xref: /freebsd/share/man/man5/pf.conf.5 (revision af23369a6deaaeb612ab266eb88b8bb8d560c322)
1.\"    $FreeBSD$
2.\"	$OpenBSD: pf.conf.5,v 1.406 2009/01/31 19:37:12 sobrado Exp $
3.\"
4.\" Copyright (c) 2002, Daniel Hartmeier
5.\" 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.\"
11.\"    - Redistributions of source code must retain the above copyright
12.\"      notice, this list of conditions and the following disclaimer.
13.\"    - Redistributions in binary form must reproduce the above
14.\"      copyright notice, this list of conditions and the following
15.\"      disclaimer in the documentation and/or other materials provided
16.\"      with the distribution.
17.\"
18.\" THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
19.\" "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
20.\" LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
21.\" FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
22.\" COPYRIGHT HOLDERS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
23.\" INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
24.\" BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
25.\" LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
26.\" CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
27.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
28.\" ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
29.\" POSSIBILITY OF SUCH DAMAGE.
30.\"
31.Dd November 24, 2022
32.Dt PF.CONF 5
33.Os
34.Sh NAME
35.Nm pf.conf
36.Nd packet filter configuration file
37.Sh DESCRIPTION
38The
39.Xr pf 4
40packet filter modifies, drops or passes packets according to rules or
41definitions specified in
42.Nm pf.conf .
43.Sh STATEMENT ORDER
44There are eight types of statements in
45.Nm pf.conf :
46.Bl -tag -width xxxx
47.It Cm Macros
48User-defined variables may be defined and used later, simplifying
49the configuration file.
50Macros must be defined before they are referenced in
51.Nm pf.conf .
52.It Cm Tables
53Tables provide a mechanism for increasing the performance and flexibility of
54rules with large numbers of source or destination addresses.
55.It Cm Options
56Options tune the behaviour of the packet filtering engine.
57.It Cm Ethernet Filtering
58Ethernet filtering provides rule-based blocking or passing of Ethernet packets.
59.It Cm Traffic Normalization Li (e.g. Em scrub )
60Traffic normalization protects internal machines against inconsistencies
61in Internet protocols and implementations.
62.It Cm Queueing
63Queueing provides rule-based bandwidth control.
64.It Cm Translation Li (Various forms of NAT)
65Translation rules specify how addresses are to be mapped or redirected to
66other addresses.
67.It Cm Packet Filtering
68Packet filtering provides rule-based blocking or passing of packets.
69.El
70.Pp
71With the exception of
72.Cm macros
73and
74.Cm tables ,
75the types of statements should be grouped and appear in
76.Nm pf.conf
77in the order shown above, as this matches the operation of the underlying
78packet filtering engine.
79By default
80.Xr pfctl 8
81enforces this order (see
82.Ar set require-order
83below).
84.Pp
85Comments can be put anywhere in the file using a hash mark
86.Pq Sq # ,
87and extend to the end of the current line.
88.Pp
89Additional configuration files can be included with the
90.Ic include
91keyword, for example:
92.Bd -literal -offset indent
93include "/etc/pf/sub.filter.conf"
94.Ed
95.Sh MACROS
96Macros can be defined that will later be expanded in context.
97Macro names must start with a letter, and may contain letters, digits
98and underscores.
99Macro names may not be reserved words (for example
100.Ar pass ,
101.Ar in ,
102.Ar out ) .
103Macros are not expanded inside quotes.
104.Pp
105For example,
106.Bd -literal -offset indent
107ext_if = \&"kue0\&"
108all_ifs = \&"{\&" $ext_if lo0 \&"}\&"
109pass out on $ext_if from any to any
110pass in  on $ext_if proto tcp from any to any port 25
111.Ed
112.Sh TABLES
113Tables are named structures which can hold a collection of addresses and
114networks.
115Lookups against tables in
116.Xr pf 4
117are relatively fast, making a single rule with tables much more efficient,
118in terms of
119processor usage and memory consumption, than a large number of rules which
120differ only in IP address (either created explicitly or automatically by rule
121expansion).
122.Pp
123Tables can be used as the source or destination of filter rules,
124.Ar scrub
125rules
126or
127translation rules such as
128.Ar nat
129or
130.Ar rdr
131(see below for details on the various rule types).
132Tables can also be used for the redirect address of
133.Ar nat
134and
135.Ar rdr
136rules and in the routing options of filter rules, but only for
137.Ar round-robin
138pools.
139.Pp
140Tables can be defined with any of the following
141.Xr pfctl 8
142mechanisms.
143As with macros, reserved words may not be used as table names.
144.Bl -tag -width "manually"
145.It Ar manually
146Persistent tables can be manually created with the
147.Ar add
148or
149.Ar replace
150option of
151.Xr pfctl 8 ,
152before or after the ruleset has been loaded.
153.It Pa pf.conf
154Table definitions can be placed directly in this file, and loaded at the
155same time as other rules are loaded, atomically.
156Table definitions inside
157.Nm pf.conf
158use the
159.Ar table
160statement, and are especially useful to define non-persistent tables.
161The contents of a pre-existing table defined without a list of addresses
162to initialize it is not altered when
163.Nm pf.conf
164is loaded.
165A table initialized with the empty list,
166.Li { } ,
167will be cleared on load.
168.El
169.Pp
170Tables may be defined with the following attributes:
171.Bl -tag -width persist
172.It Ar persist
173The
174.Ar persist
175flag forces the kernel to keep the table even when no rules refer to it.
176If the flag is not set, the kernel will automatically remove the table
177when the last rule referring to it is flushed.
178.It Ar const
179The
180.Ar const
181flag prevents the user from altering the contents of the table once it
182has been created.
183Without that flag,
184.Xr pfctl 8
185can be used to add or remove addresses from the table at any time, even
186when running with
187.Xr securelevel 7
188= 2.
189.It Ar counters
190The
191.Ar counters
192flag enables per-address packet and byte counters which can be displayed with
193.Xr pfctl 8 .
194Note that this feature carries significant memory overhead for large tables.
195.El
196.Pp
197For example,
198.Bd -literal -offset indent
199table \*(Ltprivate\*(Gt const { 10/8, 172.16/12, 192.168/16 }
200table \*(Ltbadhosts\*(Gt persist
201block on fxp0 from { \*(Ltprivate\*(Gt, \*(Ltbadhosts\*(Gt } to any
202.Ed
203.Pp
204creates a table called private, to hold RFC 1918 private network
205blocks, and a table called badhosts, which is initially empty.
206A filter rule is set up to block all traffic coming from addresses listed in
207either table.
208The private table cannot have its contents changed and the badhosts table
209will exist even when no active filter rules reference it.
210Addresses may later be added to the badhosts table, so that traffic from
211these hosts can be blocked by using
212.Bd -literal -offset indent
213# pfctl -t badhosts -Tadd 204.92.77.111
214.Ed
215.Pp
216A table can also be initialized with an address list specified in one or more
217external files, using the following syntax:
218.Bd -literal -offset indent
219table \*(Ltspam\*(Gt persist file \&"/etc/spammers\&" file \&"/etc/openrelays\&"
220block on fxp0 from \*(Ltspam\*(Gt to any
221.Ed
222.Pp
223The files
224.Pa /etc/spammers
225and
226.Pa /etc/openrelays
227list IP addresses, one per line.
228Any lines beginning with a # are treated as comments and ignored.
229In addition to being specified by IP address, hosts may also be
230specified by their hostname.
231When the resolver is called to add a hostname to a table,
232.Em all
233resulting IPv4 and IPv6 addresses are placed into the table.
234IP addresses can also be entered in a table by specifying a valid interface
235name, a valid interface group or the
236.Em self
237keyword, in which case all addresses assigned to the interface(s) will be
238added to the table.
239.Sh OPTIONS
240.Xr pf 4
241may be tuned for various situations using the
242.Ar set
243command.
244.Bl -tag -width xxxx
245.It Ar set timeout
246.Pp
247.Bl -tag -width "src.track" -compact
248.It Ar interval
249Interval between purging expired states and fragments.
250.It Ar frag
251Seconds before an unassembled fragment is expired.
252.It Ar src.track
253Length of time to retain a source tracking entry after the last state
254expires.
255.El
256.Pp
257When a packet matches a stateful connection, the seconds to live for the
258connection will be updated to that of the
259.Ar proto.modifier
260which corresponds to the connection state.
261Each packet which matches this state will reset the TTL.
262Tuning these values may improve the performance of the
263firewall at the risk of dropping valid idle connections.
264.Pp
265.Bl -tag -width xxxx -compact
266.It Ar tcp.first
267The state after the first packet.
268.It Ar tcp.opening
269The state before the destination host ever sends a packet.
270.It Ar tcp.established
271The fully established state.
272.It Ar tcp.closing
273The state after the first FIN has been sent.
274.It Ar tcp.finwait
275The state after both FINs have been exchanged and the connection is closed.
276Some hosts (notably web servers on Solaris) send TCP packets even after closing
277the connection.
278Increasing
279.Ar tcp.finwait
280(and possibly
281.Ar tcp.closing )
282can prevent blocking of such packets.
283.It Ar tcp.closed
284The state after one endpoint sends an RST.
285.El
286.Pp
287ICMP and UDP are handled in a fashion similar to TCP, but with a much more
288limited set of states:
289.Pp
290.Bl -tag -width xxxx -compact
291.It Ar udp.first
292The state after the first packet.
293.It Ar udp.single
294The state if the source host sends more than one packet but the destination
295host has never sent one back.
296.It Ar udp.multiple
297The state if both hosts have sent packets.
298.It Ar icmp.first
299The state after the first packet.
300.It Ar icmp.error
301The state after an ICMP error came back in response to an ICMP packet.
302.El
303.Pp
304Other protocols are handled similarly to UDP:
305.Pp
306.Bl -tag -width xxxx -compact
307.It Ar other.first
308.It Ar other.single
309.It Ar other.multiple
310.El
311.Pp
312Timeout values can be reduced adaptively as the number of state table
313entries grows.
314.Pp
315.Bl -tag -width xxxx -compact
316.It Ar adaptive.start
317When the number of state entries exceeds this value, adaptive scaling
318begins.
319All timeout values are scaled linearly with factor
320(adaptive.end - number of states) / (adaptive.end - adaptive.start).
321.It Ar adaptive.end
322When reaching this number of state entries, all timeout values become
323zero, effectively purging all state entries immediately.
324This value is used to define the scale factor, it should not actually
325be reached (set a lower state limit, see below).
326.El
327.Pp
328Adaptive timeouts are enabled by default, with an adaptive.start value
329equal to 60% of the state limit, and an adaptive.end value equal to
330120% of the state limit.
331They can be disabled by setting both adaptive.start and adaptive.end to 0.
332.Pp
333The adaptive timeout values can be defined both globally and for each rule.
334When used on a per-rule basis, the values relate to the number of
335states created by the rule, otherwise to the total number of
336states.
337.Pp
338For example:
339.Bd -literal -offset indent
340set timeout tcp.first 120
341set timeout tcp.established 86400
342set timeout { adaptive.start 6000, adaptive.end 12000 }
343set limit states 10000
344.Ed
345.Pp
346With 9000 state table entries, the timeout values are scaled to 50%
347(tcp.first 60, tcp.established 43200).
348.It Ar set loginterface
349Enable collection of packet and byte count statistics for the given
350interface or interface group.
351These statistics can be viewed using
352.Bd -literal -offset indent
353# pfctl -s info
354.Ed
355.Pp
356In this example
357.Xr pf 4
358collects statistics on the interface named dc0:
359.Bd -literal -offset indent
360set loginterface dc0
361.Ed
362.Pp
363One can disable the loginterface using:
364.Bd -literal -offset indent
365set loginterface none
366.Ed
367.It Ar set limit
368Sets hard limits on the memory pools used by the packet filter.
369See
370.Xr zone 9
371for an explanation of memory pools.
372.Pp
373For example,
374.Bd -literal -offset indent
375set limit states 20000
376.Ed
377.Pp
378sets the maximum number of entries in the memory pool used by state table
379entries (generated by
380.Ar pass
381rules which do not specify
382.Ar no state )
383to 20000.
384Using
385.Bd -literal -offset indent
386set limit frags 20000
387.Ed
388.Pp
389sets the maximum number of entries in the memory pool used for fragment
390reassembly (generated by
391.Ar scrub
392rules) to 20000.
393Using
394.Bd -literal -offset indent
395set limit src-nodes 2000
396.Ed
397.Pp
398sets the maximum number of entries in the memory pool used for tracking
399source IP addresses (generated by the
400.Ar sticky-address
401and
402.Ar src.track
403options) to 2000.
404Using
405.Bd -literal -offset indent
406set limit tables 1000
407set limit table-entries 100000
408.Ed
409.Pp
410sets limits on the memory pools used by tables.
411The first limits the number of tables that can exist to 1000.
412The second limits the overall number of addresses that can be stored
413in tables to 100000.
414.Pp
415Various limits can be combined on a single line:
416.Bd -literal -offset indent
417set limit { states 20000, frags 20000, src-nodes 2000 }
418.Ed
419.It Ar set ruleset-optimization
420.Bl -tag -width xxxxxxxx -compact
421.It Ar none
422Disable the ruleset optimizer.
423.It Ar basic
424Enable basic ruleset optimization.
425This is the default behaviour.
426Basic ruleset optimization does four things to improve the
427performance of ruleset evaluations:
428.Pp
429.Bl -enum -compact
430.It
431remove duplicate rules
432.It
433remove rules that are a subset of another rule
434.It
435combine multiple rules into a table when advantageous
436.It
437re-order the rules to improve evaluation performance
438.El
439.Pp
440.It Ar profile
441Uses the currently loaded ruleset as a feedback profile to tailor the
442ordering of quick rules to actual network traffic.
443.El
444.Pp
445It is important to note that the ruleset optimizer will modify the ruleset
446to improve performance.
447A side effect of the ruleset modification is that per-rule accounting
448statistics will have different meanings than before.
449If per-rule accounting is important for billing purposes or whatnot,
450either the ruleset optimizer should not be used or a label field should
451be added to all of the accounting rules to act as optimization barriers.
452.Pp
453Optimization can also be set as a command-line argument to
454.Xr pfctl 8 ,
455overriding the settings in
456.Nm .
457.It Ar set optimization
458Optimize state timeouts for one of the following network environments:
459.Pp
460.Bl -tag -width xxxx -compact
461.It Ar normal
462A normal network environment.
463Suitable for almost all networks.
464.It Ar high-latency
465A high-latency environment (such as a satellite connection).
466.It Ar satellite
467Alias for
468.Ar high-latency .
469.It Ar aggressive
470Aggressively expire connections.
471This can greatly reduce the memory usage of the firewall at the cost of
472dropping idle connections early.
