xref: /freebsd/share/man/man5/pf.conf.5 (revision dd41de95a84d979615a2ef11df6850622bf6184e)
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 April 19, 2021
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 seven 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 Traffic Normalization Li (e.g. Em scrub )
58Traffic normalization protects internal machines against inconsistencies
59in Internet protocols and implementations.
60.It Cm Queueing
61Queueing provides rule-based bandwidth control.
62.It Cm Translation Li (Various forms of NAT)
63Translation rules specify how addresses are to be mapped or redirected to
64other addresses.
65.It Cm Packet Filtering
66Packet filtering provides rule-based blocking or passing of packets.
67.El
68.Pp
69With the exception of
70.Cm macros
71and
72.Cm tables ,
73the types of statements should be grouped and appear in
74.Nm pf.conf
75in the order shown above, as this matches the operation of the underlying
76packet filtering engine.
77By default
78.Xr pfctl 8
79enforces this order (see
80.Ar set require-order
81below).
82.Pp
83Comments can be put anywhere in the file using a hash mark
84.Pq Sq # ,
85and extend to the end of the current line.
86.Pp
87Additional configuration files can be included with the
88.Ic include
89keyword, for example:
90.Bd -literal -offset indent
91include "/etc/pf/sub.filter.conf"
92.Ed
93.Sh MACROS
94Macros can be defined that will later be expanded in context.
95Macro names must start with a letter, and may contain letters, digits
96and underscores.
97Macro names may not be reserved words (for example
98.Ar pass ,
99.Ar in ,
100.Ar out ) .
101Macros are not expanded inside quotes.
102.Pp
103For example,
104.Bd -literal -offset indent
105ext_if = \&"kue0\&"
106all_ifs = \&"{\&" $ext_if lo0 \&"}\&"
107pass out on $ext_if from any to any
108pass in  on $ext_if proto tcp from any to any port 25
109.Ed
110.Sh TABLES
111Tables are named structures which can hold a collection of addresses and
112networks.
113Lookups against tables in
114.Xr pf 4
115are relatively fast, making a single rule with tables much more efficient,
116in terms of
117processor usage and memory consumption, than a large number of rules which
118differ only in IP address (either created explicitly or automatically by rule
119expansion).
120.Pp
121Tables can be used as the source or destination of filter rules,
122.Ar scrub
123rules
124or
125translation rules such as
126.Ar nat
127or
128.Ar rdr
129(see below for details on the various rule types).
130Tables can also be used for the redirect address of
131.Ar nat
132and
133.Ar rdr
134rules and in the routing options of filter rules, but only for
135.Ar round-robin
136pools.
137.Pp
138Tables can be defined with any of the following
139.Xr pfctl 8
140mechanisms.
141As with macros, reserved words may not be used as table names.
142.Bl -tag -width "manually"
143.It Ar manually
144Persistent tables can be manually created with the
145.Ar add
146or
147.Ar replace
148option of
149.Xr pfctl 8 ,
150before or after the ruleset has been loaded.
151.It Pa pf.conf
152Table definitions can be placed directly in this file, and loaded at the
153same time as other rules are loaded, atomically.
154Table definitions inside
155.Nm pf.conf
156use the
157.Ar table
158statement, and are especially useful to define non-persistent tables.
159The contents of a pre-existing table defined without a list of addresses
160to initialize it is not altered when
161.Nm pf.conf
162is loaded.
163A table initialized with the empty list,
164.Li { } ,
165will be cleared on load.
166.El
167.Pp
168Tables may be defined with the following attributes:
169.Bl -tag -width persist
170.It Ar persist
171The
172.Ar persist
173flag forces the kernel to keep the table even when no rules refer to it.
174If the flag is not set, the kernel will automatically remove the table
175when the last rule referring to it is flushed.
176.It Ar const
177The
178.Ar const
179flag prevents the user from altering the contents of the table once it
180has been created.
181Without that flag,
182.Xr pfctl 8
183can be used to add or remove addresses from the table at any time, even
184when running with
185.Xr securelevel 7
186= 2.
187.It Ar counters
188The
189.Ar counters
190flag enables per-address packet and byte counters which can be displayed with
191.Xr pfctl 8 .
192Note that this feature carries significant memory overhead for large tables.
193.El
194.Pp
195For example,
196.Bd -literal -offset indent
197table \*(Ltprivate\*(Gt const { 10/8, 172.16/12, 192.168/16 }
198table \*(Ltbadhosts\*(Gt persist
199block on fxp0 from { \*(Ltprivate\*(Gt, \*(Ltbadhosts\*(Gt } to any
200.Ed
201.Pp
202creates a table called private, to hold RFC 1918 private network
203blocks, and a table called badhosts, which is initially empty.
204A filter rule is set up to block all traffic coming from addresses listed in
205either table.
206The private table cannot have its contents changed and the badhosts table
207will exist even when no active filter rules reference it.
208Addresses may later be added to the badhosts table, so that traffic from
209these hosts can be blocked by using
210.Bd -literal -offset indent
211# pfctl -t badhosts -Tadd 204.92.77.111
212.Ed
213.Pp
214A table can also be initialized with an address list specified in one or more
215external files, using the following syntax:
216.Bd -literal -offset indent
217table \*(Ltspam\*(Gt persist file \&"/etc/spammers\&" file \&"/etc/openrelays\&"
218block on fxp0 from \*(Ltspam\*(Gt to any
219.Ed
220.Pp
221The files
222.Pa /etc/spammers
223and
224.Pa /etc/openrelays
225list IP addresses, one per line.
226Any lines beginning with a # are treated as comments and ignored.
227In addition to being specified by IP address, hosts may also be
228specified by their hostname.
229When the resolver is called to add a hostname to a table,
230.Em all
231resulting IPv4 and IPv6 addresses are placed into the table.
232IP addresses can also be entered in a table by specifying a valid interface
233name, a valid interface group or the
234.Em self
235keyword, in which case all addresses assigned to the interface(s) will be
236added to the table.
237.Sh OPTIONS
238.Xr pf 4
239may be tuned for various situations using the
240.Ar set
241command.
242.Bl -tag -width xxxx
243.It Ar set timeout
244.Pp
245.Bl -tag -width "src.track" -compact
246.It Ar interval
247Interval between purging expired states and fragments.
248.It Ar frag
249Seconds before an unassembled fragment is expired.
250.It Ar src.track
251Length of time to retain a source tracking entry after the last state
252expires.
253.El
254.Pp
255When a packet matches a stateful connection, the seconds to live for the
256connection will be updated to that of the
257.Ar proto.modifier
258which corresponds to the connection state.
259Each packet which matches this state will reset the TTL.
260Tuning these values may improve the performance of the
261firewall at the risk of dropping valid idle connections.
262.Pp
263.Bl -tag -width xxxx -compact
264.It Ar tcp.first
265The state after the first packet.
266.It Ar tcp.opening
267The state before the destination host ever sends a packet.
268.It Ar tcp.established
269The fully established state.
270.It Ar tcp.closing
271The state after the first FIN has been sent.
272.It Ar tcp.finwait
273The state after both FINs have been exchanged and the connection is closed.
274Some hosts (notably web servers on Solaris) send TCP packets even after closing
275the connection.
276Increasing
277.Ar tcp.finwait
278(and possibly
279.Ar tcp.closing )
280can prevent blocking of such packets.
281.It Ar tcp.closed
282The state after one endpoint sends an RST.
283.El
284.Pp
285ICMP and UDP are handled in a fashion similar to TCP, but with a much more
286limited set of states:
287.Pp
288.Bl -tag -width xxxx -compact
289.It Ar udp.first
290The state after the first packet.
291.It Ar udp.single
292The state if the source host sends more than one packet but the destination
293host has never sent one back.
294.It Ar udp.multiple
295The state if both hosts have sent packets.
296.It Ar icmp.first
297The state after the first packet.
298.It Ar icmp.error
299The state after an ICMP error came back in response to an ICMP packet.
300.El
301.Pp
302Other protocols are handled similarly to UDP:
303.Pp
304.Bl -tag -width xxxx -compact
305.It Ar other.first
306.It Ar other.single
307.It Ar other.multiple
308.El
309.Pp
310Timeout values can be reduced adaptively as the number of state table
311entries grows.
312.Pp
313.Bl -tag -width xxxx -compact
314.It Ar adaptive.start
315When the number of state entries exceeds this value, adaptive scaling
316begins.
317All timeout values are scaled linearly with factor
318(adaptive.end - number of states) / (adaptive.end - adaptive.start).
319.It Ar adaptive.end
320When reaching this number of state entries, all timeout values become
321zero, effectively purging all state entries immediately.
322This value is used to define the scale factor, it should not actually
323be reached (set a lower state limit, see below).
324.El
325.Pp
326Adaptive timeouts are enabled by default, with an adaptive.start value
327equal to 60% of the state limit, and an adaptive.end value equal to
328120% of the state limit.
329They can be disabled by setting both adaptive.start and adaptive.end to 0.
330.Pp
331The adaptive timeout values can be defined both globally and for each rule.
332When used on a per-rule basis, the values relate to the number of
333states created by the rule, otherwise to the total number of
334states.
335.Pp
336For example:
337.Bd -literal -offset indent
338set timeout tcp.first 120
339set timeout tcp.established 86400
340set timeout { adaptive.start 6000, adaptive.end 12000 }
341set limit states 10000
342.Ed
343.Pp
344With 9000 state table entries, the timeout values are scaled to 50%
345(tcp.first 60, tcp.established 43200).
346.It Ar set loginterface
347Enable collection of packet and byte count statistics for the given
348interface or interface group.
349These statistics can be viewed using
350.Bd -literal -offset indent
351# pfctl -s info
352.Ed
353.Pp
354In this example
355.Xr pf 4
356collects statistics on the interface named dc0:
357.Bd -literal -offset indent
358set loginterface dc0
359.Ed
360.Pp
361One can disable the loginterface using:
362.Bd -literal -offset indent
363set loginterface none
364.Ed
365.It Ar set limit
366Sets hard limits on the memory pools used by the packet filter.
367See
368.Xr zone 9
369for an explanation of memory pools.
370.Pp
371For example,
372.Bd -literal -offset indent
373set limit states 20000
374.Ed
375.Pp
376sets the maximum number of entries in the memory pool used by state table
377entries (generated by
378.Ar pass
379rules which do not specify
380.Ar no state )
381to 20000.
382Using
383.Bd -literal -offset indent
384set limit frags 20000
385.Ed
386.Pp
387sets the maximum number of entries in the memory pool used for fragment
388reassembly (generated by
389.Ar scrub
390rules) to 20000.
391Using
392.Bd -literal -offset indent
393set limit src-nodes 2000
394.Ed
395.Pp
396sets the maximum number of entries in the memory pool used for tracking
397source IP addresses (generated by the
398.Ar sticky-address
399and
400.Ar src.track
401options) to 2000.
402Using
403.Bd -literal -offset indent
404set limit tables 1000
405set limit table-entries 100000
406.Ed
407.Pp
408sets limits on the memory pools used by tables.
409The first limits the number of tables that can exist to 1000.
410The second limits the overall number of addresses that can be stored
411in tables to 100000.
412.Pp
413Various limits can be combined on a single line:
414.Bd -literal -offset indent
415set limit { states 20000, frags 20000, src-nodes 2000 }
416.Ed
417.It Ar set ruleset-optimization
418.Bl -tag -width xxxxxxxx -compact
419.It Ar none
420Disable the ruleset optimizer.
421.It Ar basic
422Enable basic ruleset optimization.
423This is the default behaviour.
424Basic ruleset optimization does four things to improve the
425performance of ruleset evaluations:
426.Pp
427.Bl -enum -compact
428.It
429remove duplicate rules
430.It
431remove rules that are a subset of another rule
432.It
433combine multiple rules into a table when advantageous
434.It
435re-order the rules to improve evaluation performance
436.El
437.Pp
438.It Ar profile
439Uses the currently loaded ruleset as a feedback profile to tailor the
440ordering of quick rules to actual network traffic.
441.El
442.Pp
443It is important to note that the ruleset optimizer will modify the ruleset
444to improve performance.
445A side effect of the ruleset modification is that per-rule accounting
446statistics will have different meanings than before.
447If per-rule accounting is important for billing purposes or whatnot,
448either the ruleset optimizer should not be used or a label field should
449be added to all of the accounting rules to act as optimization barriers.
450.Pp
451Optimization can also be set as a command-line argument to
452.Xr pfctl 8 ,
453overriding the settings in
454.Nm .
455.It Ar set optimization
456Optimize state timeouts for one of the following network environments:
457.Pp
458.Bl -tag -width xxxx -compact
459.It Ar normal
460A normal network environment.
461Suitable for almost all networks.
462.It Ar high-latency
463A high-latency environment (such as a satellite connection).
464.It Ar satellite
465Alias for
466.Ar high-latency .
467.It Ar aggressive
468Aggressively expire connections.
469This can greatly reduce the memory usage of the firewall at the cost of
470dropping idle connections early.
471.It Ar conservative
472Extremely conservative settings.
473Avoid dropping legitimate connections at the
474expense of greater memory utilization (possibly much greater on a busy
475network) and slightly increased processor utilization.
476.El
477.Pp
478For example:
479.Bd -literal -offset indent
480set optimization aggressive
481.Ed
482.It Ar set block-policy
483The
484.Ar block-policy
485option sets the default behaviour for the packet
486.Ar block
487action:
488.Pp
489.Bl -tag -width xxxxxxxx -compact
490.It Ar drop
491Packet is silently dropped.
