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 July 25, 2015 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 . 192.El 193.Pp 194For example, 195.Bd -literal -offset indent 196table \*(Ltprivate\*(Gt const { 10/8, 172.16/12, 192.168/16 } 197table \*(Ltbadhosts\*(Gt persist 198block on fxp0 from { \*(Ltprivate\*(Gt, \*(Ltbadhosts\*(Gt } to any 199.Ed 200.Pp 201creates a table called private, to hold RFC 1918 private network 202blocks, and a table called badhosts, which is initially empty. 203A filter rule is set up to block all traffic coming from addresses listed in 204either table. 205The private table cannot have its contents changed and the badhosts table 206will exist even when no active filter rules reference it. 207Addresses may later be added to the badhosts table, so that traffic from 208these hosts can be blocked by using 209.Bd -literal -offset indent 210# pfctl -t badhosts -Tadd 204.92.77.111 211.Ed 212.Pp 213A table can also be initialized with an address list specified in one or more 214external files, using the following syntax: 215.Bd -literal -offset indent 216table \*(Ltspam\*(Gt persist file \&"/etc/spammers\&" file \&"/etc/openrelays\&" 217block on fxp0 from \*(Ltspam\*(Gt to any 218.Ed 219.Pp 220The files 221.Pa /etc/spammers 222and 223.Pa /etc/openrelays 224list IP addresses, one per line. 225Any lines beginning with a # are treated as comments and ignored. 226In addition to being specified by IP address, hosts may also be 227specified by their hostname. 228When the resolver is called to add a hostname to a table, 229.Em all 230resulting IPv4 and IPv6 addresses are placed into the table. 231IP addresses can also be entered in a table by specifying a valid interface 232name, a valid interface group or the 233.Em self 234keyword, in which case all addresses assigned to the interface(s) will be 235added to the table. 236.Sh OPTIONS 237.Xr pf 4 238may be tuned for various situations using the 239.Ar set 240command. 241.Bl -tag -width xxxx 242.It Ar set timeout 243.Pp 244.Bl -tag -width "src.track" -compact 245.It Ar interval 246Interval between purging expired states and fragments. 247.It Ar frag 248Seconds before an unassembled fragment is expired. 249.It Ar src.track 250Length of time to retain a source tracking entry after the last state 251expires. 252.El 253.Pp 254When a packet matches a stateful connection, the seconds to live for the 255connection will be updated to that of the 256.Ar proto.modifier 257which corresponds to the connection state. 258Each packet which matches this state will reset the TTL. 259Tuning these values may improve the performance of the 260firewall at the risk of dropping valid idle connections. 261.Pp 262.Bl -tag -width xxxx -compact 263.It Ar tcp.first 264The state after the first packet. 265.It Ar tcp.opening 266The state before the destination host ever sends a packet. 267.It Ar tcp.established 268The fully established state. 269.It Ar tcp.closing 270The state after the first FIN has been sent. 271.It Ar tcp.finwait 272The state after both FINs have been exchanged and the connection is closed. 273Some hosts (notably web servers on Solaris) send TCP packets even after closing 274the connection. 275Increasing 276.Ar tcp.finwait 277(and possibly 278.Ar tcp.closing ) 279can prevent blocking of such packets. 280.It Ar tcp.closed 281The state after one endpoint sends an RST. 282.El 283.Pp 284ICMP and UDP are handled in a fashion similar to TCP, but with a much more 285limited set of states: 286.Pp 287.Bl -tag -width xxxx -compact 288.It Ar udp.first 289The state after the first packet. 290.It Ar udp.single 291The state if the source host sends more than one packet but the destination 292host has never sent one back. 293.It Ar udp.multiple 294The state if both hosts have sent packets. 295.It Ar icmp.first 296The state after the first packet. 297.It Ar icmp.error 298The state after an ICMP error came back in response to an ICMP packet. 299.El 300.Pp 301Other protocols are handled similarly to UDP: 302.Pp 303.Bl -tag -width xxxx -compact 304.It Ar other.first 305.It Ar other.single 306.It Ar other.multiple 307.El 308.Pp 309Timeout values can be reduced adaptively as the number of state table 310entries grows. 311.Pp 312.Bl -tag -width xxxx -compact 313.It Ar adaptive.start 314When the number of state entries exceeds this value, adaptive scaling 315begins. 316All timeout values are scaled linearly with factor 317(adaptive.end - number of states) / (adaptive.end - adaptive.start). 318.It Ar adaptive.end 319When reaching this number of state entries, all timeout values become 320zero, effectively purging all state entries immediately. 321This value is used to define the scale factor, it should not actually 322be reached (set a lower state limit, see below). 323.El 324.Pp 325Adaptive timeouts are enabled by default, with an adaptive.start value 326equal to 60% of the state limit, and an adaptive.end value equal to 327120% of the state limit. 328They can be disabled by setting both adaptive.start and adaptive.end to 0. 329.Pp 330The adaptive timeout values can be defined both globally and for each rule. 331When used on a per-rule basis, the values relate to the number of 332states created by the rule, otherwise to the total number of 333states. 334.Pp 335For example: 336.Bd -literal -offset indent 337set timeout tcp.first 120 338set timeout tcp.established 86400 339set timeout { adaptive.start 6000, adaptive.end 12000 } 340set limit states 10000 341.Ed 342.Pp 343With 9000 state table entries, the timeout values are scaled to 50% 344(tcp.first 60, tcp.established 43200). 345.It Ar set loginterface 346Enable collection of packet and byte count statistics for the given 347interface or interface group. 348These statistics can be viewed using 349.Bd -literal -offset indent 350# pfctl -s info 351.Ed 352.Pp 353In this example 354.Xr pf 4 355collects statistics on the interface named dc0: 356.Bd -literal -offset indent 357set loginterface dc0 358.Ed 359.Pp 360One can disable the loginterface using: 361.Bd -literal -offset indent 362set loginterface none 363.Ed 364.It Ar set limit 365Sets hard limits on the memory pools used by the packet filter. 366See 367.Xr zone 9 368for an explanation of memory pools. 369.Pp 370For example, 371.Bd -literal -offset indent 372set limit states 20000 373.Ed 374.Pp 375sets the maximum number of entries in the memory pool used by state table 376entries (generated by 377.Ar pass 378rules which do not specify 379.Ar no state ) 380to 20000. 381Using 382.Bd -literal -offset indent 383set limit frags 20000 384.Ed 385.Pp 386sets the maximum number of entries in the memory pool used for fragment 387reassembly (generated by 388.Ar scrub 389rules) to 20000. 390Using 391.Bd -literal -offset indent 392set limit src-nodes 2000 393.Ed 394.Pp 395sets the maximum number of entries in the memory pool used for tracking 396source IP addresses (generated by the 397.Ar sticky-address 398and 399.Ar src.track 400options) to 2000. 401Using 402.Bd -literal -offset indent 403set limit tables 1000 404set limit table-entries 100000 405.Ed 406.Pp 407sets limits on the memory pools used by tables. 408The first limits the number of tables that can exist to 1000. 409The second limits the overall number of addresses that can be stored 410in tables to 100000. 411.Pp 412Various limits can be combined on a single line: 413.Bd -literal -offset indent 414set limit { states 20000, frags 20000, src-nodes 2000 } 415.Ed 416.It Ar set ruleset-optimization 417.Bl -tag -width xxxxxxxx -compact 418.It Ar none 419Disable the ruleset optimizer. 420.It Ar basic 421Enable basic ruleset optimization. 422This is the default behaviour. 423Basic ruleset optimization does four things to improve the 424performance of ruleset evaluations: 425.Pp 426.Bl -enum -compact 427.It 428remove duplicate rules 429.It 430remove rules that are a subset of another rule 431.It 432combine multiple rules into a table when advantageous 433.It 434re-order the rules to improve evaluation performance 435.El 436.Pp 437.It Ar profile 438Uses the currently loaded ruleset as a feedback profile to tailor the 439ordering of quick rules to actual network traffic. 440.El 441.Pp 442It is important to note that the ruleset optimizer will modify the ruleset 443to improve performance. 444A side effect of the ruleset modification is that per-rule accounting 445statistics will have different meanings than before. 446If per-rule accounting is important for billing purposes or whatnot, 447either the ruleset optimizer should not be used or a label field should 448be added to all of the accounting rules to act as optimization barriers. 449.Pp 450Optimization can also be set as a command-line argument to 451.Xr pfctl 8 , 452overriding the settings in 453.Nm . 454.It Ar set optimization 455Optimize state timeouts for one of the following network environments: 456.Pp 457.Bl -tag -width xxxx -compact 458.It Ar normal 459A normal network environment. 460Suitable for almost all networks. 461.It Ar high-latency 462A high-latency environment (such as a satellite connection). 463.It Ar satellite 464Alias for 465.Ar high-latency . 466.It Ar aggressive 467Aggressively expire connections. 468This can greatly reduce the memory usage of the firewall at the cost of 469dropping idle connections early. 470.It Ar conservative 471Extremely conservative settings. 472Avoid dropping legitimate connections at the 473expense of greater memory utilization (possibly much greater on a busy 474network) and slightly increased processor utilization. 475.El 476.Pp 477For example: 478.Bd -literal -offset indent 479set optimization aggressive 480.Ed 481.It Ar set block-policy 482The 483.Ar block-policy 484option sets the default behaviour for the packet 485.Ar block 486action: 487.Pp 488.Bl -tag -width xxxxxxxx -compact 489.It Ar drop 490Packet is silently dropped. 491.It Ar return 492A TCP RST is returned for blocked TCP packets, 493an ICMP UNREACHABLE is returned for blocked UDP packets, 494and all other packets are silently dropped. 495.El 496.Pp 497For example: 498.Bd -literal -offset indent 499set block-policy return 500.Ed 501.It Ar set state-policy 502The 503.Ar state-policy 504option sets the default behaviour for states: 505.Pp 506.Bl -tag -width group-bound -compact 507.It Ar if-bound 508States are bound to interface. 509.It Ar floating 510States can match packets on any interfaces (the default). 511.El 512.Pp 513For example: 514.Bd -literal -offset indent 515set state-policy if-bound 516.Ed 517.It Ar set state-defaults 518The 519.Ar state-defaults 520option sets the state options for states created from rules 521without an explicit 522.Ar keep state . 523For example: 524.Bd -literal -offset indent 525set state-defaults pflow, no-sync 526.Ed 527.It Ar set hostid 528The 32-bit 529.Ar hostid 530identifies this firewall's state table entries to other firewalls 531in a 532.Xr pfsync 4 533failover cluster. 534By default the hostid is set to a pseudo-random value, however it may be 535desirable to manually configure it, for example to more easily identify the 536source of state table entries. 537.Bd -literal -offset indent 538set hostid 1 539.Ed 540.Pp 541The hostid may be specified in either decimal or hexadecimal. 542.It Ar set require-order 543By default 544.Xr pfctl 8 545enforces an ordering of the statement types in the ruleset to: 546.Em options , 547.Em normalization , 548.Em queueing , 549.Em translation , 550.Em filtering . 551Setting this option to 552.Ar no 553disables this enforcement. 554There may be non-trivial and non-obvious implications to an out of 555order ruleset. 556Consider carefully before disabling the order enforcement. 557.It Ar set fingerprints 558Load fingerprints of known operating systems from the given filename. 559By default fingerprints of known operating systems are automatically 560loaded from 561.Xr pf.os 5 562in 563.Pa /etc 564but can be overridden via this option. 565Setting this option may leave a small period of time where the fingerprints 566referenced by the currently active ruleset are inconsistent until the new 567ruleset finishes loading. 