1.\" 2.\" $FreeBSD: src/sbin/ipfw/ipfw.8,v 1.63.2.33 2003/02/04 01:36:02 brueffer Exp $ 3.\" $DragonFly: src/sbin/ipfw/ipfw.8,v 1.3 2004/03/11 12:28:54 hmp Exp $ 4.\" 5.Dd August 13, 2002 6.Dt IPFW 8 7.Os 8.Sh NAME 9.Nm ipfw 10.Nd IP firewall and traffic shaper control program 11.Sh SYNOPSIS 12.Nm 13.Op Fl cq 14.Cm add 15.Ar rule 16.Nm 17.Op Fl acdeftNS 18.Brq Cm list | show 19.Op Ar number ... 20.Nm 21.Op Fl f | q 22.Cm flush 23.Nm 24.Op Fl q 25.Brq Cm delete | zero | resetlog 26.Op Cm set 27.Op Ar number ... 28.Nm 29.Cm enable 30.Brq Cm firewall | one_pass | debug | verbose | dyn_keepalive 31.Nm 32.Cm disable 33.Brq Cm firewall | one_pass | debug | verbose | dyn_keepalive 34.Pp 35.Nm 36.Cm set Oo Cm disable Ar number ... Oc Op Cm enable Ar number ... 37.Nm 38.Cm set move 39.Op Cm rule 40.Ar number Cm to Ar number 41.Nm 42.Cm set swap Ar number number 43.Nm 44.Cm set show 45.Pp 46.Nm 47.Brq Cm pipe | queue 48.Ar number 49.Cm config 50.Ar config-options 51.Nm 52.Op Fl s Op Ar field 53.Brq Cm pipe | queue 54.Brq Cm delete | list | show 55.Op Ar number ... 56.Pp 57.Nm 58.Op Fl q 59.Oo 60.Fl p Ar preproc 61.Oo Fl D 62.Ar macro Ns Op = Ns Ar value 63.Oc 64.Op Fl U Ar macro 65.Oc 66.Ar pathname 67.Sh DESCRIPTION 68The 69.Nm 70utility is the user interface for controlling the 71.Xr ipfw 4 72firewall and the 73.Xr dummynet 4 74traffic shaper in 75.Dx . 76.Pp 77.Bd -ragged -offset XXXX 78.Em NOTE: 79this manual page documents the newer version of 80.Nm 81introduced in 82.Fx 83CURRENT in July 2002, also known as 84.Nm ipfw2 . 85.Nm ipfw2 86is a superset of the old firewall, 87.Nm ipfw1 . 88The differences between the two are listed in Section 89.Sx IPFW2 ENHANCEMENTS , 90which you are encouraged to read to revise older rulesets and possibly 91write them more efficiently. 92See Section 93.Sx USING IPFW2 IN FreeBSD-STABLE 94for instructions on how to run 95.Nm ipfw2 96on 97.Fx 98STABLE. 99.Ed 100.Pp 101An 102.Nm 103configuration, or 104.Em ruleset , 105is made of a list of 106.Em rules 107numbered from 1 to 65535. 108Packets are passed to 109.Nm 110from a number of different places in the protocol stack 111(depending on the source and destination of the packet, 112it is possible that 113.Nm 114is invoked multiple times on the same packet). 115The packet passed to the firewall is compared 116against each of the rules in the firewall 117.Em ruleset . 118When a match is found, the action corresponding to the 119matching rule is performed. 120.Pp 121Depending on the action and certain system settings, packets 122can be reinjected into the firewall at some rule after the 123matching one for further processing. 124.Pp 125An 126.Nm 127ruleset always includes a 128.Em default 129rule (numbered 65535) which cannot be modified, 130and matches all packets. 131The action associated with the 132.Em default 133rule can be either 134.Cm deny 135or 136.Cm allow 137depending on how the kernel is configured. 138.Pp 139If the ruleset includes one or more rules with the 140.Cm keep-state 141or 142.Cm limit 143option, then 144.Nm 145assumes a 146.Em stateful 147behaviour, i.e. upon a match it will create dynamic rules matching 148the exact parameters (addresses and ports) of the matching packet. 149.Pp 150These dynamic rules, which have a limited lifetime, are checked 151at the first occurrence of a 152.Cm check-state , 153.Cm keep-state 154or 155.Cm limit 156rule, and are typically used to open the firewall on-demand to 157legitimate traffic only. 158See the 159.Sx STATEFUL FIREWALL 160and 161.Sx EXAMPLES 162Sections below for more information on the stateful behaviour of 163.Nm . 164.Pp 165All rules (including dynamic ones) have a few associated counters: 166a packet count, a byte count, a log count and a timestamp 167indicating the time of the last match. 168Counters can be displayed or reset with 169.Nm 170commands. 171.Pp 172Rules can be added with the 173.Cm add 174command; deleted individually or in groups with the 175.Cm delete 176command, and globally with the 177.Cm flush 178command; displayed, optionally with the content of the 179counters, using the 180.Cm show 181and 182.Cm list 183commands. 