492.It Ar return
493A TCP RST is returned for blocked TCP packets,
494an ICMP UNREACHABLE is returned for blocked UDP packets,
495and all other packets are silently dropped.
496.El
497.Pp
498For example:
499.Bd -literal -offset indent
500set block-policy return
501.Ed
502.It Ar set fail-policy
503The
504.Ar fail-policy
505option sets the behaviour of rules which should pass a packet but were
506unable to do so.
507This might happen when a nat or route-to rule uses an empty table as list
508of targets or if a rule fails to create state or source node.
509The following
510.Ar block
511actions are possible:
512.Pp
513.Bl -tag -width xxxxxxxx -compact
514.It Ar drop
515Incoming packet is silently dropped.
516.It Ar return
517Incoming packet is dropped and TCP RST is returned for TCP packets,
518an ICMP UNREACHABLE is returned for UDP packets,
519and no response is sent for other packets.
520.El
521.Pp
522For example:
523.Bd -literal -offset indent
524set fail-policy return
525.Ed
526.It Ar set state-policy
527The
528.Ar state-policy
529option sets the default behaviour for states:
530.Pp
531.Bl -tag -width group-bound -compact
532.It Ar if-bound
533States are bound to interface.
534.It Ar floating
535States can match packets on any interfaces (the default).
536.El
537.Pp
538For example:
539.Bd -literal -offset indent
540set state-policy if-bound
541.Ed
542.It Ar set state-defaults
543The
544.Ar state-defaults
545option sets the state options for states created from rules
546without an explicit
547.Ar keep state .
548For example:
549.Bd -literal -offset indent
550set state-defaults no-sync
551.Ed
552.It Ar set hostid
553The 32-bit
554.Ar hostid
555identifies this firewall's state table entries to other firewalls
556in a
557.Xr pfsync 4
558failover cluster.
559By default the hostid is set to a pseudo-random value, however it may be
560desirable to manually configure it, for example to more easily identify the
561source of state table entries.
562.Bd -literal -offset indent
563set hostid 1
564.Ed
565.Pp
566The hostid may be specified in either decimal or hexadecimal.
567.It Ar set require-order
568By default
569.Xr pfctl 8
570enforces an ordering of the statement types in the ruleset to:
571.Em options ,
572.Em normalization ,
573.Em queueing ,
574.Em translation ,
575.Em filtering .
576Setting this option to
577.Ar no
578disables this enforcement.
579There may be non-trivial and non-obvious implications to an out of
580order ruleset.
581Consider carefully before disabling the order enforcement.
582.It Ar set fingerprints
583Load fingerprints of known operating systems from the given filename.
584By default fingerprints of known operating systems are automatically
585loaded from
586.Xr pf.os 5
587in
588.Pa /etc
589but can be overridden via this option.
590Setting this option may leave a small period of time where the fingerprints
591referenced by the currently active ruleset are inconsistent until the new
592ruleset finishes loading.
593.Pp
594For example:
595.Pp
596.Dl set fingerprints \&"/etc/pf.os.devel\&"
597.It Ar set skip on Aq Ar ifspec
598List interfaces for which packets should not be filtered.
599Packets passing in or out on such interfaces are passed as if pf was
600disabled, i.e. pf does not process them in any way.
601This can be useful on loopback and other virtual interfaces, when
602packet filtering is not desired and can have unexpected effects.
603For example:
604.Pp
605.Dl set skip on lo0
606.It Ar set debug
607Set the debug
608.Ar level
609to one of the following:
610.Pp
611.Bl -tag -width xxxxxxxxxxxx -compact
612.It Ar none
613Don't generate debug messages.
614.It Ar urgent
615Generate debug messages only for serious errors.
616.It Ar misc
617Generate debug messages for various errors.
618.It Ar loud
619Generate debug messages for common conditions.
620.El
621.It Ar set keepcounters
622Preserve rule counters across rule updates.
623Usually rule counters are reset to zero on every update of the ruleset.
624With
625.Ar keepcounters
626set pf will attempt to find matching rules between old and new rulesets
627and preserve the rule counters.
628.El
629.Sh TRAFFIC NORMALIZATION
630Traffic normalization is used to sanitize packet content in such
631a way that there are no ambiguities in packet interpretation on
632the receiving side.
633The normalizer does IP fragment reassembly to prevent attacks
634that confuse intrusion detection systems by sending overlapping
635IP fragments.
636Packet normalization is invoked with the
637.Ar scrub
638directive.
639.Pp
640.Ar scrub
641has the following options:
642.Bl -tag -width xxxx
643.It Ar no-df
644Clears the
645.Ar dont-fragment
646bit from a matching IP packet.
647Some operating systems are known to generate fragmented packets with the
648.Ar dont-fragment
649bit set.
650This is particularly true with NFS.
651.Ar Scrub
652will drop such fragmented
653.Ar dont-fragment
654packets unless
655.Ar no-df
656is specified.
657.Pp
658Unfortunately some operating systems also generate their
659.Ar dont-fragment
660packets with a zero IP identification field.
661Clearing the
662.Ar dont-fragment
663bit on packets with a zero IP ID may cause deleterious results if an
664upstream router later fragments the packet.
665Using the
666.Ar random-id
667modifier (see below) is recommended in combination with the
668.Ar no-df
669modifier to ensure unique IP identifiers.
670.It Ar min-ttl Aq Ar number
671Enforces a minimum TTL for matching IP packets.
672.It Ar max-mss Aq Ar number
673Enforces a maximum MSS for matching TCP packets.
674.It Xo Ar set-tos Aq Ar string
675.No \*(Ba Aq Ar number
676.Xc
677Enforces a
678.Em TOS
679for matching IP packets.
680.Em TOS
681may be
682given as one of
683.Ar critical ,
684.Ar inetcontrol ,
685.Ar lowdelay ,
686.Ar netcontrol ,
687.Ar throughput ,
688.Ar reliability ,
689or one of the DiffServ Code Points:
690.Ar ef ,
691.Ar va ,
692.Ar af11 No ... Ar af43 ,
693.Ar cs0 No ... Ar cs7 ;
694or as either hex or decimal.
695.It Ar random-id
696Replaces the IP identification field with random values to compensate
697for predictable values generated by many hosts.
698This option only applies to packets that are not fragmented
699after the optional fragment reassembly.
700.It Ar fragment reassemble
701Using
702.Ar scrub
703rules, fragments can be reassembled by normalization.
704In this case, fragments are buffered until they form a complete
705packet, and only the completed packet is passed on to the filter.
706The advantage is that filter rules have to deal only with complete
707packets, and can ignore fragments.
708The drawback of caching fragments is the additional memory cost.
709.It Ar reassemble tcp
710Statefully normalizes TCP connections.
711.Ar scrub reassemble tcp
712rules may not have the direction (in/out) specified.
713.Ar reassemble tcp
714performs the following normalizations:
715.Pp
716.Bl -tag -width timeout -compact
717.It ttl
718Neither side of the connection is allowed to reduce their IP TTL.
719An attacker may send a packet such that it reaches the firewall, affects
720the firewall state, and expires before reaching the destination host.
721.Ar reassemble tcp
722will raise the TTL of all packets back up to the highest value seen on
723the connection.
724.It timestamp modulation
725Modern TCP stacks will send a timestamp on every TCP packet and echo
726the other endpoint's timestamp back to them.
727Many operating systems will merely start the timestamp at zero when
728first booted, and increment it several times a second.
729The uptime of the host can be deduced by reading the timestamp and multiplying
730by a constant.
731Also observing several different timestamps can be used to count hosts
732behind a NAT device.
733And spoofing TCP packets into a connection requires knowing or guessing
734valid timestamps.
735Timestamps merely need to be monotonically increasing and not derived off a
736guessable base time.
737.Ar reassemble tcp
738will cause
739.Ar scrub
740to modulate the TCP timestamps with a random number.
741.It extended PAWS checks
742There is a problem with TCP on long fat pipes, in that a packet might get
743delayed for longer than it takes the connection to wrap its 32-bit sequence
744space.
745In such an occurrence, the old packet would be indistinguishable from a
746new packet and would be accepted as such.
747The solution to this is called PAWS: Protection Against Wrapped Sequence
748numbers.
749It protects against it by making sure the timestamp on each packet does
750not go backwards.
751.Ar reassemble tcp
752also makes sure the timestamp on the packet does not go forward more
753than the RFC allows.
754By doing this,
755.Xr pf 4
756artificially extends the security of TCP sequence numbers by 10 to 18
757bits when the host uses appropriately randomized timestamps, since a
758blind attacker would have to guess the timestamp as well.
759.El
760.El
761.Pp
762For example,
763.Bd -literal -offset indent
764scrub in on $ext_if all fragment reassemble
765.Ed
766.Pp
767The
768.Ar no
769option prefixed to a scrub rule causes matching packets to remain unscrubbed,
770much in the same way as
771.Ar drop quick
772works in the packet filter (see below).
773This mechanism should be used when it is necessary to exclude specific packets
774from broader scrub rules.
775.Sh QUEUEING
776The ALTQ system is currently not available in the GENERIC kernel nor as
777loadable modules.
778In order to use the herein after called queueing options one has to use a
779custom built kernel.
780Please refer to
781.Xr altq 4
782to learn about the related kernel options.
783.Pp
784Packets can be assigned to queues for the purpose of bandwidth
785control.
786At least two declarations are required to configure queues, and later
787any packet filtering rule can reference the defined queues by name.
788During the filtering component of
789.Nm pf.conf ,
790the last referenced
791.Ar queue
792name is where any packets from
793.Ar pass
794rules will be queued, while for
795.Ar block
796rules it specifies where any resulting ICMP or TCP RST
797packets should be queued.
798The
799.Ar scheduler
800defines the algorithm used to decide which packets get delayed, dropped, or
801sent out immediately.
802There are three
803.Ar schedulers
804currently supported.
805.Bl -tag -width xxxx
806.It Ar cbq
807Class Based Queueing.
808.Ar Queues
809attached to an interface build a tree, thus each
810.Ar queue
811can have further child
812.Ar queues .
813Each queue can have a
814.Ar priority
815and a
816.Ar bandwidth
817assigned.
818.Ar Priority
819mainly controls the time packets take to get sent out, while
820.Ar bandwidth
821has primarily effects on throughput.
822.Ar cbq
823achieves both partitioning and sharing of link bandwidth
824by hierarchically structured classes.
825Each class has its own
826.Ar queue
827and is assigned its share of
828.Ar bandwidth .
829A child class can borrow bandwidth from its parent class
830as long as excess bandwidth is available
831(see the option
832.Ar borrow ,
833below).
834.It Ar priq
835Priority Queueing.
836.Ar Queues
837are flat attached to the interface, thus,
838.Ar queues
839cannot have further child
840.Ar queues .
841Each
842.Ar queue
843has a unique
844.Ar priority
845assigned, ranging from 0 to 15.
846Packets in the
847.Ar queue
848with the highest
849.Ar priority
850are processed first.
851.It Ar hfsc
852Hierarchical Fair Service Curve.
853.Ar Queues
854attached to an interface build a tree, thus each
855.Ar queue
856can have further child
857.Ar queues .
858Each queue can have a
859.Ar priority
860and a
861.Ar bandwidth
862assigned.
863.Ar Priority
864mainly controls the time packets take to get sent out, while
865.Ar bandwidth
866primarily affects throughput.
867.Ar hfsc
868supports both link-sharing and guaranteed real-time services.
869It employs a service curve based QoS model,
870and its unique feature is an ability to decouple
871.Ar delay
872and
873.Ar bandwidth
874allocation.
875.El
876.Pp
877The interfaces on which queueing should be activated are declared using
878the
879.Ar altq on
880declaration.
881.Ar altq on
882has the following keywords:
883.Bl -tag -width xxxx
884.It Aq Ar interface
885Queueing is enabled on the named interface.
886.It Aq Ar scheduler
887Specifies which queueing scheduler to use.
888Currently supported values
889are
890.Ar cbq
891for Class Based Queueing,
892.Ar priq
893for Priority Queueing and
894.Ar hfsc
895for the Hierarchical Fair Service Curve scheduler.
896.It Ar bandwidth Aq Ar bw
897The maximum bitrate for all queues on an
898interface may be specified using the
899.Ar bandwidth
900keyword.
901The value can be specified as an absolute value or as a
902percentage of the interface bandwidth.
903When using an absolute value, the suffixes
904.Ar b ,
905.Ar Kb ,
906.Ar Mb ,
907and
908.Ar Gb
909are used to represent bits, kilobits, megabits, and
910gigabits per second, respectively.
911The value must not exceed the interface bandwidth.
912If
913.Ar bandwidth
914is not specified, the interface bandwidth is used
915(but take note that some interfaces do not know their bandwidth,
916or can adapt their bandwidth rates).
917.It Ar qlimit Aq Ar limit
918The maximum number of packets held in the queue.
919The default is 50.
920.It Ar tbrsize Aq Ar size
921Adjusts the size, in bytes, of the token bucket regulator.
922If not specified, heuristics based on the
923interface bandwidth are used to determine the size.
924.It Ar queue Aq Ar list
925Defines a list of subqueues to create on an interface.
926.El
927.Pp
928In the following example, the interface dc0
929should queue up to 5Mbps in four second-level queues using
930Class Based Queueing.
931Those four queues will be shown in a later example.
932.Bd -literal -offset indent
933altq on dc0 cbq bandwidth 5Mb queue { std, http, mail, ssh }
934.Ed
935.Pp
936Once interfaces are activated for queueing using the
937.Ar altq
938directive, a sequence of
939.Ar queue
940directives may be defined.
