can be viewed using
332 .Bd -literal -offset indent
338 collects statistics on the interface named dc0:
339 .Bd -literal -offset indent
343 One can disable the loginterface using:
344 .Bd -literal -offset indent
345 set loginterface none
349 Sets hard limits on the memory pools used by the packet filter.
352 for an explanation of memory pools.
355 .Bd -literal -offset indent
356 set limit states 20000
359 sets the maximum number of entries in the memory pool used by state table
360 entries (generated by
364 .Bd -literal -offset indent
365 set limit frags 20000
368 sets the maximum number of entries in the memory pool used for fragment
369 reassembly (generated by
373 .Bd -literal -offset indent
374 set limit src-nodes 2000
377 sets the maximum number of entries in the memory pool used for tracking
378 source IP addresses (generated by the
384 These can be combined:
385 .Bd -literal -offset indent
386 set limit { states 20000, frags 20000, src-nodes 2000 }
389 .It Ar set optimization
390 Optimize the engine for one of the following network environments:
392 .Bl -tag -width xxxx -compact
394 A normal network environment.
395 Suitable for almost all networks.
397 A high-latency environment (such as a satellite connection).
402 Aggressively expire connections.
403 This can greatly reduce the memory usage of the firewall at the cost of
404 dropping idle connections early.
406 Extremely conservative settings.
407 Avoid dropping legitimate connections at the
408 expense of greater memory utilization (possibly much greater on a busy
409 network) and slightly increased processor utilization.
413 .Bd -literal -offset indent
414 set optimization aggressive
417 .It Ar set block-policy
420 option sets the default behaviour for the packet
424 .Bl -tag -width xxxxxxxx -compact
426 Packet is silently dropped.
428 A TCP RST is returned for blocked TCP packets,
429 an ICMP UNREACHABLE is returned for blocked UDP packets,
430 and all other packets are silently dropped.
434 .Bd -literal -offset indent
435 set block-policy return
437 .It Ar set state-policy
440 option sets the default behaviour for states:
442 .Bl -tag -width group-bound -compact
444 States are bound to interface.
446 States are bound to interface group (i.e. ppp)
448 States can match packets on any interfaces (the default).
452 .Bd -literal -offset indent
453 set state-policy if-bound
455 .It Ar set require-order
458 enforces an ordering of the statement types in the ruleset to:
464 Setting this option to
466 disables this enforcement.
467 There may be non-trivial and non-obvious implications to an out of
469 Consider carefully before disabling the order enforcement.
470 .It Ar set fingerprints
471 Load fingerprints of known operating systems from the given filename.
472 By default fingerprints of known operating systems are automatically
477 but can be overridden via this option.
478 Setting this option may leave a small period of time where the fingerprints
479 referenced by the currently active ruleset are inconsistent until the new
480 ruleset finishes loading.
484 .Dl set fingerprints \&"/etc/pf.os.devel\&"
489 to one of the following:
491 .Bl -tag -width xxxxxxxxxxxx -compact
493 Don't generate debug messages.
495 Generate debug messages only for serious errors.
497 Generate debug messages for various errors.
499 Generate debug messages for common conditions.
502 .Sh TRAFFIC NORMALIZATION
503 Traffic normalization is used to sanitize packet content in such
504 a way that there are no ambiguities in packet interpretation on
506 The normalizer does IP fragment reassembly to prevent attacks
507 that confuse intrusion detection systems by sending overlapping
509 Packet normalization is invoked with the
514 has the following options:
519 bit from a matching IP packet.
520 Some operating systems are known to generate fragmented packets with the
523 This is particularly true with NFS.
525 will drop such fragmented
531 Unfortunately some operating systems also generate their
533 packets with a zero IP identification field.
536 bit on packets with a zero IP ID may cause deleterious results if an
537 upstream router later fragments the packet.
540 modifier (see below) is recommended in combination with the
542 modifier to ensure unique IP identifiers.
543 .It Ar min-ttl <number>
544 Enforces a minimum TTL for matching IP packets.
545 .It Ar max-mss <number>
546 Enforces a maximum MSS for matching TCP packets.
548 Replaces the IP identification field with random values to compensate
549 for predictable values generated by many hosts.
550 This option only applies to outgoing packets that are not fragmented
551 after the optional fragment reassembly.
552 .It Ar fragment reassemble
555 rules, fragments can be reassembled by normalization.
556 In this case, fragments are buffered until they form a complete
557 packet, and only the completed packet is passed on to the filter.
558 The advantage is that filter rules have to deal only with complete
559 packets, and can ignore fragments.
560 The drawback of caching fragments is the additional memory cost.
561 But the full reassembly method is the only method that currently works
563 This is the default behavior of a
565 rule if no fragmentation modifier is supplied.
567 The default fragment reassembly method is expensive, hence the option
571 will track the fragments and cache a small range descriptor.
572 Duplicate fragments are dropped and overlaps are cropped.
573 Thus data will only occur once on the wire with ambiguities resolving to
574 the first occurrence.
576 .Ar fragment reassemble
577 modifier, fragments are not buffered, they are passed as soon as they
581 reassembly mechanism does not yet work with NAT.
583 .It Ar fragment drop-ovl
584 This option is similar to the
586 modifier except that all overlapping or duplicate fragments will be
587 dropped, and all further corresponding fragments will be
589 .It Ar reassemble tcp
590 Statefully normalizes TCP connections.
591 .Ar scrub reassemble tcp
592 rules may not have the direction (in/out) specified.
594 performs the following normalizations:
596 .Bl -tag -width timeout -compact
598 Neither side of the connection is allowed to reduce their IP TTL.
599 An attacker may send a packet such that it reaches the firewall, affects
600 the firewall state, and expires before reaching the destination host.
602 will raise the TTL of all packets back up to the highest value seen on
604 .It timeout modulation
605 Modern TCP stacks will send a timestamp on every TCP packet and echo
606 the other endpoint's timestamp back to them.
607 Many operating systems will merely start the timestamp at zero when
608 first booted, and increment it several times a second.
609 The uptime of the host can be deduced by reading the timestamp and multiplying
611 Also observing several different timestamps can be used to count hosts
613 And spoofing TCP packets into a connection requires knowing or guessing
615 Timestamps merely need to be monotonically increasing and not derived off a
620 to modulate the TCP timestamps with a random number.
625 .Bd -literal -offset indent
626 scrub in on $ext_if all fragment reassemble
629 Packets can be assigned to queues for the purpose of bandwidth
631 At least two declarations are required to configure queues, and later
632 any packet filtering rule can reference the defined queues by name.
633 During the filtering component of
637 name is where any packets from
639 rules will be queued, while for
641 rules it specifies where any resulting ICMP or TCP RST
642 packets should be queued.
645 defines the algorithm used to decide which packets get delayed, dropped, or
646 sent out immediately.
652 Class Based Queueing.
654 attached to an interface build a tree, thus each
656 can have further child
658 Each queue can have a
664 mainly controls the time packets take to get sent out, while
666 has primarily effects on throughput.
670 are flat attached to the interface, thus,
672 cannot have further child
678 assigned, ranging from 0 to 15.
685 Hierarchical Fair Service Curve.
687 attached to an interface build a tree, thus each
689 can have further child
691 Each queue can have a
697 mainly controls the time packets take to get sent out, while
699 has primarily effects on throughput.
702 The interfaces on which queueing should be activated are declared using
707 has the following keywords:
710 Queueing is enabled on the named interface.
712 Specifies which queueing scheduler to use.
713 Currently supported values
716 for Class Based Queueing,
718 for Priority Queueing and
720 for the Hierarchical Fair Service Curve scheduler.
721 .It Ar bandwidth <bw>
722 The maximum bitrate for all queues on an
723 interface may be specified using the
726 The value can be specified as an absolute value or as a
727 percentage of the interface bandwidth.