184Finally, counters can be reset with the 185.Cm zero 186and 187.Cm resetlog 188commands. 189.Pp 190Also, each rule belongs to one of 32 different 191.Em sets 192, and there are 193.Nm 194commands to atomically manipulate sets, such as enable, 195disable, swap sets, move all rules in a set to another 196one, delete all rules in a set. These can be useful to 197install temporary configurations, or to test them. 198See Section 199.Sx SETS OF RULES 200for more information on 201.Em sets . 202.Pp 203The following options are available: 204.Bl -tag -width indent 205.It Fl a 206While listing, show counter values. 207The 208.Cm show 209command just implies this option. 210.It Fl c 211When entering or showing rules, print them in compact form, 212i.e. without the optional "ip from any to any" string 213when this does not carry any additional information. 214.It Fl d 215While listing, show dynamic rules in addition to static ones. 216.It Fl e 217While listing, if the 218.Fl d 219option was specified, also show expired dynamic rules. 220.It Fl f 221Don't ask for confirmation for commands that can cause problems 222if misused, 223.No i.e. Cm flush . 224If there is no tty associated with the process, this is implied. 225.It Fl N 226Try to resolve addresses and service names in output. 227.It Fl q 228While 229.Cm add Ns ing , 230.Cm zero Ns ing , 231.Cm resetlog Ns ging 232or 233.Cm flush Ns ing , 234be quiet about actions 235(implies 236.Fl f ) . 237This is useful for adjusting rules by executing multiple 238.Nm 239commands in a script 240(e.g., 241.Ql sh\ /etc/rc.firewall ) , 242or by processing a file of many 243.Nm 244rules across a remote login session. 245If a 246.Cm flush 247is performed in normal (verbose) mode (with the default kernel 248configuration), it prints a message. 249Because all rules are flushed, the message might not be delivered 250to the login session, causing the remote login session to be closed 251and the remainder of the ruleset to not be processed. 252Access to the console would then be required to recover. 253.It Fl S 254While listing rules, show the 255.Em set 256each rule belongs to. 257If this flag is not specified, disabled rules will not be 258listed. 259.It Fl s Op Ar field 260While listing pipes, sort according to one of the four 261counters (total or current packets or bytes). 262.It Fl t 263While listing, show last match timestamp. 264.El 265.Pp 266To ease configuration, rules can be put into a file which is 267processed using 268.Nm 269as shown in the last synopsis line. 270An absolute 271.Ar pathname 272must be used. 273The file will be read line by line and applied as arguments to the 274.Nm 275utility. 276.Pp 277Optionally, a preprocessor can be specified using 278.Fl p Ar preproc 279where 280.Ar pathname 281is to be piped through. 282Useful preprocessors include 283.Xr cpp 1 284and 285.Xr m4 1 . 286If 287.Ar preproc 288doesn't start with a slash 289.Pq Ql / 290as its first character, the usual 291.Ev PATH 292name search is performed. 293Care should be taken with this in environments where not all 294file systems are mounted (yet) by the time 295.Nm 296is being run (e.g. when they are mounted over NFS). 297Once 298.Fl p 299has been specified, optional 300.Fl D 301and 302.Fl U 303specifications can follow and will be passed on to the preprocessor. 304This allows for flexible configuration files (like conditionalizing 305them on the local hostname) and the use of macros to centralize 306frequently required arguments like IP addresses. 307.Pp 308The 309.Nm 310.Cm pipe 311and 312.Cm queue 313commands are used to configure the traffic shaper, as shown in the 314.Sx TRAFFIC SHAPER (DUMMYNET) CONFIGURATION 315Section below. 316.Pp 317If the world and the kernel get out of sync the 318.Nm 319ABI may break, preventing you from being able to add any rules. This can 320adversely effect the booting process. You can use 321.Nm 322.Cm disable 323.Cm firewall 324to temporarily disable the firewall to regain access to the network, 325allowing you to fix the problem. 