473.It Ar conservative
474Extremely conservative settings.
475Avoid dropping legitimate connections at the
476expense of greater memory utilization (possibly much greater on a busy
477network) and slightly increased processor utilization.
478.El
479.Pp
480For example:
481.Bd -literal -offset indent
482set optimization aggressive
483.Ed
484.It Ar set block-policy
485The
486.Ar block-policy
487option sets the default behaviour for the packet
488.Ar block
489action:
490.Pp
491.Bl -tag -width xxxxxxxx -compact
492.It Ar drop
493Packet is silently dropped.
494.It Ar return
495A TCP RST is returned for blocked TCP packets,
496an ICMP UNREACHABLE is returned for blocked UDP packets,
497and all other packets are silently dropped.
498.El
499.Pp
500For example:
501.Bd -literal -offset indent
502set block-policy return
503.Ed
504.It Ar set fail-policy
505The
506.Ar fail-policy
507option sets the behaviour of rules which should pass a packet but were
508unable to do so.
509This might happen when a nat or route-to rule uses an empty table as list
510of targets or if a rule fails to create state or source node.
511The following
512.Ar block
513actions are possible:
514.Pp
515.Bl -tag -width xxxxxxxx -compact
516.It Ar drop
517Incoming packet is silently dropped.
518.It Ar return
519Incoming packet is dropped and TCP RST is returned for TCP packets,
520an ICMP UNREACHABLE is returned for UDP packets,
521and no response is sent for other packets.
522.El
523.Pp
524For example:
525.Bd -literal -offset indent
526set fail-policy return
527.Ed
528.It Ar set state-policy
529The
530.Ar state-policy
531option sets the default behaviour for states:
532.Pp
533.Bl -tag -width group-bound -compact
534.It Ar if-bound
535States are bound to interface.
536.It Ar floating
537States can match packets on any interfaces (the default).
538.El
539.Pp
540For example:
541.Bd -literal -offset indent
542set state-policy if-bound
543.Ed
544.It Ar set syncookies never | always | adaptive
545When
546.Cm syncookies
547are active, pf will answer each incoming TCP SYN with a syncookie SYNACK,
548without allocating any resources.
549Upon reception of the client's ACK in response to the syncookie
550SYNACK, pf will evaluate the ruleset and create state if the ruleset
551permits it, complete the three way handshake with the target host and
552continue the connection with synproxy in place.
553This allows pf to be resilient against large synflood attacks which would
554run the state table against its limits otherwise.
555Due to the blind answers to every incoming SYN syncookies share the caveats of
556synproxy, namely seemingly accepting connections that will be dropped later on.
557.Pp
558.Bl -tag -width adaptive -compact
559.It Cm never
560pf will never send syncookie SYNACKs (the default).
561.It Cm always
562pf will always send syncookie SYNACKs.
563.It Cm adaptive
564pf will enable syncookie mode when a given percentage of the state table
565is used up by half-open TCP connections, as in, those that saw the initial
566SYN but didn't finish the three way handshake.
567The thresholds for entering and leaving syncookie mode can be specified using
568.Bd -literal -offset indent
569set syncookies adaptive (start 25%, end 12%)
570.Ed
571.El
572.It Ar set state-defaults
573The
574.Ar state-defaults
575option sets the state options for states created from rules
576without an explicit
577.Ar keep state .
578For example:
579.Bd -literal -offset indent
580set state-defaults no-sync
581.Ed
582.It Ar set hostid
583The 32-bit
584.Ar hostid
585identifies this firewall's state table entries to other firewalls
586in a
587.Xr pfsync 4
588failover cluster.
589By default the hostid is set to a pseudo-random value, however it may be
590desirable to manually configure it, for example to more easily identify the
591source of state table entries.
592.Bd -literal -offset indent
593set hostid 1
594.Ed
595.Pp
596The hostid may be specified in either decimal or hexadecimal.
597.It Ar set require-order
598By default
599.Xr pfctl 8
600enforces an ordering of the statement types in the ruleset to:
601.Em options ,
602.Em normalization ,
603.Em queueing ,
604.Em translation ,
605.Em filtering .
606Setting this option to
607.Ar no
608disables this enforcement.
609There may be non-trivial and non-obvious implications to an out of
610order ruleset.
611Consider carefully before disabling the order enforcement.
612.It Ar set fingerprints
613Load fingerprints of known operating systems from the given filename.
614By default fingerprints of known operating systems are automatically
615loaded from
616.Xr pf.os 5
617in
618.Pa /etc
619but can be overridden via this option.
620Setting this option may leave a small period of time where the fingerprints
621referenced by the currently active ruleset are inconsistent until the new
622ruleset finishes loading.
623.Pp
624For example:
625.Pp
626.Dl set fingerprints \&"/etc/pf.os.devel\&"
627.It Ar set skip on Aq Ar ifspec
628List interfaces for which packets should not be filtered.
629Packets passing in or out on such interfaces are passed as if pf was
630disabled, i.e. pf does not process them in any way.
631This can be useful on loopback and other virtual interfaces, when
632packet filtering is not desired and can have unexpected effects.
633For example:
634.Pp
635.Dl set skip on lo0
636.It Ar set debug
637Set the debug
638.Ar level
639to one of the following:
640.Pp
641.Bl -tag -width xxxxxxxxxxxx -compact
642.It Ar none
643Don't generate debug messages.
644.It Ar urgent
645Generate debug messages only for serious errors.
646.It Ar misc
647Generate debug messages for various errors.
648.It Ar loud
649Generate debug messages for common conditions.
650.El
651.It Ar set keepcounters
652Preserve rule counters across rule updates.
653Usually rule counters are reset to zero on every update of the ruleset.
654With
655.Ar keepcounters
656set pf will attempt to find matching rules between old and new rulesets
657and preserve the rule counters.
658.El
659.Sh ETHERNET FILTERING
660.Xr pf 4
661has the ability to
662.Ar block
663and
664.Ar pass
665packets based on attributes of their Ethernet (layer 2) header.
666.Pp
667For each packet processed by the packet filter, the filter rules are
668evaluated in sequential order, from first to last.
669The last matching rule decides what action is taken.
670If no rule matches the packet, the default action is to pass
671the packet.
672.Pp
673The folliwing actions can be used in the filter:
674.Bl -tag -width xxxx
675.It Ar block
676The packet is blocked.
677Unlike for layer 3 traffic the packet is always silently dropped.
678.It Ar pass
679The packet is passed;
680no state is created for layer 2 traffic.
681.El
682.Sh PARAMETERS
683The rule parameters specify the packets to which a rule applies.
684A packet always comes in on, or goes out through, one interface.
685Most parameters are optional.
686If a parameter is specified, the rule only applies to packets with
687matching attributes.
688Certain parameters can be expressed as lists, in which case
689.Xr pfctl 8
690generates all needed rule combinations.
691.Bl -tag -width xxxx
692.It Ar in No or Ar out
693This rule applies to incoming or outgoing packets.
694If neither
695.Ar in
696nor
697.Ar out
698are specified, the rule will match packets in both directions.
699.It Ar quick
700If a packet matches a rule which has the
701.Ar quick
702option set, this rule
703is considered the last matching rule, and evaluation of subsequent rules
704is skipped.
705.It Ar on Aq Ar interface
706This rule applies only to packets coming in on, or going out through, this
707particular interface or interface group.
708For more information on interface groups,
709see the
710.Ic group
711keyword in
712.Xr ifconfig 8 .
713.It Ar bridge-to Aq interface
714Packets matching this rule will be sent out of the specified interface without
715futher processing.
716.It Ar proto Aq Ar protocol
717This rule applies only to packets of this protocol.
718Note that Ethernet protocol numbers are different from those used in
719.Xr ip 4
720and
721.Xr ip6 4 .
722.It Xo
723.Ar from Aq Ar source
724.Ar to Aq Ar dest
725.Xc
726This rule applies only to packets with the specified source and destination
727MAC addresses.
728.It Xo Ar queue Aq Ar queue
729.Xc
730Packets matching this rule will be assigned to the specified queue.
731See
732.Sx QUEUEING
733for setup details.
734.Pp
735.It Ar tag Aq Ar string
736Packets matching this rule will be tagged with the
737specified string.
738The tag acts as an internal marker that can be used to
739identify these packets later on.
740This can be used, for example, to provide trust between
741interfaces and to determine if packets have been
742processed by translation rules.
743Tags are
744.Qq sticky ,
745meaning that the packet will be tagged even if the rule
746is not the last matching rule.
747Further matching rules can replace the tag with a
748new one but will not remove a previously applied tag.
749A packet is only ever assigned one tag at a time.
750.It Ar tagged Aq Ar string
751Used to specify that packets must already be tagged with the given tag in order
752to match the rule.
753Inverse tag matching can also be done by specifying the !  operator before the
754tagged keyword.
755.Sh TRAFFIC NORMALIZATION
756Traffic normalization is used to sanitize packet content in such
757a way that there are no ambiguities in packet interpretation on
758the receiving side.
759The normalizer does IP fragment reassembly to prevent attacks
760that confuse intrusion detection systems by sending overlapping
761IP fragments.
762Packet normalization is invoked with the
763.Ar scrub
764directive.
765.Pp
766.Ar scrub
767has the following options:
768.Bl -tag -width xxxx
769.It Ar no-df
770Clears the
771.Ar dont-fragment
772bit from a matching IP packet.
773Some operating systems are known to generate fragmented packets with the
774.Ar dont-fragment
775bit set.
776This is particularly true with NFS.
777.Ar Scrub
778will drop such fragmented
779.Ar dont-fragment
780packets unless
781.Ar no-df
782is specified.
783.Pp
784Unfortunately some operating systems also generate their
785.Ar dont-fragment
786packets with a zero IP identification field.
787Clearing the
788.Ar dont-fragment
789bit on packets with a zero IP ID may cause deleterious results if an
790upstream router later fragments the packet.
791Using the
792.Ar random-id
793modifier (see below) is recommended in combination with the
794.Ar no-df
795modifier to ensure unique IP identifiers.
796.It Ar min-ttl Aq Ar number
797Enforces a minimum TTL for matching IP packets.
798.It Ar max-mss Aq Ar number
799Enforces a maximum MSS for matching TCP packets.
800.It Xo Ar set-tos Aq Ar string
801.No \*(Ba Aq Ar number
802.Xc
803Enforces a
804.Em TOS
805for matching IP packets.
806.Em TOS
807may be
808given as one of
809.Ar critical ,
810.Ar inetcontrol ,
811.Ar lowdelay ,
812.Ar netcontrol ,
813.Ar throughput ,
814.Ar reliability ,
815or one of the DiffServ Code Points:
816.Ar ef ,
817.Ar va ,
818.Ar af11 No ... Ar af43 ,
819.Ar cs0 No ... Ar cs7 ;
820or as either hex or decimal.
821.It Ar random-id
822Replaces the IP identification field with random values to compensate
823for predictable values generated by many hosts.
824This option only applies to packets that are not fragmented
825after the optional fragment reassembly.
826.It Ar fragment reassemble
827Using
828.Ar scrub
829rules, fragments can be reassembled by normalization.
830In this case, fragments are buffered until they form a complete
831packet, and only the completed packet is passed on to the filter.
832The advantage is that filter rules have to deal only with complete
833packets, and can ignore fragments.
834The drawback of caching fragments is the additional memory cost.
835This is the default behaviour unless no fragment reassemble is specified.
836.It Ar no fragment reassemble
837Do not reassemble fragments.
838.It Ar reassemble tcp
839Statefully normalizes TCP connections.
840.Ar scrub reassemble tcp
841rules may not have the direction (in/out) specified.
842.Ar reassemble tcp
843performs the following normalizations:
844.Pp
845.Bl -tag -width timeout -compact
846.It ttl
847Neither side of the connection is allowed to reduce their IP TTL.
848An attacker may send a packet such that it reaches the firewall, affects
849the firewall state, and expires before reaching the destination host.
850.Ar reassemble tcp
851will raise the TTL of all packets back up to the highest value seen on
852the connection.
853.It timestamp modulation
854Modern TCP stacks will send a timestamp on every TCP packet and echo
855the other endpoint's timestamp back to them.
856Many operating systems will merely start the timestamp at zero when
857first booted, and increment it several times a second.
858The uptime of the host can be deduced by reading the timestamp and multiplying
859by a constant.
860Also observing several different timestamps can be used to count hosts
861behind a NAT device.
862And spoofing TCP packets into a connection requires knowing or guessing
863valid timestamps.
864Timestamps merely need to be monotonically increasing and not derived off a
865guessable base time.
866.Ar reassemble tcp
867will cause
868.Ar scrub
869to modulate the TCP timestamps with a random number.
870.It extended PAWS checks
871There is a problem with TCP on long fat pipes, in that a packet might get
872delayed for longer than it takes the connection to wrap its 32-bit sequence
873space.
874In such an occurrence, the old packet would be indistinguishable from a
875new packet and would be accepted as such.
876The solution to this is called PAWS: Protection Against Wrapped Sequence
877numbers.
878It protects against it by making sure the timestamp on each packet does
879not go backwards.
880.Ar reassemble tcp
881also makes sure the timestamp on the packet does not go forward more
882than the RFC allows.
883By doing this,
884.Xr pf 4
885artificially extends the security of TCP sequence numbers by 10 to 18
886bits when the host uses appropriately randomized timestamps, since a
887blind attacker would have to guess the timestamp as well.
888.El
889.El
890.Pp
891For example,
892.Bd -literal -offset indent
893scrub in on $ext_if all fragment reassemble
894.Ed
895.Pp
896The
897.Ar no
898option prefixed to a scrub rule causes matching packets to remain unscrubbed,
899much in the same way as
900.Ar drop quick
901works in the packet filter (see below).
902This mechanism should be used when it is necessary to exclude specific packets
903from broader scrub rules.
904.Sh QUEUEING with ALTQ
905The ALTQ system is currently not available in the GENERIC kernel nor as
906loadable modules.
907In order to use the herein after called queueing options one has to use a
908custom built kernel.
909Please refer to
910.Xr altq 4
911to learn about the related kernel options.
912.Pp
913Packets can be assigned to queues for the purpose of bandwidth
914control.
915At least two declarations are required to configure queues, and later
916any packet filtering rule can reference the defined queues by name.
917During the filtering component of
918.Nm pf.conf ,
919the last referenced
920.Ar queue
921name is where any packets from
922.Ar pass
923rules will be queued, while for
924.Ar block
925rules it specifies where any resulting ICMP or TCP RST
926packets should be queued.
927The
928.Ar scheduler
929defines the algorithm used to decide which packets get delayed, dropped, or
930sent out immediately.
931There are three
932.Ar schedulers
933currently supported.
934.Bl -tag -width xxxx
935.It Ar cbq
936Class Based Queueing.