728 When using an absolute value, the suffixes
734 are used to represent bits, kilobits, megabits, and
735 gigabits per second, respectively.
736 The value must not exceed the interface bandwidth.
739 is not specified, the interface bandwidth is used.
740 .It Ar qlimit <limit>
741 The maximum number of packets held in the queue.
743 .It Ar tbrsize <size>
744 Adjusts the size, in bytes, of the token bucket regulator.
745 If not specified, heuristics based on the
746 interface bandwidth are used to determine the size.
748 Defines a list of subqueues to create on an interface.
751 In the following example, the interface dc0
752 should queue up to 5 Mbit/s in four second-level queues using
753 Class Based Queueing.
754 Those four queues will be shown in a later example.
755 .Bd -literal -offset indent
756 altq on dc0 cbq bandwidth 5Mb queue { std, http, mail, ssh }
759 Once interfaces are activated for queueing using the
761 directive, a sequence of
763 directives may be defined.
764 The name associated with a
766 must match a queue defined in the
768 directive (e.g. mail), or, except for the
774 The following keywords can be used:
776 .It Ar on <interface>
777 Specifies the interface the queue operates on.
778 If not given, it operates on all matching interfaces.
779 .It Ar bandwidth <bw>
780 Specifies the maximum bitrate to be processed by the queue.
781 This value must not exceed the value of the parent
783 and can be specified as an absolute value or a percentage of the parent
787 scheduler does not support bandwidth specification.
788 .It Ar priority <level>
789 Between queues a priority level can be set.
794 the range is 0 to 7 and for
796 the range is 0 to 15.
797 The default for all is 1.
799 queues with a higher priority are always served first.
803 queues with a higher priority are preferred in the case of overload.
804 .It Ar qlimit <limit>
805 The maximum number of packets held in the queue.
811 can get additional parameters with
812 .Ar <scheduler> Ns Li (\& Ar <parameters> No ) .
813 Parameters are as follows:
816 Packets not matched by another queue are assigned to this one.
817 Exactly one default queue is required.
819 Enable RED (Random Early Detection) on this queue.
820 RED drops packets with a probability proportional to the average
823 Enables RIO on this queue.
824 RIO is RED with IN/OUT, thus running
825 RED two times more than RIO would achieve the same effect.
826 RIO is currently not supported in the GENERIC kernel.
828 Enables ECN (Explicit Congestion Notification) on this queue.
835 supports an additional option:
838 The queue can borrow bandwidth from the parent.
844 supports some additional options:
847 The minimum required bandwidth for the queue.
848 .It Ar upperlimit <sc>
849 The maximum allowed bandwidth for the queue.
850 .It Ar linkshare <sc>
851 The bandwidth share of a backlogged queue.
854 <sc> is an acronym for
857 The format for service curve specifications is
858 .Ar ( m1 , d , m2 ) .
860 controls the bandwidth assigned to the queue.
864 are optional and can be used to control the initial bandwidth assignment.
867 milliseconds the queue gets the bandwidth given as
869 afterwards the value given in
876 child queues can be specified as in an
878 declaration, thus building a tree of queues using a part of
879 their parent's bandwidth.
881 Packets can be assigned to queues based on filter rules by using the
886 is specified; when a second one is specified it will instead be used for
891 and for TCP ACKs with no data payload.
893 To continue the previous example, the examples below would specify the
895 queues, plus a few child queues.
898 sessions get priority over bulk transfers like
902 The queues may then be referenced by filtering rules (see
906 queue std bandwidth 10% cbq(default)
907 queue http bandwidth 60% priority 2 cbq(borrow red) \e
908 { employees, developers }
909 queue developers bandwidth 75% cbq(borrow)
910 queue employees bandwidth 15%
911 queue mail bandwidth 10% priority 0 cbq(borrow ecn)
912 queue ssh bandwidth 20% cbq(borrow) { ssh_interactive, ssh_bulk }
913 queue ssh_interactive priority 7
914 queue ssh_bulk priority 0
916 block return out on dc0 inet all queue std
917 pass out on dc0 inet proto tcp from $developerhosts to any port 80 \e
918 keep state queue developers
919 pass out on dc0 inet proto tcp from $employeehosts to any port 80 \e
920 keep state queue employees
921 pass out on dc0 inet proto tcp from any to any port 22 \e
922 keep state queue(ssh_bulk, ssh_interactive)
923 pass out on dc0 inet proto tcp from any to any port 25 \e
924 keep state queue mail
927 Translation rules modify either the source or destination address of the
928 packets associated with a stateful connection.
929 A stateful connection is automatically created to track packets matching
930 such a rule as long as they are not blocked by the filtering section of
932 The translation engine modifies the specified address and/or port in the
933 packet, recalculates IP, TCP and UDP checksums as necessary, and passes it to
934 the packet filter for evaluation.
936 Since translation occurs before filtering the filter
937 engine will see packets as they look after any
938 addresses and ports have been translated. Filter rules
939 will therefore have to filter based on the translated
940 address and port number.
941 Packets that match a translation rule are only automatically passed if
944 modifier is given, otherwise they are
951 The state entry created permits
953 to keep track of the original address for traffic associated with that state
954 and correctly direct return traffic for that connection.
956 Various types of translation are possible with pf:
961 rule specifies a bidirectional mapping between an external IP netblock
962 and an internal IP netblock.
966 rule specifies that IP addresses are to be changed as the packet
967 traverses the given interface.
968 This technique allows one or more IP addresses
969 on the translating host to support network traffic for a larger range of
970 machines on an "inside" network.
971 Although in theory any IP address can be used on the inside, it is strongly
972 recommended that one of the address ranges defined by RFC 1918 be used.
975 10.0.0.0 - 10.255.255.255 (all of net 10, i.e., 10/8)
976 172.16.0.0 - 172.31.255.255 (i.e., 172.16/12)
977 192.168.0.0 - 192.168.255.255 (i.e., 192.168/16)
980 The packet is redirected to another destination and possibly a
983 rules can optionally specify port ranges instead of single ports.
984 rdr ... port 2000:2999 -> ... port 4000
985 redirects ports 2000 to 2999 (inclusive) to port 4000.
986 rdr ... port 2000:2999 -> ... port 4000:*
987 redirects port 2000 to 4000, 2001 to 4001, ..., 2999 to 4999.
990 In addition to modifying the address, some translation rules may modify
991 source or destination ports for
995 connections; implicitly in the case of
997 rules and explicitly in the case of
1000 Port numbers are never translated with a
1004 For each packet processed by the translator, the translation rules are
1005 evaluated in sequential order, from first to last.
1006 The first matching rule decides what action is taken.
1010 option prefixed to a translation rule causes packets to remain untranslated,
1011 much in the same way as
1013 works in the packet filter (see below).
1014 If no rule matches the packet it is passed to the filter engine unmodified.
1016 Translation rules apply only to packets that pass through
1017 the specified interface, and if no interface is specified,
1018 translation is applied to packets on all interfaces.
1019 For instance, redirecting port 80 on an external interface to an internal
1020 web server will only work for connections originating from the outside.
1021 Connections to the address of the external interface from local hosts will
1022 not be redirected, since such packets do not actually pass through the
1024 Redirections cannot reflect packets back through the interface they arrive
1025 on, they can only be redirected to hosts connected to different interfaces
1026 or to the firewall itself.
1028 Note that redirecting external incoming connections to the loopback
1030 .Bd -literal -offset indent
1031 rdr on ne3 inet proto tcp to port 8025 -> 127.0.0.1 port 25
1034 will effectively allow an external host to connect to daemons
1035 bound solely to the loopback address, circumventing the traditional
1036 blocking of such connections on a real interface.