941The name associated with a
942.Ar queue
943must match a queue defined in the
944.Ar altq
945directive (e.g. mail), or, except for the
946.Ar priq
947.Ar scheduler ,
948in a parent
949.Ar queue
950declaration.
951The following keywords can be used:
952.Bl -tag -width xxxx
953.It Ar on Aq Ar interface
954Specifies the interface the queue operates on.
955If not given, it operates on all matching interfaces.
956.It Ar bandwidth Aq Ar bw
957Specifies the maximum bitrate to be processed by the queue.
958This value must not exceed the value of the parent
959.Ar queue
960and can be specified as an absolute value or a percentage of the parent
961queue's bandwidth.
962If not specified, defaults to 100% of the parent queue's bandwidth.
963The
964.Ar priq
965scheduler does not support bandwidth specification.
966.It Ar priority Aq Ar level
967Between queues a priority level can be set.
968For
969.Ar cbq
970and
971.Ar hfsc ,
972the range is 0 to 7 and for
973.Ar priq ,
974the range is 0 to 15.
975The default for all is 1.
976.Ar Priq
977queues with a higher priority are always served first.
978.Ar Cbq
979and
980.Ar Hfsc
981queues with a higher priority are preferred in the case of overload.
982.It Ar qlimit Aq Ar limit
983The maximum number of packets held in the queue.
984The default is 50.
985.El
986.Pp
987The
988.Ar scheduler
989can get additional parameters with
990.Xo Aq Ar scheduler
991.Pf ( Aq Ar parameters ) .
992.Xc
993Parameters are as follows:
994.Bl -tag -width Fl
995.It Ar default
996Packets not matched by another queue are assigned to this one.
997Exactly one default queue is required.
998.It Ar red
999Enable RED (Random Early Detection) on this queue.
1000RED drops packets with a probability proportional to the average
1001queue length.
1002.It Ar rio
1003Enables RIO on this queue.
1004RIO is RED with IN/OUT, thus running
1005RED two times more than RIO would achieve the same effect.
1006RIO is currently not supported in the GENERIC kernel.
1007.It Ar ecn
1008Enables ECN (Explicit Congestion Notification) on this queue.
1009ECN implies RED.
1010.El
1011.Pp
1012The
1013.Ar cbq
1014.Ar scheduler
1015supports an additional option:
1016.Bl -tag -width Fl
1017.It Ar borrow
1018The queue can borrow bandwidth from the parent.
1019.El
1020.Pp
1021The
1022.Ar hfsc
1023.Ar scheduler
1024supports some additional options:
1025.Bl -tag -width Fl
1026.It Ar realtime Aq Ar sc
1027The minimum required bandwidth for the queue.
1028.It Ar upperlimit Aq Ar sc
1029The maximum allowed bandwidth for the queue.
1030.It Ar linkshare Aq Ar sc
1031The bandwidth share of a backlogged queue.
1032.El
1033.Pp
1034.Aq Ar sc
1035is an acronym for
1036.Ar service curve .
1037.Pp
1038The format for service curve specifications is
1039.Ar ( m1 , d , m2 ) .
1040.Ar m2
1041controls the bandwidth assigned to the queue.
1042.Ar m1
1043and
1044.Ar d
1045are optional and can be used to control the initial bandwidth assignment.
1046For the first
1047.Ar d
1048milliseconds the queue gets the bandwidth given as
1049.Ar m1 ,
1050afterwards the value given in
1051.Ar m2 .
1052.Pp
1053Furthermore, with
1054.Ar cbq
1055and
1056.Ar hfsc ,
1057child queues can be specified as in an
1058.Ar altq
1059declaration, thus building a tree of queues using a part of
1060their parent's bandwidth.
1061.Pp
1062Packets can be assigned to queues based on filter rules by using the
1063.Ar queue
1064keyword.
1065Normally only one
1066.Ar queue
1067is specified; when a second one is specified it will instead be used for
1068packets which have a
1069.Em TOS
1070of
1071.Em lowdelay
1072and for TCP ACKs with no data payload.
1073.Pp
1074To continue the previous example, the examples below would specify the
1075four referenced
1076queues, plus a few child queues.
1077Interactive
1078.Xr ssh 1
1079sessions get priority over bulk transfers like
1080.Xr scp 1
1081and
1082.Xr sftp 1 .
1083The queues may then be referenced by filtering rules (see
1084.Sx PACKET FILTERING
1085below).
1086.Bd -literal
1087queue std bandwidth 10% cbq(default)
1088queue http bandwidth 60% priority 2 cbq(borrow red) \e
1089      { employees, developers }
1090queue  developers bandwidth 75% cbq(borrow)
1091queue  employees bandwidth 15%
1092queue mail bandwidth 10% priority 0 cbq(borrow ecn)
1093queue ssh bandwidth 20% cbq(borrow) { ssh_interactive, ssh_bulk }
1094queue  ssh_interactive bandwidth 50% priority 7 cbq(borrow)
1095queue  ssh_bulk bandwidth 50% priority 0 cbq(borrow)
1096
1097block return out on dc0 inet all queue std
1098pass out on dc0 inet proto tcp from $developerhosts to any port 80 \e
1099      queue developers
1100pass out on dc0 inet proto tcp from $employeehosts to any port 80 \e
1101      queue employees
1102pass out on dc0 inet proto tcp from any to any port 22 \e
1103      queue(ssh_bulk, ssh_interactive)
1104pass out on dc0 inet proto tcp from any to any port 25 \e
1105      queue mail
1106.Ed
1107.Sh TRANSLATION
1108Translation rules modify either the source or destination address of the
1109packets associated with a stateful connection.
1110A stateful connection is automatically created to track packets matching
1111such a rule as long as they are not blocked by the filtering section of
1112.Nm pf.conf .
1113The translation engine modifies the specified address and/or port in the
1114packet, recalculates IP, TCP and UDP checksums as necessary, and passes it to
1115the packet filter for evaluation.
1116.Pp
1117Since translation occurs before filtering the filter
1118engine will see packets as they look after any
1119addresses and ports have been translated.
1120Filter rules will therefore have to filter based on the translated
1121address and port number.
1122Packets that match a translation rule are only automatically passed if
1123the
1124.Ar pass
1125modifier is given, otherwise they are
1126still subject to
1127.Ar block
1128and
1129.Ar pass
1130rules.
1131.Pp
1132The state entry created permits
1133.Xr pf 4
1134to keep track of the original address for traffic associated with that state
1135and correctly direct return traffic for that connection.
1136.Pp
1137Various types of translation are possible with pf:
1138.Bl -tag -width xxxx
1139.It Ar binat
1140A
1141.Ar binat
1142rule specifies a bidirectional mapping between an external IP netblock
1143and an internal IP netblock.
1144.It Ar nat
1145A
1146.Ar nat
1147rule specifies that IP addresses are to be changed as the packet
1148traverses the given interface.
1149This technique allows one or more IP addresses
1150on the translating host to support network traffic for a larger range of
1151machines on an "inside" network.
1152Although in theory any IP address can be used on the inside, it is strongly
1153recommended that one of the address ranges defined by RFC 1918 be used.
1154These netblocks are:
1155.Bd -literal
115610.0.0.0 - 10.255.255.255 (all of net 10, i.e., 10/8)
1157172.16.0.0 - 172.31.255.255 (i.e., 172.16/12)
1158192.168.0.0 - 192.168.255.255 (i.e., 192.168/16)
1159.Ed
1160.It Pa rdr
1161The packet is redirected to another destination and possibly a
1162different port.
1163.Ar rdr
1164rules can optionally specify port ranges instead of single ports.
1165rdr ... port 2000:2999 -\*(Gt ... port 4000
1166redirects ports 2000 to 2999 (inclusive) to port 4000.
1167rdr ... port 2000:2999 -\*(Gt ... port 4000:*
1168redirects port 2000 to 4000, 2001 to 4001, ..., 2999 to 4999.
1169.El
1170.Pp
1171In addition to modifying the address, some translation rules may modify
1172source or destination ports for
1173.Xr tcp 4
1174or
1175.Xr udp 4
1176connections; implicitly in the case of
1177.Ar nat
1178rules and explicitly in the case of
1179.Ar rdr
1180rules.
1181Port numbers are never translated with a
1182.Ar binat
1183rule.
1184.Pp
1185Evaluation order of the translation rules is dependent on the type
1186of the translation rules and of the direction of a packet.
1187.Ar binat
1188rules are always evaluated first.
1189Then either the
1190.Ar rdr
1191rules are evaluated on an inbound packet or the
1192.Ar nat
1193rules on an outbound packet.
1194Rules of the same type are evaluated in the same order in which they
1195appear in the ruleset.
1196The first matching rule decides what action is taken.
1197.Pp
1198The
1199.Ar no
1200option prefixed to a translation rule causes packets to remain untranslated,
1201much in the same way as
1202.Ar drop quick
1203works in the packet filter (see below).
1204If no rule matches the packet it is passed to the filter engine unmodified.
1205.Pp
1206Translation rules apply only to packets that pass through
1207the specified interface, and if no interface is specified,
1208translation is applied to packets on all interfaces.
1209For instance, redirecting port 80 on an external interface to an internal
1210web server will only work for connections originating from the outside.
1211Connections to the address of the external interface from local hosts will
1212not be redirected, since such packets do not actually pass through the
1213external interface.
1214Redirections cannot reflect packets back through the interface they arrive
1215on, they can only be redirected to hosts connected to different interfaces
1216or to the firewall itself.
1217.Pp
1218Note that redirecting external incoming connections to the loopback
1219address, as in
1220.Bd -literal -offset indent
1221rdr on ne3 inet proto tcp to port smtp -\*(Gt 127.0.0.1 port spamd
1222.Ed
1223.Pp
1224will effectively allow an external host to connect to daemons
1225bound solely to the loopback address, circumventing the traditional
1226blocking of such connections on a real interface.
1227Unless this effect is desired, any of the local non-loopback addresses
1228should be used as redirection target instead, which allows external
1229connections only to daemons bound to this address or not bound to
1230any address.
1231.Pp
1232See
1233.Sx TRANSLATION EXAMPLES
1234below.
1235.Sh PACKET FILTERING
1236.Xr pf 4
1237has the ability to
1238.Ar block
1239and
1240.Ar pass
1241packets based on attributes of their layer 3 (see
1242.Xr ip 4
1243and
1244.Xr ip6 4 )
1245and layer 4 (see
1246.Xr icmp 4 ,
1247.Xr icmp6 4 ,
1248.Xr tcp 4 ,
1249.Xr udp 4 )
1250headers.
1251In addition, packets may also be
1252assigned to queues for the purpose of bandwidth control.
1253.Pp
1254For each packet processed by the packet filter, the filter rules are
1255evaluated in sequential order, from first to last.
1256The last matching rule decides what action is taken.
1257If no rule matches the packet, the default action is to pass
1258the packet.
1259.Pp
1260The following actions can be used in the filter:
1261.Bl -tag -width xxxx
1262.It Ar block
1263The packet is blocked.
1264There are a number of ways in which a
1265.Ar block
1266rule can behave when blocking a packet.
1267The default behaviour is to
1268.Ar drop
1269packets silently, however this can be overridden or made
1270explicit either globally, by setting the
1271.Ar block-policy
1272option, or on a per-rule basis with one of the following options:
1273.Pp
1274.Bl -tag -width xxxx -compact
1275.It Ar drop
1276The packet is silently dropped.
1277.It Ar return-rst
1278This applies only to
1279.Xr tcp 4
1280packets, and issues a TCP RST which closes the
1281connection.
1282.It Ar return-icmp
1283.It Ar return-icmp6
1284This causes ICMP messages to be returned for packets which match the rule.
1285By default this is an ICMP UNREACHABLE message, however this
1286can be overridden by specifying a message as a code or number.
1287.It Ar return
1288This causes a TCP RST to be returned for
1289.Xr tcp 4
1290packets and an ICMP UNREACHABLE for UDP and other packets.
1291.El
1292.Pp
1293Options returning ICMP packets currently have no effect if
1294.Xr pf 4
1295operates on a
1296.Xr if_bridge 4 ,
1297as the code to support this feature has not yet been implemented.
1298.Pp
1299The simplest mechanism to block everything by default and only pass
1300packets that match explicit rules is specify a first filter rule of:
1301.Bd -literal -offset indent
1302block all
1303.Ed
1304.It Ar pass
1305The packet is passed;
1306state is created unless the
1307.Ar no state
1308option is specified.
1309.El
1310.Pp
1311By default
1312.Xr pf 4
1313filters packets statefully; the first time a packet matches a
1314.Ar pass
1315rule, a state entry is created; for subsequent packets the filter checks
1316whether the packet matches any state.
1317If it does, the packet is passed without evaluation of any rules.
1318After the connection is closed or times out, the state entry is automatically
1319removed.
1320.Pp
1321This has several advantages.
1322For TCP connections, comparing a packet to a state involves checking
1323its sequence numbers, as well as TCP timestamps if a
1324.Ar scrub reassemble tcp
1325rule applies to the connection.
1326If these values are outside the narrow windows of expected
1327values, the packet is dropped.
1328This prevents spoofing attacks, such as when an attacker sends packets with
1329a fake source address/port but does not know the connection's sequence
1330numbers.
1331Similarly,
1332.Xr pf 4
1333knows how to match ICMP replies to states.
1334For example,
1335.Bd -literal -offset indent
1336pass out inet proto icmp all icmp-type echoreq
1337.Ed
1338.Pp
1339allows echo requests (such as those created by
1340.Xr ping 8 )
1341out statefully, and matches incoming echo replies correctly to states.
1342.Pp
1343Also, looking up states is usually faster than evaluating rules.