937.Ar Queues
938attached to an interface build a tree, thus each
939.Ar queue
940can have further child
941.Ar queues .
942Each queue can have a
943.Ar priority
944and a
945.Ar bandwidth
946assigned.
947.Ar Priority
948mainly controls the time packets take to get sent out, while
949.Ar bandwidth
950has primarily effects on throughput.
951.Ar cbq
952achieves both partitioning and sharing of link bandwidth
953by hierarchically structured classes.
954Each class has its own
955.Ar queue
956and is assigned its share of
957.Ar bandwidth .
958A child class can borrow bandwidth from its parent class
959as long as excess bandwidth is available
960(see the option
961.Ar borrow ,
962below).
963.It Ar priq
964Priority Queueing.
965.Ar Queues
966are flat attached to the interface, thus,
967.Ar queues
968cannot have further child
969.Ar queues .
970Each
971.Ar queue
972has a unique
973.Ar priority
974assigned, ranging from 0 to 15.
975Packets in the
976.Ar queue
977with the highest
978.Ar priority
979are processed first.
980.It Ar hfsc
981Hierarchical Fair Service Curve.
982.Ar Queues
983attached to an interface build a tree, thus each
984.Ar queue
985can have further child
986.Ar queues .
987Each queue can have a
988.Ar priority
989and a
990.Ar bandwidth
991assigned.
992.Ar Priority
993mainly controls the time packets take to get sent out, while
994.Ar bandwidth
995primarily affects throughput.
996.Ar hfsc
997supports both link-sharing and guaranteed real-time services.
998It employs a service curve based QoS model,
999and its unique feature is an ability to decouple
1000.Ar delay
1001and
1002.Ar bandwidth
1003allocation.
1004.El
1005.Pp
1006The interfaces on which queueing should be activated are declared using
1007the
1008.Ar altq on
1009declaration.
1010.Ar altq on
1011has the following keywords:
1012.Bl -tag -width xxxx
1013.It Aq Ar interface
1014Queueing is enabled on the named interface.
1015.It Aq Ar scheduler
1016Specifies which queueing scheduler to use.
1017Currently supported values
1018are
1019.Ar cbq
1020for Class Based Queueing,
1021.Ar priq
1022for Priority Queueing and
1023.Ar hfsc
1024for the Hierarchical Fair Service Curve scheduler.
1025.It Ar bandwidth Aq Ar bw
1026The maximum bitrate for all queues on an
1027interface may be specified using the
1028.Ar bandwidth
1029keyword.
1030The value can be specified as an absolute value or as a
1031percentage of the interface bandwidth.
1032When using an absolute value, the suffixes
1033.Ar b ,
1034.Ar Kb ,
1035.Ar Mb ,
1036and
1037.Ar Gb
1038are used to represent bits, kilobits, megabits, and
1039gigabits per second, respectively.
1040The value must not exceed the interface bandwidth.
1041If
1042.Ar bandwidth
1043is not specified, the interface bandwidth is used
1044(but take note that some interfaces do not know their bandwidth,
1045or can adapt their bandwidth rates).
1046.It Ar qlimit Aq Ar limit
1047The maximum number of packets held in the queue.
1048The default is 50.
1049.It Ar tbrsize Aq Ar size
1050Adjusts the size, in bytes, of the token bucket regulator.
1051If not specified, heuristics based on the
1052interface bandwidth are used to determine the size.
1053.It Ar queue Aq Ar list
1054Defines a list of subqueues to create on an interface.
1055.El
1056.Pp
1057In the following example, the interface dc0
1058should queue up to 5Mbps in four second-level queues using
1059Class Based Queueing.
1060Those four queues will be shown in a later example.
1061.Bd -literal -offset indent
1062altq on dc0 cbq bandwidth 5Mb queue { std, http, mail, ssh }
1063.Ed
1064.Pp
1065Once interfaces are activated for queueing using the
1066.Ar altq
1067directive, a sequence of
1068.Ar queue
1069directives may be defined.
1070The name associated with a
1071.Ar queue
1072must match a queue defined in the
1073.Ar altq
1074directive (e.g. mail), or, except for the
1075.Ar priq
1076.Ar scheduler ,
1077in a parent
1078.Ar queue
1079declaration.
1080The following keywords can be used:
1081.Bl -tag -width xxxx
1082.It Ar on Aq Ar interface
1083Specifies the interface the queue operates on.
1084If not given, it operates on all matching interfaces.
1085.It Ar bandwidth Aq Ar bw
1086Specifies the maximum bitrate to be processed by the queue.
1087This value must not exceed the value of the parent
1088.Ar queue
1089and can be specified as an absolute value or a percentage of the parent
1090queue's bandwidth.
1091If not specified, defaults to 100% of the parent queue's bandwidth.
1092The
1093.Ar priq
1094scheduler does not support bandwidth specification.
1095.It Ar priority Aq Ar level
1096Between queues a priority level can be set.
1097For
1098.Ar cbq
1099and
1100.Ar hfsc ,
1101the range is 0 to 7 and for
1102.Ar priq ,
1103the range is 0 to 15.
1104The default for all is 1.
1105.Ar Priq
1106queues with a higher priority are always served first.
1107.Ar Cbq
1108and
1109.Ar Hfsc
1110queues with a higher priority are preferred in the case of overload.
1111.It Ar qlimit Aq Ar limit
1112The maximum number of packets held in the queue.
1113The default is 50.
1114.El
1115.Pp
1116The
1117.Ar scheduler
1118can get additional parameters with
1119.Xo Aq Ar scheduler
1120.Pf ( Aq Ar parameters ) .
1121.Xc
1122Parameters are as follows:
1123.Bl -tag -width Fl
1124.It Ar default
1125Packets not matched by another queue are assigned to this one.
1126Exactly one default queue is required.
1127.It Ar red
1128Enable RED (Random Early Detection) on this queue.
1129RED drops packets with a probability proportional to the average
1130queue length.
1131.It Ar rio
1132Enables RIO on this queue.
1133RIO is RED with IN/OUT, thus running
1134RED two times more than RIO would achieve the same effect.
1135RIO is currently not supported in the GENERIC kernel.
1136.It Ar ecn
1137Enables ECN (Explicit Congestion Notification) on this queue.
1138ECN implies RED.
1139.El
1140.Pp
1141The
1142.Ar cbq
1143.Ar scheduler
1144supports an additional option:
1145.Bl -tag -width Fl
1146.It Ar borrow
1147The queue can borrow bandwidth from the parent.
1148.El
1149.Pp
1150The
1151.Ar hfsc
1152.Ar scheduler
1153supports some additional options:
1154.Bl -tag -width Fl
1155.It Ar realtime Aq Ar sc
1156The minimum required bandwidth for the queue.
1157.It Ar upperlimit Aq Ar sc
1158The maximum allowed bandwidth for the queue.
1159.It Ar linkshare Aq Ar sc
1160The bandwidth share of a backlogged queue.
1161.El
1162.Pp
1163.Aq Ar sc
1164is an acronym for
1165.Ar service curve .
1166.Pp
1167The format for service curve specifications is
1168.Ar ( m1 , d , m2 ) .
1169.Ar m2
1170controls the bandwidth assigned to the queue.
1171.Ar m1
1172and
1173.Ar d
1174are optional and can be used to control the initial bandwidth assignment.
1175For the first
1176.Ar d
1177milliseconds the queue gets the bandwidth given as
1178.Ar m1 ,
1179afterwards the value given in
1180.Ar m2 .
1181.Pp
1182Furthermore, with
1183.Ar cbq
1184and
1185.Ar hfsc ,
1186child queues can be specified as in an
1187.Ar altq
1188declaration, thus building a tree of queues using a part of
1189their parent's bandwidth.
1190.Pp
1191Packets can be assigned to queues based on filter rules by using the
1192.Ar queue
1193keyword.
1194Normally only one
1195.Ar queue
1196is specified; when a second one is specified it will instead be used for
1197packets which have a
1198.Em TOS
1199of
1200.Em lowdelay
1201and for TCP ACKs with no data payload.
1202.Pp
1203To continue the previous example, the examples below would specify the
1204four referenced
1205queues, plus a few child queues.
1206Interactive
1207.Xr ssh 1
1208sessions get priority over bulk transfers like
1209.Xr scp 1
1210and
1211.Xr sftp 1 .
1212The queues may then be referenced by filtering rules (see
1213.Sx PACKET FILTERING
1214below).
1215.Bd -literal
1216queue std bandwidth 10% cbq(default)
1217queue http bandwidth 60% priority 2 cbq(borrow red) \e
1218      { employees, developers }
1219queue  developers bandwidth 75% cbq(borrow)
1220queue  employees bandwidth 15%
1221queue mail bandwidth 10% priority 0 cbq(borrow ecn)
1222queue ssh bandwidth 20% cbq(borrow) { ssh_interactive, ssh_bulk }
1223queue  ssh_interactive bandwidth 50% priority 7 cbq(borrow)
1224queue  ssh_bulk bandwidth 50% priority 0 cbq(borrow)
1225
1226block return out on dc0 inet all queue std
1227pass out on dc0 inet proto tcp from $developerhosts to any port 80 \e
1228      queue developers
1229pass out on dc0 inet proto tcp from $employeehosts to any port 80 \e
1230      queue employees
1231pass out on dc0 inet proto tcp from any to any port 22 \e
1232      queue(ssh_bulk, ssh_interactive)
1233pass out on dc0 inet proto tcp from any to any port 25 \e
1234      queue mail
1235.Ed
1236.Sh QUEUEING with dummynet
1237Queueing can also be done with
1238.Xr dummynet 4 .
1239Queues and pipes can be created with
1240.Xr dnctl 8 .
1241.Pp
1242Packets can be assigned to queues and pipes using
1243.Ar dnqueue
1244and
1245.Ar dnpipe
1246respectively.
1247.Pp
1248Both
1249.Ar dnqueue
1250and
1251.Ar dnpipe
1252take either a single pipe or queue number or two numbers as arguments.
1253The first pipe or queue number will be used to shape the traffic in the rule
1254direction, the second will be used to shape the traffic in the reverse
1255direction.
1256If the rule does not specify a direction the first packet to create state will
1257be shaped according to the first number, and the response traffic according to
1258the second.
1259.Pp
1260If the
1261.Xr dummynet 4
1262module is not loaded any traffic sent into a queue or pipe will be dropped.
1263.Sh TRANSLATION
1264Translation rules modify either the source or destination address of the
1265packets associated with a stateful connection.
1266A stateful connection is automatically created to track packets matching
1267such a rule as long as they are not blocked by the filtering section of
1268.Nm pf.conf .
1269The translation engine modifies the specified address and/or port in the
1270packet, recalculates IP, TCP and UDP checksums as necessary, and passes it to
1271the packet filter for evaluation.
1272.Pp
1273Since translation occurs before filtering the filter
1274engine will see packets as they look after any
1275addresses and ports have been translated.
1276Filter rules will therefore have to filter based on the translated
1277address and port number.
1278Packets that match a translation rule are only automatically passed if
1279the
1280.Ar pass
1281modifier is given, otherwise they are
1282still subject to
1283.Ar block
1284and
1285.Ar pass
1286rules.
1287.Pp
1288The state entry created permits
1289.Xr pf 4
1290to keep track of the original address for traffic associated with that state
1291and correctly direct return traffic for that connection.
1292.Pp
1293Various types of translation are possible with pf:
1294.Bl -tag -width xxxx
1295.It Ar binat
1296A
1297.Ar binat
1298rule specifies a bidirectional mapping between an external IP netblock
1299and an internal IP netblock.
1300.It Ar nat
1301A
1302.Ar nat
1303rule specifies that IP addresses are to be changed as the packet
1304traverses the given interface.
1305This technique allows one or more IP addresses
1306on the translating host to support network traffic for a larger range of
1307machines on an "inside" network.
1308Although in theory any IP address can be used on the inside, it is strongly
1309recommended that one of the address ranges defined by RFC 1918 be used.
1310These netblocks are:
1311.Bd -literal
131210.0.0.0 - 10.255.255.255 (all of net 10, i.e., 10/8)
1313172.16.0.0 - 172.31.255.255 (i.e., 172.16/12)
1314192.168.0.0 - 192.168.255.255 (i.e., 192.168/16)
1315.Ed
1316.It Pa rdr
1317The packet is redirected to another destination and possibly a
1318different port.
1319.Ar rdr
1320rules can optionally specify port ranges instead of single ports.
1321rdr ... port 2000:2999 -\*(Gt ... port 4000
1322redirects ports 2000 to 2999 (inclusive) to port 4000.
1323rdr ... port 2000:2999 -\*(Gt ... port 4000:*
1324redirects port 2000 to 4000, 2001 to 4001, ..., 2999 to 4999.
1325.El
1326.Pp
1327In addition to modifying the address, some translation rules may modify
1328source or destination ports for
1329.Xr tcp 4
1330or
1331.Xr udp 4
1332connections; implicitly in the case of
1333.Ar nat
1334rules and explicitly in the case of
1335.Ar rdr
1336rules.
1337Port numbers are never translated with a
1338.Ar binat
1339rule.
1340.Pp
1341Evaluation order of the translation rules is dependent on the type
1342of the translation rules and of the direction of a packet.
1343.Ar binat
1344rules are always evaluated first.
1345Then either the
1346.Ar rdr
1347rules are evaluated on an inbound packet or the
1348.Ar nat
1349rules on an outbound packet.
1350Rules of the same type are evaluated in the same order in which they
1351appear in the ruleset.
1352The first matching rule decides what action is taken.
1353.Pp
1354The
1355.Ar no
1356option prefixed to a translation rule causes packets to remain untranslated,
1357much in the same way as
1358.Ar drop quick
1359works in the packet filter (see below).
1360If no rule matches the packet it is passed to the filter engine unmodified.
1361.Pp
1362Translation rules apply only to packets that pass through
1363the specified interface, and if no interface is specified,
1364translation is applied to packets on all interfaces.
1365For instance, redirecting port 80 on an external interface to an internal
1366web server will only work for connections originating from the outside.
1367Connections to the address of the external interface from local hosts will
1368not be redirected, since such packets do not actually pass through the
1369external interface.
1370Redirections cannot reflect packets back through the interface they arrive
1371on, they can only be redirected to hosts connected to different interfaces
1372or to the firewall itself.
1373.Pp
1374Note that redirecting external incoming connections to the loopback
1375address, as in
1376.Bd -literal -offset indent
1377rdr on ne3 inet proto tcp to port smtp -\*(Gt 127.0.0.1 port spamd
1378.Ed
1379.Pp
1380will effectively allow an external host to connect to daemons
1381bound solely to the loopback address, circumventing the traditional
1382blocking of such connections on a real interface.