568.Pp 569For example: 570.Pp 571.Dl set fingerprints \&"/etc/pf.os.devel\&" 572.It Ar set skip on Aq Ar ifspec 573List interfaces for which packets should not be filtered. 574Packets passing in or out on such interfaces are passed as if pf was 575disabled, i.e. pf does not process them in any way. 576This can be useful on loopback and other virtual interfaces, when 577packet filtering is not desired and can have unexpected effects. 578For example: 579.Pp 580.Dl set skip on lo0 581.It Ar set debug 582Set the debug 583.Ar level 584to one of the following: 585.Pp 586.Bl -tag -width xxxxxxxxxxxx -compact 587.It Ar none 588Don't generate debug messages. 589.It Ar urgent 590Generate debug messages only for serious errors. 591.It Ar misc 592Generate debug messages for various errors. 593.It Ar loud 594Generate debug messages for common conditions. 595.El 596.El 597.Sh TRAFFIC NORMALIZATION 598Traffic normalization is used to sanitize packet content in such 599a way that there are no ambiguities in packet interpretation on 600the receiving side. 601The normalizer does IP fragment reassembly to prevent attacks 602that confuse intrusion detection systems by sending overlapping 603IP fragments. 604Packet normalization is invoked with the 605.Ar scrub 606directive. 607.Pp 608.Ar scrub 609has the following options: 610.Bl -tag -width xxxx 611.It Ar no-df 612Clears the 613.Ar dont-fragment 614bit from a matching IP packet. 615Some operating systems are known to generate fragmented packets with the 616.Ar dont-fragment 617bit set. 618This is particularly true with NFS. 619.Ar Scrub 620will drop such fragmented 621.Ar dont-fragment 622packets unless 623.Ar no-df 624is specified. 625.Pp 626Unfortunately some operating systems also generate their 627.Ar dont-fragment 628packets with a zero IP identification field. 629Clearing the 630.Ar dont-fragment 631bit on packets with a zero IP ID may cause deleterious results if an 632upstream router later fragments the packet. 633Using the 634.Ar random-id 635modifier (see below) is recommended in combination with the 636.Ar no-df 637modifier to ensure unique IP identifiers. 638.It Ar min-ttl Aq Ar number 639Enforces a minimum TTL for matching IP packets. 640.It Ar max-mss Aq Ar number 641Enforces a maximum MSS for matching TCP packets. 642.It Xo Ar set-tos Aq Ar string 643.No \*(Ba Aq Ar number 644.Xc 645Enforces a 646.Em TOS 647for matching IP packets. 648.Em TOS 649may be 650given as one of 651.Ar lowdelay , 652.Ar throughput , 653.Ar reliability , 654or as either hex or decimal. 655.It Ar random-id 656Replaces the IP identification field with random values to compensate 657for predictable values generated by many hosts. 658This option only applies to packets that are not fragmented 659after the optional fragment reassembly. 660.It Ar fragment reassemble 661Using 662.Ar scrub 663rules, fragments can be reassembled by normalization. 664In this case, fragments are buffered until they form a complete 665packet, and only the completed packet is passed on to the filter. 666The advantage is that filter rules have to deal only with complete 667packets, and can ignore fragments. 668The drawback of caching fragments is the additional memory cost. 669.It Ar reassemble tcp 670Statefully normalizes TCP connections. 671.Ar scrub reassemble tcp 672rules may not have the direction (in/out) specified. 673.Ar reassemble tcp 674performs the following normalizations: 675.Pp 676.Bl -tag -width timeout -compact 677.It ttl 678Neither side of the connection is allowed to reduce their IP TTL. 679An attacker may send a packet such that it reaches the firewall, affects 680the firewall state, and expires before reaching the destination host. 681.Ar reassemble tcp 682will raise the TTL of all packets back up to the highest value seen on 683the connection. 684.It timestamp modulation 685Modern TCP stacks will send a timestamp on every TCP packet and echo 686the other endpoint's timestamp back to them. 687Many operating systems will merely start the timestamp at zero when 688first booted, and increment it several times a second. 689The uptime of the host can be deduced by reading the timestamp and multiplying 690by a constant. 691Also observing several different timestamps can be used to count hosts 692behind a NAT device. 693And spoofing TCP packets into a connection requires knowing or guessing 694valid timestamps. 695Timestamps merely need to be monotonically increasing and not derived off a 696guessable base time. 697.Ar reassemble tcp 698will cause 699.Ar scrub 700to modulate the TCP timestamps with a random number. 701.It extended PAWS checks 702There is a problem with TCP on long fat pipes, in that a packet might get 703delayed for longer than it takes the connection to wrap its 32-bit sequence 704space. 705In such an occurrence, the old packet would be indistinguishable from a 706new packet and would be accepted as such. 707The solution to this is called PAWS: Protection Against Wrapped Sequence 708numbers. 709It protects against it by making sure the timestamp on each packet does 710not go backwards. 711.Ar reassemble tcp 712also makes sure the timestamp on the packet does not go forward more 713than the RFC allows. 714By doing this, 715.Xr pf 4 716artificially extends the security of TCP sequence numbers by 10 to 18 717bits when the host uses appropriately randomized timestamps, since a 718blind attacker would have to guess the timestamp as well. 719.El 720.El 721.Pp 722For example, 723.Bd -literal -offset indent 724scrub in on $ext_if all fragment reassemble 725.Ed 726.Pp 727The 728.Ar no 729option prefixed to a scrub rule causes matching packets to remain unscrubbed, 730much in the same way as 731.Ar drop quick 732works in the packet filter (see below). 733This mechanism should be used when it is necessary to exclude specific packets 734from broader scrub rules. 735.Sh QUEUEING 736The ALTQ system is currently not available in the GENERIC kernel nor as 737loadable modules. 738In order to use the herein after called queueing options one has to use a 739custom built kernel. 740Please refer to 741.Xr altq 4 742to learn about the related kernel options. 743.Pp 744Packets can be assigned to queues for the purpose of bandwidth 745control. 746At least two declarations are required to configure queues, and later 747any packet filtering rule can reference the defined queues by name. 748During the filtering component of 749.Nm pf.conf , 750the last referenced 751.Ar queue 752name is where any packets from 753.Ar pass 754rules will be queued, while for 755.Ar block 756rules it specifies where any resulting ICMP or TCP RST 757packets should be queued. 758The 759.Ar scheduler 760defines the algorithm used to decide which packets get delayed, dropped, or 761sent out immediately. 762There are three 763.Ar schedulers 764currently supported. 765.Bl -tag -width xxxx 766.It Ar cbq 767Class Based Queueing. 768.Ar Queues 769attached to an interface build a tree, thus each 770.Ar queue 771can have further child 772.Ar queues . 773Each queue can have a 774.Ar priority 775and a 776.Ar bandwidth 777assigned. 778.Ar Priority 779mainly controls the time packets take to get sent out, while 780.Ar bandwidth 781has primarily effects on throughput. 782.Ar cbq 783achieves both partitioning and sharing of link bandwidth 784by hierarchically structured classes. 785Each class has its own 786.Ar queue 787and is assigned its share of 788.Ar bandwidth . 789A child class can borrow bandwidth from its parent class 790as long as excess bandwidth is available 791(see the option 792.Ar borrow , 793below). 794.It Ar priq 795Priority Queueing. 796.Ar Queues 797are flat attached to the interface, thus, 798.Ar queues 799cannot have further child 800.Ar queues . 801Each 802.Ar queue 803has a unique 804.Ar priority 805assigned, ranging from 0 to 15. 806Packets in the 807.Ar queue 808with the highest 809.Ar priority 810are processed first. 811.It Ar hfsc 812Hierarchical Fair Service Curve. 813.Ar Queues 814attached to an interface build a tree, thus each 815.Ar queue 816can have further child 817.Ar queues . 818Each queue can have a 819.Ar priority 820and a 821.Ar bandwidth 822assigned. 823.Ar Priority 824mainly controls the time packets take to get sent out, while 825.Ar bandwidth 826primarily affects throughput. 827.Ar hfsc 828supports both link-sharing and guaranteed real-time services. 829It employs a service curve based QoS model, 830and its unique feature is an ability to decouple 831.Ar delay 832and 833.Ar bandwidth 834allocation. 835.El 836.Pp 837The interfaces on which queueing should be activated are declared using 838the 839.Ar altq on 840declaration. 841.Ar altq on 842has the following keywords: 843.Bl -tag -width xxxx 844.It Aq Ar interface 845Queueing is enabled on the named interface. 846.It Aq Ar scheduler 847Specifies which queueing scheduler to use. 848Currently supported values 849are 850.Ar cbq 851for Class Based Queueing, 852.Ar priq 853for Priority Queueing and 854.Ar hfsc 855for the Hierarchical Fair Service Curve scheduler. 856.It Ar bandwidth Aq Ar bw 857The maximum bitrate for all queues on an 858interface may be specified using the 859.Ar bandwidth 860keyword. 861The value can be specified as an absolute value or as a 862percentage of the interface bandwidth. 863When using an absolute value, the suffixes 864.Ar b , 865.Ar Kb , 866.Ar Mb , 867and 868.Ar Gb 869are used to represent bits, kilobits, megabits, and 870gigabits per second, respectively. 871The value must not exceed the interface bandwidth. 872If 873.Ar bandwidth 874is not specified, the interface bandwidth is used 875(but take note that some interfaces do not know their bandwidth, 876or can adapt their bandwidth rates). 877.It Ar qlimit Aq Ar limit 878The maximum number of packets held in the queue. 879The default is 50. 880.It Ar tbrsize Aq Ar size 881Adjusts the size, in bytes, of the token bucket regulator. 882If not specified, heuristics based on the 883interface bandwidth are used to determine the size. 884.It Ar queue Aq Ar list 885Defines a list of subqueues to create on an interface. 886.El 887.Pp 888In the following example, the interface dc0 889should queue up to 5Mbps in four second-level queues using 890Class Based Queueing. 891Those four queues will be shown in a later example. 892.Bd -literal -offset indent 893altq on dc0 cbq bandwidth 5Mb queue { std, http, mail, ssh } 894.Ed 895.Pp 896Once interfaces are activated for queueing using the 897.Ar altq 898directive, a sequence of 899.Ar queue 900directives may be defined. 901The name associated with a 902.Ar queue 903must match a queue defined in the 904.Ar altq 905directive (e.g. mail), or, except for the 906.Ar priq 907.Ar scheduler , 908in a parent 909.Ar queue 910declaration. 911The following keywords can be used: 912.Bl -tag -width xxxx 913.It Ar on Aq Ar interface 914Specifies the interface the queue operates on. 915If not given, it operates on all matching interfaces. 916.It Ar bandwidth Aq Ar bw 917Specifies the maximum bitrate to be processed by the queue. 918This value must not exceed the value of the parent 919.Ar queue 920and can be specified as an absolute value or a percentage of the parent 921queue's bandwidth. 922If not specified, defaults to 100% of the parent queue's bandwidth. 923The 924.Ar priq 925scheduler does not support bandwidth specification. 926.It Ar priority Aq Ar level 927Between queues a priority level can be set. 928For 929.Ar cbq 930and 931.Ar hfsc , 932the range is 0 to 7 and for 933.Ar priq , 934the range is 0 to 15. 935The default for all is 1. 936.Ar Priq 937queues with a higher priority are always served first. 938.Ar Cbq 939and 940.Ar Hfsc 941queues with a higher priority are preferred in the case of overload. 942.It Ar qlimit Aq Ar limit 943The maximum number of packets held in the queue. 944The default is 50. 945.El 946.Pp 947The 948.Ar scheduler 949can get additional parameters with 950.Xo Aq Ar scheduler 951.Pf ( Aq Ar parameters ) . 952.Xc 953Parameters are as follows: 954.Bl -tag -width Fl 955.It Ar default 956Packets not matched by another queue are assigned to this one. 957Exactly one default queue is required. 