326.Sh PACKET FLOW 327A packet is checked against the active ruleset in multiple places 328in the protocol stack, under control of several sysctl variables. 329These places and variables are shown below, and it is important to 330have this picture in mind in order to design a correct ruleset. 331.Bd -literal -offset indent 332 ^ to upper layers V 333 | | 334 +----------->-----------+ 335 ^ V 336 [ip_input] [ip_output] net.inet.ip.fw.enable=1 337 | | 338 ^ V 339[ether_demux] [ether_output_frame] net.link.ether.ipfw=1 340 | | 341 +-->--[bdg_forward]-->--+ net.link.ether.bridge_ipfw=1 342 ^ V 343 | to devices | 344.Ed 345.Pp 346As can be noted from the above picture, the number of 347times the same packet goes through the firewall can 348vary between 0 and 4 depending on packet source and 349destination, and system configuration. 350.Pp 351Note that as packets flow through the stack, headers can be 352stripped or added to it, and so they may or may not be available 353for inspection. 354E.g., incoming packets will include the MAC header when 355.Nm 356is invoked from 357.Cm ether_demux() , 358but the same packets will have the MAC header stripped off when 359.Nm 360is invoked from 361.Cm ip_input() . 362.Pp 363Also note that each packet is always checked against the complete ruleset, 364irrespective of the place where the check occurs, or the source of the packet. 365If a rule contains some match patterns or actions which are not valid 366for the place of invocation (e.g. trying to match a MAC header within 367.Cm ip_input() 368), the match pattern will not match, but a 369.Cm not 370operator in front of such patterns 371.Em will 372cause the pattern to 373.Em always 374match on those packets. 375It is thus the responsibility of 376the programmer, if necessary, to write a suitable ruleset to 377differentiate among the possible places. 378.Cm skipto 379rules can be useful here, as an example: 380.Bd -literal -offset indent 381# packets from ether_demux or bdg_forward 382ipfw add 10 skipto 1000 all from any to any layer2 in 383# packets from ip_input 384ipfw add 10 skipto 2000 all from any to any not layer2 in 385# packets from ip_output 386ipfw add 10 skipto 3000 all from any to any not layer2 out 387# packets from ether_output_frame 388ipfw add 10 skipto 4000 all from any to any layer2 out 389.Ed 390.Pp 391(yes, at the moment there is no way to differentiate between 392ether_demux and bdg_forward). 393.Sh RULE FORMAT 394The format of 395.Nm 396rules is the following: 397.Bd -ragged -offset indent 398.Op Ar rule_number 399.Op Cm set Ar set_number 400.Op Cm prob Ar match_probability 401.br 402.Ar " " action 403.Op Cm log Op Cm logamount Ar number 404.Ar body 405.Ed 406.Pp 407where the body of the rule specifies which information is used 408for filtering packets, among the following: 409.Pp 410.Bl -tag -width "Source and dest. addresses and ports" -offset XXX -compact 411.It Layer-2 header fields 412When available 413.It IPv4 Protocol 414TCP, UDP, ICMP, etc. 415.It Source and dest. addresses and ports 416.It Direction 417See Section 418.Sx PACKET FLOW 419.It Transmit and receive interface 420By name or address 421.It Misc. IP header fields 422Version, type of service, datagram length, identification, 423fragment flag (non-zero IP offset), 424Time To Live 425.It IP options 426.It Misc. TCP header fields 427TCP flags (SYN, FIN, ACK, RST, etc.), 428sequence number, acknowledgment number, 429window 430.It TCP options 431.It ICMP types 432for ICMP packets 433.It User/group ID 434When the packet can be associated with a local socket. 435.El 436.Pp 437Note that some of the above information, e.g. source MAC or IP addresses and 438TCP/UDP ports, could easily be spoofed, so filtering on those fields 439alone might not guarantee the desired results. 440.Bl -tag -width indent 441.It Ar rule_number 442Each rule is associated with a 443.Ar rule_number 444in the range 1..65535, with the latter reserved for the 445.Em default 446rule. 447Rules are checked sequentially by rule number. 