1383Unless this effect is desired, any of the local non-loopback addresses
1384should be used as redirection target instead, which allows external
1385connections only to daemons bound to this address or not bound to
1386any address.
1387.Pp
1388See
1389.Sx TRANSLATION EXAMPLES
1390below.
1391.Sh PACKET FILTERING
1392.Xr pf 4
1393has the ability to
1394.Ar block
1395,
1396.Ar pass
1397and
1398.Ar match
1399packets based on attributes of their layer 3 (see
1400.Xr ip 4
1401and
1402.Xr ip6 4 )
1403and layer 4 (see
1404.Xr icmp 4 ,
1405.Xr icmp6 4 ,
1406.Xr tcp 4 ,
1407.Xr udp 4 )
1408headers.
1409In addition, packets may also be
1410assigned to queues for the purpose of bandwidth control.
1411.Pp
1412For each packet processed by the packet filter, the filter rules are
1413evaluated in sequential order, from first to last.
1414For
1415.Ar block
1416and
1417.Ar pass
1418, the last matching rule decides what action is taken.
1419For
1420.Ar match
1421, rules are evaulated every time they match; the pass/block state of a packet
1422remains unchanged.
1423If no rule matches the packet, the default action is to pass
1424the packet.
1425.Pp
1426The following actions can be used in the filter:
1427.Bl -tag -width xxxx
1428.It Ar block
1429The packet is blocked.
1430There are a number of ways in which a
1431.Ar block
1432rule can behave when blocking a packet.
1433The default behaviour is to
1434.Ar drop
1435packets silently, however this can be overridden or made
1436explicit either globally, by setting the
1437.Ar block-policy
1438option, or on a per-rule basis with one of the following options:
1439.Pp
1440.Bl -tag -width xxxx -compact
1441.It Ar drop
1442The packet is silently dropped.
1443.It Ar return-rst
1444This applies only to
1445.Xr tcp 4
1446packets, and issues a TCP RST which closes the
1447connection.
1448.It Ar return-icmp
1449.It Ar return-icmp6
1450This causes ICMP messages to be returned for packets which match the rule.
1451By default this is an ICMP UNREACHABLE message, however this
1452can be overridden by specifying a message as a code or number.
1453.It Ar return
1454This causes a TCP RST to be returned for
1455.Xr tcp 4
1456packets and an ICMP UNREACHABLE for UDP and other packets.
1457.El
1458.Pp
1459Options returning ICMP packets currently have no effect if
1460.Xr pf 4
1461operates on a
1462.Xr if_bridge 4 ,
1463as the code to support this feature has not yet been implemented.
1464.Pp
1465The simplest mechanism to block everything by default and only pass
1466packets that match explicit rules is specify a first filter rule of:
1467.Bd -literal -offset indent
1468block all
1469.Ed
1470.It Ar match
1471The packet is matched.
1472This mechanism is used to provide fine grained filtering without altering the
1473block/pass state of a packet.
1474.Ar match
1475rules differ from
1476.Ar block
1477and
1478.Ar pass
1479rules in that parameters are set every time a packet matches the rule, not only
1480on the last matching rule.
1481For the following parameters, this means that the parameter effectively becomes
1482"sticky" until explicitly overridden:
1483.Ar queue ,
1484.Ar dnpipe ,
1485.Ar dnqueue
1486.
1487.It Ar pass
1488The packet is passed;
1489state is created unless the
1490.Ar no state
1491option is specified.
1492.El
1493.Pp
1494By default
1495.Xr pf 4
1496filters packets statefully; the first time a packet matches a
1497.Ar pass
1498rule, a state entry is created; for subsequent packets the filter checks
1499whether the packet matches any state.
1500If it does, the packet is passed without evaluation of any rules.
1501After the connection is closed or times out, the state entry is automatically
1502removed.
1503.Pp
1504This has several advantages.
1505For TCP connections, comparing a packet to a state involves checking
1506its sequence numbers, as well as TCP timestamps if a
1507.Ar scrub reassemble tcp
1508rule applies to the connection.
1509If these values are outside the narrow windows of expected
1510values, the packet is dropped.
1511This prevents spoofing attacks, such as when an attacker sends packets with
1512a fake source address/port but does not know the connection's sequence
1513numbers.
1514Similarly,
1515.Xr pf 4
1516knows how to match ICMP replies to states.
1517For example,
1518.Bd -literal -offset indent
1519pass out inet proto icmp all icmp-type echoreq
1520.Ed
1521.Pp
1522allows echo requests (such as those created by
1523.Xr ping 8 )
1524out statefully, and matches incoming echo replies correctly to states.
1525.Pp
1526Also, looking up states is usually faster than evaluating rules.
1527If there are 50 rules, all of them are evaluated sequentially in O(n).
1528Even with 50000 states, only 16 comparisons are needed to match a
1529state, since states are stored in a binary search tree that allows
1530searches in O(log2 n).
1531.Pp
1532Furthermore, correct handling of ICMP error messages is critical to
1533many protocols, particularly TCP.
1534.Xr pf 4
1535matches ICMP error messages to the correct connection, checks them against
1536connection parameters, and passes them if appropriate.
1537For example if an ICMP source quench message referring to a stateful TCP
1538connection arrives, it will be matched to the state and get passed.
1539.Pp
1540Finally, state tracking is required for
1541.Ar nat , binat No and Ar rdr
1542rules, in order to track address and port translations and reverse the
1543translation on returning packets.
1544.Pp
1545.Xr pf 4
1546will also create state for other protocols which are effectively stateless by
1547nature.
1548UDP packets are matched to states using only host addresses and ports,
1549and other protocols are matched to states using only the host addresses.
1550.Pp
1551If stateless filtering of individual packets is desired,
1552the
1553.Ar no state
1554keyword can be used to specify that state will not be created
1555if this is the last matching rule.
1556A number of parameters can also be set to affect how
1557.Xr pf 4
1558handles state tracking.
1559See
1560.Sx STATEFUL TRACKING OPTIONS
1561below for further details.
1562.Sh PARAMETERS
1563The rule parameters specify the packets to which a rule applies.
1564A packet always comes in on, or goes out through, one interface.
1565Most parameters are optional.
1566If a parameter is specified, the rule only applies to packets with
1567matching attributes.
1568Certain parameters can be expressed as lists, in which case
1569.Xr pfctl 8
1570generates all needed rule combinations.
1571.Bl -tag -width xxxx
1572.It Ar in No or Ar out
1573This rule applies to incoming or outgoing packets.
1574If neither
1575.Ar in
1576nor
1577.Ar out
1578are specified, the rule will match packets in both directions.
1579.It Ar log
1580In addition to the action specified, a log message is generated.
1581Only the packet that establishes the state is logged,
1582unless the
1583.Ar no state
1584option is specified.
1585The logged packets are sent to a
1586.Xr pflog 4
1587interface, by default
1588.Ar pflog0 .
1589This interface is monitored by the
1590.Xr pflogd 8
1591logging daemon, which dumps the logged packets to the file
1592.Pa /var/log/pflog
1593in
1594.Xr pcap 3
1595binary format.
1596.It Ar log (all)
1597Used to force logging of all packets for a connection.
1598This is not necessary when
1599.Ar no state
1600is explicitly specified.
1601As with
1602.Ar log ,
1603packets are logged to
1604.Xr pflog 4 .
1605.It Ar log (user)
1606Logs the
1607.Ux
1608user ID of the user that owns the socket and the PID of the process that
1609has the socket open where the packet is sourced from or destined to
1610(depending on which socket is local).
1611This is in addition to the normal information logged.
1612.Pp
1613Only the first packet
1614logged via
1615.Ar log (all, user)
1616will have the user credentials logged when using stateful matching.
1617.It Ar log (to Aq Ar interface )
1618Send logs to the specified
1619.Xr pflog 4
1620interface instead of
1621.Ar pflog0 .
1622.It Ar quick
1623If a packet matches a rule which has the
1624.Ar quick
1625option set, this rule
1626is considered the last matching rule, and evaluation of subsequent rules
1627is skipped.
1628.It Ar on Aq Ar interface
1629This rule applies only to packets coming in on, or going out through, this
1630particular interface or interface group.
1631For more information on interface groups,
1632see the
1633.Ic group
1634keyword in
1635.Xr ifconfig 8 .
1636.It Aq Ar af
1637This rule applies only to packets of this address family.
1638Supported values are
1639.Ar inet
1640and
1641.Ar inet6 .
1642.It Ar proto Aq Ar protocol
1643This rule applies only to packets of this protocol.
1644Common protocols are
1645.Xr icmp 4 ,
1646.Xr icmp6 4 ,
1647.Xr tcp 4 ,
1648and
1649.Xr udp 4 .
1650For a list of all the protocol name to number mappings used by
1651.Xr pfctl 8 ,
1652see the file
1653.Pa /etc/protocols .
1654.It Xo
1655.Ar from Aq Ar source
1656.Ar port Aq Ar source
1657.Ar os Aq Ar source
1658.Ar to Aq Ar dest
1659.Ar port Aq Ar dest
1660.Xc
1661This rule applies only to packets with the specified source and destination
1662addresses and ports.
1663.Pp
1664Addresses can be specified in CIDR notation (matching netblocks), as
1665symbolic host names, interface names or interface group names, or as any
1666of the following keywords:
1667.Pp
1668.Bl -tag -width xxxxxxxxxxxxxx -compact
1669.It Ar any
1670Any address.
1671.It Ar no-route
1672Any address which is not currently routable.
1673.It Ar urpf-failed
1674Any source address that fails a unicast reverse path forwarding (URPF)
1675check, i.e. packets coming in on an interface other than that which holds
1676the route back to the packet's source address.
1677.It Aq Ar table
1678Any address that matches the given table.
1679.El
1680.Pp
1681Ranges of addresses are specified by using the
1682.Sq -
1683operator.
1684For instance:
1685.Dq 10.1.1.10 - 10.1.1.12
1686means all addresses from 10.1.1.10 to 10.1.1.12,
1687hence addresses 10.1.1.10, 10.1.1.11, and 10.1.1.12.
1688.Pp
1689Interface names and interface group names can have modifiers appended:
1690.Pp
1691.Bl -tag -width xxxxxxxxxxxx -compact
1692.It Ar :network
1693Translates to the network(s) attached to the interface.
1694.It Ar :broadcast
1695Translates to the interface's broadcast address(es).
1696.It Ar :peer
1697Translates to the point-to-point interface's peer address(es).
1698.It Ar :0
1699Do not include interface aliases.
1700.El
1701.Pp
1702Host names may also have the
1703.Ar :0
1704option appended to restrict the name resolution to the first of each
1705v4 and non-link-local v6 address found.
1706.Pp
1707Host name resolution and interface to address translation are done at
1708ruleset load-time.
1709When the address of an interface (or host name) changes (under DHCP or PPP,
1710for instance), the ruleset must be reloaded for the change to be reflected
1711in the kernel.
1712Surrounding the interface name (and optional modifiers) in parentheses
1713changes this behaviour.
1714When the interface name is surrounded by parentheses, the rule is
1715automatically updated whenever the interface changes its address.
1716The ruleset does not need to be reloaded.
1717This is especially useful with
1718.Ar nat .
1719.Pp
1720Ports can be specified either by number or by name.
1721For example, port 80 can be specified as
1722.Em www .
1723For a list of all port name to number mappings used by
1724.Xr pfctl 8 ,
1725see the file
1726.Pa /etc/services .
1727.Pp
1728Ports and ranges of ports are specified by using these operators:
1729.Bd -literal -offset indent
1730=	(equal)
1731!=	(unequal)
1732\*(Lt	(less than)
1733\*(Le	(less than or equal)
1734\*(Gt	(greater than)
1735\*(Ge	(greater than or equal)
1736:	(range including boundaries)
1737\*(Gt\*(Lt	(range excluding boundaries)
1738\*(Lt\*(Gt	(except range)
1739.Ed
1740.Pp
1741.Sq \*(Gt\*(Lt ,
1742.Sq \*(Lt\*(Gt
1743and
1744.Sq \&:
1745are binary operators (they take two arguments).
1746For instance:
1747.Bl -tag -width Fl
1748.It Ar port 2000:2004
1749means
1750.Sq all ports \*(Ge 2000 and \*(Le 2004 ,
1751hence ports 2000, 2001, 2002, 2003 and 2004.
1752.It Ar port 2000 \*(Gt\*(Lt 2004
1753means
1754.Sq all ports \*(Gt 2000 and \*(Lt 2004 ,
1755hence ports 2001, 2002 and 2003.
1756.It Ar port 2000 \*(Lt\*(Gt 2004
1757means
1758.Sq all ports \*(Lt 2000 or \*(Gt 2004 ,
1759hence ports 1-1999 and 2005-65535.
1760.El
1761.Pp
1762The operating system of the source host can be specified in the case of TCP
1763rules with the
1764.Ar OS
1765modifier.
1766See the
1767.Sx OPERATING SYSTEM FINGERPRINTING
1768section for more information.
1769.Pp
1770The host, port and OS specifications are optional, as in the following examples:
1771.Bd -literal -offset indent
1772pass in all
1773pass in from any to any
1774pass in proto tcp from any port \*(Le 1024 to any
1775pass in proto tcp from any to any port 25
1776pass in proto tcp from 10.0.0.0/8 port \*(Gt 1024 \e
1777      to ! 10.1.2.3 port != ssh
1778pass in proto tcp from any os "OpenBSD"
1779.Ed
1780.It Ar all
1781This is equivalent to "from any to any".
1782.It Ar group Aq Ar group
1783Similar to
1784.Ar user ,
1785this rule only applies to packets of sockets owned by the specified group.
1786.It Ar user Aq Ar user
1787This rule only applies to packets of sockets owned by the specified user.
1788For outgoing connections initiated from the firewall, this is the user
1789that opened the connection.
1790For incoming connections to the firewall itself, this is the user that
1791listens on the destination port.
1792For forwarded connections, where the firewall is not a connection endpoint,
1793the user and group are
1794.Em unknown .
1795.Pp
1796All packets, both outgoing and incoming, of one connection are associated
1797with the same user and group.
1798Only TCP and UDP packets can be associated with users; for other protocols
1799these parameters are ignored.
1800.Pp
1801User and group refer to the effective (as opposed to the real) IDs, in
1802case the socket is created by a setuid/setgid process.
1803User and group IDs are stored when a socket is created;
1804when a process creates a listening socket as root (for instance, by
1805binding to a privileged port) and subsequently changes to another
1806user ID (to drop privileges), the credentials will remain root.
1807.Pp
1808User and group IDs can be specified as either numbers or names.
1809The syntax is similar to the one for ports.
1810The value
1811.Em unknown
1812matches packets of forwarded connections.
1813.Em unknown
1814can only be used with the operators
1815.Cm =
1816and
1817.Cm != .