958.It Ar red 959Enable RED (Random Early Detection) on this queue. 960RED drops packets with a probability proportional to the average 961queue length. 962.It Ar rio 963Enables RIO on this queue. 964RIO is RED with IN/OUT, thus running 965RED two times more than RIO would achieve the same effect. 966RIO is currently not supported in the GENERIC kernel. 967.It Ar ecn 968Enables ECN (Explicit Congestion Notification) on this queue. 969ECN implies RED. 970.El 971.Pp 972The 973.Ar cbq 974.Ar scheduler 975supports an additional option: 976.Bl -tag -width Fl 977.It Ar borrow 978The queue can borrow bandwidth from the parent. 979.El 980.Pp 981The 982.Ar hfsc 983.Ar scheduler 984supports some additional options: 985.Bl -tag -width Fl 986.It Ar realtime Aq Ar sc 987The minimum required bandwidth for the queue. 988.It Ar upperlimit Aq Ar sc 989The maximum allowed bandwidth for the queue. 990.It Ar linkshare Aq Ar sc 991The bandwidth share of a backlogged queue. 992.El 993.Pp 994.Aq Ar sc 995is an acronym for 996.Ar service curve . 997.Pp 998The format for service curve specifications is 999.Ar ( m1 , d , m2 ) . 1000.Ar m2 1001controls the bandwidth assigned to the queue. 1002.Ar m1 1003and 1004.Ar d 1005are optional and can be used to control the initial bandwidth assignment. 1006For the first 1007.Ar d 1008milliseconds the queue gets the bandwidth given as 1009.Ar m1 , 1010afterwards the value given in 1011.Ar m2 . 1012.Pp 1013Furthermore, with 1014.Ar cbq 1015and 1016.Ar hfsc , 1017child queues can be specified as in an 1018.Ar altq 1019declaration, thus building a tree of queues using a part of 1020their parent's bandwidth. 1021.Pp 1022Packets can be assigned to queues based on filter rules by using the 1023.Ar queue 1024keyword. 1025Normally only one 1026.Ar queue 1027is specified; when a second one is specified it will instead be used for 1028packets which have a 1029.Em TOS 1030of 1031.Em lowdelay 1032and for TCP ACKs with no data payload. 1033.Pp 1034To continue the previous example, the examples below would specify the 1035four referenced 1036queues, plus a few child queues. 1037Interactive 1038.Xr ssh 1 1039sessions get priority over bulk transfers like 1040.Xr scp 1 1041and 1042.Xr sftp 1 . 1043The queues may then be referenced by filtering rules (see 1044.Sx PACKET FILTERING 1045below). 1046.Bd -literal 1047queue std bandwidth 10% cbq(default) 1048queue http bandwidth 60% priority 2 cbq(borrow red) \e 1049 { employees, developers } 1050queue developers bandwidth 75% cbq(borrow) 1051queue employees bandwidth 15% 1052queue mail bandwidth 10% priority 0 cbq(borrow ecn) 1053queue ssh bandwidth 20% cbq(borrow) { ssh_interactive, ssh_bulk } 1054queue ssh_interactive bandwidth 50% priority 7 cbq(borrow) 1055queue ssh_bulk bandwidth 50% priority 0 cbq(borrow) 1056 1057block return out on dc0 inet all queue std 1058pass out on dc0 inet proto tcp from $developerhosts to any port 80 \e 1059 queue developers 1060pass out on dc0 inet proto tcp from $employeehosts to any port 80 \e 1061 queue employees 1062pass out on dc0 inet proto tcp from any to any port 22 \e 1063 queue(ssh_bulk, ssh_interactive) 1064pass out on dc0 inet proto tcp from any to any port 25 \e 1065 queue mail 1066.Ed 1067.Sh TRANSLATION 1068Translation rules modify either the source or destination address of the 1069packets associated with a stateful connection. 1070A stateful connection is automatically created to track packets matching 1071such a rule as long as they are not blocked by the filtering section of 1072.Nm pf.conf . 1073The translation engine modifies the specified address and/or port in the 1074packet, recalculates IP, TCP and UDP checksums as necessary, and passes it to 1075the packet filter for evaluation. 1076.Pp 1077Since translation occurs before filtering the filter 1078engine will see packets as they look after any 1079addresses and ports have been translated. 1080Filter rules will therefore have to filter based on the translated 1081address and port number. 1082Packets that match a translation rule are only automatically passed if 1083the 1084.Ar pass 1085modifier is given, otherwise they are 1086still subject to 1087.Ar block 1088and 1089.Ar pass 1090rules. 1091.Pp 1092The state entry created permits 1093.Xr pf 4 1094to keep track of the original address for traffic associated with that state 1095and correctly direct return traffic for that connection. 1096.Pp 1097Various types of translation are possible with pf: 1098.Bl -tag -width xxxx 1099.It Ar binat 1100A 1101.Ar binat 1102rule specifies a bidirectional mapping between an external IP netblock 1103and an internal IP netblock. 1104.It Ar nat 1105A 1106.Ar nat 1107rule specifies that IP addresses are to be changed as the packet 1108traverses the given interface. 1109This technique allows one or more IP addresses 1110on the translating host to support network traffic for a larger range of 1111machines on an "inside" network. 1112Although in theory any IP address can be used on the inside, it is strongly 1113recommended that one of the address ranges defined by RFC 1918 be used. 1114These netblocks are: 1115.Bd -literal 111610.0.0.0 - 10.255.255.255 (all of net 10, i.e., 10/8) 1117172.16.0.0 - 172.31.255.255 (i.e., 172.16/12) 1118192.168.0.0 - 192.168.255.255 (i.e., 192.168/16) 1119.Ed 1120.It Pa rdr 1121The packet is redirected to another destination and possibly a 1122different port. 1123.Ar rdr 1124rules can optionally specify port ranges instead of single ports. 1125rdr ... port 2000:2999 -\*(Gt ... port 4000 1126redirects ports 2000 to 2999 (inclusive) to port 4000. 1127rdr ... port 2000:2999 -\*(Gt ... port 4000:* 1128redirects port 2000 to 4000, 2001 to 4001, ..., 2999 to 4999. 1129.El 1130.Pp 1131In addition to modifying the address, some translation rules may modify 1132source or destination ports for 1133.Xr tcp 4 1134or 1135.Xr udp 4 1136connections; implicitly in the case of 1137.Ar nat 1138rules and explicitly in the case of 1139.Ar rdr 1140rules. 1141Port numbers are never translated with a 1142.Ar binat 1143rule. 1144.Pp 1145Evaluation order of the translation rules is dependent on the type 1146of the translation rules and of the direction of a packet. 1147.Ar binat 1148rules are always evaluated first. 1149Then either the 1150.Ar rdr 1151rules are evaluated on an inbound packet or the 1152.Ar nat 1153rules on an outbound packet. 1154Rules of the same type are evaluated in the same order in which they 1155appear in the ruleset. 1156The first matching rule decides what action is taken. 1157.Pp 1158The 1159.Ar no 1160option prefixed to a translation rule causes packets to remain untranslated, 1161much in the same way as 1162.Ar drop quick 1163works in the packet filter (see below). 1164If no rule matches the packet it is passed to the filter engine unmodified. 1165.Pp 1166Translation rules apply only to packets that pass through 1167the specified interface, and if no interface is specified, 1168translation is applied to packets on all interfaces. 1169For instance, redirecting port 80 on an external interface to an internal 1170web server will only work for connections originating from the outside. 1171Connections to the address of the external interface from local hosts will 1172not be redirected, since such packets do not actually pass through the 1173external interface. 1174Redirections cannot reflect packets back through the interface they arrive 1175on, they can only be redirected to hosts connected to different interfaces 1176or to the firewall itself. 1177.Pp 1178Note that redirecting external incoming connections to the loopback 1179address, as in 1180.Bd -literal -offset indent 1181rdr on ne3 inet proto tcp to port smtp -\*(Gt 127.0.0.1 port spamd 1182.Ed 1183.Pp 1184will effectively allow an external host to connect to daemons 1185bound solely to the loopback address, circumventing the traditional 1186blocking of such connections on a real interface. 1187Unless this effect is desired, any of the local non-loopback addresses 1188should be used as redirection target instead, which allows external 1189connections only to daemons bound to this address or not bound to 1190any address. 1191.Pp 1192See 1193.Sx TRANSLATION EXAMPLES 1194below. 1195.Sh PACKET FILTERING 1196.Xr pf 4 1197has the ability to 1198.Ar block 1199and 1200.Ar pass 1201packets based on attributes of their layer 3 (see 1202.Xr ip 4 1203and 1204.Xr ip6 4 ) 1205and layer 4 (see 1206.Xr icmp 4 , 1207.Xr icmp6 4 , 1208.Xr tcp 4 , 1209.Xr udp 4 ) 1210headers. 1211In addition, packets may also be 1212assigned to queues for the purpose of bandwidth control. 1213.Pp 1214For each packet processed by the packet filter, the filter rules are 1215evaluated in sequential order, from first to last. 1216The last matching rule decides what action is taken. 1217If no rule matches the packet, the default action is to pass 1218the packet. 1219.Pp 1220The following actions can be used in the filter: 1221.Bl -tag -width xxxx 1222.It Ar block 1223The packet is blocked. 1224There are a number of ways in which a 1225.Ar block 1226rule can behave when blocking a packet. 1227The default behaviour is to 1228.Ar drop 1229packets silently, however this can be overridden or made 1230explicit either globally, by setting the 1231.Ar block-policy 1232option, or on a per-rule basis with one of the following options: 1233.Pp 1234.Bl -tag -width xxxx -compact 1235.It Ar drop 1236The packet is silently dropped. 1237.It Ar return-rst 1238This applies only to 1239.Xr tcp 4 1240packets, and issues a TCP RST which closes the 1241connection. 1242.It Ar return-icmp 1243.It Ar return-icmp6 1244This causes ICMP messages to be returned for packets which match the rule. 1245By default this is an ICMP UNREACHABLE message, however this 1246can be overridden by specifying a message as a code or number. 1247.It Ar return 1248This causes a TCP RST to be returned for 1249.Xr tcp 4 1250packets and an ICMP UNREACHABLE for UDP and other packets. 1251.El 1252.Pp 1253Options returning ICMP packets currently have no effect if 1254.Xr pf 4 1255operates on a 1256.Xr if_bridge 4 , 1257as the code to support this feature has not yet been implemented. 1258.Pp 1259The simplest mechanism to block everything by default and only pass 1260packets that match explicit rules is specify a first filter rule of: 1261.Bd -literal -offset indent 1262block all 1263.Ed 1264.It Ar pass 1265The packet is passed; 1266state is created unless the 1267.Ar no state 1268option is specified. 1269.El 1270.Pp 1271By default 1272.Xr pf 4 1273filters packets statefully; the first time a packet matches a 1274.Ar pass 1275rule, a state entry is created; for subsequent packets the filter checks 1276whether the packet matches any state. 1277If it does, the packet is passed without evaluation of any rules. 1278After the connection is closed or times out, the state entry is automatically 1279removed. 1280.Pp 1281This has several advantages. 1282For TCP connections, comparing a packet to a state involves checking 1283its sequence numbers, as well as TCP timestamps if a 1284.Ar scrub reassemble tcp 1285rule applies to the connection. 1286If these values are outside the narrow windows of expected 1287values, the packet is dropped. 1288This prevents spoofing attacks, such as when an attacker sends packets with 1289a fake source address/port but does not know the connection's sequence 1290numbers. 1291Similarly, 1292.Xr pf 4 1293knows how to match ICMP replies to states. 1294For example, 1295.Bd -literal -offset indent 1296pass out inet proto icmp all icmp-type echoreq 1297.Ed 1298.Pp 1299allows echo requests (such as those created by 1300.Xr ping 8 ) 1301out statefully, and matches incoming echo replies correctly to states. 1302.Pp 1303Also, looking up states is usually faster than evaluating rules. 1304If there are 50 rules, all of them are evaluated sequentially in O(n). 1305Even with 50000 states, only 16 comparisons are needed to match a 1306state, since states are stored in a binary search tree that allows 1307searches in O(log2 n). 