1344If there are 50 rules, all of them are evaluated sequentially in O(n).
1345Even with 50000 states, only 16 comparisons are needed to match a
1346state, since states are stored in a binary search tree that allows
1347searches in O(log2 n).
1348.Pp
1349Furthermore, correct handling of ICMP error messages is critical to
1350many protocols, particularly TCP.
1351.Xr pf 4
1352matches ICMP error messages to the correct connection, checks them against
1353connection parameters, and passes them if appropriate.
1354For example if an ICMP source quench message referring to a stateful TCP
1355connection arrives, it will be matched to the state and get passed.
1356.Pp
1357Finally, state tracking is required for
1358.Ar nat , binat No and Ar rdr
1359rules, in order to track address and port translations and reverse the
1360translation on returning packets.
1361.Pp
1362.Xr pf 4
1363will also create state for other protocols which are effectively stateless by
1364nature.
1365UDP packets are matched to states using only host addresses and ports,
1366and other protocols are matched to states using only the host addresses.
1367.Pp
1368If stateless filtering of individual packets is desired,
1369the
1370.Ar no state
1371keyword can be used to specify that state will not be created
1372if this is the last matching rule.
1373A number of parameters can also be set to affect how
1374.Xr pf 4
1375handles state tracking.
1376See
1377.Sx STATEFUL TRACKING OPTIONS
1378below for further details.
1379.Sh PARAMETERS
1380The rule parameters specify the packets to which a rule applies.
1381A packet always comes in on, or goes out through, one interface.
1382Most parameters are optional.
1383If a parameter is specified, the rule only applies to packets with
1384matching attributes.
1385Certain parameters can be expressed as lists, in which case
1386.Xr pfctl 8
1387generates all needed rule combinations.
1388.Bl -tag -width xxxx
1389.It Ar in No or Ar out
1390This rule applies to incoming or outgoing packets.
1391If neither
1392.Ar in
1393nor
1394.Ar out
1395are specified, the rule will match packets in both directions.
1396.It Ar log
1397In addition to the action specified, a log message is generated.
1398Only the packet that establishes the state is logged,
1399unless the
1400.Ar no state
1401option is specified.
1402The logged packets are sent to a
1403.Xr pflog 4
1404interface, by default
1405.Ar pflog0 .
1406This interface is monitored by the
1407.Xr pflogd 8
1408logging daemon, which dumps the logged packets to the file
1409.Pa /var/log/pflog
1410in
1411.Xr pcap 3
1412binary format.
1413.It Ar log (all)
1414Used to force logging of all packets for a connection.
1415This is not necessary when
1416.Ar no state
1417is explicitly specified.
1418As with
1419.Ar log ,
1420packets are logged to
1421.Xr pflog 4 .
1422.It Ar log (user)
1423Logs the
1424.Ux
1425user ID of the user that owns the socket and the PID of the process that
1426has the socket open where the packet is sourced from or destined to
1427(depending on which socket is local).
1428This is in addition to the normal information logged.
1429.Pp
1430Only the first packet
1431logged via
1432.Ar log (all, user)
1433will have the user credentials logged when using stateful matching.
1434.It Ar log (to Aq Ar interface )
1435Send logs to the specified
1436.Xr pflog 4
1437interface instead of
1438.Ar pflog0 .
1439.It Ar quick
1440If a packet matches a rule which has the
1441.Ar quick
1442option set, this rule
1443is considered the last matching rule, and evaluation of subsequent rules
1444is skipped.
1445.It Ar on Aq Ar interface
1446This rule applies only to packets coming in on, or going out through, this
1447particular interface or interface group.
1448For more information on interface groups,
1449see the
1450.Ic group
1451keyword in
1452.Xr ifconfig 8 .
1453.It Aq Ar af
1454This rule applies only to packets of this address family.
1455Supported values are
1456.Ar inet
1457and
1458.Ar inet6 .
1459.It Ar proto Aq Ar protocol
1460This rule applies only to packets of this protocol.
1461Common protocols are
1462.Xr icmp 4 ,
1463.Xr icmp6 4 ,
1464.Xr tcp 4 ,
1465and
1466.Xr udp 4 .
1467For a list of all the protocol name to number mappings used by
1468.Xr pfctl 8 ,
1469see the file
1470.Pa /etc/protocols .
1471.It Xo
1472.Ar from Aq Ar source
1473.Ar port Aq Ar source
1474.Ar os Aq Ar source
1475.Ar to Aq Ar dest
1476.Ar port Aq Ar dest
1477.Xc
1478This rule applies only to packets with the specified source and destination
1479addresses and ports.
1480.Pp
1481Addresses can be specified in CIDR notation (matching netblocks), as
1482symbolic host names, interface names or interface group names, or as any
1483of the following keywords:
1484.Pp
1485.Bl -tag -width xxxxxxxxxxxxxx -compact
1486.It Ar any
1487Any address.
1488.It Ar no-route
1489Any address which is not currently routable.
1490.It Ar urpf-failed
1491Any source address that fails a unicast reverse path forwarding (URPF)
1492check, i.e. packets coming in on an interface other than that which holds
1493the route back to the packet's source address.
1494.It Aq Ar table
1495Any address that matches the given table.
1496.El
1497.Pp
1498Ranges of addresses are specified by using the
1499.Sq -
1500operator.
1501For instance:
1502.Dq 10.1.1.10 - 10.1.1.12
1503means all addresses from 10.1.1.10 to 10.1.1.12,
1504hence addresses 10.1.1.10, 10.1.1.11, and 10.1.1.12.
1505.Pp
1506Interface names and interface group names can have modifiers appended:
1507.Pp
1508.Bl -tag -width xxxxxxxxxxxx -compact
1509.It Ar :network
1510Translates to the network(s) attached to the interface.
1511.It Ar :broadcast
1512Translates to the interface's broadcast address(es).
1513.It Ar :peer
1514Translates to the point-to-point interface's peer address(es).
1515.It Ar :0
1516Do not include interface aliases.
1517.El
1518.Pp
1519Host names may also have the
1520.Ar :0
1521option appended to restrict the name resolution to the first of each
1522v4 and non-link-local v6 address found.
1523.Pp
1524Host name resolution and interface to address translation are done at
1525ruleset load-time.
1526When the address of an interface (or host name) changes (under DHCP or PPP,
1527for instance), the ruleset must be reloaded for the change to be reflected
1528in the kernel.
1529Surrounding the interface name (and optional modifiers) in parentheses
1530changes this behaviour.
1531When the interface name is surrounded by parentheses, the rule is
1532automatically updated whenever the interface changes its address.
1533The ruleset does not need to be reloaded.
1534This is especially useful with
1535.Ar nat .
1536.Pp
1537Ports can be specified either by number or by name.
1538For example, port 80 can be specified as
1539.Em www .
1540For a list of all port name to number mappings used by
1541.Xr pfctl 8 ,
1542see the file
1543.Pa /etc/services .
1544.Pp
1545Ports and ranges of ports are specified by using these operators:
1546.Bd -literal -offset indent
1547=	(equal)
1548!=	(unequal)
1549\*(Lt	(less than)
1550\*(Le	(less than or equal)
1551\*(Gt	(greater than)
1552\*(Ge	(greater than or equal)
1553:	(range including boundaries)
1554\*(Gt\*(Lt	(range excluding boundaries)
1555\*(Lt\*(Gt	(except range)
1556.Ed
1557.Pp
1558.Sq \*(Gt\*(Lt ,
1559.Sq \*(Lt\*(Gt
1560and
1561.Sq \&:
1562are binary operators (they take two arguments).
1563For instance:
1564.Bl -tag -width Fl
1565.It Ar port 2000:2004
1566means
1567.Sq all ports \*(Ge 2000 and \*(Le 2004 ,
1568hence ports 2000, 2001, 2002, 2003 and 2004.
1569.It Ar port 2000 \*(Gt\*(Lt 2004
1570means
1571.Sq all ports \*(Gt 2000 and \*(Lt 2004 ,
1572hence ports 2001, 2002 and 2003.
1573.It Ar port 2000 \*(Lt\*(Gt 2004
1574means
1575.Sq all ports \*(Lt 2000 or \*(Gt 2004 ,
1576hence ports 1-1999 and 2005-65535.
1577.El
1578.Pp
1579The operating system of the source host can be specified in the case of TCP
1580rules with the
1581.Ar OS
1582modifier.
1583See the
1584.Sx OPERATING SYSTEM FINGERPRINTING
1585section for more information.
1586.Pp
1587The host, port and OS specifications are optional, as in the following examples:
1588.Bd -literal -offset indent
1589pass in all
1590pass in from any to any
1591pass in proto tcp from any port \*(Le 1024 to any
1592pass in proto tcp from any to any port 25
1593pass in proto tcp from 10.0.0.0/8 port \*(Gt 1024 \e
1594      to ! 10.1.2.3 port != ssh
1595pass in proto tcp from any os "OpenBSD"
1596.Ed
1597.It Ar all
1598This is equivalent to "from any to any".
1599.It Ar group Aq Ar group
1600Similar to
1601.Ar user ,
1602this rule only applies to packets of sockets owned by the specified group.
1603.It Ar user Aq Ar user
1604This rule only applies to packets of sockets owned by the specified user.
1605For outgoing connections initiated from the firewall, this is the user
1606that opened the connection.
1607For incoming connections to the firewall itself, this is the user that
1608listens on the destination port.
1609For forwarded connections, where the firewall is not a connection endpoint,
1610the user and group are
1611.Em unknown .
1612.Pp
1613All packets, both outgoing and incoming, of one connection are associated
1614with the same user and group.
1615Only TCP and UDP packets can be associated with users; for other protocols
1616these parameters are ignored.
1617.Pp
1618User and group refer to the effective (as opposed to the real) IDs, in
1619case the socket is created by a setuid/setgid process.
1620User and group IDs are stored when a socket is created;
1621when a process creates a listening socket as root (for instance, by
1622binding to a privileged port) and subsequently changes to another
1623user ID (to drop privileges), the credentials will remain root.
1624.Pp
1625User and group IDs can be specified as either numbers or names.
1626The syntax is similar to the one for ports.
1627The value
1628.Em unknown
1629matches packets of forwarded connections.
1630.Em unknown
1631can only be used with the operators
1632.Cm =
1633and
1634.Cm != .
1635Other constructs like
1636.Cm user \*(Ge unknown
1637are invalid.
1638Forwarded packets with unknown user and group ID match only rules
1639that explicitly compare against
1640.Em unknown
1641with the operators
1642.Cm =
1643or
1644.Cm != .
1645For instance
1646.Cm user \*(Ge 0
1647does not match forwarded packets.
1648The following example allows only selected users to open outgoing
1649connections:
1650.Bd -literal -offset indent
1651block out proto { tcp, udp } all
1652pass  out proto { tcp, udp } all user { \*(Lt 1000, dhartmei }
1653.Ed
1654.It Xo Ar flags Aq Ar a
1655.Pf / Ns Aq Ar b
1656.No \*(Ba / Ns Aq Ar b
1657.No \*(Ba any
1658.Xc
1659This rule only applies to TCP packets that have the flags
1660.Aq Ar a
1661set out of set
1662.Aq Ar b .
1663Flags not specified in
1664.Aq Ar b
1665are ignored.
1666For stateful connections, the default is
1667.Ar flags S/SA .
1668To indicate that flags should not be checked at all, specify
1669.Ar flags any .
1670The flags are: (F)IN, (S)YN, (R)ST, (P)USH, (A)CK, (U)RG, (E)CE, and C(W)R.
1671.Bl -tag -width Fl
1672.It Ar flags S/S
1673Flag SYN is set.
1674The other flags are ignored.
1675.It Ar flags S/SA
1676This is the default setting for stateful connections.
1677Out of SYN and ACK, exactly SYN may be set.
1678SYN, SYN+PSH and SYN+RST match, but SYN+ACK, ACK and ACK+RST do not.
1679This is more restrictive than the previous example.
1680.It Ar flags /SFRA
1681If the first set is not specified, it defaults to none.
1682All of SYN, FIN, RST and ACK must be unset.
1683.El
1684.Pp
1685Because
1686.Ar flags S/SA
1687is applied by default (unless
1688.Ar no state
1689is specified), only the initial SYN packet of a TCP handshake will create
1690a state for a TCP connection.
1691It is possible to be less restrictive, and allow state creation from
1692intermediate
1693.Pq non-SYN
1694packets, by specifying
1695.Ar flags any .
1696This will cause
1697.Xr pf 4
1698to synchronize to existing connections, for instance
1699if one flushes the state table.
1700However, states created from such intermediate packets may be missing
1701connection details such as the TCP window scaling factor.
1702States which modify the packet flow, such as those affected by
1703.Ar nat , binat No or Ar rdr
1704rules,
1705.Ar modulate No or Ar synproxy state
1706options, or scrubbed with
1707.Ar reassemble tcp
1708will also not be recoverable from intermediate packets.
1709Such connections will stall and time out.
1710.It Xo Ar icmp-type Aq Ar type
1711.Ar code Aq Ar code
1712.Xc
1713.It Xo Ar icmp6-type Aq Ar type
1714.Ar code Aq Ar code
1715.Xc
1716This rule only applies to ICMP or ICMPv6 packets with the specified type
1717and code.
1718Text names for ICMP types and codes are listed in
1719.Xr icmp 4
1720and
1721.Xr icmp6 4 .
1722This parameter is only valid for rules that cover protocols ICMP or
1723ICMP6.
1724The protocol and the ICMP type indicator
1725.Po
1726.Ar icmp-type
1727or
1728.Ar icmp6-type
1729.Pc
1730must match.