1818Other constructs like
1819.Cm user \*(Ge unknown
1820are invalid.
1821Forwarded packets with unknown user and group ID match only rules
1822that explicitly compare against
1823.Em unknown
1824with the operators
1825.Cm =
1826or
1827.Cm != .
1828For instance
1829.Cm user \*(Ge 0
1830does not match forwarded packets.
1831The following example allows only selected users to open outgoing
1832connections:
1833.Bd -literal -offset indent
1834block out proto { tcp, udp } all
1835pass  out proto { tcp, udp } all user { \*(Lt 1000, dhartmei }
1836.Ed
1837.It Xo Ar flags Aq Ar a
1838.Pf / Ns Aq Ar b
1839.No \*(Ba / Ns Aq Ar b
1840.No \*(Ba any
1841.Xc
1842This rule only applies to TCP packets that have the flags
1843.Aq Ar a
1844set out of set
1845.Aq Ar b .
1846Flags not specified in
1847.Aq Ar b
1848are ignored.
1849For stateful connections, the default is
1850.Ar flags S/SA .
1851To indicate that flags should not be checked at all, specify
1852.Ar flags any .
1853The flags are: (F)IN, (S)YN, (R)ST, (P)USH, (A)CK, (U)RG, (E)CE, and C(W)R.
1854.Bl -tag -width Fl
1855.It Ar flags S/S
1856Flag SYN is set.
1857The other flags are ignored.
1858.It Ar flags S/SA
1859This is the default setting for stateful connections.
1860Out of SYN and ACK, exactly SYN may be set.
1861SYN, SYN+PSH and SYN+RST match, but SYN+ACK, ACK and ACK+RST do not.
1862This is more restrictive than the previous example.
1863.It Ar flags /SFRA
1864If the first set is not specified, it defaults to none.
1865All of SYN, FIN, RST and ACK must be unset.
1866.El
1867.Pp
1868Because
1869.Ar flags S/SA
1870is applied by default (unless
1871.Ar no state
1872is specified), only the initial SYN packet of a TCP handshake will create
1873a state for a TCP connection.
1874It is possible to be less restrictive, and allow state creation from
1875intermediate
1876.Pq non-SYN
1877packets, by specifying
1878.Ar flags any .
1879This will cause
1880.Xr pf 4
1881to synchronize to existing connections, for instance
1882if one flushes the state table.
1883However, states created from such intermediate packets may be missing
1884connection details such as the TCP window scaling factor.
1885States which modify the packet flow, such as those affected by
1886.Ar nat , binat No or Ar rdr
1887rules,
1888.Ar modulate No or Ar synproxy state
1889options, or scrubbed with
1890.Ar reassemble tcp
1891will also not be recoverable from intermediate packets.
1892Such connections will stall and time out.
1893.It Xo Ar icmp-type Aq Ar type
1894.Ar code Aq Ar code
1895.Xc
1896.It Xo Ar icmp6-type Aq Ar type
1897.Ar code Aq Ar code
1898.Xc
1899This rule only applies to ICMP or ICMPv6 packets with the specified type
1900and code.
1901Text names for ICMP types and codes are listed in
1902.Xr icmp 4
1903and
1904.Xr icmp6 4 .
1905This parameter is only valid for rules that cover protocols ICMP or
1906ICMP6.
1907The protocol and the ICMP type indicator
1908.Po
1909.Ar icmp-type
1910or
1911.Ar icmp6-type
1912.Pc
1913must match.
1914.It Xo Ar tos Aq Ar string
1915.No \*(Ba Aq Ar number
1916.Xc
1917This rule applies to packets with the specified
1918.Em TOS
1919bits set.
1920.Em TOS
1921may be
1922given as one of
1923.Ar critical ,
1924.Ar inetcontrol ,
1925.Ar lowdelay ,
1926.Ar netcontrol ,
1927.Ar throughput ,
1928.Ar reliability ,
1929or one of the DiffServ Code Points:
1930.Ar ef ,
1931.Ar va ,
1932.Ar af11 No ... Ar af43 ,
1933.Ar cs0 No ... Ar cs7 ;
1934or as either hex or decimal.
1935.Pp
1936For example, the following rules are identical:
1937.Bd -literal -offset indent
1938pass all tos lowdelay
1939pass all tos 0x10
1940pass all tos 16
1941.Ed
1942.It Ar allow-opts
1943By default, IPv4 packets with IP options or IPv6 packets with routing
1944extension headers are blocked.
1945When
1946.Ar allow-opts
1947is specified for a
1948.Ar pass
1949rule, packets that pass the filter based on that rule (last matching)
1950do so even if they contain IP options or routing extension headers.
1951For packets that match state, the rule that initially created the
1952state is used.
1953The implicit
1954.Ar pass
1955rule that is used when a packet does not match any rules does not
1956allow IP options.
1957.It Ar label Aq Ar string
1958Adds a label (name) to the rule, which can be used to identify the rule.
1959For instance,
1960pfctl -s labels
1961shows per-rule statistics for rules that have labels.
1962.Pp
1963The following macros can be used in labels:
1964.Pp
1965.Bl -tag -width $srcaddr -compact -offset indent
1966.It Ar $if
1967The interface.
1968.It Ar $srcaddr
1969The source IP address.
1970.It Ar $dstaddr
1971The destination IP address.
1972.It Ar $srcport
1973The source port specification.
1974.It Ar $dstport
1975The destination port specification.
1976.It Ar $proto
1977The protocol name.
1978.It Ar $nr
1979The rule number.
1980.El
1981.Pp
1982For example:
1983.Bd -literal -offset indent
1984ips = \&"{ 1.2.3.4, 1.2.3.5 }\&"
1985pass in proto tcp from any to $ips \e
1986      port \*(Gt 1023 label \&"$dstaddr:$dstport\&"
1987.Ed
1988.Pp
1989expands to
1990.Bd -literal -offset indent
1991pass in inet proto tcp from any to 1.2.3.4 \e
1992      port \*(Gt 1023 label \&"1.2.3.4:\*(Gt1023\&"
1993pass in inet proto tcp from any to 1.2.3.5 \e
1994      port \*(Gt 1023 label \&"1.2.3.5:\*(Gt1023\&"
1995.Ed
1996.Pp
1997The macro expansion for the
1998.Ar label
1999directive occurs only at configuration file parse time, not during runtime.
2000.It Ar ridentifier Aq Ar number
2001Add an identifier (number) to the rule, which can be used to correlate the rule
2002to pflog entries, even after ruleset updates.
2003.It Xo Ar queue Aq Ar queue
2004.No \*(Ba ( Aq Ar queue ,
2005.Aq Ar queue )
2006.Xc
2007Packets matching this rule will be assigned to the specified queue.
2008If two queues are given, packets which have a
2009.Em TOS
2010of
2011.Em lowdelay
2012and TCP ACKs with no data payload will be assigned to the second one.
2013See
2014.Sx QUEUEING
2015for setup details.
2016.Pp
2017For example:
2018.Bd -literal -offset indent
2019pass in proto tcp to port 25 queue mail
2020pass in proto tcp to port 22 queue(ssh_bulk, ssh_prio)
2021.Ed
2022.It Cm set prio Ar priority | Pq Ar priority , priority
2023Packets matching this rule will be assigned a specific queueing priority.
2024Priorities are assigned as integers 0 through 7.
2025If the packet is transmitted on a
2026.Xr vlan 4
2027interface, the queueing priority will be written as the priority
2028code point in the 802.1Q VLAN header.
2029If two priorities are given, packets which have a TOS of
2030.Cm lowdelay
2031and TCP ACKs with no data payload will be assigned to the second one.
2032.Pp
2033For example:
2034.Bd -literal -offset indent
2035pass in proto tcp to port 25 set prio 2
2036pass in proto tcp to port 22 set prio (2, 5)
2037.Ed
2038.It Ar tag Aq Ar string
2039Packets matching this rule will be tagged with the
2040specified string.
2041The tag acts as an internal marker that can be used to
2042identify these packets later on.
2043This can be used, for example, to provide trust between
2044interfaces and to determine if packets have been
2045processed by translation rules.
2046Tags are
2047.Qq sticky ,
2048meaning that the packet will be tagged even if the rule
2049is not the last matching rule.
2050Further matching rules can replace the tag with a
2051new one but will not remove a previously applied tag.
2052A packet is only ever assigned one tag at a time.
2053Packet tagging can be done during
2054.Ar nat ,
2055.Ar rdr ,
2056.Ar binat
2057or
2058.Ar ether
2059rules in addition to filter rules.
2060Tags take the same macros as labels (see above).
2061.It Ar tagged Aq Ar string
2062Used with filter, translation or scrub rules
2063to specify that packets must already
2064be tagged with the given tag in order to match the rule.
2065Inverse tag matching can also be done
2066by specifying the
2067.Cm !\&
2068operator before the
2069.Ar tagged
2070keyword.
2071.It Ar rtable Aq Ar number
2072Used to select an alternate routing table for the routing lookup.
2073Only effective before the route lookup happened, i.e. when filtering inbound.
2074.It Xo Ar divert-to Aq Ar host
2075.Ar port Aq Ar port
2076.Xc
2077Used to redirect packets to a local socket bound to
2078.Ar host
2079and
2080.Ar port .
2081The packets will not be modified, so
2082.Xr getsockname 2
2083on the socket will return the original destination address of the packet.
2084.It Ar divert-reply
2085Used to receive replies for sockets that are bound to addresses
2086which are not local to the machine.
2087See
2088.Xr setsockopt 2
2089for information on how to bind these sockets.
2090.It Ar probability Aq Ar number
2091A probability attribute can be attached to a rule, with a value set between
20920 and 1, bounds not included.
2093In that case, the rule will be honoured using the given probability value
2094only.
2095For example, the following rule will drop 20% of incoming ICMP packets:
2096.Bd -literal -offset indent
2097block in proto icmp probability 20%
2098.Ed
2099.It Ar prio Aq Ar number
2100Only match packets which have the given queueing priority assigned.
2101.El
2102.Sh ROUTING
2103If a packet matches a rule with a route option set, the packet filter will
2104route the packet according to the type of route option.
2105When such a rule creates state, the route option is also applied to all
2106packets matching the same connection.
2107.Bl -tag -width xxxx
2108.It Ar route-to
2109The
2110.Ar route-to
2111option routes the packet to the specified interface with an optional address
2112for the next hop.
2113When a
2114.Ar route-to
2115rule creates state, only packets that pass in the same direction as the
2116filter rule specifies will be routed in this way.
2117Packets passing in the opposite direction (replies) are not affected
2118and are routed normally.
2119.It Ar reply-to
2120The
2121.Ar reply-to
2122option is similar to
2123.Ar route-to ,
2124but routes packets that pass in the opposite direction (replies) to the
2125specified interface.
2126Opposite direction is only defined in the context of a state entry, and
2127.Ar reply-to
2128is useful only in rules that create state.
2129It can be used on systems with multiple external connections to
2130route all outgoing packets of a connection through the interface
2131the incoming connection arrived through (symmetric routing enforcement).
2132.It Ar dup-to
2133The
2134.Ar dup-to
2135option creates a duplicate of the packet and routes it like
2136.Ar route-to .
2137The original packet gets routed as it normally would.
2138.El
2139.Sh POOL OPTIONS
2140For
2141.Ar nat
2142and
2143.Ar rdr
2144rules, (as well as for the
2145.Ar route-to ,
2146.Ar reply-to
2147and
2148.Ar dup-to
2149rule options) for which there is a single redirection address which has a
2150subnet mask smaller than 32 for IPv4 or 128 for IPv6 (more than one IP
2151address), a variety of different methods for assigning this address can be
2152used:
2153.Bl -tag -width xxxx
2154.It Ar bitmask
2155The
2156.Ar bitmask
2157option applies the network portion of the redirection address to the address
2158to be modified (source with
2159.Ar nat ,
2160destination with
2161.Ar rdr ) .
2162.It Ar random
2163The
2164.Ar random
2165option selects an address at random within the defined block of addresses.
2166.It Ar source-hash
2167The
2168.Ar source-hash
2169option uses a hash of the source address to determine the redirection address,
2170ensuring that the redirection address is always the same for a given source.
2171An optional key can be specified after this keyword either in hex or as a
2172string; by default
2173.Xr pfctl 8
2174randomly generates a key for source-hash every time the
2175ruleset is reloaded.
2176.It Ar round-robin
2177The
2178.Ar round-robin
2179option loops through the redirection address(es).
2180.Pp
2181When more than one redirection address is specified,
2182.Ar round-robin
2183is the only permitted pool type.
2184.It Ar static-port
2185With
2186.Ar nat
2187rules, the
2188.Ar static-port
2189option prevents
2190.Xr pf 4
2191from modifying the source port on TCP and UDP packets.
2192.It Xo Ar map-e-portset Aq Ar psid-offset
2193.No / Aq Ar psid-len
2194.No / Aq Ar psid
2195.Xc
2196With
2197.Ar nat
2198rules, the
2199.Ar map-e-portset
2200option enables the source port translation of MAP-E (RFC 7597) Customer Edge.
2201In order to make the host act as a MAP-E Customer Edge, setting up a tunneling
2202interface and pass rules for encapsulated packets are required in addition
2203to the map-e-portset nat rule.
2204.Pp
2205For example:
2206.Bd -literal -offset indent
2207nat on $gif_mape_if from $int_if:network to any \e
2208      -> $ipv4_mape_src map-e-portset 6/8/0x34
2209.Ed
2210.Pp
2211sets PSID offset 6, PSID length 8, PSID 0x34.
2212.El
2213.Pp
2214Additionally, the
2215.Ar sticky-address
2216option can be specified to help ensure that multiple connections from the
2217same source are mapped to the same redirection address.
2218This option can be used with the
2219.Ar random
2220and
2221.Ar round-robin
2222pool options.
2223Note that by default these associations are destroyed as soon as there are
2224no longer states which refer to them; in order to make the mappings last
2225beyond the lifetime of the states, increase the global options with
2226.Ar set timeout src.track .
2227See
2228.Sx STATEFUL TRACKING OPTIONS
2229for more ways to control the source tracking.
2230.Sh STATE MODULATION
2231Much of the security derived from TCP is attributable to how well the
2232initial sequence numbers (ISNs) are chosen.
2233Some popular stack implementations choose
2234.Em very
2235poor ISNs and thus are normally susceptible to ISN prediction exploits.
2236By applying a
2237.Ar modulate state
2238rule to a TCP connection,
2239.Xr pf 4
2240will create a high quality random sequence number for each connection
2241endpoint.
2242.Pp
2243The
2244.Ar modulate state
2245directive implicitly keeps state on the rule and is
2246only applicable to TCP connections.