1308.Pp 1309Furthermore, correct handling of ICMP error messages is critical to 1310many protocols, particularly TCP. 1311.Xr pf 4 1312matches ICMP error messages to the correct connection, checks them against 1313connection parameters, and passes them if appropriate. 1314For example if an ICMP source quench message referring to a stateful TCP 1315connection arrives, it will be matched to the state and get passed. 1316.Pp 1317Finally, state tracking is required for 1318.Ar nat , binat No and Ar rdr 1319rules, in order to track address and port translations and reverse the 1320translation on returning packets. 1321.Pp 1322.Xr pf 4 1323will also create state for other protocols which are effectively stateless by 1324nature. 1325UDP packets are matched to states using only host addresses and ports, 1326and other protocols are matched to states using only the host addresses. 1327.Pp 1328If stateless filtering of individual packets is desired, 1329the 1330.Ar no state 1331keyword can be used to specify that state will not be created 1332if this is the last matching rule. 1333A number of parameters can also be set to affect how 1334.Xr pf 4 1335handles state tracking. 1336See 1337.Sx STATEFUL TRACKING OPTIONS 1338below for further details. 1339.Sh PARAMETERS 1340The rule parameters specify the packets to which a rule applies. 1341A packet always comes in on, or goes out through, one interface. 1342Most parameters are optional. 1343If a parameter is specified, the rule only applies to packets with 1344matching attributes. 1345Certain parameters can be expressed as lists, in which case 1346.Xr pfctl 8 1347generates all needed rule combinations. 1348.Bl -tag -width xxxx 1349.It Ar in No or Ar out 1350This rule applies to incoming or outgoing packets. 1351If neither 1352.Ar in 1353nor 1354.Ar out 1355are specified, the rule will match packets in both directions. 1356.It Ar log 1357In addition to the action specified, a log message is generated. 1358Only the packet that establishes the state is logged, 1359unless the 1360.Ar no state 1361option is specified. 1362The logged packets are sent to a 1363.Xr pflog 4 1364interface, by default 1365.Ar pflog0 . 1366This interface is monitored by the 1367.Xr pflogd 8 1368logging daemon, which dumps the logged packets to the file 1369.Pa /var/log/pflog 1370in 1371.Xr pcap 3 1372binary format. 1373.It Ar log (all) 1374Used to force logging of all packets for a connection. 1375This is not necessary when 1376.Ar no state 1377is explicitly specified. 1378As with 1379.Ar log , 1380packets are logged to 1381.Xr pflog 4 . 1382.It Ar log (user) 1383Logs the 1384.Ux 1385user ID of the user that owns the socket and the PID of the process that 1386has the socket open where the packet is sourced from or destined to 1387(depending on which socket is local). 1388This is in addition to the normal information logged. 1389.Pp 1390Only the first packet 1391logged via 1392.Ar log (all, user) 1393will have the user credentials logged when using stateful matching. 1394.It Ar log (to Aq Ar interface ) 1395Send logs to the specified 1396.Xr pflog 4 1397interface instead of 1398.Ar pflog0 . 1399.It Ar quick 1400If a packet matches a rule which has the 1401.Ar quick 1402option set, this rule 1403is considered the last matching rule, and evaluation of subsequent rules 1404is skipped. 1405.It Ar on Aq Ar interface 1406This rule applies only to packets coming in on, or going out through, this 1407particular interface or interface group. 1408For more information on interface groups, 1409see the 1410.Ic group 1411keyword in 1412.Xr ifconfig 8 . 1413.It Aq Ar af 1414This rule applies only to packets of this address family. 1415Supported values are 1416.Ar inet 1417and 1418.Ar inet6 . 1419.It Ar proto Aq Ar protocol 1420This rule applies only to packets of this protocol. 1421Common protocols are 1422.Xr icmp 4 , 1423.Xr icmp6 4 , 1424.Xr tcp 4 , 1425and 1426.Xr udp 4 . 1427For a list of all the protocol name to number mappings used by 1428.Xr pfctl 8 , 1429see the file 1430.Pa /etc/protocols . 1431.It Xo 1432.Ar from Aq Ar source 1433.Ar port Aq Ar source 1434.Ar os Aq Ar source 1435.Ar to Aq Ar dest 1436.Ar port Aq Ar dest 1437.Xc 1438This rule applies only to packets with the specified source and destination 1439addresses and ports. 1440.Pp 1441Addresses can be specified in CIDR notation (matching netblocks), as 1442symbolic host names, interface names or interface group names, or as any 1443of the following keywords: 1444.Pp 1445.Bl -tag -width xxxxxxxxxxxxxx -compact 1446.It Ar any 1447Any address. 1448.It Ar no-route 1449Any address which is not currently routable. 1450.It Ar urpf-failed 1451Any source address that fails a unicast reverse path forwarding (URPF) 1452check, i.e. packets coming in on an interface other than that which holds 1453the route back to the packet's source address. 1454.It Aq Ar table 1455Any address that matches the given table. 1456.El 1457.Pp 1458Ranges of addresses are specified by using the 1459.Sq - 1460operator. 1461For instance: 1462.Dq 10.1.1.10 - 10.1.1.12 1463means all addresses from 10.1.1.10 to 10.1.1.12, 1464hence addresses 10.1.1.10, 10.1.1.11, and 10.1.1.12. 1465.Pp 1466Interface names and interface group names can have modifiers appended: 1467.Pp 1468.Bl -tag -width xxxxxxxxxxxx -compact 1469.It Ar :network 1470Translates to the network(s) attached to the interface. 1471.It Ar :broadcast 1472Translates to the interface's broadcast address(es). 1473.It Ar :peer 1474Translates to the point-to-point interface's peer address(es). 1475.It Ar :0 1476Do not include interface aliases. 1477.El 1478.Pp 1479Host names may also have the 1480.Ar :0 1481option appended to restrict the name resolution to the first of each 1482v4 and v6 address found. 1483.Pp 1484Host name resolution and interface to address translation are done at 1485ruleset load-time. 1486When the address of an interface (or host name) changes (under DHCP or PPP, 1487for instance), the ruleset must be reloaded for the change to be reflected 1488in the kernel. 1489Surrounding the interface name (and optional modifiers) in parentheses 1490changes this behaviour. 1491When the interface name is surrounded by parentheses, the rule is 1492automatically updated whenever the interface changes its address. 1493The ruleset does not need to be reloaded. 1494This is especially useful with 1495.Ar nat . 1496.Pp 1497Ports can be specified either by number or by name. 1498For example, port 80 can be specified as 1499.Em www . 1500For a list of all port name to number mappings used by 1501.Xr pfctl 8 , 1502see the file 1503.Pa /etc/services . 1504.Pp 1505Ports and ranges of ports are specified by using these operators: 1506.Bd -literal -offset indent 1507= (equal) 1508!= (unequal) 1509\*(Lt (less than) 1510\*(Le (less than or equal) 1511\*(Gt (greater than) 1512\*(Ge (greater than or equal) 1513: (range including boundaries) 1514\*(Gt\*(Lt (range excluding boundaries) 1515\*(Lt\*(Gt (except range) 1516.Ed 1517.Pp 1518.Sq \*(Gt\*(Lt , 1519.Sq \*(Lt\*(Gt 1520and 1521.Sq \&: 1522are binary operators (they take two arguments). 1523For instance: 1524.Bl -tag -width Fl 1525.It Ar port 2000:2004 1526means 1527.Sq all ports \*(Ge 2000 and \*(Le 2004 , 1528hence ports 2000, 2001, 2002, 2003 and 2004. 1529.It Ar port 2000 \*(Gt\*(Lt 2004 1530means 1531.Sq all ports \*(Gt 2000 and \*(Lt 2004 , 1532hence ports 2001, 2002 and 2003. 1533.It Ar port 2000 \*(Lt\*(Gt 2004 1534means 1535.Sq all ports \*(Lt 2000 or \*(Gt 2004 , 1536hence ports 1-1999 and 2005-65535. 1537.El 1538.Pp 1539The operating system of the source host can be specified in the case of TCP 1540rules with the 1541.Ar OS 1542modifier. 1543See the 1544.Sx OPERATING SYSTEM FINGERPRINTING 1545section for more information. 1546.Pp 1547The host, port and OS specifications are optional, as in the following examples: 1548.Bd -literal -offset indent 1549pass in all 1550pass in from any to any 1551pass in proto tcp from any port \*(Le 1024 to any 1552pass in proto tcp from any to any port 25 1553pass in proto tcp from 10.0.0.0/8 port \*(Gt 1024 \e 1554 to ! 10.1.2.3 port != ssh 1555pass in proto tcp from any os "OpenBSD" 1556.Ed 1557.It Ar all 1558This is equivalent to "from any to any". 1559.It Ar group Aq Ar group 1560Similar to 1561.Ar user , 1562this rule only applies to packets of sockets owned by the specified group. 1563.It Ar user Aq Ar user 1564This rule only applies to packets of sockets owned by the specified user. 1565For outgoing connections initiated from the firewall, this is the user 1566that opened the connection. 1567For incoming connections to the firewall itself, this is the user that 1568listens on the destination port. 1569For forwarded connections, where the firewall is not a connection endpoint, 1570the user and group are 1571.Em unknown . 1572.Pp 1573All packets, both outgoing and incoming, of one connection are associated 1574with the same user and group. 1575Only TCP and UDP packets can be associated with users; for other protocols 1576these parameters are ignored. 1577.Pp 1578User and group refer to the effective (as opposed to the real) IDs, in 1579case the socket is created by a setuid/setgid process. 1580User and group IDs are stored when a socket is created; 1581when a process creates a listening socket as root (for instance, by 1582binding to a privileged port) and subsequently changes to another 1583user ID (to drop privileges), the credentials will remain root. 1584.Pp 1585User and group IDs can be specified as either numbers or names. 1586The syntax is similar to the one for ports. 1587The value 1588.Em unknown 1589matches packets of forwarded connections. 1590.Em unknown 1591can only be used with the operators 1592.Cm = 1593and 1594.Cm != . 1595Other constructs like 1596.Cm user \*(Ge unknown 1597are invalid. 1598Forwarded packets with unknown user and group ID match only rules 1599that explicitly compare against 1600.Em unknown 1601with the operators 1602.Cm = 1603or 1604.Cm != . 1605For instance 1606.Cm user \*(Ge 0 1607does not match forwarded packets. 1608The following example allows only selected users to open outgoing 1609connections: 1610.Bd -literal -offset indent 1611block out proto { tcp, udp } all 1612pass out proto { tcp, udp } all user { \*(Lt 1000, dhartmei } 1613.Ed 1614.It Xo Ar flags Aq Ar a 1615.Pf / Ns Aq Ar b 1616.No \*(Ba / Ns Aq Ar b 1617.No \*(Ba any 1618.Xc 1619This rule only applies to TCP packets that have the flags 1620.Aq Ar a 1621set out of set 1622.Aq Ar b . 1623Flags not specified in 1624.Aq Ar b 1625are ignored. 1626For stateful connections, the default is 1627.Ar flags S/SA . 1628To indicate that flags should not be checked at all, specify 1629.Ar flags any . 1630The flags are: (F)IN, (S)YN, (R)ST, (P)USH, (A)CK, (U)RG, (E)CE, and C(W)R. 1631.Bl -tag -width Fl 1632.It Ar flags S/S 1633Flag SYN is set. 1634The other flags are ignored. 1635.It Ar flags S/SA 1636This is the default setting for stateful connections. 1637Out of SYN and ACK, exactly SYN may be set. 1638SYN, SYN+PSH and SYN+RST match, but SYN+ACK, ACK and ACK+RST do not. 1639This is more restrictive than the previous example. 1640.It Ar flags /SFRA 1641If the first set is not specified, it defaults to none. 1642All of SYN, FIN, RST and ACK must be unset. 1643.El 1644.Pp 1645Because 1646.Ar flags S/SA 1647is applied by default (unless 1648.Ar no state 1649is specified), only the initial SYN packet of a TCP handshake will create 1650a state for a TCP connection. 1651It is possible to be less restrictive, and allow state creation from 1652intermediate 1653.Pq non-SYN 1654packets, by specifying 1655.Ar flags any . 1656This will cause 1657.Xr pf 4 1658to synchronize to existing connections, for instance 1659if one flushes the state table. 1660However, states created from such intermediate packets may be missing 1661connection details such as the TCP window scaling factor. 1662States which modify the packet flow, such as those affected by 1663.Ar nat , binat No or Ar rdr 1664rules, 1665.Ar modulate No or Ar synproxy state 1666options, or scrubbed with 1667.Ar reassemble tcp 1668will also not be recoverable from intermediate packets. 