1731.It Xo Ar tos Aq Ar string
1732.No \*(Ba Aq Ar number
1733.Xc
1734This rule applies to packets with the specified
1735.Em TOS
1736bits set.
1737.Em TOS
1738may be
1739given as one of
1740.Ar critical ,
1741.Ar inetcontrol ,
1742.Ar lowdelay ,
1743.Ar netcontrol ,
1744.Ar throughput ,
1745.Ar reliability ,
1746or one of the DiffServ Code Points:
1747.Ar ef ,
1748.Ar va ,
1749.Ar af11 No ... Ar af43 ,
1750.Ar cs0 No ... Ar cs7 ;
1751or as either hex or decimal.
1752.Pp
1753For example, the following rules are identical:
1754.Bd -literal -offset indent
1755pass all tos lowdelay
1756pass all tos 0x10
1757pass all tos 16
1758.Ed
1759.It Ar allow-opts
1760By default, IPv4 packets with IP options or IPv6 packets with routing
1761extension headers are blocked.
1762When
1763.Ar allow-opts
1764is specified for a
1765.Ar pass
1766rule, packets that pass the filter based on that rule (last matching)
1767do so even if they contain IP options or routing extension headers.
1768For packets that match state, the rule that initially created the
1769state is used.
1770The implicit
1771.Ar pass
1772rule that is used when a packet does not match any rules does not
1773allow IP options.
1774.It Ar label Aq Ar string
1775Adds a label (name) to the rule, which can be used to identify the rule.
1776For instance,
1777pfctl -s labels
1778shows per-rule statistics for rules that have labels.
1779.Pp
1780The following macros can be used in labels:
1781.Pp
1782.Bl -tag -width $srcaddr -compact -offset indent
1783.It Ar $if
1784The interface.
1785.It Ar $srcaddr
1786The source IP address.
1787.It Ar $dstaddr
1788The destination IP address.
1789.It Ar $srcport
1790The source port specification.
1791.It Ar $dstport
1792The destination port specification.
1793.It Ar $proto
1794The protocol name.
1795.It Ar $nr
1796The rule number.
1797.El
1798.Pp
1799For example:
1800.Bd -literal -offset indent
1801ips = \&"{ 1.2.3.4, 1.2.3.5 }\&"
1802pass in proto tcp from any to $ips \e
1803      port \*(Gt 1023 label \&"$dstaddr:$dstport\&"
1804.Ed
1805.Pp
1806expands to
1807.Bd -literal -offset indent
1808pass in inet proto tcp from any to 1.2.3.4 \e
1809      port \*(Gt 1023 label \&"1.2.3.4:\*(Gt1023\&"
1810pass in inet proto tcp from any to 1.2.3.5 \e
1811      port \*(Gt 1023 label \&"1.2.3.5:\*(Gt1023\&"
1812.Ed
1813.Pp
1814The macro expansion for the
1815.Ar label
1816directive occurs only at configuration file parse time, not during runtime.
1817.It Xo Ar queue Aq Ar queue
1818.No \*(Ba ( Aq Ar queue ,
1819.Aq Ar queue )
1820.Xc
1821Packets matching this rule will be assigned to the specified queue.
1822If two queues are given, packets which have a
1823.Em TOS
1824of
1825.Em lowdelay
1826and TCP ACKs with no data payload will be assigned to the second one.
1827See
1828.Sx QUEUEING
1829for setup details.
1830.Pp
1831For example:
1832.Bd -literal -offset indent
1833pass in proto tcp to port 25 queue mail
1834pass in proto tcp to port 22 queue(ssh_bulk, ssh_prio)
1835.Ed
1836.Pp
1837.It Cm set prio Ar priority | Pq Ar priority , priority
1838Packets matching this rule will be assigned a specific queueing priority.
1839Priorities are assigned as integers 0 through 7.
1840If the packet is transmitted on a
1841.Xr vlan 4
1842interface, the queueing priority will be written as the priority
1843code point in the 802.1Q VLAN header.
1844If two priorities are given, packets which have a TOS of
1845.Cm lowdelay
1846and TCP ACKs with no data payload will be assigned to the second one.
1847.Pp
1848For example:
1849.Bd -literal -offset indent
1850pass in proto tcp to port 25 set prio 2
1851pass in proto tcp to port 22 set prio (2, 5)
1852.Ed
1853.Pp
1854.It Ar tag Aq Ar string
1855Packets matching this rule will be tagged with the
1856specified string.
1857The tag acts as an internal marker that can be used to
1858identify these packets later on.
1859This can be used, for example, to provide trust between
1860interfaces and to determine if packets have been
1861processed by translation rules.
1862Tags are
1863.Qq sticky ,
1864meaning that the packet will be tagged even if the rule
1865is not the last matching rule.
1866Further matching rules can replace the tag with a
1867new one but will not remove a previously applied tag.
1868A packet is only ever assigned one tag at a time.
1869Packet tagging can be done during
1870.Ar nat ,
1871.Ar rdr ,
1872or
1873.Ar binat
1874rules in addition to filter rules.
1875Tags take the same macros as labels (see above).
1876.It Ar tagged Aq Ar string
1877Used with filter, translation or scrub rules
1878to specify that packets must already
1879be tagged with the given tag in order to match the rule.
1880Inverse tag matching can also be done
1881by specifying the
1882.Cm !\&
1883operator before the
1884.Ar tagged
1885keyword.
1886.It Ar rtable Aq Ar number
1887Used to select an alternate routing table for the routing lookup.
1888Only effective before the route lookup happened, i.e. when filtering inbound.
1889.It Xo Ar divert-to Aq Ar host
1890.Ar port Aq Ar port
1891.Xc
1892Used to redirect packets to a local socket bound to
1893.Ar host
1894and
1895.Ar port .
1896The packets will not be modified, so
1897.Xr getsockname 2
1898on the socket will return the original destination address of the packet.
1899.It Ar divert-reply
1900Used to receive replies for sockets that are bound to addresses
1901which are not local to the machine.
1902See
1903.Xr setsockopt 2
1904for information on how to bind these sockets.
1905.It Ar probability Aq Ar number
1906A probability attribute can be attached to a rule, with a value set between
19070 and 1, bounds not included.
1908In that case, the rule will be honoured using the given probability value
1909only.
1910For example, the following rule will drop 20% of incoming ICMP packets:
1911.Bd -literal -offset indent
1912block in proto icmp probability 20%
1913.Ed
1914.It Ar prio Aq Ar number
1915Only match packets which have the given queueing priority assigned.
1916.Pp
1917.El
1918.Sh ROUTING
1919If a packet matches a rule with a route option set, the packet filter will
1920route the packet according to the type of route option.
1921When such a rule creates state, the route option is also applied to all
1922packets matching the same connection.
1923.Bl -tag -width xxxx
1924.It Ar route-to
1925The
1926.Ar route-to
1927option routes the packet to the specified interface with an optional address
1928for the next hop.
1929When a
1930.Ar route-to
1931rule creates state, only packets that pass in the same direction as the
1932filter rule specifies will be routed in this way.
1933Packets passing in the opposite direction (replies) are not affected
1934and are routed normally.
1935.It Ar reply-to
1936The
1937.Ar reply-to
1938option is similar to
1939.Ar route-to ,
1940but routes packets that pass in the opposite direction (replies) to the
1941specified interface.
1942Opposite direction is only defined in the context of a state entry, and
1943.Ar reply-to
1944is useful only in rules that create state.
1945It can be used on systems with multiple external connections to
1946route all outgoing packets of a connection through the interface
1947the incoming connection arrived through (symmetric routing enforcement).
1948.It Ar dup-to
1949The
1950.Ar dup-to
1951option creates a duplicate of the packet and routes it like
1952.Ar route-to .
1953The original packet gets routed as it normally would.
1954.El
1955.Sh POOL OPTIONS
1956For
1957.Ar nat
1958and
1959.Ar rdr
1960rules, (as well as for the
1961.Ar route-to ,
1962.Ar reply-to
1963and
1964.Ar dup-to
1965rule options) for which there is a single redirection address which has a
1966subnet mask smaller than 32 for IPv4 or 128 for IPv6 (more than one IP
1967address), a variety of different methods for assigning this address can be
1968used:
1969.Bl -tag -width xxxx
1970.It Ar bitmask
1971The
1972.Ar bitmask
1973option applies the network portion of the redirection address to the address
1974to be modified (source with
1975.Ar nat ,
1976destination with
1977.Ar rdr ) .
1978.It Ar random
1979The
1980.Ar random
1981option selects an address at random within the defined block of addresses.
1982.It Ar source-hash
1983The
1984.Ar source-hash
1985option uses a hash of the source address to determine the redirection address,
1986ensuring that the redirection address is always the same for a given source.
1987An optional key can be specified after this keyword either in hex or as a
1988string; by default
1989.Xr pfctl 8
1990randomly generates a key for source-hash every time the
1991ruleset is reloaded.
1992.It Ar round-robin
1993The
1994.Ar round-robin
1995option loops through the redirection address(es).
1996.Pp
1997When more than one redirection address is specified,
1998.Ar round-robin
1999is the only permitted pool type.
2000.It Ar static-port
2001With
2002.Ar nat
2003rules, the
2004.Ar static-port
2005option prevents
2006.Xr pf 4
2007from modifying the source port on TCP and UDP packets.
2008.It Xo Ar map-e-portset Aq Ar psid-offset
2009.No / Aq Ar psid-len
2010.No / Aq Ar psid
2011.Xc
2012With
2013.Ar nat
2014rules, the
2015.Ar map-e-portset
2016option enables the source port translation of MAP-E (RFC 7597) Customer Edge.
2017In order to make the host act as a MAP-E Customer Edge, setting up a tunneling
2018interface and pass rules for encapsulated packets are required in addition
2019to the map-e-portset nat rule.
2020.Pp
2021For example:
2022.Bd -literal -offset indent
2023nat on $gif_mape_if from $int_if:network to any \e
2024      -> $ipv4_mape_src map-e-portset 6/8/0x34
2025.Ed
2026.Pp
2027sets PSID offset 6, PSID length 8, PSID 0x34.
2028.Ed
2029.El
2030.Pp
2031Additionally, the
2032.Ar sticky-address
2033option can be specified to help ensure that multiple connections from the
2034same source are mapped to the same redirection address.
2035This option can be used with the
2036.Ar random
2037and
2038.Ar round-robin
2039pool options.
2040Note that by default these associations are destroyed as soon as there are
2041no longer states which refer to them; in order to make the mappings last
2042beyond the lifetime of the states, increase the global options with
2043.Ar set timeout src.track .
2044See
2045.Sx STATEFUL TRACKING OPTIONS
2046for more ways to control the source tracking.
2047.Sh STATE MODULATION
2048Much of the security derived from TCP is attributable to how well the
2049initial sequence numbers (ISNs) are chosen.
2050Some popular stack implementations choose
2051.Em very
2052poor ISNs and thus are normally susceptible to ISN prediction exploits.
2053By applying a
2054.Ar modulate state
2055rule to a TCP connection,
2056.Xr pf 4
2057will create a high quality random sequence number for each connection
2058endpoint.
2059.Pp
2060The
2061.Ar modulate state
2062directive implicitly keeps state on the rule and is
2063only applicable to TCP connections.
2064.Pp
2065For instance:
2066.Bd -literal -offset indent
2067block all
2068pass out proto tcp from any to any modulate state
2069pass in  proto tcp from any to any port 25 flags S/SFRA modulate state
2070.Ed
2071.Pp
2072Note that modulated connections will not recover when the state table
2073is lost (firewall reboot, flushing the state table, etc...).
2074.Xr pf 4
2075will not be able to infer a connection again after the state table flushes
2076the connection's modulator.
2077When the state is lost, the connection may be left dangling until the
2078respective endpoints time out the connection.
2079It is possible on a fast local network for the endpoints to start an ACK
2080storm while trying to resynchronize after the loss of the modulator.
2081The default
2082.Ar flags
2083settings (or a more strict equivalent) should be used on
2084.Ar modulate state
2085rules to prevent ACK storms.
2086.Pp
2087Note that alternative methods are available
2088to prevent loss of the state table
2089and allow for firewall failover.
2090See
2091.Xr carp 4
2092and
2093.Xr pfsync 4
2094for further information.
2095.Sh SYN PROXY
2096By default,
2097.Xr pf 4
2098passes packets that are part of a
2099.Xr tcp 4
2100handshake between the endpoints.
2101The
2102.Ar synproxy state
2103option can be used to cause
2104.Xr pf 4
2105itself to complete the handshake with the active endpoint, perform a handshake
2106with the passive endpoint, and then forward packets between the endpoints.
2107.Pp
2108No packets are sent to the passive endpoint before the active endpoint has
2109completed the handshake, hence so-called SYN floods with spoofed source
2110addresses will not reach the passive endpoint, as the sender can't complete the
2111handshake.
2112.Pp
2113The proxy is transparent to both endpoints, they each see a single
2114connection from/to the other endpoint.
2115.Xr pf 4
2116chooses random initial sequence numbers for both handshakes.
2117Once the handshakes are completed, the sequence number modulators
2118(see previous section) are used to translate further packets of the
2119connection.
2120.Ar synproxy state
2121includes
2122.Ar modulate state .
2123.Pp
2124Rules with
2125.Ar synproxy
2126will not work if
2127.Xr pf 4
2128operates on a
2129.Xr bridge 4 .
2130.Pp
2131Example:
2132.Bd -literal -offset indent
2133pass in proto tcp from any to any port www synproxy state
2134.Ed
2135.Sh STATEFUL TRACKING OPTIONS
2136A number of options related to stateful tracking can be applied on a
2137per-rule basis.