2247.Pp
2248For instance:
2249.Bd -literal -offset indent
2250block all
2251pass out proto tcp from any to any modulate state
2252pass in  proto tcp from any to any port 25 flags S/SFRA modulate state
2253.Ed
2254.Pp
2255Note that modulated connections will not recover when the state table
2256is lost (firewall reboot, flushing the state table, etc...).
2257.Xr pf 4
2258will not be able to infer a connection again after the state table flushes
2259the connection's modulator.
2260When the state is lost, the connection may be left dangling until the
2261respective endpoints time out the connection.
2262It is possible on a fast local network for the endpoints to start an ACK
2263storm while trying to resynchronize after the loss of the modulator.
2264The default
2265.Ar flags
2266settings (or a more strict equivalent) should be used on
2267.Ar modulate state
2268rules to prevent ACK storms.
2269.Pp
2270Note that alternative methods are available
2271to prevent loss of the state table
2272and allow for firewall failover.
2273See
2274.Xr carp 4
2275and
2276.Xr pfsync 4
2277for further information.
2278.Sh SYN PROXY
2279By default,
2280.Xr pf 4
2281passes packets that are part of a
2282.Xr tcp 4
2283handshake between the endpoints.
2284The
2285.Ar synproxy state
2286option can be used to cause
2287.Xr pf 4
2288itself to complete the handshake with the active endpoint, perform a handshake
2289with the passive endpoint, and then forward packets between the endpoints.
2290.Pp
2291No packets are sent to the passive endpoint before the active endpoint has
2292completed the handshake, hence so-called SYN floods with spoofed source
2293addresses will not reach the passive endpoint, as the sender can't complete the
2294handshake.
2295.Pp
2296The proxy is transparent to both endpoints, they each see a single
2297connection from/to the other endpoint.
2298.Xr pf 4
2299chooses random initial sequence numbers for both handshakes.
2300Once the handshakes are completed, the sequence number modulators
2301(see previous section) are used to translate further packets of the
2302connection.
2303.Ar synproxy state
2304includes
2305.Ar modulate state .
2306.Pp
2307Rules with
2308.Ar synproxy
2309will not work if
2310.Xr pf 4
2311operates on a
2312.Xr bridge 4 .
2313.Pp
2314Example:
2315.Bd -literal -offset indent
2316pass in proto tcp from any to any port www synproxy state
2317.Ed
2318.Sh STATEFUL TRACKING OPTIONS
2319A number of options related to stateful tracking can be applied on a
2320per-rule basis.
2321.Ar keep state ,
2322.Ar modulate state
2323and
2324.Ar synproxy state
2325support these options, and
2326.Ar keep state
2327must be specified explicitly to apply options to a rule.
2328.Pp
2329.Bl -tag -width xxxx -compact
2330.It Ar max Aq Ar number
2331Limits the number of concurrent states the rule may create.
2332When this limit is reached, further packets that would create
2333state will not match this rule until existing states time out.
2334.It Ar no-sync
2335Prevent state changes for states created by this rule from appearing on the
2336.Xr pfsync 4
2337interface.
2338.It Xo Aq Ar timeout
2339.Aq Ar seconds
2340.Xc
2341Changes the timeout values used for states created by this rule.
2342For a list of all valid timeout names, see
2343.Sx OPTIONS
2344above.
2345.It Ar sloppy
2346Uses a sloppy TCP connection tracker that does not check sequence
2347numbers at all, which makes insertion and ICMP teardown attacks way
2348easier.
2349This is intended to be used in situations where one does not see all
2350packets of a connection, e.g. in asymmetric routing situations.
2351Cannot be used with modulate or synproxy state.
2352.El
2353.Pp
2354Multiple options can be specified, separated by commas:
2355.Bd -literal -offset indent
2356pass in proto tcp from any to any \e
2357      port www keep state \e
2358      (max 100, source-track rule, max-src-nodes 75, \e
2359      max-src-states 3, tcp.established 60, tcp.closing 5)
2360.Ed
2361.Pp
2362When the
2363.Ar source-track
2364keyword is specified, the number of states per source IP is tracked.
2365.Pp
2366.Bl -tag -width xxxx -compact
2367.It Ar source-track rule
2368The maximum number of states created by this rule is limited by the rule's
2369.Ar max-src-nodes
2370and
2371.Ar max-src-states
2372options.
2373Only state entries created by this particular rule count toward the rule's
2374limits.
2375.It Ar source-track global
2376The number of states created by all rules that use this option is limited.
2377Each rule can specify different
2378.Ar max-src-nodes
2379and
2380.Ar max-src-states
2381options, however state entries created by any participating rule count towards
2382each individual rule's limits.
2383.El
2384.Pp
2385The following limits can be set:
2386.Pp
2387.Bl -tag -width xxxx -compact
2388.It Ar max-src-nodes Aq Ar number
2389Limits the maximum number of source addresses which can simultaneously
2390have state table entries.
2391.It Ar max-src-states Aq Ar number
2392Limits the maximum number of simultaneous state entries that a single
2393source address can create with this rule.
2394.El
2395.Pp
2396For stateful TCP connections, limits on established connections (connections
2397which have completed the TCP 3-way handshake) can also be enforced
2398per source IP.
2399.Pp
2400.Bl -tag -width xxxx -compact
2401.It Ar max-src-conn Aq Ar number
2402Limits the maximum number of simultaneous TCP connections which have
2403completed the 3-way handshake that a single host can make.
2404.It Xo Ar max-src-conn-rate Aq Ar number
2405.No / Aq Ar seconds
2406.Xc
2407Limit the rate of new connections over a time interval.
2408The connection rate is an approximation calculated as a moving average.
2409.El
2410.Pp
2411Because the 3-way handshake ensures that the source address is not being
2412spoofed, more aggressive action can be taken based on these limits.
2413With the
2414.Ar overload Aq Ar table
2415state option, source IP addresses which hit either of the limits on
2416established connections will be added to the named table.
2417This table can be used in the ruleset to block further activity from
2418the offending host, redirect it to a tarpit process, or restrict its
2419bandwidth.
2420.Pp
2421The optional
2422.Ar flush
2423keyword kills all states created by the matching rule which originate
2424from the host which exceeds these limits.
2425The
2426.Ar global
2427modifier to the flush command kills all states originating from the
2428offending host, regardless of which rule created the state.
2429.Pp
2430For example, the following rules will protect the webserver against
2431hosts making more than 100 connections in 10 seconds.
2432Any host which connects faster than this rate will have its address added
2433to the
2434.Aq bad_hosts
2435table and have all states originating from it flushed.
2436Any new packets arriving from this host will be dropped unconditionally
2437by the block rule.
2438.Bd -literal -offset indent
2439block quick from \*(Ltbad_hosts\*(Gt
2440pass in on $ext_if proto tcp to $webserver port www keep state \e
2441	(max-src-conn-rate 100/10, overload \*(Ltbad_hosts\*(Gt flush global)
2442.Ed
2443.Sh OPERATING SYSTEM FINGERPRINTING
2444Passive OS Fingerprinting is a mechanism to inspect nuances of a TCP
2445connection's initial SYN packet and guess at the host's operating system.
2446Unfortunately these nuances are easily spoofed by an attacker so the
2447fingerprint is not useful in making security decisions.
2448But the fingerprint is typically accurate enough to make policy decisions
2449upon.
2450.Pp
2451The fingerprints may be specified by operating system class, by
2452version, or by subtype/patchlevel.
2453The class of an operating system is typically the vendor or genre
2454and would be
2455.Ox
2456for the
2457.Xr pf 4
2458firewall itself.
2459The version of the oldest available
2460.Ox
2461release on the main FTP site
2462would be 2.6 and the fingerprint would be written
2463.Pp
2464.Dl \&"OpenBSD 2.6\&"
2465.Pp
2466The subtype of an operating system is typically used to describe the
2467patchlevel if that patch led to changes in the TCP stack behavior.
2468In the case of
2469.Ox ,
2470the only subtype is for a fingerprint that was
2471normalized by the
2472.Ar no-df
2473scrub option and would be specified as
2474.Pp
2475.Dl \&"OpenBSD 3.3 no-df\&"
2476.Pp
2477Fingerprints for most popular operating systems are provided by
2478.Xr pf.os 5 .
2479Once
2480.Xr pf 4
2481is running, a complete list of known operating system fingerprints may
2482be listed by running:
2483.Pp
2484.Dl # pfctl -so
2485.Pp
2486Filter rules can enforce policy at any level of operating system specification
2487assuming a fingerprint is present.
2488Policy could limit traffic to approved operating systems or even ban traffic
2489from hosts that aren't at the latest service pack.
2490.Pp
2491The
2492.Ar unknown
2493class can also be used as the fingerprint which will match packets for
2494which no operating system fingerprint is known.
2495.Pp
2496Examples:
2497.Bd -literal -offset indent
2498pass  out proto tcp from any os OpenBSD
2499block out proto tcp from any os Doors
2500block out proto tcp from any os "Doors PT"
2501block out proto tcp from any os "Doors PT SP3"
2502block out from any os "unknown"
2503pass on lo0 proto tcp from any os "OpenBSD 3.3 lo0"
2504.Ed
2505.Pp
2506Operating system fingerprinting is limited only to the TCP SYN packet.
2507This means that it will not work on other protocols and will not match
2508a currently established connection.
2509.Pp
2510Caveat: operating system fingerprints are occasionally wrong.
2511There are three problems: an attacker can trivially craft his packets to
2512appear as any operating system he chooses;
2513an operating system patch could change the stack behavior and no fingerprints
2514will match it until the database is updated;
2515and multiple operating systems may have the same fingerprint.
2516.Sh BLOCKING SPOOFED TRAFFIC
2517"Spoofing" is the faking of IP addresses, typically for malicious
2518purposes.
2519The
2520.Ar antispoof
2521directive expands to a set of filter rules which will block all
2522traffic with a source IP from the network(s) directly connected
2523to the specified interface(s) from entering the system through
2524any other interface.
2525.Pp
2526For example, the line
2527.Bd -literal -offset indent
2528antispoof for lo0
2529.Ed
2530.Pp
2531expands to
2532.Bd -literal -offset indent
2533block drop in on ! lo0 inet from 127.0.0.1/8 to any
2534block drop in on ! lo0 inet6 from ::1 to any
2535.Ed
2536.Pp
2537For non-loopback interfaces, there are additional rules to block incoming
2538packets with a source IP address identical to the interface's IP(s).
2539For example, assuming the interface wi0 had an IP address of 10.0.0.1 and a
2540netmask of 255.255.255.0,
2541the line
2542.Bd -literal -offset indent
2543antispoof for wi0 inet
2544.Ed
2545.Pp
2546expands to
2547.Bd -literal -offset indent
2548block drop in on ! wi0 inet from 10.0.0.0/24 to any
2549block drop in inet from 10.0.0.1 to any
2550.Ed
2551.Pp
2552Caveat: Rules created by the
2553.Ar antispoof
2554directive interfere with packets sent over loopback interfaces
2555to local addresses.
2556One should pass these explicitly.
2557.Sh FRAGMENT HANDLING
2558The size of IP datagrams (packets) can be significantly larger than the
2559maximum transmission unit (MTU) of the network.
2560In cases when it is necessary or more efficient to send such large packets,
2561the large packet will be fragmented into many smaller packets that will each
2562fit onto the wire.
2563Unfortunately for a firewalling device, only the first logical fragment will
2564contain the necessary header information for the subprotocol that allows
2565.Xr pf 4
2566to filter on things such as TCP ports or to perform NAT.
2567.Pp
2568Besides the use of
2569.Ar scrub
2570rules as described in
2571.Sx TRAFFIC NORMALIZATION
2572above, there are three options for handling fragments in the packet filter.
2573.Pp
2574One alternative is to filter individual fragments with filter rules.
2575If no
2576.Ar scrub
2577rule applies to a fragment, it is passed to the filter.
2578Filter rules with matching IP header parameters decide whether the
2579fragment is passed or blocked, in the same way as complete packets
2580are filtered.
2581Without reassembly, fragments can only be filtered based on IP header
2582fields (source/destination address, protocol), since subprotocol header
2583fields are not available (TCP/UDP port numbers, ICMP code/type).
2584The
2585.Ar fragment
2586option can be used to restrict filter rules to apply only to
2587fragments, but not complete packets.
2588Filter rules without the
2589.Ar fragment
2590option still apply to fragments, if they only specify IP header fields.
2591For instance, the rule
2592.Bd -literal -offset indent
2593pass in proto tcp from any to any port 80
2594.Ed
2595.Pp
2596never applies to a fragment, even if the fragment is part of a TCP
2597packet with destination port 80, because without reassembly this information
2598is not available for each fragment.
2599This also means that fragments cannot create new or match existing
2600state table entries, which makes stateful filtering and address
2601translation (NAT, redirection) for fragments impossible.
2602.Pp
2603It's also possible to reassemble only certain fragments by specifying
2604source or destination addresses or protocols as parameters in
2605.Ar scrub
2606rules.
2607.Pp
2608In most cases, the benefits of reassembly outweigh the additional
2609memory cost, and it's recommended to use
2610.Ar scrub
2611rules to reassemble
2612all fragments via the
2613.Ar fragment reassemble
2614modifier.
2615.Pp
2616The memory allocated for fragment caching can be limited using
2617.Xr pfctl 8 .
2618Once this limit is reached, fragments that would have to be cached
2619are dropped until other entries time out.
2620The timeout value can also be adjusted.
2621.Pp
2622When forwarding reassembled IPv6 packets, pf refragments them with
2623the original maximum fragment size.
2624This allows the sender to determine the optimal fragment size by
2625path MTU discovery.
2626.Sh ANCHORS
2627Besides the main ruleset,
2628.Xr pfctl 8
2629can load rulesets into
2630.Ar anchor
2631attachment points.
2632An
2633.Ar anchor
2634is a container that can hold rules, address tables, and other anchors.
2635.Pp
2636An
2637.Ar anchor
2638has a name which specifies the path where
2639.Xr pfctl 8
2640can be used to access the anchor to perform operations on it, such as
2641attaching child anchors to it or loading rules into it.
2642Anchors may be nested, with components separated by
2643.Sq /
2644characters, similar to how file system hierarchies are laid out.
2645The main ruleset is actually the default anchor, so filter and
2646translation rules, for example, may also be contained in any anchor.
2647.Pp
2648An anchor can reference another
2649.Ar anchor
2650attachment point
2651using the following kinds
2652of rules:
2653.Bl -tag -width xxxx
2654.It Ar nat-anchor Aq Ar name
2655Evaluates the
2656.Ar nat
2657rules in the specified
2658.Ar anchor .
2659.It Ar rdr-anchor Aq Ar name
2660Evaluates the
2661.Ar rdr
2662rules in the specified
2663.Ar anchor .
2664.It Ar binat-anchor Aq Ar name
2665Evaluates the
2666.Ar binat
2667rules in the specified
2668.Ar anchor .