1669Such connections will stall and time out. 1670.It Xo Ar icmp-type Aq Ar type 1671.Ar code Aq Ar code 1672.Xc 1673.It Xo Ar icmp6-type Aq Ar type 1674.Ar code Aq Ar code 1675.Xc 1676This rule only applies to ICMP or ICMPv6 packets with the specified type 1677and code. 1678Text names for ICMP types and codes are listed in 1679.Xr icmp 4 1680and 1681.Xr icmp6 4 . 1682This parameter is only valid for rules that cover protocols ICMP or 1683ICMP6. 1684The protocol and the ICMP type indicator 1685.Po 1686.Ar icmp-type 1687or 1688.Ar icmp6-type 1689.Pc 1690must match. 1691.It Xo Ar tos Aq Ar string 1692.No \*(Ba Aq Ar number 1693.Xc 1694This rule applies to packets with the specified 1695.Em TOS 1696bits set. 1697.Em TOS 1698may be 1699given as one of 1700.Ar lowdelay , 1701.Ar throughput , 1702.Ar reliability , 1703or as either hex or decimal. 1704.Pp 1705For example, the following rules are identical: 1706.Bd -literal -offset indent 1707pass all tos lowdelay 1708pass all tos 0x10 1709pass all tos 16 1710.Ed 1711.It Ar allow-opts 1712By default, IPv4 packets with IP options or IPv6 packets with routing 1713extension headers are blocked. 1714When 1715.Ar allow-opts 1716is specified for a 1717.Ar pass 1718rule, packets that pass the filter based on that rule (last matching) 1719do so even if they contain IP options or routing extension headers. 1720For packets that match state, the rule that initially created the 1721state is used. 1722The implicit 1723.Ar pass 1724rule that is used when a packet does not match any rules does not 1725allow IP options. 1726.It Ar label Aq Ar string 1727Adds a label (name) to the rule, which can be used to identify the rule. 1728For instance, 1729pfctl -s labels 1730shows per-rule statistics for rules that have labels. 1731.Pp 1732The following macros can be used in labels: 1733.Pp 1734.Bl -tag -width $srcaddr -compact -offset indent 1735.It Ar $if 1736The interface. 1737.It Ar $srcaddr 1738The source IP address. 1739.It Ar $dstaddr 1740The destination IP address. 1741.It Ar $srcport 1742The source port specification. 1743.It Ar $dstport 1744The destination port specification. 1745.It Ar $proto 1746The protocol name. 1747.It Ar $nr 1748The rule number. 1749.El 1750.Pp 1751For example: 1752.Bd -literal -offset indent 1753ips = \&"{ 1.2.3.4, 1.2.3.5 }\&" 1754pass in proto tcp from any to $ips \e 1755 port \*(Gt 1023 label \&"$dstaddr:$dstport\&" 1756.Ed 1757.Pp 1758expands to 1759.Bd -literal -offset indent 1760pass in inet proto tcp from any to 1.2.3.4 \e 1761 port \*(Gt 1023 label \&"1.2.3.4:\*(Gt1023\&" 1762pass in inet proto tcp from any to 1.2.3.5 \e 1763 port \*(Gt 1023 label \&"1.2.3.5:\*(Gt1023\&" 1764.Ed 1765.Pp 1766The macro expansion for the 1767.Ar label 1768directive occurs only at configuration file parse time, not during runtime. 1769.It Xo Ar queue Aq Ar queue 1770.No \*(Ba ( Aq Ar queue , 1771.Aq Ar queue ) 1772.Xc 1773Packets matching this rule will be assigned to the specified queue. 1774If two queues are given, packets which have a 1775.Em TOS 1776of 1777.Em lowdelay 1778and TCP ACKs with no data payload will be assigned to the second one. 1779See 1780.Sx QUEUEING 1781for setup details. 1782.Pp 1783For example: 1784.Bd -literal -offset indent 1785pass in proto tcp to port 25 queue mail 1786pass in proto tcp to port 22 queue(ssh_bulk, ssh_prio) 1787.Ed 1788.It Ar tag Aq Ar string 1789Packets matching this rule will be tagged with the 1790specified string. 1791The tag acts as an internal marker that can be used to 1792identify these packets later on. 1793This can be used, for example, to provide trust between 1794interfaces and to determine if packets have been 1795processed by translation rules. 1796Tags are 1797.Qq sticky , 1798meaning that the packet will be tagged even if the rule 1799is not the last matching rule. 1800Further matching rules can replace the tag with a 1801new one but will not remove a previously applied tag. 1802A packet is only ever assigned one tag at a time. 1803Packet tagging can be done during 1804.Ar nat , 1805.Ar rdr , 1806or 1807.Ar binat 1808rules in addition to filter rules. 1809Tags take the same macros as labels (see above). 1810.It Ar tagged Aq Ar string 1811Used with filter, translation or scrub rules 1812to specify that packets must already 1813be tagged with the given tag in order to match the rule. 1814Inverse tag matching can also be done 1815by specifying the 1816.Cm !\& 1817operator before the 1818.Ar tagged 1819keyword. 1820.It Ar rtable Aq Ar number 1821Used to select an alternate routing table for the routing lookup. 1822Only effective before the route lookup happened, i.e. when filtering inbound. 1823.It Xo Ar divert-to Aq Ar host 1824.Ar port Aq Ar port 1825.Xc 1826Used to redirect packets to a local socket bound to 1827.Ar host 1828and 1829.Ar port . 1830The packets will not be modified, so 1831.Xr getsockname 2 1832on the socket will return the original destination address of the packet. 1833.It Ar divert-reply 1834Used to receive replies for sockets that are bound to addresses 1835which are not local to the machine. 1836See 1837.Xr setsockopt 2 1838for information on how to bind these sockets. 1839.It Ar probability Aq Ar number 1840A probability attribute can be attached to a rule, with a value set between 18410 and 1, bounds not included. 1842In that case, the rule will be honoured using the given probability value 1843only. 1844For example, the following rule will drop 20% of incoming ICMP packets: 1845.Bd -literal -offset indent 1846block in proto icmp probability 20% 1847.Ed 1848.El 1849.Sh ROUTING 1850If a packet matches a rule with a route option set, the packet filter will 1851route the packet according to the type of route option. 1852When such a rule creates state, the route option is also applied to all 1853packets matching the same connection. 1854.Bl -tag -width xxxx 1855.It Ar fastroute 1856The 1857.Ar fastroute 1858option does a normal route lookup to find the next hop for the packet. 1859.It Ar route-to 1860The 1861.Ar route-to 1862option routes the packet to the specified interface with an optional address 1863for the next hop. 1864When a 1865.Ar route-to 1866rule creates state, only packets that pass in the same direction as the 1867filter rule specifies will be routed in this way. 1868Packets passing in the opposite direction (replies) are not affected 1869and are routed normally. 1870.It Ar reply-to 1871The 1872.Ar reply-to 1873option is similar to 1874.Ar route-to , 1875but routes packets that pass in the opposite direction (replies) to the 1876specified interface. 1877Opposite direction is only defined in the context of a state entry, and 1878.Ar reply-to 1879is useful only in rules that create state. 1880It can be used on systems with multiple external connections to 1881route all outgoing packets of a connection through the interface 1882the incoming connection arrived through (symmetric routing enforcement). 1883.It Ar dup-to 1884The 1885.Ar dup-to 1886option creates a duplicate of the packet and routes it like 1887.Ar route-to . 1888The original packet gets routed as it normally would. 1889.El 1890.Sh POOL OPTIONS 1891For 1892.Ar nat 1893and 1894.Ar rdr 1895rules, (as well as for the 1896.Ar route-to , 1897.Ar reply-to 1898and 1899.Ar dup-to 1900rule options) for which there is a single redirection address which has a 1901subnet mask smaller than 32 for IPv4 or 128 for IPv6 (more than one IP 1902address), a variety of different methods for assigning this address can be 1903used: 1904.Bl -tag -width xxxx 1905.It Ar bitmask 1906The 1907.Ar bitmask 1908option applies the network portion of the redirection address to the address 1909to be modified (source with 1910.Ar nat , 1911destination with 1912.Ar rdr ) . 1913.It Ar random 1914The 1915.Ar random 1916option selects an address at random within the defined block of addresses. 1917.It Ar source-hash 1918The 1919.Ar source-hash 1920option uses a hash of the source address to determine the redirection address, 1921ensuring that the redirection address is always the same for a given source. 1922An optional key can be specified after this keyword either in hex or as a 1923string; by default 1924.Xr pfctl 8 1925randomly generates a key for source-hash every time the 1926ruleset is reloaded. 1927.It Ar round-robin 1928The 1929.Ar round-robin 1930option loops through the redirection address(es). 1931.Pp 1932When more than one redirection address is specified, 1933.Ar round-robin 1934is the only permitted pool type. 1935.It Ar static-port 1936With 1937.Ar nat 1938rules, the 1939.Ar static-port 1940option prevents 1941.Xr pf 4 1942from modifying the source port on TCP and UDP packets. 1943.El 1944.Pp 1945Additionally, the 1946.Ar sticky-address 1947option can be specified to help ensure that multiple connections from the 1948same source are mapped to the same redirection address. 1949This option can be used with the 1950.Ar random 1951and 1952.Ar round-robin 1953pool options. 1954Note that by default these associations are destroyed as soon as there are 1955no longer states which refer to them; in order to make the mappings last 1956beyond the lifetime of the states, increase the global options with 1957.Ar set timeout src.track . 1958See 1959.Sx STATEFUL TRACKING OPTIONS 1960for more ways to control the source tracking. 1961.Sh STATE MODULATION 1962Much of the security derived from TCP is attributable to how well the 1963initial sequence numbers (ISNs) are chosen. 1964Some popular stack implementations choose 1965.Em very 1966poor ISNs and thus are normally susceptible to ISN prediction exploits. 1967By applying a 1968.Ar modulate state 1969rule to a TCP connection, 1970.Xr pf 4 1971will create a high quality random sequence number for each connection 1972endpoint. 1973.Pp 1974The 1975.Ar modulate state 1976directive implicitly keeps state on the rule and is 1977only applicable to TCP connections. 1978.Pp 1979For instance: 1980.Bd -literal -offset indent 1981block all 1982pass out proto tcp from any to any modulate state 1983pass in proto tcp from any to any port 25 flags S/SFRA modulate state 1984.Ed 1985.Pp 1986Note that modulated connections will not recover when the state table 1987is lost (firewall reboot, flushing the state table, etc...). 1988.Xr pf 4 1989will not be able to infer a connection again after the state table flushes 1990the connection's modulator. 1991When the state is lost, the connection may be left dangling until the 1992respective endpoints time out the connection. 1993It is possible on a fast local network for the endpoints to start an ACK 1994storm while trying to resynchronize after the loss of the modulator. 1995The default 1996.Ar flags 1997settings (or a more strict equivalent) should be used on 1998.Ar modulate state 1999rules to prevent ACK storms. 2000.Pp 2001Note that alternative methods are available 2002to prevent loss of the state table 2003and allow for firewall failover. 2004See 2005.Xr carp 4 2006and 2007.Xr pfsync 4 2008for further information. 2009.Sh SYN PROXY 2010By default, 2011.Xr pf 4 2012passes packets that are part of a 2013.Xr tcp 4 2014handshake between the endpoints. 2015The 2016.Ar synproxy state 2017option can be used to cause 2018.Xr pf 4 2019itself to complete the handshake with the active endpoint, perform a handshake 2020with the passive endpoint, and then forward packets between the endpoints. 2021.Pp 2022No packets are sent to the passive endpoint before the active endpoint has 2023completed the handshake, hence so-called SYN floods with spoofed source 2024addresses will not reach the passive endpoint, as the sender can't complete the 2025handshake. 2026.Pp 2027The proxy is transparent to both endpoints, they each see a single 2028connection from/to the other endpoint. 2029.Xr pf 4 2030chooses random initial sequence numbers for both handshakes. 2031Once the handshakes are completed, the sequence number modulators 2032(see previous section) are used to translate further packets of the 2033connection. 2034.Ar synproxy state 2035includes 2036.Ar modulate state . 2037.Pp 2038Rules with 2039.Ar synproxy 2040will not work if 2041.Xr pf 4 2042operates on a 2043.Xr bridge 4 . 