1037 Unless this effect is desired, any of the local non-loopback addresses
1038 should be used as redirection target instead, which allows external
1039 connections only to daemons bound to this address or not bound to
1043 .Sx TRANSLATION EXAMPLES
1045 .Sh PACKET FILTERING
1051 packets based on attributes of their layer 3 (see
1061 In addition, packets may also be
1062 assigned to queues for the purpose of bandwidth control.
1064 For each packet processed by the packet filter, the filter rules are
1065 evaluated in sequential order, from first to last.
1066 The last matching rule decides what action is taken.
1068 The following actions can be used in the filter:
1069 .Bl -tag -width xxxx
1071 The packet is blocked.
1072 There are a number of ways in which a
1074 rule can behave when blocking a packet.
1075 The default behaviour is to
1077 packets silently, however this can be overridden or made
1078 explicit either globally, by setting the
1080 option, or on a per-rule basis with one of the following options:
1082 .Bl -tag -width xxxx -compact
1084 The packet is silently dropped.
1086 This applies only to
1088 packets, and issues a TCP RST which closes the
1092 This causes ICMP messages to be returned for packets which match the rule.
1093 By default this is an ICMP UNREACHABLE message, however this
1094 can be overridden by specifying a message as a code or number.
1096 This causes a TCP RST to be returned for
1098 packets and an ICMP UNREACHABLE for UDP and other packets.
1101 Options returning packets have no effect if
1106 The packet is passed.
1109 If no rule matches the packet, the default action is
1112 To block everything by default and only pass packets
1113 that match explicit rules, one uses
1114 .Bd -literal -offset indent
1118 as the first filter rule.
1124 The rule parameters specify the packets to which a rule applies.
1125 A packet always comes in on, or goes out through, one interface.
1126 Most parameters are optional.
1127 If a parameter is specified, the rule only applies to packets with
1128 matching attributes.
1129 Certain parameters can be expressed as lists, in which case
1131 generates all needed rule combinations.
1132 .Bl -tag -width xxxx
1133 .It Ar in No or Ar out
1134 This rule applies to incoming or outgoing packets.
1139 are specified, the rule will match packets in both directions.
1141 In addition to the action specified, a log message is generated.
1142 All packets for that connection are logged, unless the
1147 options are specified, in which case only the
1148 packet that establishes the state is logged.
1155 The logged packets are sent to the
1158 This interface is monitored by the
1160 logging daemon, which dumps the logged packets to the file
1171 rules to force logging of all packets for a connection.
1174 packets are logged to
1177 If a packet matches a rule which has the
1179 option set, this rule
1180 is considered the last matching rule, and evaluation of subsequent rules
1182 .It Ar on <interface>
1183 This rule applies only to packets coming in on, or going out through, this
1184 particular interface.
1185 It is also possible to simply give the interface driver name, like ppp or fxp,
1186 to make the rule match packets flowing through a group of interfaces.
1188 This rule applies only to packets of this address family.
1189 Supported values are
1193 .It Ar proto <protocol>
1194 This rule applies only to packets of this protocol.
1195 Common protocols are
1201 For a list of all the protocol name to number mappings used by
1204 .Em /etc/protocols .
1206 .Ar from <source> port <source> os <source>
1207 .Ar to <dest> port <dest>
1209 This rule applies only to packets with the specified source and destination
1210 addresses and ports.
1212 Addresses can be specified in CIDR notation (matching netblocks), as
1213 symbolic host names or interface names, or as any of the following keywords:
1215 .Bl -tag -width xxxxxxxxxxxx -compact
1219 Any address which is not currently routable.
1221 Any address that matches the given table.
1224 Interface names can have modifiers appended:
1226 .Bl -tag -width xxxxxxxxxxxx -compact
1228 Translates to the network(s) attached to the interface.
1230 Translates to the interface's broadcast address(es).
1232 Translates to the point to point interface's peer address(es).
1234 Do not include interface aliases.
1237 Host names may also have the
1239 option appended to restrict the name resolution to the first of each
1240 v4 and v6 address found.
1242 Host name resolution and interface to address translation are done at
1244 When the address of an interface (or host name) changes (under DHCP or PPP,
1245 for instance), the ruleset must be reloaded for the change to be reflected
1247 Surrounding the interface name (and optional modifiers) in parentheses
1248 changes this behaviour.
1249 When the interface name is surrounded by parentheses, the rule is
1250 automatically updated whenever the interface changes its address.
1251 The ruleset does not need to be reloaded.
1252 This is especially useful with
1255 Ports can be specified either by number or by name.
1256 For example, port 80 can be specified as
1258 For a list of all port name to number mappings used by
1263 Ports and ranges of ports are specified by using these operators:
1264 .Bd -literal -offset indent
1268 <= (less than or equal)
1270 >= (greater than or equal)
1271 : (range including boundaries)
1272 >< (range excluding boundaries)
1277 are binary operators (they take two arguments).
1280 .It Ar port 2000:2004
1282 .Sq all ports >= 2000 and <= 2004 ,
1283 hence ports 2000, 2001, 2002, 2003 and 2004.
1284 .It Ar port 2000 >< 2004
1286 .Sq all ports > 2000 and < 2004 ,
1287 hence ports 2001, 2002 and 2003.
1288 .It Ar port 2000 <> 2004
1290 .Sq all ports < 2000 or > 2004 ,
1291 hence ports 1-1999 and 2005-65535.
1294 The operating system of the source host can be specified in the case of TCP
1299 .Sx OPERATING SYSTEM FINGERPRINTING
1300 section for more information.
1302 The host, port and OS specifications are optional, as in the following examples:
1303 .Bd -literal -offset indent
1305 pass in from any to any
1306 pass in proto tcp from any port <= 1024 to any
1307 pass in proto tcp from any to any port 25
1308 pass in proto tcp from 10.0.0.0/8 port > 1024 \e
1309 to ! 10.1.2.3 port != ssh
1310 pass in proto tcp from any os "OpenBSD" flags S/SA
1313 This is equivalent to "from any to any".
1314 .It Ar group <group>
1317 this rule only applies to packets of sockets owned by the specified group.
1319 This rule only applies to packets of sockets owned by the specified user.
1320 For outgoing connections initiated from the firewall, this is the user
1321 that opened the connection.
1322 For incoming connections to the firewall itself, this is the user that
1323 listens on the destination port.
1324 For forwarded connections, where the firewall is not a connection endpoint,
1325 the user and group are
1328 All packets, both outgoing and incoming, of one connection are associated
1329 with the same user and group.
1330 Only TCP and UDP packets can be associated with users; for other protocols
1331 these parameters are ignored.
1333 User and group refer to the effective (as opposed to the real) IDs, in
1334 case the socket is created by a setuid/setgid process.
1335 User and group IDs are stored when a socket is created;
1336 when a process creates a listening socket as root (for instance, by
1337 binding to a privileged port) and subsequently changes to another
1338 user ID (to drop privileges), the credentials will remain root.
1340 User and group IDs can be specified as either numbers or names.
1341 The syntax is similar to the one for ports.
1344 matches packets of forwarded connections.
1346 can only be used with the operators
1350 Other constructs like
1353 Forwarded packets with unknown user and group ID match only rules
1354 that explicitly compare against
1362 does not match forwarded packets.
1363 The following example allows only selected users to open outgoing
1365 .Bd -literal -offset indent
1366 block out proto { tcp, udp } all
1367 pass out proto { tcp, udp } all \e
1368 user { < 1000, dhartmei } keep state
1370 .It Ar flags <a>/<b> | /<b>
1371 This rule only applies to TCP packets that have the flags
1375 Flags not specified in
1378 The flags are: (F)IN, (S)YN, (R)ST, (P)USH, (A)CK, (U)RG, (E)CE, and C(W)R.
1382 The other flags are ignored.
1384 Out of SYN and ACK, exactly SYN may be set.
1385 SYN, SYN+PSH and SYN+RST match, but SYN+ACK, ACK and ACK+RST do not.
1386 This is more restrictive than the previous example.
1388 If the first set is not specified, it defaults to none.
1389 All of SYN, FIN, RST and ACK must be unset.
1391 .It Ar icmp-type <type> code <code>
1392 .It Ar icmp6-type <type> code <code>
1393 This rule only applies to ICMP or ICMPv6 packets with the specified type
1395 This parameter is only valid for rules that cover protocols ICMP or
1397 The protocol and the ICMP type indicator (icmp-type or icmp6-type)
1400 By default, packets which contain IP options are blocked.
1405 rule, packets that pass the filter based on that rule (last matching)
1406 do so even if they contain IP options.