2669.It Ar anchor Aq Ar name
2670Evaluates the filter rules in the specified
2671.Ar anchor .
2672.It Xo Ar load anchor
2673.Aq Ar name
2674.Ar from Aq Ar file
2675.Xc
2676Loads the rules from the specified file into the
2677anchor
2678.Ar name .
2679.El
2680.Pp
2681When evaluation of the main ruleset reaches an
2682.Ar anchor
2683rule,
2684.Xr pf 4
2685will proceed to evaluate all rules specified in that anchor.
2686.Pp
2687Matching filter and translation rules marked with the
2688.Ar quick
2689option are final and abort the evaluation of the rules in other
2690anchors and the main ruleset.
2691If the
2692.Ar anchor
2693itself is marked with the
2694.Ar quick
2695option,
2696ruleset evaluation will terminate when the anchor is exited if the packet is
2697matched by any rule within the anchor.
2698.Pp
2699.Ar anchor
2700rules are evaluated relative to the anchor in which they are contained.
2701For example, all
2702.Ar anchor
2703rules specified in the main ruleset will reference anchor
2704attachment points underneath the main ruleset, and
2705.Ar anchor
2706rules specified in a file loaded from a
2707.Ar load anchor
2708rule will be attached under that anchor point.
2709.Pp
2710Rules may be contained in
2711.Ar anchor
2712attachment points which do not contain any rules when the main ruleset
2713is loaded, and later such anchors can be manipulated through
2714.Xr pfctl 8
2715without reloading the main ruleset or other anchors.
2716For example,
2717.Bd -literal -offset indent
2718ext_if = \&"kue0\&"
2719block on $ext_if all
2720anchor spam
2721pass out on $ext_if all
2722pass in on $ext_if proto tcp from any \e
2723      to $ext_if port smtp
2724.Ed
2725.Pp
2726blocks all packets on the external interface by default, then evaluates
2727all rules in the
2728.Ar anchor
2729named "spam", and finally passes all outgoing connections and
2730incoming connections to port 25.
2731.Bd -literal -offset indent
2732# echo \&"block in quick from 1.2.3.4 to any\&" \&| \e
2733      pfctl -a spam -f -
2734.Ed
2735.Pp
2736This loads a single rule into the
2737.Ar anchor ,
2738which blocks all packets from a specific address.
2739.Pp
2740The anchor can also be populated by adding a
2741.Ar load anchor
2742rule after the
2743.Ar anchor
2744rule:
2745.Bd -literal -offset indent
2746anchor spam
2747load anchor spam from "/etc/pf-spam.conf"
2748.Ed
2749.Pp
2750When
2751.Xr pfctl 8
2752loads
2753.Nm pf.conf ,
2754it will also load all the rules from the file
2755.Pa /etc/pf-spam.conf
2756into the anchor.
2757.Pp
2758Optionally,
2759.Ar anchor
2760rules can specify packet filtering parameters using the same syntax as
2761filter rules.
2762When parameters are used, the
2763.Ar anchor
2764rule is only evaluated for matching packets.
2765This allows conditional evaluation of anchors, like:
2766.Bd -literal -offset indent
2767block on $ext_if all
2768anchor spam proto tcp from any to any port smtp
2769pass out on $ext_if all
2770pass in on $ext_if proto tcp from any to $ext_if port smtp
2771.Ed
2772.Pp
2773The rules inside
2774.Ar anchor
2775spam are only evaluated for
2776.Ar tcp
2777packets with destination port 25.
2778Hence,
2779.Bd -literal -offset indent
2780# echo \&"block in quick from 1.2.3.4 to any" \&| \e
2781      pfctl -a spam -f -
2782.Ed
2783.Pp
2784will only block connections from 1.2.3.4 to port 25.
2785.Pp
2786Anchors may end with the asterisk
2787.Pq Sq *
2788character, which signifies that all anchors attached at that point
2789should be evaluated in the alphabetical ordering of their anchor name.
2790For example,
2791.Bd -literal -offset indent
2792anchor "spam/*"
2793.Ed
2794.Pp
2795will evaluate each rule in each anchor attached to the
2796.Li spam
2797anchor.
2798Note that it will only evaluate anchors that are directly attached to the
2799.Li spam
2800anchor, and will not descend to evaluate anchors recursively.
2801.Pp
2802Since anchors are evaluated relative to the anchor in which they are
2803contained, there is a mechanism for accessing the parent and ancestor
2804anchors of a given anchor.
2805Similar to file system path name resolution, if the sequence
2806.Dq ..
2807appears as an anchor path component, the parent anchor of the current
2808anchor in the path evaluation at that point will become the new current
2809anchor.
2810As an example, consider the following:
2811.Bd -literal -offset indent
2812# echo ' anchor "spam/allowed" ' | pfctl -f -
2813# echo -e ' anchor "../banned" \en pass' | \e
2814      pfctl -a spam/allowed -f -
2815.Ed
2816.Pp
2817Evaluation of the main ruleset will lead into the
2818.Li spam/allowed
2819anchor, which will evaluate the rules in the
2820.Li spam/banned
2821anchor, if any, before finally evaluating the
2822.Ar pass
2823rule.
2824.Pp
2825Filter rule
2826.Ar anchors
2827can also be loaded inline in the ruleset within a brace ('{' '}') delimited
2828block.
2829Brace delimited blocks may contain rules or other brace-delimited blocks.
2830When anchors are loaded this way the anchor name becomes optional.
2831.Bd -literal -offset indent
2832anchor "external" on $ext_if {
2833	block
2834	anchor out {
2835		pass proto tcp from any to port { 25, 80, 443 }
2836	}
2837	pass in proto tcp to any port 22
2838}
2839.Ed
2840.Pp
2841Since the parser specification for anchor names is a string, any
2842reference to an anchor name containing
2843.Sq /
2844characters will require double quote
2845.Pq Sq \&"
2846characters around the anchor name.
2847.Sh TRANSLATION EXAMPLES
2848This example maps incoming requests on port 80 to port 8080, on
2849which a daemon is running (because, for example, it is not run as root,
2850and therefore lacks permission to bind to port 80).
2851.Bd -literal
2852# use a macro for the interface name, so it can be changed easily
2853ext_if = \&"ne3\&"
2854
2855# map daemon on 8080 to appear to be on 80
2856rdr on $ext_if proto tcp from any to any port 80 -\*(Gt 127.0.0.1 port 8080
2857.Ed
2858.Pp
2859If the
2860.Ar pass
2861modifier is given, packets matching the translation rule are passed without
2862inspecting the filter rules:
2863.Bd -literal
2864rdr pass on $ext_if proto tcp from any to any port 80 -\*(Gt 127.0.0.1 \e
2865      port 8080
2866.Ed
2867.Pp
2868In the example below, vlan12 is configured as 192.168.168.1;
2869the machine translates all packets coming from 192.168.168.0/24 to 204.92.77.111
2870when they are going out any interface except vlan12.
2871This has the net effect of making traffic from the 192.168.168.0/24
2872network appear as though it is the Internet routable address
2873204.92.77.111 to nodes behind any interface on the router except
2874for the nodes on vlan12.
2875(Thus, 192.168.168.1 can talk to the 192.168.168.0/24 nodes.)
2876.Bd -literal
2877nat on ! vlan12 from 192.168.168.0/24 to any -\*(Gt 204.92.77.111
2878.Ed
2879.Pp
2880In the example below, the machine sits between a fake internal 144.19.74.*
2881network, and a routable external IP of 204.92.77.100.
2882The
2883.Ar no nat
2884rule excludes protocol AH from being translated.
2885.Bd -literal
2886# NO NAT
2887no nat on $ext_if proto ah from 144.19.74.0/24 to any
2888nat on $ext_if from 144.19.74.0/24 to any -\*(Gt 204.92.77.100
2889.Ed
2890.Pp
2891In the example below, packets bound for one specific server, as well as those
2892generated by the sysadmins are not proxied; all other connections are.
2893.Bd -literal
2894# NO RDR
2895no rdr on $int_if proto { tcp, udp } from any to $server port 80
2896no rdr on $int_if proto { tcp, udp } from $sysadmins to any port 80
2897rdr on $int_if proto { tcp, udp } from any to any port 80 -\*(Gt 127.0.0.1 \e
2898      port 80
2899.Ed
2900.Pp
2901This longer example uses both a NAT and a redirection.
2902The external interface has the address 157.161.48.183.
2903On localhost, we are running
2904.Xr ftp-proxy 8 ,
2905waiting for FTP sessions to be redirected to it.
2906The three mandatory anchors for
2907.Xr ftp-proxy 8
2908are omitted from this example; see the
2909.Xr ftp-proxy 8
2910manpage.
2911.Bd -literal
2912# NAT
2913# Translate outgoing packets' source addresses (any protocol).
2914# In this case, any address but the gateway's external address is mapped.
2915nat on $ext_if inet from ! ($ext_if) to any -\*(Gt ($ext_if)
2916
2917# NAT PROXYING
2918# Map outgoing packets' source port to an assigned proxy port instead of
2919# an arbitrary port.
2920# In this case, proxy outgoing isakmp with port 500 on the gateway.
2921nat on $ext_if inet proto udp from any port = isakmp to any -\*(Gt ($ext_if) \e
2922      port 500
2923
2924# BINAT
2925# Translate outgoing packets' source address (any protocol).
2926# Translate incoming packets' destination address to an internal machine
2927# (bidirectional).
2928binat on $ext_if from 10.1.2.150 to any -\*(Gt $ext_if
2929
2930# Translate packets arriving on $peer_if addressed to 172.22.16.0/20
2931# to the corresponding address in 172.21.16.0/20 (bidirectional).
2932binat on $peer_if from 172.21.16.0/20 to any -> 172.22.16.0/20
2933
2934# RDR
2935# Translate incoming packets' destination addresses.
2936# As an example, redirect a TCP and UDP port to an internal machine.
2937rdr on $ext_if inet proto tcp from any to ($ext_if) port 8080 \e
2938      -\*(Gt 10.1.2.151 port 22
2939rdr on $ext_if inet proto udp from any to ($ext_if) port 8080 \e
2940      -\*(Gt 10.1.2.151 port 53
2941
2942# RDR
2943# Translate outgoing ftp control connections to send them to localhost
2944# for proxying with ftp-proxy(8) running on port 8021.
2945rdr on $int_if proto tcp from any to any port 21 -\*(Gt 127.0.0.1 port 8021
2946.Ed
2947.Pp
2948In this example, a NAT gateway is set up to translate internal addresses
2949using a pool of public addresses (192.0.2.16/28) and to redirect
2950incoming web server connections to a group of web servers on the internal
2951network.
2952.Bd -literal
2953# NAT LOAD BALANCE
2954# Translate outgoing packets' source addresses using an address pool.
2955# A given source address is always translated to the same pool address by
2956# using the source-hash keyword.
2957nat on $ext_if inet from any to any -\*(Gt 192.0.2.16/28 source-hash
2958
2959# RDR ROUND ROBIN
2960# Translate incoming web server connections to a group of web servers on
2961# the internal network.
2962rdr on $ext_if proto tcp from any to any port 80 \e
2963      -\*(Gt { 10.1.2.155, 10.1.2.160, 10.1.2.161 } round-robin
2964.Ed
2965.Sh FILTER EXAMPLES
2966.Bd -literal
2967# The external interface is kue0
2968# (157.161.48.183, the only routable address)
2969# and the private network is 10.0.0.0/8, for which we are doing NAT.
2970
2971# use a macro for the interface name, so it can be changed easily
2972ext_if = \&"kue0\&"
2973
2974# normalize all incoming traffic
2975scrub in on $ext_if all fragment reassemble
2976
2977# block and log everything by default
2978block return log on $ext_if all
2979
2980# block anything coming from source we have no back routes for
2981block in from no-route to any
2982
2983# block packets whose ingress interface does not match the one in
2984# the route back to their source address
2985block in from urpf-failed to any
2986
2987# block and log outgoing packets that do not have our address as source,
2988# they are either spoofed or something is misconfigured (NAT disabled,
2989# for instance), we want to be nice and do not send out garbage.
2990block out log quick on $ext_if from ! 157.161.48.183 to any
2991
2992# silently drop broadcasts (cable modem noise)
2993block in quick on $ext_if from any to 255.255.255.255
2994
2995# block and log incoming packets from reserved address space and invalid
2996# addresses, they are either spoofed or misconfigured, we cannot reply to
2997# them anyway (hence, no return-rst).
2998block in log quick on $ext_if from { 10.0.0.0/8, 172.16.0.0/12, \e
2999      192.168.0.0/16, 255.255.255.255/32 } to any
3000
3001# ICMP
3002
3003# pass out/in certain ICMP queries and keep state (ping)
3004# state matching is done on host addresses and ICMP id (not type/code),
3005# so replies (like 0/0 for 8/0) will match queries
3006# ICMP error messages (which always refer to a TCP/UDP packet) are
3007# handled by the TCP/UDP states
3008pass on $ext_if inet proto icmp all icmp-type 8 code 0
3009
3010# UDP
3011
3012# pass out all UDP connections and keep state
3013pass out on $ext_if proto udp all
3014
3015# pass in certain UDP connections and keep state (DNS)
3016pass in on $ext_if proto udp from any to any port domain
3017
3018# TCP
3019
3020# pass out all TCP connections and modulate state
3021pass out on $ext_if proto tcp all modulate state
3022
3023# pass in certain TCP connections and keep state (SSH, SMTP, DNS, IDENT)
3024pass in on $ext_if proto tcp from any to any port { ssh, smtp, domain, \e
3025      auth }
3026
3027# Do not allow Windows 9x SMTP connections since they are typically
3028# a viral worm. Alternately we could limit these OSes to 1 connection each.
3029block in on $ext_if proto tcp from any os {"Windows 95", "Windows 98"} \e
3030      to any port smtp
3031
3032# IPv6
3033# pass in/out all IPv6 traffic: note that we have to enable this in two
3034# different ways, on both our physical interface and our tunnel
3035pass quick on gif0 inet6
3036pass quick on $ext_if proto ipv6
3037
3038# Packet Tagging
3039
3040# three interfaces: $int_if, $ext_if, and $wifi_if (wireless). NAT is
3041# being done on $ext_if for all outgoing packets. tag packets in on
3042# $int_if and pass those tagged packets out on $ext_if.  all other
3043# outgoing packets (i.e., packets from the wireless network) are only
3044# permitted to access port 80.
3045
3046pass in on $int_if from any to any tag INTNET
3047pass in on $wifi_if from any to any
3048
3049block out on $ext_if from any to any
3050pass out quick on $ext_if tagged INTNET
3051pass out on $ext_if proto tcp from any to any port 80
3052
3053# tag incoming packets as they are redirected to spamd(8). use the tag
3054# to pass those packets through the packet filter.