2044.Pp 2045Example: 2046.Bd -literal -offset indent 2047pass in proto tcp from any to any port www synproxy state 2048.Ed 2049.Sh STATEFUL TRACKING OPTIONS 2050A number of options related to stateful tracking can be applied on a 2051per-rule basis. 2052.Ar keep state , 2053.Ar modulate state 2054and 2055.Ar synproxy state 2056support these options, and 2057.Ar keep state 2058must be specified explicitly to apply options to a rule. 2059.Pp 2060.Bl -tag -width xxxx -compact 2061.It Ar max Aq Ar number 2062Limits the number of concurrent states the rule may create. 2063When this limit is reached, further packets that would create 2064state will not match this rule until existing states time out. 2065.It Ar no-sync 2066Prevent state changes for states created by this rule from appearing on the 2067.Xr pfsync 4 2068interface. 2069.It Xo Aq Ar timeout 2070.Aq Ar seconds 2071.Xc 2072Changes the timeout values used for states created by this rule. 2073For a list of all valid timeout names, see 2074.Sx OPTIONS 2075above. 2076.It Ar sloppy 2077Uses a sloppy TCP connection tracker that does not check sequence 2078numbers at all, which makes insertion and ICMP teardown attacks way 2079easier. 2080This is intended to be used in situations where one does not see all 2081packets of a connection, e.g. in asymmetric routing situations. 2082Cannot be used with modulate or synproxy state. 2083.It Ar pflow 2084States created by this rule are exported on the 2085.Xr pflow 4 2086interface. 2087.El 2088.Pp 2089Multiple options can be specified, separated by commas: 2090.Bd -literal -offset indent 2091pass in proto tcp from any to any \e 2092 port www keep state \e 2093 (max 100, source-track rule, max-src-nodes 75, \e 2094 max-src-states 3, tcp.established 60, tcp.closing 5) 2095.Ed 2096.Pp 2097When the 2098.Ar source-track 2099keyword is specified, the number of states per source IP is tracked. 2100.Pp 2101.Bl -tag -width xxxx -compact 2102.It Ar source-track rule 2103The maximum number of states created by this rule is limited by the rule's 2104.Ar max-src-nodes 2105and 2106.Ar max-src-states 2107options. 2108Only state entries created by this particular rule count toward the rule's 2109limits. 2110.It Ar source-track global 2111The number of states created by all rules that use this option is limited. 2112Each rule can specify different 2113.Ar max-src-nodes 2114and 2115.Ar max-src-states 2116options, however state entries created by any participating rule count towards 2117each individual rule's limits. 2118.El 2119.Pp 2120The following limits can be set: 2121.Pp 2122.Bl -tag -width xxxx -compact 2123.It Ar max-src-nodes Aq Ar number 2124Limits the maximum number of source addresses which can simultaneously 2125have state table entries. 2126.It Ar max-src-states Aq Ar number 2127Limits the maximum number of simultaneous state entries that a single 2128source address can create with this rule. 2129.El 2130.Pp 2131For stateful TCP connections, limits on established connections (connections 2132which have completed the TCP 3-way handshake) can also be enforced 2133per source IP. 2134.Pp 2135.Bl -tag -width xxxx -compact 2136.It Ar max-src-conn Aq Ar number 2137Limits the maximum number of simultaneous TCP connections which have 2138completed the 3-way handshake that a single host can make. 2139.It Xo Ar max-src-conn-rate Aq Ar number 2140.No / Aq Ar seconds 2141.Xc 2142Limit the rate of new connections over a time interval. 2143The connection rate is an approximation calculated as a moving average. 2144.El 2145.Pp 2146Because the 3-way handshake ensures that the source address is not being 2147spoofed, more aggressive action can be taken based on these limits. 2148With the 2149.Ar overload Aq Ar table 2150state option, source IP addresses which hit either of the limits on 2151established connections will be added to the named table. 2152This table can be used in the ruleset to block further activity from 2153the offending host, redirect it to a tarpit process, or restrict its 2154bandwidth. 2155.Pp 2156The optional 2157.Ar flush 2158keyword kills all states created by the matching rule which originate 2159from the host which exceeds these limits. 2160The 2161.Ar global 2162modifier to the flush command kills all states originating from the 2163offending host, regardless of which rule created the state. 2164.Pp 2165For example, the following rules will protect the webserver against 2166hosts making more than 100 connections in 10 seconds. 2167Any host which connects faster than this rate will have its address added 2168to the 2169.Aq bad_hosts 2170table and have all states originating from it flushed. 2171Any new packets arriving from this host will be dropped unconditionally 2172by the block rule. 2173.Bd -literal -offset indent 2174block quick from \*(Ltbad_hosts\*(Gt 2175pass in on $ext_if proto tcp to $webserver port www keep state \e 2176 (max-src-conn-rate 100/10, overload \*(Ltbad_hosts\*(Gt flush global) 2177.Ed 2178.Sh OPERATING SYSTEM FINGERPRINTING 2179Passive OS Fingerprinting is a mechanism to inspect nuances of a TCP 2180connection's initial SYN packet and guess at the host's operating system. 2181Unfortunately these nuances are easily spoofed by an attacker so the 2182fingerprint is not useful in making security decisions. 2183But the fingerprint is typically accurate enough to make policy decisions 2184upon. 2185.Pp 2186The fingerprints may be specified by operating system class, by 2187version, or by subtype/patchlevel. 2188The class of an operating system is typically the vendor or genre 2189and would be 2190.Ox 2191for the 2192.Xr pf 4 2193firewall itself. 2194The version of the oldest available 2195.Ox 2196release on the main FTP site 2197would be 2.6 and the fingerprint would be written 2198.Pp 2199.Dl \&"OpenBSD 2.6\&" 2200.Pp 2201The subtype of an operating system is typically used to describe the 2202patchlevel if that patch led to changes in the TCP stack behavior. 2203In the case of 2204.Ox , 2205the only subtype is for a fingerprint that was 2206normalized by the 2207.Ar no-df 2208scrub option and would be specified as 2209.Pp 2210.Dl \&"OpenBSD 3.3 no-df\&" 2211.Pp 2212Fingerprints for most popular operating systems are provided by 2213.Xr pf.os 5 . 2214Once 2215.Xr pf 4 2216is running, a complete list of known operating system fingerprints may 2217be listed by running: 2218.Pp 2219.Dl # pfctl -so 2220.Pp 2221Filter rules can enforce policy at any level of operating system specification 2222assuming a fingerprint is present. 2223Policy could limit traffic to approved operating systems or even ban traffic 2224from hosts that aren't at the latest service pack. 2225.Pp 2226The 2227.Ar unknown 2228class can also be used as the fingerprint which will match packets for 2229which no operating system fingerprint is known. 2230.Pp 2231Examples: 2232.Bd -literal -offset indent 2233pass out proto tcp from any os OpenBSD 2234block out proto tcp from any os Doors 2235block out proto tcp from any os "Doors PT" 2236block out proto tcp from any os "Doors PT SP3" 2237block out from any os "unknown" 2238pass on lo0 proto tcp from any os "OpenBSD 3.3 lo0" 2239.Ed 2240.Pp 2241Operating system fingerprinting is limited only to the TCP SYN packet. 2242This means that it will not work on other protocols and will not match 2243a currently established connection. 2244.Pp 2245Caveat: operating system fingerprints are occasionally wrong. 2246There are three problems: an attacker can trivially craft his packets to 2247appear as any operating system he chooses; 2248an operating system patch could change the stack behavior and no fingerprints 2249will match it until the database is updated; 2250and multiple operating systems may have the same fingerprint. 2251.Sh BLOCKING SPOOFED TRAFFIC 2252"Spoofing" is the faking of IP addresses, typically for malicious 2253purposes. 2254The 2255.Ar antispoof 2256directive expands to a set of filter rules which will block all 2257traffic with a source IP from the network(s) directly connected 2258to the specified interface(s) from entering the system through 2259any other interface. 2260.Pp 2261For example, the line 2262.Bd -literal -offset indent 2263antispoof for lo0 2264.Ed 2265.Pp 2266expands to 2267.Bd -literal -offset indent 2268block drop in on ! lo0 inet from 127.0.0.1/8 to any 2269block drop in on ! lo0 inet6 from ::1 to any 2270.Ed 2271.Pp 2272For non-loopback interfaces, there are additional rules to block incoming 2273packets with a source IP address identical to the interface's IP(s). 2274For example, assuming the interface wi0 had an IP address of 10.0.0.1 and a 2275netmask of 255.255.255.0, 2276the line 2277.Bd -literal -offset indent 2278antispoof for wi0 inet 2279.Ed 2280.Pp 2281expands to 2282.Bd -literal -offset indent 2283block drop in on ! wi0 inet from 10.0.0.0/24 to any 2284block drop in inet from 10.0.0.1 to any 2285.Ed 2286.Pp 2287Caveat: Rules created by the 2288.Ar antispoof 2289directive interfere with packets sent over loopback interfaces 2290to local addresses. 2291One should pass these explicitly. 2292.Sh FRAGMENT HANDLING 2293The size of IP datagrams (packets) can be significantly larger than the 2294maximum transmission unit (MTU) of the network. 2295In cases when it is necessary or more efficient to send such large packets, 2296the large packet will be fragmented into many smaller packets that will each 2297fit onto the wire. 2298Unfortunately for a firewalling device, only the first logical fragment will 2299contain the necessary header information for the subprotocol that allows 2300.Xr pf 4 2301to filter on things such as TCP ports or to perform NAT. 2302.Pp 2303Besides the use of 2304.Ar scrub 2305rules as described in 2306.Sx TRAFFIC NORMALIZATION 2307above, there are three options for handling fragments in the packet filter. 2308.Pp 2309One alternative is to filter individual fragments with filter rules. 2310If no 2311.Ar scrub 2312rule applies to a fragment, it is passed to the filter. 2313Filter rules with matching IP header parameters decide whether the 2314fragment is passed or blocked, in the same way as complete packets 2315are filtered. 2316Without reassembly, fragments can only be filtered based on IP header 2317fields (source/destination address, protocol), since subprotocol header 2318fields are not available (TCP/UDP port numbers, ICMP code/type). 2319The 2320.Ar fragment 2321option can be used to restrict filter rules to apply only to 2322fragments, but not complete packets. 2323Filter rules without the 2324.Ar fragment 2325option still apply to fragments, if they only specify IP header fields. 2326For instance, the rule 2327.Bd -literal -offset indent 2328pass in proto tcp from any to any port 80 2329.Ed 2330.Pp 2331never applies to a fragment, even if the fragment is part of a TCP 2332packet with destination port 80, because without reassembly this information 2333is not available for each fragment. 2334This also means that fragments cannot create new or match existing 2335state table entries, which makes stateful filtering and address 2336translation (NAT, redirection) for fragments impossible. 2337.Pp 2338It's also possible to reassemble only certain fragments by specifying 2339source or destination addresses or protocols as parameters in 2340.Ar scrub 2341rules. 2342.Pp 2343In most cases, the benefits of reassembly outweigh the additional 2344memory cost, and it's recommended to use 2345.Ar scrub 2346rules to reassemble 2347all fragments via the 2348.Ar fragment reassemble 2349modifier. 2350.Pp 2351The memory allocated for fragment caching can be limited using 2352.Xr pfctl 8 . 2353Once this limit is reached, fragments that would have to be cached 2354are dropped until other entries time out. 2355The timeout value can also be adjusted. 2356.Pp 2357When forwarding reassembled IPv6 packets, pf refragments them with 2358the original maximum fragment size. 2359This allows the sender to determine the optimal fragment size by 2360path MTU discovery. 2361.Sh ANCHORS 2362Besides the main ruleset, 2363.Xr pfctl 8 2364can load rulesets into 2365.Ar anchor 2366attachment points. 2367An 2368.Ar anchor 2369is a container that can hold rules, address tables, and other anchors. 