2138.Ar keep state ,
2139.Ar modulate state
2140and
2141.Ar synproxy state
2142support these options, and
2143.Ar keep state
2144must be specified explicitly to apply options to a rule.
2145.Pp
2146.Bl -tag -width xxxx -compact
2147.It Ar max Aq Ar number
2148Limits the number of concurrent states the rule may create.
2149When this limit is reached, further packets that would create
2150state will not match this rule until existing states time out.
2151.It Ar no-sync
2152Prevent state changes for states created by this rule from appearing on the
2153.Xr pfsync 4
2154interface.
2155.It Xo Aq Ar timeout
2156.Aq Ar seconds
2157.Xc
2158Changes the timeout values used for states created by this rule.
2159For a list of all valid timeout names, see
2160.Sx OPTIONS
2161above.
2162.It Ar sloppy
2163Uses a sloppy TCP connection tracker that does not check sequence
2164numbers at all, which makes insertion and ICMP teardown attacks way
2165easier.
2166This is intended to be used in situations where one does not see all
2167packets of a connection, e.g. in asymmetric routing situations.
2168Cannot be used with modulate or synproxy state.
2169.El
2170.Pp
2171Multiple options can be specified, separated by commas:
2172.Bd -literal -offset indent
2173pass in proto tcp from any to any \e
2174      port www keep state \e
2175      (max 100, source-track rule, max-src-nodes 75, \e
2176      max-src-states 3, tcp.established 60, tcp.closing 5)
2177.Ed
2178.Pp
2179When the
2180.Ar source-track
2181keyword is specified, the number of states per source IP is tracked.
2182.Pp
2183.Bl -tag -width xxxx -compact
2184.It Ar source-track rule
2185The maximum number of states created by this rule is limited by the rule's
2186.Ar max-src-nodes
2187and
2188.Ar max-src-states
2189options.
2190Only state entries created by this particular rule count toward the rule's
2191limits.
2192.It Ar source-track global
2193The number of states created by all rules that use this option is limited.
2194Each rule can specify different
2195.Ar max-src-nodes
2196and
2197.Ar max-src-states
2198options, however state entries created by any participating rule count towards
2199each individual rule's limits.
2200.El
2201.Pp
2202The following limits can be set:
2203.Pp
2204.Bl -tag -width xxxx -compact
2205.It Ar max-src-nodes Aq Ar number
2206Limits the maximum number of source addresses which can simultaneously
2207have state table entries.
2208.It Ar max-src-states Aq Ar number
2209Limits the maximum number of simultaneous state entries that a single
2210source address can create with this rule.
2211.El
2212.Pp
2213For stateful TCP connections, limits on established connections (connections
2214which have completed the TCP 3-way handshake) can also be enforced
2215per source IP.
2216.Pp
2217.Bl -tag -width xxxx -compact
2218.It Ar max-src-conn Aq Ar number
2219Limits the maximum number of simultaneous TCP connections which have
2220completed the 3-way handshake that a single host can make.
2221.It Xo Ar max-src-conn-rate Aq Ar number
2222.No / Aq Ar seconds
2223.Xc
2224Limit the rate of new connections over a time interval.
2225The connection rate is an approximation calculated as a moving average.
2226.El
2227.Pp
2228Because the 3-way handshake ensures that the source address is not being
2229spoofed, more aggressive action can be taken based on these limits.
2230With the
2231.Ar overload Aq Ar table
2232state option, source IP addresses which hit either of the limits on
2233established connections will be added to the named table.
2234This table can be used in the ruleset to block further activity from
2235the offending host, redirect it to a tarpit process, or restrict its
2236bandwidth.
2237.Pp
2238The optional
2239.Ar flush
2240keyword kills all states created by the matching rule which originate
2241from the host which exceeds these limits.
2242The
2243.Ar global
2244modifier to the flush command kills all states originating from the
2245offending host, regardless of which rule created the state.
2246.Pp
2247For example, the following rules will protect the webserver against
2248hosts making more than 100 connections in 10 seconds.
2249Any host which connects faster than this rate will have its address added
2250to the
2251.Aq bad_hosts
2252table and have all states originating from it flushed.
2253Any new packets arriving from this host will be dropped unconditionally
2254by the block rule.
2255.Bd -literal -offset indent
2256block quick from \*(Ltbad_hosts\*(Gt
2257pass in on $ext_if proto tcp to $webserver port www keep state \e
2258	(max-src-conn-rate 100/10, overload \*(Ltbad_hosts\*(Gt flush global)
2259.Ed
2260.Sh OPERATING SYSTEM FINGERPRINTING
2261Passive OS Fingerprinting is a mechanism to inspect nuances of a TCP
2262connection's initial SYN packet and guess at the host's operating system.
2263Unfortunately these nuances are easily spoofed by an attacker so the
2264fingerprint is not useful in making security decisions.
2265But the fingerprint is typically accurate enough to make policy decisions
2266upon.
2267.Pp
2268The fingerprints may be specified by operating system class, by
2269version, or by subtype/patchlevel.
2270The class of an operating system is typically the vendor or genre
2271and would be
2272.Ox
2273for the
2274.Xr pf 4
2275firewall itself.
2276The version of the oldest available
2277.Ox
2278release on the main FTP site
2279would be 2.6 and the fingerprint would be written
2280.Pp
2281.Dl \&"OpenBSD 2.6\&"
2282.Pp
2283The subtype of an operating system is typically used to describe the
2284patchlevel if that patch led to changes in the TCP stack behavior.
2285In the case of
2286.Ox ,
2287the only subtype is for a fingerprint that was
2288normalized by the
2289.Ar no-df
2290scrub option and would be specified as
2291.Pp
2292.Dl \&"OpenBSD 3.3 no-df\&"
2293.Pp
2294Fingerprints for most popular operating systems are provided by
2295.Xr pf.os 5 .
2296Once
2297.Xr pf 4
2298is running, a complete list of known operating system fingerprints may
2299be listed by running:
2300.Pp
2301.Dl # pfctl -so
2302.Pp
2303Filter rules can enforce policy at any level of operating system specification
2304assuming a fingerprint is present.
2305Policy could limit traffic to approved operating systems or even ban traffic
2306from hosts that aren't at the latest service pack.
2307.Pp
2308The
2309.Ar unknown
2310class can also be used as the fingerprint which will match packets for
2311which no operating system fingerprint is known.
2312.Pp
2313Examples:
2314.Bd -literal -offset indent
2315pass  out proto tcp from any os OpenBSD
2316block out proto tcp from any os Doors
2317block out proto tcp from any os "Doors PT"
2318block out proto tcp from any os "Doors PT SP3"
2319block out from any os "unknown"
2320pass on lo0 proto tcp from any os "OpenBSD 3.3 lo0"
2321.Ed
2322.Pp
2323Operating system fingerprinting is limited only to the TCP SYN packet.
2324This means that it will not work on other protocols and will not match
2325a currently established connection.
2326.Pp
2327Caveat: operating system fingerprints are occasionally wrong.
2328There are three problems: an attacker can trivially craft his packets to
2329appear as any operating system he chooses;
2330an operating system patch could change the stack behavior and no fingerprints
2331will match it until the database is updated;
2332and multiple operating systems may have the same fingerprint.
2333.Sh BLOCKING SPOOFED TRAFFIC
2334"Spoofing" is the faking of IP addresses, typically for malicious
2335purposes.
2336The
2337.Ar antispoof
2338directive expands to a set of filter rules which will block all
2339traffic with a source IP from the network(s) directly connected
2340to the specified interface(s) from entering the system through
2341any other interface.
2342.Pp
2343For example, the line
2344.Bd -literal -offset indent
2345antispoof for lo0
2346.Ed
2347.Pp
2348expands to
2349.Bd -literal -offset indent
2350block drop in on ! lo0 inet from 127.0.0.1/8 to any
2351block drop in on ! lo0 inet6 from ::1 to any
2352.Ed
2353.Pp
2354For non-loopback interfaces, there are additional rules to block incoming
2355packets with a source IP address identical to the interface's IP(s).
2356For example, assuming the interface wi0 had an IP address of 10.0.0.1 and a
2357netmask of 255.255.255.0,
2358the line
2359.Bd -literal -offset indent
2360antispoof for wi0 inet
2361.Ed
2362.Pp
2363expands to
2364.Bd -literal -offset indent
2365block drop in on ! wi0 inet from 10.0.0.0/24 to any
2366block drop in inet from 10.0.0.1 to any
2367.Ed
2368.Pp
2369Caveat: Rules created by the
2370.Ar antispoof
2371directive interfere with packets sent over loopback interfaces
2372to local addresses.
2373One should pass these explicitly.
2374.Sh FRAGMENT HANDLING
2375The size of IP datagrams (packets) can be significantly larger than the
2376maximum transmission unit (MTU) of the network.
2377In cases when it is necessary or more efficient to send such large packets,
2378the large packet will be fragmented into many smaller packets that will each
2379fit onto the wire.
2380Unfortunately for a firewalling device, only the first logical fragment will
2381contain the necessary header information for the subprotocol that allows
2382.Xr pf 4
2383to filter on things such as TCP ports or to perform NAT.
2384.Pp
2385Besides the use of
2386.Ar scrub
2387rules as described in
2388.Sx TRAFFIC NORMALIZATION
2389above, there are three options for handling fragments in the packet filter.
2390.Pp
2391One alternative is to filter individual fragments with filter rules.
2392If no
2393.Ar scrub
2394rule applies to a fragment, it is passed to the filter.
2395Filter rules with matching IP header parameters decide whether the
2396fragment is passed or blocked, in the same way as complete packets
2397are filtered.
2398Without reassembly, fragments can only be filtered based on IP header
2399fields (source/destination address, protocol), since subprotocol header
2400fields are not available (TCP/UDP port numbers, ICMP code/type).
2401The
2402.Ar fragment
2403option can be used to restrict filter rules to apply only to
2404fragments, but not complete packets.
2405Filter rules without the
2406.Ar fragment
2407option still apply to fragments, if they only specify IP header fields.
2408For instance, the rule
2409.Bd -literal -offset indent
2410pass in proto tcp from any to any port 80
2411.Ed
2412.Pp
2413never applies to a fragment, even if the fragment is part of a TCP
2414packet with destination port 80, because without reassembly this information
2415is not available for each fragment.
2416This also means that fragments cannot create new or match existing
2417state table entries, which makes stateful filtering and address
2418translation (NAT, redirection) for fragments impossible.
2419.Pp
2420It's also possible to reassemble only certain fragments by specifying
2421source or destination addresses or protocols as parameters in
2422.Ar scrub
2423rules.
2424.Pp
2425In most cases, the benefits of reassembly outweigh the additional
2426memory cost, and it's recommended to use
2427.Ar scrub
2428rules to reassemble
2429all fragments via the
2430.Ar fragment reassemble
2431modifier.
2432.Pp
2433The memory allocated for fragment caching can be limited using
2434.Xr pfctl 8 .
2435Once this limit is reached, fragments that would have to be cached
2436are dropped until other entries time out.
2437The timeout value can also be adjusted.
2438.Pp
2439When forwarding reassembled IPv6 packets, pf refragments them with
2440the original maximum fragment size.
2441This allows the sender to determine the optimal fragment size by
2442path MTU discovery.
2443.Sh ANCHORS
2444Besides the main ruleset,
2445.Xr pfctl 8
2446can load rulesets into
2447.Ar anchor
2448attachment points.
2449An
2450.Ar anchor
2451is a container that can hold rules, address tables, and other anchors.
2452.Pp
2453An
2454.Ar anchor
2455has a name which specifies the path where
2456.Xr pfctl 8
2457can be used to access the anchor to perform operations on it, such as
2458attaching child anchors to it or loading rules into it.
2459Anchors may be nested, with components separated by
2460.Sq /
2461characters, similar to how file system hierarchies are laid out.
2462The main ruleset is actually the default anchor, so filter and
2463translation rules, for example, may also be contained in any anchor.
2464.Pp
2465An anchor can reference another
2466.Ar anchor
2467attachment point
2468using the following kinds
2469of rules:
2470.Bl -tag -width xxxx
2471.It Ar nat-anchor Aq Ar name
2472Evaluates the
2473.Ar nat
2474rules in the specified
2475.Ar anchor .
2476.It Ar rdr-anchor Aq Ar name
2477Evaluates the
2478.Ar rdr
2479rules in the specified
2480.Ar anchor .
2481.It Ar binat-anchor Aq Ar name
2482Evaluates the
2483.Ar binat
2484rules in the specified
2485.Ar anchor .
2486.It Ar anchor Aq Ar name
2487Evaluates the filter rules in the specified
2488.Ar anchor .
2489.It Xo Ar load anchor
2490.Aq Ar name
2491.Ar from Aq Ar file
2492.Xc
2493Loads the rules from the specified file into the
2494anchor
2495.Ar name .
2496.El
2497.Pp
2498When evaluation of the main ruleset reaches an
2499.Ar anchor
2500rule,
2501.Xr pf 4
2502will proceed to evaluate all rules specified in that anchor.
2503.Pp
2504Matching filter and translation rules marked with the
2505.Ar quick
2506option are final and abort the evaluation of the rules in other
2507anchors and the main ruleset.
2508If the
2509.Ar anchor
2510itself is marked with the
2511.Ar quick
2512option,
2513ruleset evaluation will terminate when the anchor is exited if the packet is
2514matched by any rule within the anchor.
2515.Pp
2516.Ar anchor
2517rules are evaluated relative to the anchor in which they are contained.