3055
3056rdr on $ext_if inet proto tcp from \*(Ltspammers\*(Gt to port smtp \e
3057	tag SPAMD -\*(Gt 127.0.0.1 port spamd
3058
3059block in on $ext_if
3060pass in on $ext_if inet proto tcp tagged SPAMD
3061.Ed
3062.Sh GRAMMAR
3063Syntax for
3064.Nm
3065in BNF:
3066.Bd -literal
3067line           = ( option | ether-rule | pf-rule | nat-rule | binat-rule |
3068                 rdr-rule | antispoof-rule | altq-rule | queue-rule |
3069                 trans-anchors | anchor-rule | anchor-close | load-anchor |
3070                 table-rule | include )
3071
3072option         = "set" ( [ "timeout" ( timeout | "{" timeout-list "}" ) ] |
3073                 [ "ruleset-optimization" [ "none" | "basic" | "profile" ]] |
3074                 [ "optimization" [ "default" | "normal" |
3075                 "high-latency" | "satellite" |
3076                 "aggressive" | "conservative" ] ]
3077                 [ "limit" ( limit-item | "{" limit-list "}" ) ] |
3078                 [ "loginterface" ( interface-name | "none" ) ] |
3079                 [ "block-policy" ( "drop" | "return" ) ] |
3080                 [ "state-policy" ( "if-bound" | "floating" ) ]
3081                 [ "state-defaults" state-opts ]
3082                 [ "require-order" ( "yes" | "no" ) ]
3083                 [ "fingerprints" filename ] |
3084                 [ "skip on" ifspec ] |
3085                 [ "debug" ( "none" | "urgent" | "misc" | "loud" ) ]
3086                 [ "keepcounters" ] )
3087
3088ether-rule     = "ether" etheraction [ ( "in" | "out" ) ]
3089                 [ "quick" ] [ "on" ifspec ] [ "bridge-to" interface-name ]
3090                 [ etherprotospec ] etherhosts [ "l3" hosts ]
3091                 [ etherfilteropt-list ]
3092
3093pf-rule        = action [ ( "in" | "out" ) ]
3094                 [ "log" [ "(" logopts ")"] ] [ "quick" ]
3095                 [ "on" ifspec ] [ route ] [ af ] [ protospec ]
3096                 hosts [ filteropt-list ]
3097
3098logopts        = logopt [ "," logopts ]
3099logopt         = "all" | "user" | "to" interface-name
3100
3101etherfilteropt-list = etherfilteropt-list etherfilteropt | etherfilteropt
3102etherfilteropt = "tag" string | "tagged" string | "queue" ( string )
3103
3104filteropt-list = filteropt-list filteropt | filteropt
3105filteropt      = user | group | flags | icmp-type | icmp6-type | "tos" tos |
3106                 ( "no" | "keep" | "modulate" | "synproxy" ) "state"
3107                 [ "(" state-opts ")" ] |
3108                 "fragment" | "no-df" | "min-ttl" number | "set-tos" tos |
3109                 "max-mss" number | "random-id" | "reassemble tcp" |
3110                 fragmentation | "allow-opts" |
3111                 "label" string | "tag" string | [ ! ] "tagged" string |
3112                 "set prio" ( number | "(" number [ [ "," ] number ] ")" ) |
3113                 "queue" ( string | "(" string [ [ "," ] string ] ")" ) |
3114                 "rtable" number | "probability" number"%" | "prio" number |
3115                 "dnpipe" ( number | "(" number "," number ")" ) |
3116                 "dnqueue" ( number | "(" number "," number ")" ) |
3117                 "ridentifier" number
3118
3119nat-rule       = [ "no" ] "nat" [ "pass" [ "log" [ "(" logopts ")" ] ] ]
3120                 [ "on" ifspec ] [ af ]
3121                 [ protospec ] hosts [ "tag" string ] [ "tagged" string ]
3122                 [ "-\*(Gt" ( redirhost | "{" redirhost-list "}" )
3123                 [ portspec ] [ pooltype ] [ "static-port" ]
3124                 [ "map-e-portset" number "/" number "/" number ] ]
3125
3126binat-rule     = [ "no" ] "binat" [ "pass" [ "log" [ "(" logopts ")" ] ] ]
3127                 [ "on" interface-name ] [ af ]
3128                 [ "proto" ( proto-name | proto-number ) ]
3129                 "from" address [ "/" mask-bits ] "to" ipspec
3130                 [ "tag" string ] [ "tagged" string ]
3131                 [ "-\*(Gt" address [ "/" mask-bits ] ]
3132
3133rdr-rule       = [ "no" ] "rdr" [ "pass" [ "log" [ "(" logopts ")" ] ] ]
3134                 [ "on" ifspec ] [ af ]
3135                 [ protospec ] hosts [ "tag" string ] [ "tagged" string ]
3136                 [ "-\*(Gt" ( redirhost | "{" redirhost-list "}" )
3137                 [ portspec ] [ pooltype ] ]
3138
3139antispoof-rule = "antispoof" [ "log" ] [ "quick" ]
3140                 "for" ifspec [ af ] [ "label" string ]
3141                 [ "ridentifier" number ]
3142
3143table-rule     = "table" "\*(Lt" string "\*(Gt" [ tableopts-list ]
3144tableopts-list = tableopts-list tableopts | tableopts
3145tableopts      = "persist" | "const" | "counters" | "file" string |
3146                 "{" [ tableaddr-list ] "}"
3147tableaddr-list = tableaddr-list [ "," ] tableaddr-spec | tableaddr-spec
3148tableaddr-spec = [ "!" ] tableaddr [ "/" mask-bits ]
3149tableaddr      = hostname | ifspec | "self" |
3150                 ipv4-dotted-quad | ipv6-coloned-hex
3151
3152altq-rule      = "altq on" interface-name queueopts-list
3153                 "queue" subqueue
3154queue-rule     = "queue" string [ "on" interface-name ] queueopts-list
3155                 subqueue
3156
3157anchor-rule    = "anchor" [ string ] [ ( "in" | "out" ) ] [ "on" ifspec ]
3158                 [ af ] [ protospec ] [ hosts ] [ filteropt-list ] [ "{" ]
3159
3160anchor-close   = "}"
3161
3162trans-anchors  = ( "nat-anchor" | "rdr-anchor" | "binat-anchor" ) string
3163                 [ "on" ifspec ] [ af ] [ "proto" ] [ protospec ] [ hosts ]
3164
3165load-anchor    = "load anchor" string "from" filename
3166
3167queueopts-list = queueopts-list queueopts | queueopts
3168queueopts      = [ "bandwidth" bandwidth-spec ] |
3169                 [ "qlimit" number ] | [ "tbrsize" number ] |
3170                 [ "priority" number ] | [ schedulers ]
3171schedulers     = ( cbq-def | priq-def | hfsc-def )
3172bandwidth-spec = "number" ( "b" | "Kb" | "Mb" | "Gb" | "%" )
3173
3174etheraction    = "pass" | "block"
3175action         = "pass" | "block" [ return ] | [ "no" ] "scrub"
3176return         = "drop" | "return" | "return-rst" [ "( ttl" number ")" ] |
3177                 "return-icmp" [ "(" icmpcode [ [ "," ] icmp6code ] ")" ] |
3178                 "return-icmp6" [ "(" icmp6code ")" ]
3179icmpcode       = ( icmp-code-name | icmp-code-number )
3180icmp6code      = ( icmp6-code-name | icmp6-code-number )
3181
3182ifspec         = ( [ "!" ] ( interface-name | interface-group ) ) |
3183                 "{" interface-list "}"
3184interface-list = [ "!" ] ( interface-name | interface-group )
3185                 [ [ "," ] interface-list ]
3186route          = ( "route-to" | "reply-to" | "dup-to" )
3187                 ( routehost | "{" routehost-list "}" )
3188                 [ pooltype ]
3189af             = "inet" | "inet6"
3190
3191etherprotospec = "proto" ( proto-number | "{" proto-list "}" )
3192protospec      = "proto" ( proto-name | proto-number |
3193                 "{" proto-list "}" )
3194proto-list     = ( proto-name | proto-number ) [ [ "," ] proto-list ]
3195
3196etherhosts     = "from" macaddress "to" macaddress
3197macaddress     = mac | mac "/" masklen | mac "&" mask
3198
3199hosts          = "all" |
3200                 "from" ( "any" | "no-route" | "urpf-failed" | "self" | host |
3201                 "{" host-list "}" ) [ port ] [ os ]
3202                 "to"   ( "any" | "no-route" | "self" | host |
3203                 "{" host-list "}" ) [ port ]
3204
3205ipspec         = "any" | host | "{" host-list "}"
3206host           = [ "!" ] ( address [ "/" mask-bits ] | "\*(Lt" string "\*(Gt" )
3207redirhost      = address [ "/" mask-bits ]
3208routehost      = "(" interface-name [ address [ "/" mask-bits ] ] ")"
3209address        = ( interface-name | interface-group |
3210                 "(" ( interface-name | interface-group ) ")" |
3211                 hostname | ipv4-dotted-quad | ipv6-coloned-hex )
3212host-list      = host [ [ "," ] host-list ]
3213redirhost-list = redirhost [ [ "," ] redirhost-list ]
3214routehost-list = routehost [ [ "," ] routehost-list ]
3215
3216port           = "port" ( unary-op | binary-op | "{" op-list "}" )
3217portspec       = "port" ( number | name ) [ ":" ( "*" | number | name ) ]
3218os             = "os"  ( os-name | "{" os-list "}" )
3219user           = "user" ( unary-op | binary-op | "{" op-list "}" )
3220group          = "group" ( unary-op | binary-op | "{" op-list "}" )
3221
3222unary-op       = [ "=" | "!=" | "\*(Lt" | "\*(Le" | "\*(Gt" | "\*(Ge" ]
3223                 ( name | number )
3224binary-op      = number ( "\*(Lt\*(Gt" | "\*(Gt\*(Lt" | ":" ) number
3225op-list        = ( unary-op | binary-op ) [ [ "," ] op-list ]
3226
3227os-name        = operating-system-name
3228os-list        = os-name [ [ "," ] os-list ]
3229
3230flags          = "flags" ( [ flag-set ] "/"  flag-set | "any" )
3231flag-set       = [ "F" ] [ "S" ] [ "R" ] [ "P" ] [ "A" ] [ "U" ] [ "E" ]
3232                 [ "W" ]
3233
3234icmp-type      = "icmp-type" ( icmp-type-code | "{" icmp-list "}" )
3235icmp6-type     = "icmp6-type" ( icmp-type-code | "{" icmp-list "}" )
3236icmp-type-code = ( icmp-type-name | icmp-type-number )
3237                 [ "code" ( icmp-code-name | icmp-code-number ) ]
3238icmp-list      = icmp-type-code [ [ "," ] icmp-list ]
3239
3240tos            = ( "lowdelay" | "throughput" | "reliability" |
3241                 [ "0x" ] number )
3242
3243state-opts     = state-opt [ [ "," ] state-opts ]
3244state-opt      = ( "max" number | "no-sync" | timeout | "sloppy" |
3245                 "source-track" [ ( "rule" | "global" ) ] |
3246                 "max-src-nodes" number | "max-src-states" number |
3247                 "max-src-conn" number |
3248                 "max-src-conn-rate" number "/" number |
3249                 "overload" "\*(Lt" string "\*(Gt" [ "flush" ] |
3250                 "if-bound" | "floating" )
3251
3252fragmentation  = [ "fragment reassemble" ]
3253
3254timeout-list   = timeout [ [ "," ] timeout-list ]
3255timeout        = ( "tcp.first" | "tcp.opening" | "tcp.established" |
3256                 "tcp.closing" | "tcp.finwait" | "tcp.closed" |
3257                 "udp.first" | "udp.single" | "udp.multiple" |
3258                 "icmp.first" | "icmp.error" |
3259                 "other.first" | "other.single" | "other.multiple" |
3260                 "frag" | "interval" | "src.track" |
3261                 "adaptive.start" | "adaptive.end" ) number
3262
3263limit-list     = limit-item [ [ "," ] limit-list ]
3264limit-item     = ( "states" | "frags" | "src-nodes" ) number
3265
3266pooltype       = ( "bitmask" | "random" |
3267                 "source-hash" [ ( hex-key | string-key ) ] |
3268                 "round-robin" ) [ sticky-address ]
3269
3270subqueue       = string | "{" queue-list "}"
3271queue-list     = string [ [ "," ] string ]
3272cbq-def        = "cbq" [ "(" cbq-opt [ [ "," ] cbq-opt ] ")" ]
3273priq-def       = "priq" [ "(" priq-opt [ [ "," ] priq-opt ] ")" ]
3274hfsc-def       = "hfsc" [ "(" hfsc-opt [ [ "," ] hfsc-opt ] ")" ]
3275cbq-opt        = ( "default" | "borrow" | "red" | "ecn" | "rio" )
3276priq-opt       = ( "default" | "red" | "ecn" | "rio" )
3277hfsc-opt       = ( "default" | "red" | "ecn" | "rio" |
3278                 linkshare-sc | realtime-sc | upperlimit-sc )
3279linkshare-sc   = "linkshare" sc-spec
3280realtime-sc    = "realtime" sc-spec
3281upperlimit-sc  = "upperlimit" sc-spec
3282sc-spec        = ( bandwidth-spec |
3283                 "(" bandwidth-spec number bandwidth-spec ")" )
3284include        = "include" filename
3285.Ed
3286.Sh FILES
3287.Bl -tag -width "/etc/protocols" -compact
3288.It Pa /etc/hosts
3289Host name database.
3290.It Pa /etc/pf.conf
3291Default location of the ruleset file.
3292The file has to be created manually as it is not installed with a
3293standard installation.
3294.It Pa /etc/pf.os
3295Default location of OS fingerprints.
3296.It Pa /etc/protocols
3297Protocol name database.
3298.It Pa /etc/services
3299Service name database.
3300.El
3301.Sh SEE ALSO
3302.Xr altq 4 ,
3303.Xr carp 4 ,
3304.Xr icmp 4 ,
3305.Xr icmp6 4 ,
3306.Xr ip 4 ,
3307.Xr ip6 4 ,
3308.Xr pf 4 ,
3309.Xr pfsync 4 ,
3310.Xr tcp 4 ,
3311.Xr udp 4 ,
3312.Xr hosts 5 ,
3313.Xr pf.os 5 ,
3314.Xr protocols 5 ,
3315.Xr services 5 ,
3316.Xr ftp-proxy 8 ,
3317.Xr pfctl 8 ,
3318.Xr pflogd 8
3319.Sh HISTORY
3320The
3321.Nm
3322file format first appeared in
3323.Ox 3.0 .
3324