2370.Pp 2371An 2372.Ar anchor 2373has a name which specifies the path where 2374.Xr pfctl 8 2375can be used to access the anchor to perform operations on it, such as 2376attaching child anchors to it or loading rules into it. 2377Anchors may be nested, with components separated by 2378.Sq / 2379characters, similar to how file system hierarchies are laid out. 2380The main ruleset is actually the default anchor, so filter and 2381translation rules, for example, may also be contained in any anchor. 2382.Pp 2383An anchor can reference another 2384.Ar anchor 2385attachment point 2386using the following kinds 2387of rules: 2388.Bl -tag -width xxxx 2389.It Ar nat-anchor Aq Ar name 2390Evaluates the 2391.Ar nat 2392rules in the specified 2393.Ar anchor . 2394.It Ar rdr-anchor Aq Ar name 2395Evaluates the 2396.Ar rdr 2397rules in the specified 2398.Ar anchor . 2399.It Ar binat-anchor Aq Ar name 2400Evaluates the 2401.Ar binat 2402rules in the specified 2403.Ar anchor . 2404.It Ar anchor Aq Ar name 2405Evaluates the filter rules in the specified 2406.Ar anchor . 2407.It Xo Ar load anchor 2408.Aq Ar name 2409.Ar from Aq Ar file 2410.Xc 2411Loads the rules from the specified file into the 2412anchor 2413.Ar name . 2414.El 2415.Pp 2416When evaluation of the main ruleset reaches an 2417.Ar anchor 2418rule, 2419.Xr pf 4 2420will proceed to evaluate all rules specified in that anchor. 2421.Pp 2422Matching filter and translation rules marked with the 2423.Ar quick 2424option are final and abort the evaluation of the rules in other 2425anchors and the main ruleset. 2426If the 2427.Ar anchor 2428itself is marked with the 2429.Ar quick 2430option, 2431ruleset evaluation will terminate when the anchor is exited if the packet is 2432matched by any rule within the anchor. 2433.Pp 2434.Ar anchor 2435rules are evaluated relative to the anchor in which they are contained. 2436For example, all 2437.Ar anchor 2438rules specified in the main ruleset will reference anchor 2439attachment points underneath the main ruleset, and 2440.Ar anchor 2441rules specified in a file loaded from a 2442.Ar load anchor 2443rule will be attached under that anchor point. 2444.Pp 2445Rules may be contained in 2446.Ar anchor 2447attachment points which do not contain any rules when the main ruleset 2448is loaded, and later such anchors can be manipulated through 2449.Xr pfctl 8 2450without reloading the main ruleset or other anchors. 2451For example, 2452.Bd -literal -offset indent 2453ext_if = \&"kue0\&" 2454block on $ext_if all 2455anchor spam 2456pass out on $ext_if all 2457pass in on $ext_if proto tcp from any \e 2458 to $ext_if port smtp 2459.Ed 2460.Pp 2461blocks all packets on the external interface by default, then evaluates 2462all rules in the 2463.Ar anchor 2464named "spam", and finally passes all outgoing connections and 2465incoming connections to port 25. 2466.Bd -literal -offset indent 2467# echo \&"block in quick from 1.2.3.4 to any\&" \&| \e 2468 pfctl -a spam -f - 2469.Ed 2470.Pp 2471This loads a single rule into the 2472.Ar anchor , 2473which blocks all packets from a specific address. 2474.Pp 2475The anchor can also be populated by adding a 2476.Ar load anchor 2477rule after the 2478.Ar anchor 2479rule: 2480.Bd -literal -offset indent 2481anchor spam 2482load anchor spam from "/etc/pf-spam.conf" 2483.Ed 2484.Pp 2485When 2486.Xr pfctl 8 2487loads 2488.Nm pf.conf , 2489it will also load all the rules from the file 2490.Pa /etc/pf-spam.conf 2491into the anchor. 2492.Pp 2493Optionally, 2494.Ar anchor 2495rules can specify packet filtering parameters using the same syntax as 2496filter rules. 2497When parameters are used, the 2498.Ar anchor 2499rule is only evaluated for matching packets. 2500This allows conditional evaluation of anchors, like: 2501.Bd -literal -offset indent 2502block on $ext_if all 2503anchor spam proto tcp from any to any port smtp 2504pass out on $ext_if all 2505pass in on $ext_if proto tcp from any to $ext_if port smtp 2506.Ed 2507.Pp 2508The rules inside 2509.Ar anchor 2510spam are only evaluated for 2511.Ar tcp 2512packets with destination port 25. 2513Hence, 2514.Bd -literal -offset indent 2515# echo \&"block in quick from 1.2.3.4 to any" \&| \e 2516 pfctl -a spam -f - 2517.Ed 2518.Pp 2519will only block connections from 1.2.3.4 to port 25. 2520.Pp 2521Anchors may end with the asterisk 2522.Pq Sq * 2523character, which signifies that all anchors attached at that point 2524should be evaluated in the alphabetical ordering of their anchor name. 2525For example, 2526.Bd -literal -offset indent 2527anchor "spam/*" 2528.Ed 2529.Pp 2530will evaluate each rule in each anchor attached to the 2531.Li spam 2532anchor. 2533Note that it will only evaluate anchors that are directly attached to the 2534.Li spam 2535anchor, and will not descend to evaluate anchors recursively. 2536.Pp 2537Since anchors are evaluated relative to the anchor in which they are 2538contained, there is a mechanism for accessing the parent and ancestor 2539anchors of a given anchor. 2540Similar to file system path name resolution, if the sequence 2541.Dq .. 2542appears as an anchor path component, the parent anchor of the current 2543anchor in the path evaluation at that point will become the new current 2544anchor. 2545As an example, consider the following: 2546.Bd -literal -offset indent 2547# echo ' anchor "spam/allowed" ' | pfctl -f - 2548# echo -e ' anchor "../banned" \en pass' | \e 2549 pfctl -a spam/allowed -f - 2550.Ed 2551.Pp 2552Evaluation of the main ruleset will lead into the 2553.Li spam/allowed 2554anchor, which will evaluate the rules in the 2555.Li spam/banned 2556anchor, if any, before finally evaluating the 2557.Ar pass 2558rule. 2559.Pp 2560Filter rule 2561.Ar anchors 2562can also be loaded inline in the ruleset within a brace ('{' '}') delimited 2563block. 2564Brace delimited blocks may contain rules or other brace-delimited blocks. 2565When anchors are loaded this way the anchor name becomes optional. 2566.Bd -literal -offset indent 2567anchor "external" on egress { 2568 block 2569 anchor out { 2570 pass proto tcp from any to port { 25, 80, 443 } 2571 } 2572 pass in proto tcp to any port 22 2573} 2574.Ed 2575.Pp 2576Since the parser specification for anchor names is a string, any 2577reference to an anchor name containing 2578.Sq / 2579characters will require double quote 2580.Pq Sq \&" 2581characters around the anchor name. 2582.Sh TRANSLATION EXAMPLES 2583This example maps incoming requests on port 80 to port 8080, on 2584which a daemon is running (because, for example, it is not run as root, 2585and therefore lacks permission to bind to port 80). 2586.Bd -literal 2587# use a macro for the interface name, so it can be changed easily 2588ext_if = \&"ne3\&" 2589 2590# map daemon on 8080 to appear to be on 80 2591rdr on $ext_if proto tcp from any to any port 80 -\*(Gt 127.0.0.1 port 8080 2592.Ed 2593.Pp 2594If the 2595.Ar pass 2596modifier is given, packets matching the translation rule are passed without 2597inspecting the filter rules: 2598.Bd -literal 2599rdr pass on $ext_if proto tcp from any to any port 80 -\*(Gt 127.0.0.1 \e 2600 port 8080 2601.Ed 2602.Pp 2603In the example below, vlan12 is configured as 192.168.168.1; 2604the machine translates all packets coming from 192.168.168.0/24 to 204.92.77.111 2605when they are going out any interface except vlan12. 2606This has the net effect of making traffic from the 192.168.168.0/24 2607network appear as though it is the Internet routable address 2608204.92.77.111 to nodes behind any interface on the router except 2609for the nodes on vlan12. 2610(Thus, 192.168.168.1 can talk to the 192.168.168.0/24 nodes.) 2611.Bd -literal 2612nat on ! vlan12 from 192.168.168.0/24 to any -\*(Gt 204.92.77.111 2613.Ed 2614.Pp 2615In the example below, the machine sits between a fake internal 144.19.74.* 2616network, and a routable external IP of 204.92.77.100. 2617The 2618.Ar no nat 2619rule excludes protocol AH from being translated. 2620.Bd -literal 2621# NO NAT 2622no nat on $ext_if proto ah from 144.19.74.0/24 to any 2623nat on $ext_if from 144.19.74.0/24 to any -\*(Gt 204.92.77.100 2624.Ed 2625.Pp 2626In the example below, packets bound for one specific server, as well as those 2627generated by the sysadmins are not proxied; all other connections are. 2628.Bd -literal 2629# NO RDR 2630no rdr on $int_if proto { tcp, udp } from any to $server port 80 2631no rdr on $int_if proto { tcp, udp } from $sysadmins to any port 80 2632rdr on $int_if proto { tcp, udp } from any to any port 80 -\*(Gt 127.0.0.1 \e 2633 port 80 2634.Ed 2635.Pp 2636This longer example uses both a NAT and a redirection. 2637The external interface has the address 157.161.48.183. 2638On localhost, we are running 2639.Xr ftp-proxy 8 , 2640waiting for FTP sessions to be redirected to it. 2641The three mandatory anchors for 2642.Xr ftp-proxy 8 2643are omitted from this example; see the 2644.Xr ftp-proxy 8 2645manpage. 2646.Bd -literal 2647# NAT 2648# Translate outgoing packets' source addresses (any protocol). 2649# In this case, any address but the gateway's external address is mapped. 2650nat on $ext_if inet from ! ($ext_if) to any -\*(Gt ($ext_if) 2651 2652# NAT PROXYING 2653# Map outgoing packets' source port to an assigned proxy port instead of 2654# an arbitrary port. 2655# In this case, proxy outgoing isakmp with port 500 on the gateway. 2656nat on $ext_if inet proto udp from any port = isakmp to any -\*(Gt ($ext_if) \e 2657 port 500 2658 2659# BINAT 2660# Translate outgoing packets' source address (any protocol). 2661# Translate incoming packets' destination address to an internal machine 2662# (bidirectional). 2663binat on $ext_if from 10.1.2.150 to any -\*(Gt $ext_if 2664 2665# RDR 2666# Translate incoming packets' destination addresses. 2667# As an example, redirect a TCP and UDP port to an internal machine. 2668rdr on $ext_if inet proto tcp from any to ($ext_if) port 8080 \e 2669 -\*(Gt 10.1.2.151 port 22 2670rdr on $ext_if inet proto udp from any to ($ext_if) port 8080 \e 2671 -\*(Gt 10.1.2.151 port 53 2672 2673# RDR 2674# Translate outgoing ftp control connections to send them to localhost 2675# for proxying with ftp-proxy(8) running on port 8021. 2676rdr on $int_if proto tcp from any to any port 21 -\*(Gt 127.0.0.1 port 8021 2677.Ed 2678.Pp 2679In this example, a NAT gateway is set up to translate internal addresses 2680using a pool of public addresses (192.0.2.16/28) and to redirect 2681incoming web server connections to a group of web servers on the internal 2682network. 2683.Bd -literal 2684# NAT LOAD BALANCE 2685# Translate outgoing packets' source addresses using an address pool. 2686# A given source address is always translated to the same pool address by 2687# using the source-hash keyword. 2688nat on $ext_if inet from any to any -\*(Gt 192.0.2.16/28 source-hash 2689 2690# RDR ROUND ROBIN 2691# Translate incoming web server connections to a group of web servers on 2692# the internal network. 2693rdr on $ext_if proto tcp from any to any port 80 \e 2694 -\*(Gt { 10.1.2.155, 10.1.2.160, 10.1.2.161 } round-robin 2695.Ed 2696.Sh FILTER EXAMPLES 2697.Bd -literal 2698# The external interface is kue0 2699# (157.161.48.183, the only routable address) 2700# and the private network is 10.0.0.0/8, for which we are doing NAT. 2701 2702# use a macro for the interface name, so it can be changed easily 2703ext_if = \&"kue0\&" 2704 2705# normalize all incoming traffic 2706scrub in on $ext_if all fragment reassemble 2707 2708# block and log everything by default 2709block return log on $ext_if all 2710 2711# block anything coming from source we have no back routes for 2712block in from no-route to any 2713 2714# block packets whose ingress interface does not match the one in 2715# the route back to their source address 2716block in from urpf-failed to any 2717 2718# block and log outgoing packets that do not have our address as source, 2719# they are either spoofed or something is misconfigured (NAT disabled, 2720# for instance), we want to be nice and do not send out garbage. 2721block out log quick on $ext_if from ! 