2518For example, all
2519.Ar anchor
2520rules specified in the main ruleset will reference anchor
2521attachment points underneath the main ruleset, and
2522.Ar anchor
2523rules specified in a file loaded from a
2524.Ar load anchor
2525rule will be attached under that anchor point.
2526.Pp
2527Rules may be contained in
2528.Ar anchor
2529attachment points which do not contain any rules when the main ruleset
2530is loaded, and later such anchors can be manipulated through
2531.Xr pfctl 8
2532without reloading the main ruleset or other anchors.
2533For example,
2534.Bd -literal -offset indent
2535ext_if = \&"kue0\&"
2536block on $ext_if all
2537anchor spam
2538pass out on $ext_if all
2539pass in on $ext_if proto tcp from any \e
2540      to $ext_if port smtp
2541.Ed
2542.Pp
2543blocks all packets on the external interface by default, then evaluates
2544all rules in the
2545.Ar anchor
2546named "spam", and finally passes all outgoing connections and
2547incoming connections to port 25.
2548.Bd -literal -offset indent
2549# echo \&"block in quick from 1.2.3.4 to any\&" \&| \e
2550      pfctl -a spam -f -
2551.Ed
2552.Pp
2553This loads a single rule into the
2554.Ar anchor ,
2555which blocks all packets from a specific address.
2556.Pp
2557The anchor can also be populated by adding a
2558.Ar load anchor
2559rule after the
2560.Ar anchor
2561rule:
2562.Bd -literal -offset indent
2563anchor spam
2564load anchor spam from "/etc/pf-spam.conf"
2565.Ed
2566.Pp
2567When
2568.Xr pfctl 8
2569loads
2570.Nm pf.conf ,
2571it will also load all the rules from the file
2572.Pa /etc/pf-spam.conf
2573into the anchor.
2574.Pp
2575Optionally,
2576.Ar anchor
2577rules can specify packet filtering parameters using the same syntax as
2578filter rules.
2579When parameters are used, the
2580.Ar anchor
2581rule is only evaluated for matching packets.
2582This allows conditional evaluation of anchors, like:
2583.Bd -literal -offset indent
2584block on $ext_if all
2585anchor spam proto tcp from any to any port smtp
2586pass out on $ext_if all
2587pass in on $ext_if proto tcp from any to $ext_if port smtp
2588.Ed
2589.Pp
2590The rules inside
2591.Ar anchor
2592spam are only evaluated for
2593.Ar tcp
2594packets with destination port 25.
2595Hence,
2596.Bd -literal -offset indent
2597# echo \&"block in quick from 1.2.3.4 to any" \&| \e
2598      pfctl -a spam -f -
2599.Ed
2600.Pp
2601will only block connections from 1.2.3.4 to port 25.
2602.Pp
2603Anchors may end with the asterisk
2604.Pq Sq *
2605character, which signifies that all anchors attached at that point
2606should be evaluated in the alphabetical ordering of their anchor name.
2607For example,
2608.Bd -literal -offset indent
2609anchor "spam/*"
2610.Ed
2611.Pp
2612will evaluate each rule in each anchor attached to the
2613.Li spam
2614anchor.
2615Note that it will only evaluate anchors that are directly attached to the
2616.Li spam
2617anchor, and will not descend to evaluate anchors recursively.
2618.Pp
2619Since anchors are evaluated relative to the anchor in which they are
2620contained, there is a mechanism for accessing the parent and ancestor
2621anchors of a given anchor.
2622Similar to file system path name resolution, if the sequence
2623.Dq ..
2624appears as an anchor path component, the parent anchor of the current
2625anchor in the path evaluation at that point will become the new current
2626anchor.
2627As an example, consider the following:
2628.Bd -literal -offset indent
2629# echo ' anchor "spam/allowed" ' | pfctl -f -
2630# echo -e ' anchor "../banned" \en pass' | \e
2631      pfctl -a spam/allowed -f -
2632.Ed
2633.Pp
2634Evaluation of the main ruleset will lead into the
2635.Li spam/allowed
2636anchor, which will evaluate the rules in the
2637.Li spam/banned
2638anchor, if any, before finally evaluating the
2639.Ar pass
2640rule.
2641.Pp
2642Filter rule
2643.Ar anchors
2644can also be loaded inline in the ruleset within a brace ('{' '}') delimited
2645block.
2646Brace delimited blocks may contain rules or other brace-delimited blocks.
2647When anchors are loaded this way the anchor name becomes optional.
2648.Bd -literal -offset indent
2649anchor "external" on $ext_if {
2650	block
2651	anchor out {
2652		pass proto tcp from any to port { 25, 80, 443 }
2653	}
2654	pass in proto tcp to any port 22
2655}
2656.Ed
2657.Pp
2658Since the parser specification for anchor names is a string, any
2659reference to an anchor name containing
2660.Sq /
2661characters will require double quote
2662.Pq Sq \&"
2663characters around the anchor name.
2664.Sh TRANSLATION EXAMPLES
2665This example maps incoming requests on port 80 to port 8080, on
2666which a daemon is running (because, for example, it is not run as root,
2667and therefore lacks permission to bind to port 80).
2668.Bd -literal
2669# use a macro for the interface name, so it can be changed easily
2670ext_if = \&"ne3\&"
2671
2672# map daemon on 8080 to appear to be on 80
2673rdr on $ext_if proto tcp from any to any port 80 -\*(Gt 127.0.0.1 port 8080
2674.Ed
2675.Pp
2676If the
2677.Ar pass
2678modifier is given, packets matching the translation rule are passed without
2679inspecting the filter rules:
2680.Bd -literal
2681rdr pass on $ext_if proto tcp from any to any port 80 -\*(Gt 127.0.0.1 \e
2682      port 8080
2683.Ed
2684.Pp
2685In the example below, vlan12 is configured as 192.168.168.1;
2686the machine translates all packets coming from 192.168.168.0/24 to 204.92.77.111
2687when they are going out any interface except vlan12.
2688This has the net effect of making traffic from the 192.168.168.0/24
2689network appear as though it is the Internet routable address
2690204.92.77.111 to nodes behind any interface on the router except
2691for the nodes on vlan12.
2692(Thus, 192.168.168.1 can talk to the 192.168.168.0/24 nodes.)
2693.Bd -literal
2694nat on ! vlan12 from 192.168.168.0/24 to any -\*(Gt 204.92.77.111
2695.Ed
2696.Pp
2697In the example below, the machine sits between a fake internal 144.19.74.*
2698network, and a routable external IP of 204.92.77.100.
2699The
2700.Ar no nat
2701rule excludes protocol AH from being translated.
2702.Bd -literal
2703# NO NAT
2704no nat on $ext_if proto ah from 144.19.74.0/24 to any
2705nat on $ext_if from 144.19.74.0/24 to any -\*(Gt 204.92.77.100
2706.Ed
2707.Pp
2708In the example below, packets bound for one specific server, as well as those
2709generated by the sysadmins are not proxied; all other connections are.
2710.Bd -literal
2711# NO RDR
2712no rdr on $int_if proto { tcp, udp } from any to $server port 80
2713no rdr on $int_if proto { tcp, udp } from $sysadmins to any port 80
2714rdr on $int_if proto { tcp, udp } from any to any port 80 -\*(Gt 127.0.0.1 \e
2715      port 80
2716.Ed
2717.Pp
2718This longer example uses both a NAT and a redirection.
2719The external interface has the address 157.161.48.183.
2720On localhost, we are running
2721.Xr ftp-proxy 8 ,
2722waiting for FTP sessions to be redirected to it.
2723The three mandatory anchors for
2724.Xr ftp-proxy 8
2725are omitted from this example; see the
2726.Xr ftp-proxy 8
2727manpage.
2728.Bd -literal
2729# NAT
2730# Translate outgoing packets' source addresses (any protocol).
2731# In this case, any address but the gateway's external address is mapped.
2732nat on $ext_if inet from ! ($ext_if) to any -\*(Gt ($ext_if)
2733
2734# NAT PROXYING
2735# Map outgoing packets' source port to an assigned proxy port instead of
2736# an arbitrary port.
2737# In this case, proxy outgoing isakmp with port 500 on the gateway.
2738nat on $ext_if inet proto udp from any port = isakmp to any -\*(Gt ($ext_if) \e
2739      port 500
2740
2741# BINAT
2742# Translate outgoing packets' source address (any protocol).
2743# Translate incoming packets' destination address to an internal machine
2744# (bidirectional).
2745binat on $ext_if from 10.1.2.150 to any -\*(Gt $ext_if
2746
2747# RDR
2748# Translate incoming packets' destination addresses.
2749# As an example, redirect a TCP and UDP port to an internal machine.
2750rdr on $ext_if inet proto tcp from any to ($ext_if) port 8080 \e
2751      -\*(Gt 10.1.2.151 port 22
2752rdr on $ext_if inet proto udp from any to ($ext_if) port 8080 \e
2753      -\*(Gt 10.1.2.151 port 53
2754
2755# RDR
2756# Translate outgoing ftp control connections to send them to localhost
2757# for proxying with ftp-proxy(8) running on port 8021.
2758rdr on $int_if proto tcp from any to any port 21 -\*(Gt 127.0.0.1 port 8021
2759.Ed
2760.Pp
2761In this example, a NAT gateway is set up to translate internal addresses
2762using a pool of public addresses (192.0.2.16/28) and to redirect
2763incoming web server connections to a group of web servers on the internal
2764network.
2765.Bd -literal
2766# NAT LOAD BALANCE
2767# Translate outgoing packets' source addresses using an address pool.
2768# A given source address is always translated to the same pool address by
2769# using the source-hash keyword.
2770nat on $ext_if inet from any to any -\*(Gt 192.0.2.16/28 source-hash
2771
2772# RDR ROUND ROBIN
2773# Translate incoming web server connections to a group of web servers on
2774# the internal network.
2775rdr on $ext_if proto tcp from any to any port 80 \e
2776      -\*(Gt { 10.1.2.155, 10.1.2.160, 10.1.2.161 } round-robin
2777.Ed
2778.Sh FILTER EXAMPLES
2779.Bd -literal
2780# The external interface is kue0
2781# (157.161.48.183, the only routable address)
2782# and the private network is 10.0.0.0/8, for which we are doing NAT.
2783
2784# use a macro for the interface name, so it can be changed easily
2785ext_if = \&"kue0\&"
2786
2787# normalize all incoming traffic
2788scrub in on $ext_if all fragment reassemble
2789
2790# block and log everything by default
2791block return log on $ext_if all
2792
2793# block anything coming from source we have no back routes for
2794block in from no-route to any
2795
2796# block packets whose ingress interface does not match the one in
2797# the route back to their source address
2798block in from urpf-failed to any
2799
2800# block and log outgoing packets that do not have our address as source,
2801# they are either spoofed or something is misconfigured (NAT disabled,
2802# for instance), we want to be nice and do not send out garbage.
2803block out log quick on $ext_if from ! 157.161.48.183 to any
2804
2805# silently drop broadcasts (cable modem noise)
2806block in quick on $ext_if from any to 255.255.255.255
2807
2808# block and log incoming packets from reserved address space and invalid
2809# addresses, they are either spoofed or misconfigured, we cannot reply to
2810# them anyway (hence, no return-rst).
2811block in log quick on $ext_if from { 10.0.0.0/8, 172.16.0.0/12, \e
2812      192.168.0.0/16, 255.255.255.255/32 } to any
2813
2814# ICMP
2815
2816# pass out/in certain ICMP queries and keep state (ping)
2817# state matching is done on host addresses and ICMP id (not type/code),
2818# so replies (like 0/0 for 8/0) will match queries
2819# ICMP error messages (which always refer to a TCP/UDP packet) are
2820# handled by the TCP/UDP states
2821pass on $ext_if inet proto icmp all icmp-type 8 code 0
2822
2823# UDP
2824
2825# pass out all UDP connections and keep state
2826pass out on $ext_if proto udp all
2827
2828# pass in certain UDP connections and keep state (DNS)
2829pass in on $ext_if proto udp from any to any port domain
2830
2831# TCP
2832
2833# pass out all TCP connections and modulate state
2834pass out on $ext_if proto tcp all modulate state
2835
2836# pass in certain TCP connections and keep state (SSH, SMTP, DNS, IDENT)
2837pass in on $ext_if proto tcp from any to any port { ssh, smtp, domain, \e
2838      auth }
2839
2840# Do not allow Windows 9x SMTP connections since they are typically
2841# a viral worm. Alternately we could limit these OSes to 1 connection each.
2842block in on $ext_if proto tcp from any os {"Windows 95", "Windows 98"} \e
2843      to any port smtp
2844
2845# IPv6
2846# pass in/out all IPv6 traffic: note that we have to enable this in two
2847# different ways, on both our physical interface and our tunnel
2848pass quick on gif0 inet6
2849pass quick on $ext_if proto ipv6
2850
2851# Packet Tagging
2852
2853# three interfaces: $int_if, $ext_if, and $wifi_if (wireless). NAT is
2854# being done on $ext_if for all outgoing packets. tag packets in on
2855# $int_if and pass those tagged packets out on $ext_if.  all other
2856# outgoing packets (i.e., packets from the wireless network) are only
2857# permitted to access port 80.
2858
2859pass in on $int_if from any to any tag INTNET
2860pass in on $wifi_if from any to any
2861
2862block out on $ext_if from any to any
2863pass out quick on $ext_if tagged INTNET
2864pass out on $ext_if proto tcp from any to any port 80
2865
2866# tag incoming packets as they are redirected to spamd(8). use the tag
2867# to pass those packets through the packet filter.