1407 For packets that match state, the rule that initially created the
1411 rule that is used when a packet does not match any rules does not
1413 .It Ar label <string>
1414 Adds a label (name) to the rule, which can be used to identify the rule.
1417 shows per-rule statistics for rules that have labels.
1419 The following macros can be used in labels:
1421 .Bl -tag -width $srcaddr -compact -offset indent
1425 The source IP address.
1427 The destination IP address.
1429 The source port specification.
1431 The destination port specification.
1439 .Bd -literal -offset indent
1440 ips = \&"{ 1.2.3.4, 1.2.3.5 }\&"
1441 pass in proto tcp from any to $ips \e
1442 port > 1023 label \&"$dstaddr:$dstport\&"
1446 .Bd -literal -offset indent
1447 pass in inet proto tcp from any to 1.2.3.4 \e
1448 port > 1023 label \&"1.2.3.4:>1023\&"
1449 pass in inet proto tcp from any to 1.2.3.5 \e
1450 port > 1023 label \&"1.2.3.5:>1023\&"
1453 The macro expansion for the
1455 directive occurs only at configuration file parse time, not during runtime.
1456 .It Ar queue <queue> | ( <queue> , <queue> )
1457 Packets matching this rule will be assigned to the specified queue.
1458 If two queues are given, packets which have a
1462 and TCP ACKs with no data payload will be assigned to the second one.
1468 .Bd -literal -offset indent
1469 pass in proto tcp to port 25 queue mail
1470 pass in proto tcp to port 22 queue(ssh_bulk, ssh_prio)
1473 Packets matching this rule will be tagged with the
1475 The tag acts as an internal marker that can be used to
1476 identify these packets later on.
1477 This can be used, for example, to provide trust between
1478 interfaces and to determine if packets have been
1479 processed by translation rules.
1482 meaning that the packet will be tagged even if the rule
1483 is not the last matching rule.
1484 Further matching rules can replace the tag with a
1485 new one but will not remove a previously applied tag.
1486 A packet is only ever assigned one tag at a time.
1490 keyword must also use
1494 .Ar synproxy state .
1495 Packet tagging can be done during
1500 rules in addition to filter rules.
1501 Tags take the same macros as labels (see above).
1502 .It Ar tagged <string>
1503 Used with filter rules to specify that packets must already
1504 be tagged with the given tag in order to match the rule.
1505 Inverse tag matching can also be done
1513 If a packet matches a rule with a route option set, the packet filter will
1514 route the packet according to the type of route option.
1515 When such a rule creates state, the route option is also applied to all
1516 packets matching the same connection.
1517 .Bl -tag -width xxxx
1521 option does a normal route lookup to find the next hop for the packet.
1525 option routes the packet to the specified interface with an optional address
1529 rule creates state, only packets that pass in the same direction as the
1530 filter rule specifies will be routed in this way.
1531 Packets passing in the opposite direction (replies) are not affected
1532 and are routed normally.
1536 option is similar to
1538 but routes packets that pass in the opposite direction (replies) to the
1539 specified interface.
1540 Opposite direction is only defined in the context of a state entry, and
1542 is useful only in rules that create state.
1543 It can be used on systems with multiple external connections to
1544 route all outgoing packets of a connection through the interface
1545 the incoming connection arrived through (symmetric routing enforcement).
1549 option creates a duplicate of the packet and routes it like
1551 The original packet gets routed as it normally would.
1558 rules, (as well as for the
1563 rule options) for which there is a single redirection address which has a
1564 subnet mask smaller than 32 for IPv4 or 128 for IPv6 (more than one IP
1565 address), a variety of different methods for assigning this address can be
1567 .Bl -tag -width xxxx
1571 option applies the network portion of the redirection address to the address
1572 to be modified (source with
1579 option selects an address at random within the defined block of addresses.
1583 option uses a hash of the source address to determine the redirection address,
1584 ensuring that the redirection address is always the same for a given source.
1585 An optional key can be specified after this keyword either in hex or as a
1588 randomly generates a key for source-hash every time the
1589 ruleset is reloaded.
1593 option loops through the redirection address(es).
1595 When more than one redirection address is specified,
1597 is the only permitted pool type.
1605 from modifying the source port on TCP and UDP packets.
1610 option can be specified to help ensure that multiple connections from the
1611 same source are mapped to the same redirection address.
1612 This option can be used with the
1617 Note that by default these associations are destroyed as soon as there are
1618 no longer states which refer to them; in order to make the mappings last
1619 beyond the lifetime of the states, increase the global options with
1620 .Ar set timeout source-track
1622 .Sx STATEFUL TRACKING OPTIONS
1623 for more ways to control the source tracking.
1624 .Sh STATEFUL INSPECTION
1626 is a stateful packet filter, which means it can track the state of
1628 Instead of passing all traffic to port 25, for instance, it is possible
1629 to pass only the initial packet, and then begin to keep state.
1630 Subsequent traffic will flow because the filter is aware of the connection.
1632 If a packet matches a
1633 .Ar pass ... keep state
1634 rule, the filter creates a state for this connection and automatically
1635 lets pass all subsequent packets of that connection.
1637 Before any rules are evaluated, the filter checks whether the packet
1639 If it does, the packet is passed without evaluation of any rules.
1641 States are removed after the connection is closed or has timed out.
1643 This has several advantages.
1644 Comparing a packet to a state involves checking its sequence numbers.
1645 If the sequence numbers are outside the narrow windows of expected
1646 values, the packet is dropped.
1647 This prevents spoofing attacks, such as when an attacker sends packets with
1648 a fake source address/port but does not know the connection's sequence
1651 Also, looking up states is usually faster than evaluating rules.
1652 If there are 50 rules, all of them are evaluated sequentially in O(n).
1653 Even with 50000 states, only 16 comparisons are needed to match a
1654 state, since states are stored in a binary search tree that allows
1655 searches in O(log2 n).
1658 .Bd -literal -offset indent
1660 pass out proto tcp from any to any flags S/SA keep state
1661 pass in proto tcp from any to any port 25 flags S/SA keep state
1664 This ruleset blocks everything by default.
1665 Only outgoing connections and incoming connections to port 25 are allowed.
1666 The initial packet of each connection has the SYN
1667 flag set, will be passed and creates state.
1668 All further packets of these connections are passed if they match a state.
1670 By default, packets coming in and out of any interface can match a state,
1671 but it is also possible to change that behaviour by assigning states to a
1672 single interface or a group of interfaces.
1674 The default policy is specified by the
1676 global option, but this can be adjusted on a per-rule basis by adding one
1685 For example, if a rule is defined as:
1686 .Bd -literal -offset indent
1687 pass out on ppp from any to 10.12/16 keep state (group-bound)
1690 A state created on ppp0 would match packets an all PPP interfaces,
1691 but not packets flowing through fxp0 or any other interface.
1695 is the more flexible option when the firewall is in a dynamic routing
1697 However, this has some security implications since a state created by one
1698 trusted network could allow potentially hostile packets coming in from other
1703 restricts state creation to the initial SYN
1704 packet of the TCP handshake.
1705 One can also be less restrictive, and allow state creation from
1711 to synchronize to existing connections, for instance
1712 if one flushes the state table.