448Multiple rules can have the same number, in which case they are 449checked (and listed) according to the order in which they have 450been added. 451If a rule is entered without specifying a number, the kernel will 452assign one in such a way that the rule becomes the last one 453before the 454.Em default 455rule. 456Automatic rule numbers are assigned by incrementing the last 457non-default rule number by the value of the sysctl variable 458.Ar net.inet.ip.fw.autoinc_step 459which defaults to 100. 460If this is not possible (e.g. because we would go beyond the 461maximum allowed rule number), the number of the last 462non-default value is used instead. 463.It Cm set Ar set_number 464Each rule is associated with a 465.Ar set_number 466in the range 0..31, with the latter reserved for the 467.Em default 468rule. 469Sets can be individually disabled and enabled, so this parameter 470is of fundamental importance for atomic ruleset manipulation. 471It can be also used to simplify deletion of groups of rules. 472If a rule is entered without specifying a set number, 473set 0 will be used. 474.It Cm prob Ar match_probability 475A match is only declared with the specified probability 476(floating point number between 0 and 1). 477This can be useful for a number of applications such as 478random packet drop or 479(in conjunction with 480.Xr dummynet 4 ) 481to simulate the effect of multiple paths leading to out-of-order 482packet delivery. 483.It Cm log Op Cm logamount Ar number 484When a packet matches a rule with the 485.Cm log 486keyword, a message will be 487logged to 488.Xr syslogd 8 489with a 490.Dv LOG_SECURITY 491facility. 492The logging only occurs if the sysctl variable 493.Em net.inet.ip.fw.verbose 494is set to 1 495(which is the default when the kernel is compiled with 496.Dv IPFIREWALL_VERBOSE 497) and the number of packets logged so far for that 498particular rule does not exceed the 499.Cm logamount 500parameter. 501If no 502.Cm logamount 503is specified, the limit is taken from the sysctl variable 504.Em net.inet.ip.fw.verbose_limit . 505In both cases, a value of 0 removes the logging limit. 506.Pp 507Once the limit is reached, logging can be re-enabled by 508clearing the logging counter or the packet counter for that entry, see the 509.Cm resetlog 510command. 511.Pp 512.El 513.Ss RULE ACTIONS 514A rule can be associated with one of the following actions, which 515will be executed when the packet matches the body of the rule. 516.Bl -tag -width indent 517.It Cm allow | accept | pass | permit 518Allow packets that match rule. 519The search terminates. 520.It Cm check-state 521Checks the packet against the dynamic ruleset. 522If a match is found, execute the action associated with 523the rule which generated this dynamic rule, otherwise 524move to the next rule. 525.br 526.Cm Check-state 527rules do not have a body. 528If no 529.Cm check-state 530rule is found, the dynamic ruleset is checked at the first 531.Cm keep-state 532or 533.Cm limit 534rule. 535.It Cm count 536Update counters for all packets that match rule. 537The search continues with the next rule. 538.It Cm deny | drop 539Discard packets that match this rule. 540The search terminates. 541.It Cm divert Ar port 542Divert packets that match this rule to the 543.Xr divert 4 544socket bound to port 545.Ar port . 546The search terminates. 547.It Cm fwd | forward Ar ipaddr Ns Op , Ns Ar port 548Change the next-hop on matching packets to 549.Ar ipaddr , 550which can be an IP address in dotted quad format or a host name. 551The search terminates if this rule matches. 552.Pp 553If 554.Ar ipaddr 555is a local address, then matching packets will be forwarded to 556.Ar port 557(or the port number in the packet if one is not specified in the rule) 558on the local machine. 559.br 560If 561.Ar ipaddr 562is not a local address, then the port number 563(if specified) is ignored, and the packet will be 564forwarded to the remote address, using the route as found in 565the local routing table for that IP. 566.br 567A 568.Ar fwd 569rule will not match layer-2 packets (those received 570on ether_input, ether_output, or bridged). 