2868
2869rdr on $ext_if inet proto tcp from \*(Ltspammers\*(Gt to port smtp \e
2870	tag SPAMD -\*(Gt 127.0.0.1 port spamd
2871
2872block in on $ext_if
2873pass in on $ext_if inet proto tcp tagged SPAMD
2874.Ed
2875.Sh GRAMMAR
2876Syntax for
2877.Nm
2878in BNF:
2879.Bd -literal
2880line           = ( option | pf-rule | nat-rule | binat-rule | rdr-rule |
2881                 antispoof-rule | altq-rule | queue-rule | trans-anchors |
2882                 anchor-rule | anchor-close | load-anchor | table-rule |
2883                 include )
2884
2885option         = "set" ( [ "timeout" ( timeout | "{" timeout-list "}" ) ] |
2886                 [ "ruleset-optimization" [ "none" | "basic" | "profile" ]] |
2887                 [ "optimization" [ "default" | "normal" |
2888                 "high-latency" | "satellite" |
2889                 "aggressive" | "conservative" ] ]
2890                 [ "limit" ( limit-item | "{" limit-list "}" ) ] |
2891                 [ "loginterface" ( interface-name | "none" ) ] |
2892                 [ "block-policy" ( "drop" | "return" ) ] |
2893                 [ "state-policy" ( "if-bound" | "floating" ) ]
2894                 [ "state-defaults" state-opts ]
2895                 [ "require-order" ( "yes" | "no" ) ]
2896                 [ "fingerprints" filename ] |
2897                 [ "skip on" ifspec ] |
2898                 [ "debug" ( "none" | "urgent" | "misc" | "loud" ) ]
2899                 [ "keepcounters" ] )
2900
2901pf-rule        = action [ ( "in" | "out" ) ]
2902                 [ "log" [ "(" logopts ")"] ] [ "quick" ]
2903                 [ "on" ifspec ] [ route ] [ af ] [ protospec ]
2904                 hosts [ filteropt-list ]
2905
2906logopts        = logopt [ "," logopts ]
2907logopt         = "all" | "user" | "to" interface-name
2908
2909filteropt-list = filteropt-list filteropt | filteropt
2910filteropt      = user | group | flags | icmp-type | icmp6-type | "tos" tos |
2911                 ( "no" | "keep" | "modulate" | "synproxy" ) "state"
2912                 [ "(" state-opts ")" ] |
2913                 "fragment" | "no-df" | "min-ttl" number | "set-tos" tos |
2914                 "max-mss" number | "random-id" | "reassemble tcp" |
2915                 fragmentation | "allow-opts" |
2916                 "label" string | "tag" string | [ ! ] "tagged" string |
2917                 "set prio" ( number | "(" number [ [ "," ] number ] ")" ) |
2918                 "queue" ( string | "(" string [ [ "," ] string ] ")" ) |
2919                 "rtable" number | "probability" number"%" | "prio" number
2920
2921nat-rule       = [ "no" ] "nat" [ "pass" [ "log" [ "(" logopts ")" ] ] ]
2922                 [ "on" ifspec ] [ af ]
2923                 [ protospec ] hosts [ "tag" string ] [ "tagged" string ]
2924                 [ "-\*(Gt" ( redirhost | "{" redirhost-list "}" )
2925                 [ portspec ] [ pooltype ] [ "static-port" ]
2926                 [ "map-e-portset" number "/" number "/" number ] ]
2927
2928binat-rule     = [ "no" ] "binat" [ "pass" [ "log" [ "(" logopts ")" ] ] ]
2929                 [ "on" interface-name ] [ af ]
2930                 [ "proto" ( proto-name | proto-number ) ]
2931                 "from" address [ "/" mask-bits ] "to" ipspec
2932                 [ "tag" string ] [ "tagged" string ]
2933                 [ "-\*(Gt" address [ "/" mask-bits ] ]
2934
2935rdr-rule       = [ "no" ] "rdr" [ "pass" [ "log" [ "(" logopts ")" ] ] ]
2936                 [ "on" ifspec ] [ af ]
2937                 [ protospec ] hosts [ "tag" string ] [ "tagged" string ]
2938                 [ "-\*(Gt" ( redirhost | "{" redirhost-list "}" )
2939                 [ portspec ] [ pooltype ] ]
2940
2941antispoof-rule = "antispoof" [ "log" ] [ "quick" ]
2942                 "for" ifspec [ af ] [ "label" string ]
2943
2944table-rule     = "table" "\*(Lt" string "\*(Gt" [ tableopts-list ]
2945tableopts-list = tableopts-list tableopts | tableopts
2946tableopts      = "persist" | "const" | "counters" | "file" string |
2947                 "{" [ tableaddr-list ] "}"
2948tableaddr-list = tableaddr-list [ "," ] tableaddr-spec | tableaddr-spec
2949tableaddr-spec = [ "!" ] tableaddr [ "/" mask-bits ]
2950tableaddr      = hostname | ifspec | "self" |
2951                 ipv4-dotted-quad | ipv6-coloned-hex
2952
2953altq-rule      = "altq on" interface-name queueopts-list
2954                 "queue" subqueue
2955queue-rule     = "queue" string [ "on" interface-name ] queueopts-list
2956                 subqueue
2957
2958anchor-rule    = "anchor" [ string ] [ ( "in" | "out" ) ] [ "on" ifspec ]
2959                 [ af ] [ protospec ] [ hosts ] [ filteropt-list ] [ "{" ]
2960
2961anchor-close   = "}"
2962
2963trans-anchors  = ( "nat-anchor" | "rdr-anchor" | "binat-anchor" ) string
2964                 [ "on" ifspec ] [ af ] [ "proto" ] [ protospec ] [ hosts ]
2965
2966load-anchor    = "load anchor" string "from" filename
2967
2968queueopts-list = queueopts-list queueopts | queueopts
2969queueopts      = [ "bandwidth" bandwidth-spec ] |
2970                 [ "qlimit" number ] | [ "tbrsize" number ] |
2971                 [ "priority" number ] | [ schedulers ]
2972schedulers     = ( cbq-def | priq-def | hfsc-def )
2973bandwidth-spec = "number" ( "b" | "Kb" | "Mb" | "Gb" | "%" )
2974
2975action         = "pass" | "block" [ return ] | [ "no" ] "scrub"
2976return         = "drop" | "return" | "return-rst" [ "( ttl" number ")" ] |
2977                 "return-icmp" [ "(" icmpcode [ [ "," ] icmp6code ] ")" ] |
2978                 "return-icmp6" [ "(" icmp6code ")" ]
2979icmpcode       = ( icmp-code-name | icmp-code-number )
2980icmp6code      = ( icmp6-code-name | icmp6-code-number )
2981
2982ifspec         = ( [ "!" ] ( interface-name | interface-group ) ) |
2983                 "{" interface-list "}"
2984interface-list = [ "!" ] ( interface-name | interface-group )
2985                 [ [ "," ] interface-list ]
2986route          = ( "route-to" | "reply-to" | "dup-to" )
2987                 ( routehost | "{" routehost-list "}" )
2988                 [ pooltype ]
2989af             = "inet" | "inet6"
2990
2991protospec      = "proto" ( proto-name | proto-number |
2992                 "{" proto-list "}" )
2993proto-list     = ( proto-name | proto-number ) [ [ "," ] proto-list ]
2994
2995hosts          = "all" |
2996                 "from" ( "any" | "no-route" | "urpf-failed" | "self" | host |
2997                 "{" host-list "}" ) [ port ] [ os ]
2998                 "to"   ( "any" | "no-route" | "self" | host |
2999                 "{" host-list "}" ) [ port ]
3000
3001ipspec         = "any" | host | "{" host-list "}"
3002host           = [ "!" ] ( address [ "/" mask-bits ] | "\*(Lt" string "\*(Gt" )
3003redirhost      = address [ "/" mask-bits ]
3004routehost      = "(" interface-name [ address [ "/" mask-bits ] ] ")"
3005address        = ( interface-name | interface-group |
3006                 "(" ( interface-name | interface-group ) ")" |
3007                 hostname | ipv4-dotted-quad | ipv6-coloned-hex )
3008host-list      = host [ [ "," ] host-list ]
3009redirhost-list = redirhost [ [ "," ] redirhost-list ]
3010routehost-list = routehost [ [ "," ] routehost-list ]
3011
3012port           = "port" ( unary-op | binary-op | "{" op-list "}" )
3013portspec       = "port" ( number | name ) [ ":" ( "*" | number | name ) ]
3014os             = "os"  ( os-name | "{" os-list "}" )
3015user           = "user" ( unary-op | binary-op | "{" op-list "}" )
3016group          = "group" ( unary-op | binary-op | "{" op-list "}" )
3017
3018unary-op       = [ "=" | "!=" | "\*(Lt" | "\*(Le" | "\*(Gt" | "\*(Ge" ]
3019                 ( name | number )
3020binary-op      = number ( "\*(Lt\*(Gt" | "\*(Gt\*(Lt" | ":" ) number
3021op-list        = ( unary-op | binary-op ) [ [ "," ] op-list ]
3022
3023os-name        = operating-system-name
3024os-list        = os-name [ [ "," ] os-list ]
3025
3026flags          = "flags" ( [ flag-set ] "/"  flag-set | "any" )
3027flag-set       = [ "F" ] [ "S" ] [ "R" ] [ "P" ] [ "A" ] [ "U" ] [ "E" ]
3028                 [ "W" ]
3029
3030icmp-type      = "icmp-type" ( icmp-type-code | "{" icmp-list "}" )
3031icmp6-type     = "icmp6-type" ( icmp-type-code | "{" icmp-list "}" )
3032icmp-type-code = ( icmp-type-name | icmp-type-number )
3033                 [ "code" ( icmp-code-name | icmp-code-number ) ]
3034icmp-list      = icmp-type-code [ [ "," ] icmp-list ]
3035
3036tos            = ( "lowdelay" | "throughput" | "reliability" |
3037                 [ "0x" ] number )
3038
3039state-opts     = state-opt [ [ "," ] state-opts ]
3040state-opt      = ( "max" number | "no-sync" | timeout | "sloppy" |
3041                 "source-track" [ ( "rule" | "global" ) ] |
3042                 "max-src-nodes" number | "max-src-states" number |
3043                 "max-src-conn" number |
3044                 "max-src-conn-rate" number "/" number |
3045                 "overload" "\*(Lt" string "\*(Gt" [ "flush" ] |
3046                 "if-bound" | "floating" )
3047
3048fragmentation  = [ "fragment reassemble" ]
3049
3050timeout-list   = timeout [ [ "," ] timeout-list ]
3051timeout        = ( "tcp.first" | "tcp.opening" | "tcp.established" |
3052                 "tcp.closing" | "tcp.finwait" | "tcp.closed" |
3053                 "udp.first" | "udp.single" | "udp.multiple" |
3054                 "icmp.first" | "icmp.error" |
3055                 "other.first" | "other.single" | "other.multiple" |
3056                 "frag" | "interval" | "src.track" |
3057                 "adaptive.start" | "adaptive.end" ) number
3058
3059limit-list     = limit-item [ [ "," ] limit-list ]
3060limit-item     = ( "states" | "frags" | "src-nodes" ) number
3061
3062pooltype       = ( "bitmask" | "random" |
3063                 "source-hash" [ ( hex-key | string-key ) ] |
3064                 "round-robin" ) [ sticky-address ]
3065
3066subqueue       = string | "{" queue-list "}"
3067queue-list     = string [ [ "," ] string ]
3068cbq-def        = "cbq" [ "(" cbq-opt [ [ "," ] cbq-opt ] ")" ]
3069priq-def       = "priq" [ "(" priq-opt [ [ "," ] priq-opt ] ")" ]
3070hfsc-def       = "hfsc" [ "(" hfsc-opt [ [ "," ] hfsc-opt ] ")" ]
3071cbq-opt        = ( "default" | "borrow" | "red" | "ecn" | "rio" )
3072priq-opt       = ( "default" | "red" | "ecn" | "rio" )
3073hfsc-opt       = ( "default" | "red" | "ecn" | "rio" |
3074                 linkshare-sc | realtime-sc | upperlimit-sc )
3075linkshare-sc   = "linkshare" sc-spec
3076realtime-sc    = "realtime" sc-spec
3077upperlimit-sc  = "upperlimit" sc-spec
3078sc-spec        = ( bandwidth-spec |
3079                 "(" bandwidth-spec number bandwidth-spec ")" )
3080include        = "include" filename
3081.Ed
3082.Sh FILES
3083.Bl -tag -width "/etc/protocols" -compact
3084.It Pa /etc/hosts
3085Host name database.
3086.It Pa /etc/pf.conf
3087Default location of the ruleset file.
3088The file has to be created manually as it is not installed with a
3089standard installation.
3090.It Pa /etc/pf.os
3091Default location of OS fingerprints.
3092.It Pa /etc/protocols
3093Protocol name database.
3094.It Pa /etc/services
3095Service name database.
3096.El
3097.Sh SEE ALSO
3098.Xr altq 4 ,
3099.Xr carp 4 ,
3100.Xr icmp 4 ,
3101.Xr icmp6 4 ,
3102.Xr ip 4 ,
3103.Xr ip6 4 ,
3104.Xr pf 4 ,
3105.Xr pfsync 4 ,
3106.Xr tcp 4 ,
3107.Xr udp 4 ,
3108.Xr hosts 5 ,
3109.Xr pf.os 5 ,
3110.Xr protocols 5 ,
3111.Xr services 5 ,
3112.Xr ftp-proxy 8 ,
3113.Xr pfctl 8 ,
3114.Xr pflogd 8
3115.Sh HISTORY
3116The
3117.Nm
3118file format first appeared in
3119.Ox 3.0 .
3120