1714 For UDP, which is stateless by nature,
1716 will create state as well.
1717 UDP packets are matched to states using only host addresses and ports.
1719 ICMP messages fall into two categories: ICMP error messages, which always
1720 refer to a TCP or UDP packet, are matched against the referred to connection.
1721 If one keeps state on a TCP connection, and an ICMP source quench message
1722 referring to this TCP connection arrives, it will be matched to the right
1723 state and get passed.
1727 creates an ICMP state, and
1729 knows how to match ICMP replies to states.
1731 .Bd -literal -offset indent
1732 pass out inet proto icmp all icmp-type echoreq keep state
1735 allows echo requests (such as those created by
1737 out, creates state, and matches incoming echo replies correctly to states.
1740 .Ar nat , binat No and Ar rdr
1741 rules implicitly create state for connections.
1742 .Sh STATE MODULATION
1743 Much of the security derived from TCP is attributable to how well the
1744 initial sequence numbers (ISNs) are chosen.
1745 Some popular stack implementations choose
1747 poor ISNs and thus are normally susceptible to ISN prediction exploits.
1750 rule to a TCP connection,
1752 will create a high quality random sequence number for each connection
1757 directive implicitly keeps state on the rule and is
1758 only applicable to TCP connections.
1761 .Bd -literal -offset indent
1763 pass out proto tcp from any to any modulate state
1764 pass in proto tcp from any to any port 25 flags S/SA modulate state
1767 There are two caveats associated with state modulation:
1770 rule can not be applied to a pre-existing but unmodulated connection.
1771 Such an application would desynchronize TCP's strict
1772 sequencing between the two endpoints.
1779 modifier and the pre-existing connection will be inferred without
1780 the protection conferred by modulation.
1782 The other caveat affects currently modulated states when the state table
1783 is lost (firewall reboot, flushing the state table, etc...).
1785 will not be able to infer a connection again after the state table flushes
1786 the connection's modulator.
1787 When the state is lost, the connection may be left dangling until the
1788 respective endpoints time out the connection.
1789 It is possible on a fast local network for the endpoints to start an ACK
1790 storm while trying to resynchronize after the loss of the modulator.
1795 rules between fast networks is suggested to prevent ACK storms.
1799 passes packets that are part of a
1801 handshake between the endpoints.
1804 option can be used to cause
1806 itself to complete the handshake with the active endpoint, perform a handshake
1807 with the passive endpoint, and then forward packets between the endpoints.
1809 No packets are sent to the passive endpoint before the active endpoint has
1810 completed the handshake, hence so-called SYN floods with spoofed source
1811 addresses will not reach the passive endpoint, as the sender can't complete the
1814 The proxy is transparent to both endpoints, they each see a single
1815 connection from/to the other endpoint.
1817 chooses random initial sequence numbers for both handshakes.
1818 Once the handshakes are completed, the sequence number modulators
1819 (see previous section) are used to translate further packets of the
1836 .Bd -literal -offset indent
1837 pass in proto tcp from any to any port www flags S/SA synproxy state
1839 .Sh STATEFUL TRACKING OPTIONS
1845 support the following options:
1847 .Bl -tag -width xxxx -compact
1849 Limits the number of concurrent states the rule may create.
1850 When this limit is reached, further packets matching the rule that would
1851 create state are dropped, until existing states time out.
1853 Prevent state changes for states created by this rule from appearing on the
1856 .It Ar <timeout> <seconds>
1857 Changes the timeout values used for states created by this rule.
1861 keyword is specified, the number of states per source IP is tracked.
1862 The following limits can be set:
1864 .Bl -tag -width xxxx -compact
1865 .It Ar max-src-nodes
1866 Limits the maximum number of source addresses which can simultaneously
1867 have state table entries.
1868 .It Ar max-src-states
1869 Limits the maximum number of simultaneous state entries that a single
1870 source address can create with this rule.
1872 For a list of all valid timeout names, see
1876 Multiple options can be specified, separated by commas:
1878 pass in proto tcp from any to any \e
1879 port www flags S/SA keep state \e
1880 (max 100, source-track rule, max-src-nodes 75, \e
1881 max-src-states 3, tcp.established 60, tcp.closing 5)
1884 .Sh OPERATING SYSTEM FINGERPRINTING
1885 Passive OS Fingerprinting is a mechanism to inspect nuances of a TCP
1886 connection's initial SYN packet and guess at the host's operating system.
1887 Unfortunately these nuances are easily spoofed by an attacker so the
1888 fingerprint is not useful in making security decisions.
1889 But the fingerprint is typically accurate enough to make policy decisions
1892 The fingerprints may be specified by operating system class, by
1893 version, or by subtype/patchlevel.
1894 The class of an operating system is typically the vender or genre
1895 and would be OpenBSD for the
1898 The version of the oldest available OpenBSD release on the main ftp site
1899 would be 2.6 and the fingerprint would be written
1901 .Dl \&"OpenBSD 2.6\&"
1903 The subtype of an operating system is typically used to describe the
1904 patchlevel if that patch led to changes in the TCP stack behavior.
1905 In the case of OpenBSD, the only subtype is for a fingerprint that was
1908 scrub option and would be specified as
1910 .Dl \&"OpenBSD 3.3 no-df\&"
1912 Fingerprints for most popular operating systems are provided by
1916 is running, a complete list of known operating system fingerprints may
1917 be listed by running:
1921 Filter rules can enforce policy at any level of operating system specification
1922 assuming a fingerprint is present.
1923 Policy could limit traffic to approved operating systems or even ban traffic
1924 from hosts that aren't at the latest service pack.
1928 class can also be used as the fingerprint which will match packets for
1929 which no operating system fingerprint is known.
1932 .Bd -literal -offset indent
1933 pass out proto tcp from any os OpenBSD keep state
1934 block out proto tcp from any os Doors
1935 block out proto tcp from any os "Doors PT"
1936 block out proto tcp from any os "Doors PT SP3"
1937 block out from any os "unknown"
1938 pass on lo0 proto tcp from any os "OpenBSD 3.3 lo0" keep state
1941 Operating system fingerprinting is limited only to the TCP SYN packet.
1942 This means that it will not work on other protocols and will not match
1943 a currently established connection.
1945 Caveat: operating system fingerprints are occasionally wrong.
1946 There are three problems: an attacker can trivially craft his packets to
1947 appear as any operating system he chooses;
1948 an operating system patch could change the stack behavior and no fingerprints
1949 will match it until the database is updated;
1950 and multiple operating systems may have the same fingerprint.
1951 .Sh BLOCKING SPOOFED TRAFFIC
1952 "Spoofing" is the faking of IP addresses, typically for malicious
1956 directive expands to a set of filter rules which will block all
1957 traffic with a source IP from the network(s) directly connected
1958 to the specified interface(s) from entering the system through
1959 any other interface.
1961 For example, the line
1962 .Bd -literal -offset indent
1967 .Bd -literal -offset indent
1968 block drop in on ! lo0 inet from 127.0.0.1/8 to any
1969 block drop in on ! lo0 inet6 from ::1 to any
1972 For non-loopback interfaces, there are additional rules to block incoming
1973 packets with a source IP address identical to the interface's IP(s).
1974 For example, assuming the interface wi0 had an IP address of 10.0.0.1 and a
1975 netmask of 255.255.255.0,
1977 .Bd -literal -offset indent
1978 antispoof for wi0 inet
1982 .Bd -literal -offset indent
1983 block drop in on ! wi0 inet from 10.0.0.0/24 to any
1984 block drop in inet from 10.0.0.1 to any
1987 Caveat: Rules created by the
1989 directive interfere with packets sent over loopback interfaces
1991 One should pass these explicitly.
1992 .Sh FRAGMENT HANDLING
1993 The size of IP datagrams (packets) can be significantly larger than the
1994 maximum transmission unit (MTU) of the network.