571.br 572The 573.Cm fwd 574action does not change the contents of the packet at all. 575In particular, the destination address remains unmodified, so 576packets forwarded to another system will usually be rejected by that system 577unless there is a matching rule on that system to capture them. 578For packets forwarded locally, 579the local address of the socket will be 580set to the original destination address of the packet. 581This makes the 582.Xr netstat 1 583entry look rather weird but is intended for 584use with transparent proxy servers. 585.It Cm pipe Ar pipe_nr 586Pass packet to a 587.Xr dummynet 4 588.Dq pipe 589(for bandwidth limitation, delay, etc.). 590See the 591.Sx TRAFFIC SHAPER (DUMMYNET) CONFIGURATION 592Section for further information. 593The search terminates; however, on exit from the pipe and if 594the 595.Xr sysctl 8 596variable 597.Em net.inet.ip.fw.one_pass 598is not set, the packet is passed again to the firewall code 599starting from the next rule. 600.It Cm queue Ar queue_nr 601Pass packet to a 602.Xr dummynet 4 603.Dq queue 604(for bandwidth limitation using WF2Q+). 605.It Cm reject 606(Deprecated). 607Synonym for 608.Cm unreach host . 609.It Cm reset 610Discard packets that match this rule, and if the 611packet is a TCP packet, try to send a TCP reset (RST) notice. 612The search terminates. 613.It Cm skipto Ar number 614Skip all subsequent rules numbered less than 615.Ar number . 616The search continues with the first rule numbered 617.Ar number 618or higher. 619.It Cm tee Ar port 620Send a copy of packets matching this rule to the 621.Xr divert 4 622socket bound to port 623.Ar port . 624The search terminates and the original packet is accepted 625(but see Section 626.Sx BUGS 627below). 628.It Cm unreach Ar code 629Discard packets that match this rule, and try to send an ICMP 630unreachable notice with code 631.Ar code , 632where 633.Ar code 634is a number from 0 to 255, or one of these aliases: 635.Cm net , host , protocol , port , 636.Cm needfrag , srcfail , net-unknown , host-unknown , 637.Cm isolated , net-prohib , host-prohib , tosnet , 638.Cm toshost , filter-prohib , host-precedence 639or 640.Cm precedence-cutoff . 641The search terminates. 642.El 643.Ss RULE BODY 644The body of a rule contains zero or more patterns (such as 645specific source and destination addresses or ports, 646protocol options, incoming or outgoing interfaces, etc.) 647that the packet must match in order to be recognised. 648In general, the patterns are connected by (implicit) 649.Cm and 650operators -- i.e. all must match in order for the 651rule to match. 652Individual patterns can be prefixed by the 653.Cm not 654operator to reverse the result of the match, as in 655.Pp 656.Dl "ipfw add 100 allow ip from not 1.2.3.4 to any" 657.Pp 658Additionally, sets of alternative match patterns ( 659.Em or-blocks 660) can be constructed by putting the patterns in 661lists enclosed between parentheses ( ) or braces { }, and 662using the 663.Cm or 664operator as follows: 665.Pp 666.Dl "ipfw add 100 allow ip from { x or not y or z } to any" 667.Pp 668Only one level of parentheses is allowed. 669Beware that most shells have special meanings for parentheses 670or braces, so it is advisable to put a backslash \\ in front of them 671to prevent such interpretations. 672.Pp 673The body of a rule must in general include a source and destination 674address specifier. 675The keyword 676.Ar any 677can be used in various places to specify that the content of 678a required field is irrelevant. 679.Pp 680The rule body has the following format: 681.Bd -ragged -offset indent 682.Op Ar proto Cm from Ar src Cm to Ar dst 683.Op Ar options 684.Ed 685.Pp 686The first part (protocol from src to dst) is for backward 687compatibility with 688.Nm ipfw1 . 689In 690.Nm ipfw2 691any match pattern (including MAC headers, IPv4 protocols, 692addresses and ports) can be specified in the 693.Ar options 694section. 695.Pp 696Rule fields have the following meaning: 697.Bl -tag -width indent 698.It Ar proto : protocol | Cm { Ar protocol Cm or ... } 699An IPv4 protocol (or an 700.Em or-block 701with multiple protocols) specified by number or name 702(for a complete list see 703.Pa /etc/protocols ) . 