157.161.48.183 to any 2722 2723# silently drop broadcasts (cable modem noise) 2724block in quick on $ext_if from any to 255.255.255.255 2725 2726# block and log incoming packets from reserved address space and invalid 2727# addresses, they are either spoofed or misconfigured, we cannot reply to 2728# them anyway (hence, no return-rst). 2729block in log quick on $ext_if from { 10.0.0.0/8, 172.16.0.0/12, \e 2730 192.168.0.0/16, 255.255.255.255/32 } to any 2731 2732# ICMP 2733 2734# pass out/in certain ICMP queries and keep state (ping) 2735# state matching is done on host addresses and ICMP id (not type/code), 2736# so replies (like 0/0 for 8/0) will match queries 2737# ICMP error messages (which always refer to a TCP/UDP packet) are 2738# handled by the TCP/UDP states 2739pass on $ext_if inet proto icmp all icmp-type 8 code 0 2740 2741# UDP 2742 2743# pass out all UDP connections and keep state 2744pass out on $ext_if proto udp all 2745 2746# pass in certain UDP connections and keep state (DNS) 2747pass in on $ext_if proto udp from any to any port domain 2748 2749# TCP 2750 2751# pass out all TCP connections and modulate state 2752pass out on $ext_if proto tcp all modulate state 2753 2754# pass in certain TCP connections and keep state (SSH, SMTP, DNS, IDENT) 2755pass in on $ext_if proto tcp from any to any port { ssh, smtp, domain, \e 2756 auth } 2757 2758# Do not allow Windows 9x SMTP connections since they are typically 2759# a viral worm. Alternately we could limit these OSes to 1 connection each. 2760block in on $ext_if proto tcp from any os {"Windows 95", "Windows 98"} \e 2761 to any port smtp 2762 2763# IPv6 2764# pass in/out all IPv6 traffic: note that we have to enable this in two 2765# different ways, on both our physical interface and our tunnel 2766pass quick on gif0 inet6 2767pass quick on $ext_if proto ipv6 2768 2769# Packet Tagging 2770 2771# three interfaces: $int_if, $ext_if, and $wifi_if (wireless). NAT is 2772# being done on $ext_if for all outgoing packets. tag packets in on 2773# $int_if and pass those tagged packets out on $ext_if. all other 2774# outgoing packets (i.e., packets from the wireless network) are only 2775# permitted to access port 80. 2776 2777pass in on $int_if from any to any tag INTNET 2778pass in on $wifi_if from any to any 2779 2780block out on $ext_if from any to any 2781pass out quick on $ext_if tagged INTNET 2782pass out on $ext_if proto tcp from any to any port 80 2783 2784# tag incoming packets as they are redirected to spamd(8). use the tag 2785# to pass those packets through the packet filter. 2786 2787rdr on $ext_if inet proto tcp from \*(Ltspammers\*(Gt to port smtp \e 2788 tag SPAMD -\*(Gt 127.0.0.1 port spamd 2789 2790block in on $ext_if 2791pass in on $ext_if inet proto tcp tagged SPAMD 2792.Ed 2793.Sh GRAMMAR 2794Syntax for 2795.Nm 2796in BNF: 2797.Bd -literal 2798line = ( option | pf-rule | nat-rule | binat-rule | rdr-rule | 2799 antispoof-rule | altq-rule | queue-rule | trans-anchors | 2800 anchor-rule | anchor-close | load-anchor | table-rule | 2801 include ) 2802 2803option = "set" ( [ "timeout" ( timeout | "{" timeout-list "}" ) ] | 2804 [ "ruleset-optimization" [ "none" | "basic" | "profile" ]] | 2805 [ "optimization" [ "default" | "normal" | 2806 "high-latency" | "satellite" | 2807 "aggressive" | "conservative" ] ] 2808 [ "limit" ( limit-item | "{" limit-list "}" ) ] | 2809 [ "loginterface" ( interface-name | "none" ) ] | 2810 [ "block-policy" ( "drop" | "return" ) ] | 2811 [ "state-policy" ( "if-bound" | "floating" ) ] 2812 [ "state-defaults" state-opts ] 2813 [ "require-order" ( "yes" | "no" ) ] 2814 [ "fingerprints" filename ] | 2815 [ "skip on" ifspec ] | 2816 [ "debug" ( "none" | "urgent" | "misc" | "loud" ) ] ) 2817 2818pf-rule = action [ ( "in" | "out" ) ] 2819 [ "log" [ "(" logopts ")"] ] [ "quick" ] 2820 [ "on" ifspec ] [ "fastroute" | route ] [ af ] [ protospec ] 2821 hosts [ filteropt-list ] 2822 2823logopts = logopt [ "," logopts ] 2824logopt = "all" | "user" | "to" interface-name 2825 2826filteropt-list = filteropt-list filteropt | filteropt 2827filteropt = user | group | flags | icmp-type | icmp6-type | "tos" tos | 2828 ( "no" | "keep" | "modulate" | "synproxy" ) "state" 2829 [ "(" state-opts ")" ] | 2830 "fragment" | "no-df" | "min-ttl" number | "set-tos" tos | 2831 "max-mss" number | "random-id" | "reassemble tcp" | 2832 fragmentation | "allow-opts" | 2833 "label" string | "tag" string | [ ! ] "tagged" string | 2834 "queue" ( string | "(" string [ [ "," ] string ] ")" ) | 2835 "rtable" number | "probability" number"%" 2836 2837nat-rule = [ "no" ] "nat" [ "pass" [ "log" [ "(" logopts ")" ] ] ] 2838 [ "on" ifspec ] [ af ] 2839 [ protospec ] hosts [ "tag" string ] [ "tagged" string ] 2840 [ "-\*(Gt" ( redirhost | "{" redirhost-list "}" ) 2841 [ portspec ] [ pooltype ] [ "static-port" ] ] 2842 2843binat-rule = [ "no" ] "binat" [ "pass" [ "log" [ "(" logopts ")" ] ] ] 2844 [ "on" interface-name ] [ af ] 2845 [ "proto" ( proto-name | proto-number ) ] 2846 "from" address [ "/" mask-bits ] "to" ipspec 2847 [ "tag" string ] [ "tagged" string ] 2848 [ "-\*(Gt" address [ "/" mask-bits ] ] 2849 2850rdr-rule = [ "no" ] "rdr" [ "pass" [ "log" [ "(" logopts ")" ] ] ] 2851 [ "on" ifspec ] [ af ] 2852 [ protospec ] hosts [ "tag" string ] [ "tagged" string ] 2853 [ "-\*(Gt" ( redirhost | "{" redirhost-list "}" ) 2854 [ portspec ] [ pooltype ] ] 2855 2856antispoof-rule = "antispoof" [ "log" ] [ "quick" ] 2857 "for" ifspec [ af ] [ "label" string ] 2858 2859table-rule = "table" "\*(Lt" string "\*(Gt" [ tableopts-list ] 2860tableopts-list = tableopts-list tableopts | tableopts 2861tableopts = "persist" | "const" | "counters" | "file" string | 2862 "{" [ tableaddr-list ] "}" 2863tableaddr-list = tableaddr-list [ "," ] tableaddr-spec | tableaddr-spec 2864tableaddr-spec = [ "!" ] tableaddr [ "/" mask-bits ] 2865tableaddr = hostname | ifspec | "self" | 2866 ipv4-dotted-quad | ipv6-coloned-hex 2867 2868altq-rule = "altq on" interface-name queueopts-list 2869 "queue" subqueue 2870queue-rule = "queue" string [ "on" interface-name ] queueopts-list 2871 subqueue 2872 2873anchor-rule = "anchor" [ string ] [ ( "in" | "out" ) ] [ "on" ifspec ] 2874 [ af ] [ protospec ] [ hosts ] [ filteropt-list ] [ "{" ] 2875 2876anchor-close = "}" 2877 2878trans-anchors = ( "nat-anchor" | "rdr-anchor" | "binat-anchor" ) string 2879 [ "on" ifspec ] [ af ] [ "proto" ] [ protospec ] [ hosts ] 2880 2881load-anchor = "load anchor" string "from" filename 2882 2883queueopts-list = queueopts-list queueopts | queueopts 2884queueopts = [ "bandwidth" bandwidth-spec ] | 2885 [ "qlimit" number ] | [ "tbrsize" number ] | 2886 [ "priority" number ] | [ schedulers ] 2887schedulers = ( cbq-def | priq-def | hfsc-def ) 2888bandwidth-spec = "number" ( "b" | "Kb" | "Mb" | "Gb" | "%" ) 2889 2890action = "pass" | "block" [ return ] | [ "no" ] "scrub" 2891return = "drop" | "return" | "return-rst" [ "( ttl" number ")" ] | 2892 "return-icmp" [ "(" icmpcode [ [ "," ] icmp6code ] ")" ] | 2893 "return-icmp6" [ "(" icmp6code ")" ] 2894icmpcode = ( icmp-code-name | icmp-code-number ) 2895icmp6code = ( icmp6-code-name | icmp6-code-number ) 2896 2897ifspec = ( [ "!" ] ( interface-name | interface-group ) ) | 2898 "{" interface-list "}" 2899interface-list = [ "!" ] ( interface-name | interface-group ) 2900 [ [ "," ] interface-list ] 2901route = ( "route-to" | "reply-to" | "dup-to" ) 2902 ( routehost | "{" routehost-list "}" ) 2903 [ pooltype ] 2904af = "inet" | "inet6" 2905 2906protospec = "proto" ( proto-name | proto-number | 2907 "{" proto-list "}" ) 2908proto-list = ( proto-name | proto-number ) [ [ "," ] proto-list ] 2909 2910hosts = "all" | 2911 "from" ( "any" | "no-route" | "urpf-failed" | "self" | host | 2912 "{" host-list "}" ) [ port ] [ os ] 2913 "to" ( "any" | "no-route" | "self" | host | 2914 "{" host-list "}" ) [ port ] 2915 2916ipspec = "any" | host | "{" host-list "}" 2917host = [ "!" ] ( address [ "/" mask-bits ] | "\*(Lt" string "\*(Gt" ) 2918redirhost = address [ "/" mask-bits ] 2919routehost = "(" interface-name [ address [ "/" mask-bits ] ] ")" 2920address = ( interface-name | interface-group | 2921 "(" ( interface-name | interface-group ) ")" | 2922 hostname | ipv4-dotted-quad | ipv6-coloned-hex ) 2923host-list = host [ [ "," ] host-list ] 2924redirhost-list = redirhost [ [ "," ] redirhost-list ] 2925routehost-list = routehost [ [ "," ] routehost-list ] 2926 2927port = "port" ( unary-op | binary-op | "{" op-list "}" ) 2928portspec = "port" ( number | name ) [ ":" ( "*" | number | name ) ] 2929os = "os" ( os-name | "{" os-list "}" ) 2930user = "user" ( unary-op | binary-op | "{" op-list "}" ) 2931group = "group" ( unary-op | binary-op | "{" op-list "}" ) 2932 2933unary-op = [ "=" | "!=" | "\*(Lt" | "\*(Le" | "\*(Gt" | "\*(Ge" ] 2934 ( name | number ) 2935binary-op = number ( "\*(Lt\*(Gt" | "\*(Gt\*(Lt" | ":" ) number 2936op-list = ( unary-op | binary-op ) [ [ "," ] op-list ] 2937 2938os-name = operating-system-name 2939os-list = os-name [ [ "," ] os-list ] 2940 2941flags = "flags" ( [ flag-set ] "/" flag-set | "any" ) 2942flag-set = [ "F" ] [ "S" ] [ "R" ] [ "P" ] [ "A" ] [ "U" ] [ "E" ] 2943 [ "W" ] 2944 2945icmp-type = "icmp-type" ( icmp-type-code | "{" icmp-list "}" ) 2946icmp6-type = "icmp6-type" ( icmp-type-code | "{" icmp-list "}" ) 2947icmp-type-code = ( icmp-type-name | icmp-type-number ) 2948 [ "code" ( icmp-code-name | icmp-code-number ) ] 2949icmp-list = icmp-type-code [ [ "," ] icmp-list ] 2950 2951tos = ( "lowdelay" | "throughput" | "reliability" | 2952 [ "0x" ] number ) 2953 2954state-opts = state-opt [ [ "," ] state-opts ] 2955state-opt = ( "max" number | "no-sync" | timeout | "sloppy" | "pflow" | 2956 "source-track" [ ( "rule" | "global" ) ] | 2957 "max-src-nodes" number | "max-src-states" number | 2958 "max-src-conn" number | 2959 "max-src-conn-rate" number "/" number | 2960 "overload" "\*(Lt" string "\*(Gt" [ "flush" ] | 2961 "if-bound" | "floating" ) 2962 2963fragmentation = [ "fragment reassemble" ] 2964 2965timeout-list = timeout [ [ "," ] timeout-list ] 2966timeout = ( "tcp.first" | "tcp.opening" | "tcp.established" | 2967 "tcp.closing" | "tcp.finwait" | "tcp.closed" | 2968 "udp.first" | "udp.single" | "udp.multiple" | 2969 "icmp.first" | "icmp.error" | 2970 "other.first" | "other.single" | "other.multiple" | 2971 "frag" | "interval" | "src.track" | 2972 "adaptive.start" | "adaptive.end" ) number 2973 2974limit-list = limit-item [ [ "," ] limit-list ] 2975limit-item = ( "states" | "frags" | "src-nodes" ) number 2976 2977pooltype = ( "bitmask" | "random" | 2978 "source-hash" [ ( hex-key | string-key ) ] | 2979 "round-robin" ) [ sticky-address ] 2980 2981subqueue = string | "{" queue-list "}" 2982queue-list = string [ [ "," ] string ] 2983cbq-def = "cbq" [ "(" cbq-opt [ [ "," ] cbq-opt ] ")" ] 2984priq-def = "priq" [ "(" priq-opt [ [ "," ] priq-opt ] ")" ] 2985hfsc-def = "hfsc" [ "(" hfsc-opt [ [ "," ] hfsc-opt ] ")" ] 2986cbq-opt = ( "default" | "borrow" | "red" | "ecn" | "rio" ) 2987priq-opt = ( "default" | "red" | "ecn" | "rio" ) 2988hfsc-opt = ( "default" | "red" | "ecn" | "rio" | 2989 linkshare-sc | realtime-sc | upperlimit-sc ) 2990linkshare-sc = "linkshare" sc-spec 2991realtime-sc = "realtime" sc-spec 2992upperlimit-sc = "upperlimit" sc-spec 2993sc-spec = ( bandwidth-spec | 2994 "(" bandwidth-spec number bandwidth-spec ")" ) 2995include = "include" filename 2996.Ed 2997.Sh FILES 2998.Bl -tag -width "/etc/protocols" -compact 2999.It Pa /etc/hosts 3000Host name database. 3001.It Pa /etc/pf.conf 3002Default location of the ruleset file. 3003.It Pa /etc/pf.os 3004Default location of OS fingerprints. 3005.It Pa /etc/protocols 3006Protocol name database. 3007.It Pa /etc/services 3008Service name database. 3009.El 3010.Sh SEE ALSO 3011.Xr altq 4 , 3012.Xr carp 4 , 3013.Xr icmp 4 , 3014.Xr icmp6 4 , 3015.Xr ip 4 , 3016.Xr ip6 4 , 3017.Xr pf 4 , 3018.Xr pflow 4 , 3019.Xr pfsync 4 , 3020.Xr tcp 4 , 3021.Xr udp 4 , 3022.Xr hosts 5 , 3023.Xr pf.os 5 , 3024.Xr protocols 5 , 3025.Xr services 5 , 3026.Xr ftp-proxy 8 , 3027.Xr pfctl 8 , 3028.Xr pflogd 8 3029.Sh HISTORY 3030The 3031.Nm 3032file format first appeared in 3033.Ox 3.0 . 3034