1995 In cases when it is necessary or more efficient to send such large packets,
1996 the large packet will be fragmented into many smaller packets that will each
1998 Unfortunately for a firewalling device, only the first logical fragment will
1999 contain the necessary header information for the subprotocol that allows
2001 to filter on things such as TCP ports or to perform NAT.
2005 rules as described in
2006 .Sx TRAFFIC NORMALIZATION
2007 above, there are three options for handling fragments in the packet filter.
2009 One alternative is to filter individual fragments with filter rules.
2012 rule applies to a fragment, it is passed to the filter.
2013 Filter rules with matching IP header parameters decide whether the
2014 fragment is passed or blocked, in the same way as complete packets
2016 Without reassembly, fragments can only be filtered based on IP header
2017 fields (source/destination address, protocol), since subprotocol header
2018 fields are not available (TCP/UDP port numbers, ICMP code/type).
2021 option can be used to restrict filter rules to apply only to
2022 fragments, but not complete packets.
2023 Filter rules without the
2025 option still apply to fragments, if they only specify IP header fields.
2026 For instance, the rule
2027 .Bd -literal -offset indent
2028 pass in proto tcp from any to any port 80
2031 never applies to a fragment, even if the fragment is part of a TCP
2032 packet with destination port 80, because without reassembly this information
2033 is not available for each fragment.
2034 This also means that fragments cannot create new or match existing
2035 state table entries, which makes stateful filtering and address
2036 translation (NAT, redirection) for fragments impossible.
2038 It's also possible to reassemble only certain fragments by specifying
2039 source or destination addresses or protocols as parameters in
2043 In most cases, the benefits of reassembly outweigh the additional
2044 memory cost, and it's recommended to use
2047 all fragments via the
2048 .Ar fragment reassemble
2051 The memory allocated for fragment caching can be limited using
2053 Once this limit is reached, fragments that would have to be cached
2054 are dropped until other entries time out.
2055 The timeout value can also be adjusted.
2057 Currently, only IPv4 fragments are supported and IPv6 fragments
2058 are blocked unconditionally.
2059 .Sh ANCHORS AND NAMED RULESETS
2060 Besides the main ruleset,
2062 can load named rulesets into
2067 contains a list of named rulesets.
2070 has a name which specifies where
2072 can be used to attach sub-rulesets.
2073 A named ruleset contains filter and translation rules, like the
2075 The main ruleset can reference
2078 using the following kinds
2080 .Bl -tag -width xxxx
2081 .It Ar nat-anchor <name>
2084 rules of all named rulesets in the specified
2086 .It Ar rdr-anchor <name>
2089 rules of all named rulesets in the specified
2091 .It Ar binat-anchor <name>
2094 rules of all named rulesets in the specified
2096 .It Ar anchor <name>
2097 Evaluates the filter rules of all named rulesets in the specified
2099 .It Ar load anchor <name>:<ruleset> from <file>
2100 Loads the rules from the specified file into the named
2103 attached to the anchor
2107 When evaluation of the main ruleset reaches an
2111 will proceed to evaluate all rules specified in the
2112 named rulesets attached to that
2115 Matching filter rules in named rulesets with the
2117 option and matching translation rules are final and abort the
2118 evaluation of both the rules in the
2120 and the main ruleset.
2122 Only the main ruleset can contain
2128 contains more than one named ruleset, they are evaluated
2129 in the alphabetical order of their names.
2133 attachment points which do not contain any rules when the main ruleset
2134 is loaded, and later such named rulesets can be manipulated through
2136 without reloading the main ruleset.
2138 .Bd -literal -offset indent
2140 block on $ext_if all
2142 pass out on $ext_if all keep state
2143 pass in on $ext_if proto tcp from any \e
2144 to $ext_if port smtp keep state
2147 blocks all packets on the external interface by default, then evaluates
2150 named "spam", and finally passes all outgoing connections and
2151 incoming connections to port 25.
2152 .Bd -literal -offset indent
2153 # echo \&"block in quick from 1.2.3.4 to any\&" \&| \e
2154 pfctl -a spam:manual -f -
2157 loads a single ruleset containing a single rule into the
2159 which blocks all packets from a specific address.
2161 The named ruleset can also be populated by adding a
2166 .Bd -literal -offset indent
2168 load anchor spam:manual from "/etc/pf-spam.conf"
2175 it will also load all the rules from the file
2176 .Pa /etc/pf-spam.conf
2177 into the named ruleset.
2181 rules can specify the parameter's
2182 direction, interface, address family, protocol and source/destination
2184 using the same syntax as filter rules.
2185 When parameters are used, the
2187 rule is only evaluated for matching packets.
2188 This allows conditional evaluation of named rulesets, like:
2189 .Bd -literal -offset indent
2190 block on $ext_if all
2191 anchor spam proto tcp from any to any port smtp
2192 pass out on $ext_if all keep state
2193 pass in on $ext_if proto tcp from any to $ext_if port smtp keep state
2198 spam are only evaluated for
2200 packets with destination port 25.
2202 .Bd -literal -offset indent
2203 # echo \&"block in quick from 1.2.3.4 to any" \&| \e
2204 pfctl -a spam:manual -f -
2207 will only block connections from 1.2.3.4 to port 25.
2208 .Sh TRANSLATION EXAMPLES
2209 This example maps incoming requests on port 80 to port 8080, on
2210 which a daemon is running (because, for example, it is not run as root,
2211 and therefore lacks permission to bind to port 80).
2213 # use a macro for the interface name, so it can be changed easily
2216 # map daemon on 8080 to appear to be on 80
2217 rdr on $ext_if proto tcp from any to any port 80 -> 127.0.0.1 port 8080
2222 modifier is given, packets matching the translation rule are passed without
2223 inspecting the filter rules:
2225 rdr pass on $ext_if proto tcp from any to any port 80 -> 127.0.0.1 \e
2229 In the example below, vlan12 is configured as 192.168.168.1;
2230 the machine translates all packets coming from 192.168.168.0/24 to 204.92.77.111
2231 when they are going out any interface except vlan12.
2232 This has the net effect of making traffic from the 192.168.168.0/24
2233 network appear as though it is the Internet routable address
2234 204.92.77.111 to nodes behind any interface on the router except
2235 for the nodes on vlan12.
2236 (Thus, 192.168.168.1 can talk to the 192.168.168.0/24 nodes.)
2238 nat on ! vlan12 from 192.168.168.0/24 to any -> 204.92.77.111
2241 In the example below, the machine sits between a fake internal 144.19.74.*
2242 network, and a routable external IP of 204.92.77.100.
2245 rule excludes protocol AH from being translated.
2248 no nat on $ext_if proto ah from 144.19.74.0/24 to any
2249 nat on $ext_if from 144.19.74.0/24 to any -> 204.92.77.100
2252 In the example below, packets bound for one specific server, as well as those
2253 generated by the sysadmins are not proxied; all other connections are.
2256 no rdr on $int_if proto { tcp, udp } from any to $server port 80
2257 no rdr on $int_if proto { tcp, udp } from $sysadmins to any port 80
2258 rdr on $int_if proto { tcp, udp } from any to any port 80 -> 127.0.0.1 \e
2262 This longer example uses both a NAT and a redirection.
2263 The external interface has the address 157.161.48.183.
2264 On the internal interface, we are running
2266 listening for outbound ftp sessions captured to port 8021.
2269 # Translate outgoing packets' source addresses (any protocol).
2270 # In this case, any address but the gateway's external address is mapped.
2271 nat on $ext_if inet from ! ($ext_if) to any -> ($ext_if)
2274 # Map outgoing packets' source port to an assigned proxy port instead of
2275 # an arbitrary port.
2276 # In this case, proxy outgoing isakmp with port 500 on the gateway.