704The 705.Cm ip 706or 707.Cm all 708keywords mean any protocol will match. 709.It Ar src No and Ar dst : ip-address | Cm { Ar ip-address Cm or ... } Op Ar ports 710A single 711.Ar ip-address 712, or an 713.Em or-block 714containing one or more of them, 715optionally followed by 716.Ar ports 717specifiers. 718.It Ar ip-address : 719An address (or set of addresses) specified in one of the following 720ways, optionally preceded by a 721.Cm not 722operator: 723.Bl -tag -width indent 724.It Cm any 725matches any IP address. 726.It Cm me 727matches any IP address configured on an interface in the system. 728The address list is evaluated at the time the packet is 729analysed. 730.It Ar numeric-ip | hostname 731Matches a single IPv4 address, specified as dotted-quad or a hostname. 732Hostnames are resolved at the time the rule is added to the firewall list. 733.It Ar addr Ns / Ns Ar masklen 734Matches all addresses with base 735.Ar addr 736(specified as a dotted quad or a hostname) 737and mask width of 738.Cm masklen 739bits. 740As an example, 1.2.3.4/25 will match 741all IP numbers from 1.2.3.0 to 1.2.3.127 . 742.It Ar addr Ns / Ns Ar masklen Ns Cm { Ns Ar num,num,... Ns Cm } 743Matches all addresses with base address 744.Ar addr 745(specified as a dotted quad or a hostname) 746and whose last byte is in the list between braces { } . 747Note that there must be no spaces between braces, commas and 748numbers. 749The 750.Ar masklen 751field is used to limit the size of the set of addresses, 752and can have any value between 24 and 32. 753.br 754As an example, an address specified as 1.2.3.4/24{128,35,55,89} 755will match the following IP addresses: 756.br 7571.2.3.128 1.2.3.35 1.2.3.55 1.2.3.89 . 758.br 759This format is particularly useful to handle sparse address sets 760within a single rule. Because the matching occurs using a 761bitmask, it takes constant time and dramatically reduces 762the complexity of rulesets. 763.It Ar addr Ns : Ns Ar mask 764Matches all addresses with base 765.Ar addr 766(specified as a dotted quad or a hostname) 767and the mask of 768.Ar mask , 769specified as a dotted quad. 770As an example, 1.2.3.4/255.0.255.0 will match 7711.*.3.*. 772We suggest to use this form only for non-contiguous 773masks, and resort to the 774.Ar addr Ns / Ns Ar masklen 775format for contiguous masks, which is more compact and less 776error-prone. 777.El 778.It Ar ports : Oo Cm not Oc Bro Ar port | port Ns \&- Ns Ar port Ns Brc Op , Ns Ar ... 779For protocols which support port numbers (such as TCP and UDP), optional 780.Cm ports 781may be specified as one or more ports or port ranges, separated 782by commas but no spaces, and an optional 783.Cm not 784operator. 785The 786.Ql \&- 787notation specifies a range of ports (including boundaries). 788.Pp 789Service names (from 790.Pa /etc/services ) 791may be used instead of numeric port values. 792The length of the port list is limited to 30 ports or ranges, 793though one can specify larger ranges by using an 794.Em or-block 795in the 796.Cm options 797section of the rule. 798.Pp 799A backslash 800.Pq Ql \e 801can be used to escape the dash 802.Pq Ql - 803character in a service name (from a shell, the backslash must be 804typed twice to avoid the shell itself interpreting it as an escape 805character). 806.Pp 807.Dl "ipfw add count tcp from any ftp\e\e-data-ftp to any" 808.Pp 809Fragmented packets which have a non-zero offset (i.e. not the first 810fragment) will never match a rule which has one or more port 811specifications. 812See the 813.Cm frag 814option for details on matching fragmented packets. 815.El 816.Ss RULE OPTIONS (MATCH PATTERNS) 817Additional match patterns can be used within 818rules. Zero or more of these so-called 819.Em options 820can be present in a rule, optionally prefixed by the 821.Cm not 822operand, and possibly grouped into 823.Em or-blocks . 824.Pp 825The following match patterns can be used (listed in alphabetical order): 826.Bl -tag -width indent 827.It Cm bridged 828Matches only bridged packets. 829.It Cm dst-ip Ar ip address 830Matches IP packets whose destination IP is one of the address(es) 831specified as argument. 832.It Cm dst-port Ar source ports 833Matches IP packets whose destination port is one of the port(s) 834specified as argument. 835.It Cm established 836Matches TCP packets that have the RST or ACK bits set. 837.It Cm frag 838Matches packets that are fragments and not the first 839fragment of an IP datagram. Note that these packets will not have 840the next protocol header (e.g. TCP, UDP) so options that look into 841these headers cannot match. 842.It Cm gid Ar group 843Matches all TCP or UDP packets sent by or received for a 844.Ar group . 845A 846.Ar group 847may be specified by name or number. 848.It Cm icmptypes Ar types 849Matches ICMP packets whose ICMP type is in the list 850.Ar types . 851The list may be specified as any combination of ranges or 852individual types separated by commas. 853The supported ICMP types are: 854.Pp 855echo reply 856.Pq Cm 0 , 857destination unreachable 858.Pq Cm 3 , 859source quench 860.Pq Cm 4 , 861redirect 862.Pq Cm 5 , 863echo request 864.Pq Cm 8 , 865router advertisement 866.Pq Cm 9 , 867router solicitation 868.Pq Cm 10 , 869time-to-live exceeded 870.Pq Cm 11 , 871IP header bad 872.Pq Cm 12 , 873timestamp request 874.Pq Cm 13 , 875timestamp reply 876.Pq Cm 14 , 877information request 878.Pq Cm 15 , 879information reply 880.Pq Cm 16 , 881address mask request 882.Pq Cm 17 883and address mask reply 884.Pq Cm 18 . 885.It Cm in | out 886Matches incoming or outgoing packets, respectively. 887.Cm in 888and 889.Cm out 890are mutually exclusive (in fact, 891.Cm out 892is implemented as 893.Cm not in Ns No ). 894.It Cm ipid Ar id 895Matches IP packets whose 896.Cm ip_id 897field has value 898.Ar id . 899.It Cm iplen Ar len 900Matches IP packets whose total length, including header and data, is 901.Ar len 902bytes. 903.It Cm ipoptions Ar spec 904Matches packets whose IP header contains the comma separated list of 905options specified in 906.Ar spec . 907The supported IP options are: 908.Pp 909.Cm ssrr 910(strict source route), 911.Cm lsrr 912(loose source route), 913.Cm rr 914(record packet route) and 915.Cm ts 916(timestamp). 917The absence of a particular option may be denoted 918with a 919.Ql \&! . 920.It Cm ipprecedence Ar precedence 921Matches IP packets whose precedence field is equal to 922.Ar precedence . 923.It Cm iptos Ar spec 924Matches IP packets whose 925.Cm tos 926field contains the comma separated list of 927service types specified in 928.Ar spec . 929The supported IP types of service are: 930.Pp 931.Cm lowdelay 932.Pq Dv IPTOS_LOWDELAY , 933.Cm throughput 934.Pq Dv IPTOS_THROUGHPUT , 935.Cm reliability 936.Pq Dv IPTOS_RELIABILITY , 937.Cm mincost 938.Pq Dv IPTOS_MINCOST , 939.Cm congestion 940.Pq Dv IPTOS_CE . 941The absence of a particular type may be denoted 942with a 943.Ql \&! . 944.It Cm ipttl Ar ttl 945Matches IP packets whose time to live is 946.Ar ttl . 947.It Cm ipversion Ar ver 948Matches IP packets whose IP version field is 949.Ar ver . 950.It Cm keep-state 951Upon a match, the firewall will create a dynamic rule, whose 952default behaviour is to match bidirectional traffic between 953source and destination IP/port using the same protocol. 954The rule has a limited lifetime (controlled by a set of 955.Xr sysctl 8 956variables), and the lifetime is refreshed every time a matching 957packet is found. 958.It Cm layer2 959Matches only layer2 packets, i.e. those passed to 960.Nm 961from ether_demux() and ether_output_frame(). 962.It Cm limit Bro Cm src-addr | src-port | dst-addr | dst-port Brc Ar N 963The firewall will only allow 964.Ar N 965connections with the same 966set of parameters as specified in the rule. 967One or more 968of source and destination addresses and ports can be 969specified. 