2277 nat on $ext_if inet proto udp from any port = isakmp to any -> ($ext_if) \e
2281 # Translate outgoing packets' source address (any protocol).
2282 # Translate incoming packets' destination address to an internal machine
2284 binat on $ext_if from 10.1.2.150 to any -> ($ext_if)
2287 # Translate incoming packets' destination addresses.
2288 # As an example, redirect a TCP and UDP port to an internal machine.
2289 rdr on $ext_if inet proto tcp from any to ($ext_if) port 8080 \e
2290 -> 10.1.2.151 port 22
2291 rdr on $ext_if inet proto udp from any to ($ext_if) port 8080 \e
2292 -> 10.1.2.151 port 53
2295 # Translate outgoing ftp control connections to send them to localhost
2296 # for proxying with ftp-proxy(8) running on port 8021.
2297 rdr on $int_if proto tcp from any to any port 21 -> 127.0.0.1 port 8021
2300 In this example, a NAT gateway is set up to translate internal addresses
2301 using a pool of public addresses (192.0.2.16/28) and to redirect
2302 incoming web server connections to a group of web servers on the internal
2306 # Translate outgoing packets' source addresses using an address pool.
2307 # A given source address is always translated to the same pool address by
2308 # using the source-hash keyword.
2309 nat on $ext_if inet from any to any -> 192.0.2.16/28 source-hash
2312 # Translate incoming web server connections to a group of web servers on
2313 # the internal network.
2314 rdr on $ext_if proto tcp from any to any port 80 \e
2315 -> { 10.1.2.155, 10.1.2.160, 10.1.2.161 } round-robin
2319 # The external interface is kue0
2320 # (157.161.48.183, the only routable address)
2321 # and the private network is 10.0.0.0/8, for which we are doing NAT.
2323 # use a macro for the interface name, so it can be changed easily
2326 # normalize all incoming traffic
2327 scrub in on $ext_if all fragment reassemble
2329 # block and log everything by default
2330 block return log on $ext_if all
2332 # block anything coming from source we have no back routes for
2333 block in from no-route to any
2335 # block and log outgoing packets that do not have our address as source,
2336 # they are either spoofed or something is misconfigured (NAT disabled,
2337 # for instance), we want to be nice and do not send out garbage.
2338 block out log quick on $ext_if from ! 157.161.48.183 to any
2340 # silently drop broadcasts (cable modem noise)
2341 block in quick on $ext_if from any to 255.255.255.255
2343 # block and log incoming packets from reserved address space and invalid
2344 # addresses, they are either spoofed or misconfigured, we cannot reply to
2345 # them anyway (hence, no return-rst).
2346 block in log quick on $ext_if from { 10.0.0.0/8, 172.16.0.0/12, \e
2347 192.168.0.0/16, 255.255.255.255/32 } to any
2351 # pass out/in certain ICMP queries and keep state (ping)
2352 # state matching is done on host addresses and ICMP id (not type/code),
2353 # so replies (like 0/0 for 8/0) will match queries
2354 # ICMP error messages (which always refer to a TCP/UDP packet) are
2355 # handled by the TCP/UDP states
2356 pass on $ext_if inet proto icmp all icmp-type 8 code 0 keep state
2360 # pass out all UDP connections and keep state
2361 pass out on $ext_if proto udp all keep state
2363 # pass in certain UDP connections and keep state (DNS)
2364 pass in on $ext_if proto udp from any to any port domain keep state
2368 # pass out all TCP connections and modulate state
2369 pass out on $ext_if proto tcp all modulate state
2371 # pass in certain TCP connections and keep state (SSH, SMTP, DNS, IDENT)
2372 pass in on $ext_if proto tcp from any to any port { ssh, smtp, domain, \e
2373 auth } flags S/SA keep state
2375 # pass in data mode connections for ftp-proxy running on this host.
2376 # (see ftp-proxy(8) for details)
2377 pass in on $ext_if proto tcp from any to 157.161.48.183 port >= 49152 \e
2378 flags S/SA keep state
2380 # Do not allow Windows 9x SMTP connections since they are typically
2381 # a viral worm. Alternately we could limit these OSes to 1 connection each.
2382 block in on $ext_if proto tcp from any os {"Windows 95", "Windows 98"} \e
2387 # three interfaces: $int_if, $ext_if, and $wifi_if (wireless). NAT is
2388 # being done on $ext_if for all outgoing packets. tag packets in on
2389 # $int_if and pass those tagged packets out on $ext_if. all other
2390 # outgoing packets (i.e., packets from the wireless network) are only
2391 # permitted to access port 80.
2393 pass in on $int_if from any to any tag INTNET keep state
2394 pass in on $wifi_if from any to any keep state
2396 block out on $ext_if from any to any
2397 pass out quick on $ext_if tagged INTNET keep state
2398 pass out on $ext_if from any to any port 80 keep state
2400 # tag incoming packets as they are redirected to spamd(8). use the tag
2401 # to pass those packets through the packet filter.
2403 rdr on $ext_if inet proto tcp from <spammers> to port smtp \e
2404 tag SPAMD -> 127.0.0.1 port spamd
2407 pass in on $ext_if inet proto tcp tagged SPAMD keep state
2414 line = ( option | pf-rule | nat-rule | binat-rule | rdr-rule |
2415 antispoof-rule | altq-rule | queue-rule | anchor-rule |
2416 trans-anchors | load-anchors | table-rule )
2418 option = "set" ( [ "timeout" ( timeout | "{" timeout-list "}" ) ] |
2419 [ "optimization" [ "default" | "normal" |
2420 "high-latency" | "satellite" |
2421 "aggressive" | "conservative" ] ]
2422 [ "limit" ( limit-item | "{" limit-list "}" ) ] |
2423 [ "loginterface" ( interface-name | "none" ) ] |
2424 [ "block-policy" ( "drop" | "return" ) ] |
2425 [ "state-policy" ( "if-bound" | "group-bound" |
2427 [ "require-order" ( "yes" | "no" ) ]
2428 [ "fingerprints" filename ] |
2429 [ "debug" ( "none" | "urgent" | "misc" | "loud" ) ] )
2431 pf-rule = action [ ( "in" | "out" ) ]
2432 [ "log" | "log-all" ] [ "quick" ]
2433 [ "on" ifspec ] [ route ] [ af ] [ protospec ]
2434 hosts [ filteropt-list ]
2436 filteropt-list = filteropt-list filteropt | filteropt
2437 filteropt = user | group | flags | icmp-type | icmp6-type | tos |
2438 ( "keep" | "modulate" | "synproxy" ) "state"
2439 [ "(" state-opts ")" ] |
2440 "fragment" | "no-df" | "min-ttl" number |
2441 "max-mss" number | "random-id" | "reassemble tcp" |
2442 fragmentation | "allow-opts" |
2443 "label" string | "tag" string | [ ! ] "tagged" string
2444 "queue" ( string | "(" string [ [ "," ] string ] ")" )
2446 nat-rule = [ "no" ] "nat" [ "pass" ] [ "on" ifspec ] [ af ]
2447 [ protospec ] hosts [ "tag" string ]
2448 [ "->" ( redirhost | "{" redirhost-list "}" )
2449 [ portspec ] [ pooltype ] [ "static-port" ] ]
2451 binat-rule = [ "no" ] "binat" [ "pass" ] [ "on" interface-name ]
2452 [ af ] [ "proto" ( proto-name | proto-number ) ]
2453 "from" address [ "/" mask-bits ] "to" ipspec
2455 [ "->" address [ "/" mask-bits ] ]
2457 rdr-rule = [ "no" ] "rdr" [ "pass" ] [ "on" ifspec ] [ af ]
2458 [ protospec ] hosts [ "tag" string ]
2459 [ "->" ( redirhost | "{" redirhost-list "}" )
2460 [ portspec ] [ pooltype ] ]
2462 antispoof-rule = "antispoof" [ "log" ] [ "quick" ]
2463 "for" ( interface-name | "{" interface-list "}" )
2464 [ af ] [ "label" string ]
2466 table-rule = "table" "<" string ">" [ tableopts-list ]
2467 tableopts-list = tableopts-list tableopts | tableopts
2468 tableopts = "persist" | "const" | "file" string |
2469 "{" [ tableaddr-list ] "}"
2470 tableaddr-list = tableaddr-list [ "," ] tableaddr-spec | tableaddr-spec
2471 tableaddr-spec = [ "!" ] tableaddr [ "/" mask-bits ]
2472 tableaddr = hostname | ipv4-dotted-quad | ipv6-coloned-hex |
2473 interface-name | "self"
2475 altq-rule = "altq on" interface-name queueopts-list
2477 queue-rule = "queue" string [ "on" interface-name ] queueopts-list
2480 anchor-rule = "anchor" string [ ( "in" | "out" ) ] [ "on" ifspec ]
2481 [ af ] [ "proto" ] [ protospec ] [ hosts ]
2483 trans-anchors = ( "nat-anchor" | "rdr-anchor" | "binat-anchor" ) string
2484 [ "on" ifspec ] [ af ] [ "proto" ] [ protospec ] [ hosts ]
2486 load-anchor = "load anchor" anchorname:rulesetname "from" filename
2488 queueopts-list = queueopts-list queueopts | queueopts
2489 queueopts = [ "bandwidth" bandwidth-spec ] |
2490 [ "qlimit" number ] | [ "tbrsize" number ] |
2491 [ "priority" number ] | [ schedulers ]
2492 schedulers = ( cbq-def | priq-def | hfsc-def )
2493 bandwidth-spec = "number" ( "b" | "Kb" | "Mb" | "Gb" | "%" )
2495 action = "pass" | "block" [ return ] | "scrub"
2496 return = "drop" | "return" | "return-rst" [ "( ttl" number ")" ] |
2497 "return-icmp" [ "(" icmpcode ["," icmp6code ] ")" ] |
2498 "return-icmp6" [ "(" icmp6code ")" ]
2499 icmpcode = ( icmp-code-name | icmp-code-number )
2500 icmp6code = ( icmp6-code-name | icmp6-code-number )
2502 ifspec = ( [ "!" ] interface-name ) | "{" interface-list "}"
2503 interface-list = [ "!" ] interface-name [ [ "," ] interface-list ]
2504 route = "fastroute" |
2505 ( "route-to" | "reply-to" | "dup-to" )
2506 ( routehost | "{" routehost-list "}" )
2508 af = "inet" | "inet6"
2510 protospec = "proto" ( proto-name | proto-number |
2511 "{" proto-list "}" )
2512 proto-list = ( proto-name | proto-number ) [ [ "," ] proto-list ]
2515 "from" ( "any" | "no-route" | "self" | host |
2516 "{" host-list "}" ) [ port ] [ os ]
2517 "to" ( "any" | "no-route" | "self" | host |
2518 "{" host-list "}" ) [ port ]
2520 ipspec = "any" | host | "{" host-list "}"
2521 host = [ "!" ] ( address [ "/" mask-bits ] | "<" string ">" )
2522 redirhost = address [ "/" mask-bits ]
2523 routehost = ( interface-name [ address [ "/" mask-bits ] ] )
2524 address = ( interface-name | "(" interface-name ")" | hostname |
2525 ipv4-dotted-quad | ipv6-coloned-hex )
2526 host-list = host [ [ "," ] host-list ]
2527 redirhost-list = redirhost [ [ "," ] redirhost-list ]
2528 routehost-list = routehost [ [ "," ] routehost-list ]
2530 port = "port" ( unary-op | binary-op | "{" op-list "}" )
2531 portspec = "port" ( number | name ) [ ":" ( "*" | number | name ) ]
2532 os = "os" ( os-name | "{" os-list "}" )
2533 user = "user" ( unary-op | binary-op | "{" op-list "}" )
2534 group = "group" ( unary-op | binary-op | "{" op-list "}" )
2536 unary-op = [ "=" | "!=" | "<" | "<=" | ">" | ">=" ]
2538 binary-op = number ( "<>" | "><" | ":" ) number
2539 op-list = ( unary-op | binary-op ) [ [ "," ] op-list ]
2541 os-name = operating-system-name
2542 os-list = os-name [ [ "," ] os-list ]
2544 flags = "flags" [ flag-set ] "/" flag-set
2545 flag-set = [ "F" ] [ "S" ] [ "R" ] [ "P" ] [ "A" ] [ "U" ] [ "E" ]
2548 icmp-type = "icmp-type" ( icmp-type-code | "{" icmp-list "}" )
2549 icmp6-type = "icmp6-type" ( icmp-type-code | "{" icmp-list "}" )
2550 icmp-type-code = ( icmp-type-name | icmp-type-number )
2551 [ "code" ( icmp-code-name | icmp-code-number ) ]
2552 icmp-list = icmp-type-code [ [ "," ] icmp-list ]
2554 tos = "tos" ( "lowdelay" | "throughput" | "reliability" |
2557 state-opts = state-opt [ [ "," ] state-opts ]
2558 state-opt = ( "max" number | "no-sync" | timeout |
2559 "source-track" [ ( "rule" | "global" ) ] |
2560 "max-src-nodes" number | "max-src-states" number |
2561 "if-bound" | "group-bound" | "floating" )
2563 fragmentation = [ "fragment reassemble" | "fragment crop" |
2564 "fragment drop-ovl" ]
2566 timeout-list = timeout [ [ "," ] timeout-list ]
2567 timeout = ( "tcp.first" | "tcp.opening" | "tcp.established" |
2568 "tcp.closing" | "tcp.finwait" | "tcp.closed" |
2569 "udp.first" | "udp.single" | "udp.multiple" |
2570 "icmp.first" | "icmp.error" |
2571 "other.first" | "other.single" | "other.multiple" |
2572 "frag" | "interval" | "src.track" |
2573 "adaptive.start" | "adaptive.end" ) number
2575 limit-list = limit-item [ [ "," ] limit-list ]
2576 limit-item = ( "states" | "frags" | "src-nodes" ) number
2578 pooltype = ( "bitmask" | "random" |
2579 "source-hash" [ ( hex-key | string-key ) ] |
2580 "round-robin" ) [ sticky-address ]
2582 subqueue = string | "{" queue-list "}"
2583 queue-list = string [ [ "," ] string ]
2584 cbq-def = "cbq" [ "(" cbq-opt [ [ "," ] cbq-opt ] ")" ]
2585 priq-def = "priq" [ "(" priq-opt [ [ "," ] priq-opt ] ")" ]
2586 hfsc-def = "hfsc" [ "(" hfsc-opt [ [ "," ] hfsc-opt ] ")" ]
2587 cbq-opt = ( "default" | "borrow" | "red" | "ecn" | "rio" )
2588 priq-opt = ( "default" | "red" | "ecn" | "rio" )
2589 hfsc-opt = ( "default" | "red" | "ecn" | "rio" |
2590 linkshare-sc | realtime-sc | upperlimit-sc )
2591 linkshare-sc = "linkshare" sc-spec
2592 realtime-sc = "realtime" sc-spec
2593 upperlimit-sc = "upperlimit" sc-spec
2594 sc-spec = ( bandwidth-spec |
2595 "(" bandwidth-spec number bandwidth-spec ")" )
2598 .Bl -tag -width "/etc/protocols" -compact
2602 Default location of the ruleset file.
2604 Default location of OS fingerprints.
2605 .It Pa /etc/protocols
2606 Protocol name database.
2607 .It Pa /etc/services
2608 Service name database.
2609 .It Pa /usr/share/pf
2631 file format first appeared in