970.It Cm { MAC | mac } Ar dst-mac src-mac 971Match packets with a given 972.Ar dst-mac 973and 974.Ar src-mac 975addresses, specified as the 976.Cm any 977keyword (matching any MAC address), or six groups of hex digits 978separated by colons, 979and optionally followed by a mask indicating how many bits are 980significant, as in 981.Pp 982.Dl "MAC 10:20:30:40:50:60/33 any" 983.Pp 984Note that the order of MAC addresses (destination first, 985source second) is 986the same as on the wire, but the opposite of the one used for 987IP addresses. 988.It Cm mac-type Ar mac-type 989Matches packets whose Ethernet Type field 990corresponds to one of those specified as argument. 991.Ar mac-type 992is specified in the same way as 993.Cm port numbers 994(i.e. one or more comma-separated single values or ranges). 995You can use symbolic names for known values such as 996.Em vlan , ipv4, ipv6 . 997Values can be entered as decimal or hexadecimal (if prefixed by 0x), 998and they are always printed as hexadecimal (unless the 999.Cm -N 1000option is used, in which case symbolic resolution will be attempted). 1001.It Cm proto Ar protocol 1002Matches packets with the corresponding IPv4 protocol. 1003.It Cm recv | xmit | via Brq Ar ifX | Ar if Ns Cm * | Ar ipno | Ar any 1004Matches packets received, transmitted or going through, 1005respectively, the interface specified by exact name 1006.Ns No ( Ar ifX Ns No ), 1007by device name 1008.Ns No ( Ar if Ns Ar * Ns No ), 1009by IP address, or through some interface. 1010.Pp 1011The 1012.Cm via 1013keyword causes the interface to always be checked. 1014If 1015.Cm recv 1016or 1017.Cm xmit 1018is used instead of 1019.Cm via , 1020then only the receive or transmit interface (respectively) 1021is checked. 1022By specifying both, it is possible to match packets based on 1023both receive and transmit interface, e.g.: 1024.Pp 1025.Dl "ipfw add deny ip from any to any out recv ed0 xmit ed1" 1026.Pp 1027The 1028.Cm recv 1029interface can be tested on either incoming or outgoing packets, 1030while the 1031.Cm xmit 1032interface can only be tested on outgoing packets. 1033So 1034.Cm out 1035is required (and 1036.Cm in 1037is invalid) whenever 1038.Cm xmit 1039is used. 1040.Pp 1041A packet may not have a receive or transmit interface: packets 1042originating from the local host have no receive interface, 1043while packets destined for the local host have no transmit 1044interface. 1045.It Cm setup 1046Matches TCP packets that have the SYN bit set but no ACK bit. 1047This is the short form of 1048.Dq Li tcpflags\ syn,!ack . 1049.It Cm src-ip Ar ip-address 1050Matches IP packets whose source IP is one of the address(es) 1051specified as argument. 1052.It Cm src-port Ar ports 1053Matches IP packets whose source port is one of the port(s) 1054specified as argument. 1055.It Cm tcpack Ar ack 1056TCP packets only. 1057Match if the TCP header acknowledgment number field is set to 1058.Ar ack . 1059.It Cm tcpflags Ar spec 1060TCP packets only. 1061Match if the TCP header contains the comma separated list of 1062flags specified in 1063.Ar spec . 1064The supported TCP flags are: 1065.Pp 1066.Cm fin , 1067.Cm syn , 1068.Cm rst , 1069.Cm psh , 1070.Cm ack 1071and 1072.Cm urg . 1073The absence of a particular flag may be denoted 1074with a 1075.Ql \&! . 1076A rule which contains a 1077.Cm tcpflags 1078specification can never match a fragmented packet which has 1079a non-zero offset. 1080See the 1081.Cm frag 1082option for details on matching fragmented packets. 1083.It Cm tcpseq Ar seq 1084TCP packets only. 1085Match if the TCP header sequence number field is set to 1086.Ar seq . 1087.It Cm tcpwin Ar win 1088TCP packets only. 1089Match if the TCP header window field is set to 1090.Ar win . 1091.It Cm tcpoptions Ar spec 1092TCP packets only. 1093Match if the TCP header contains the comma separated list of 1094options specified in 1095.Ar spec . 1096The supported TCP options are: 1097.Pp 1098.Cm mss 1099(maximum segment size), 1100.Cm window 1101(tcp window advertisement), 1102.Cm sack 1103(selective ack), 1104.Cm ts 1105(rfc1323 timestamp) and 1106.Cm cc 1107(rfc1644 t/tcp connection count). 1108The absence of a particular option may be denoted 1109with a 1110.Ql \&! . 1111.It Cm uid Ar user 1112Match all TCP or UDP packets sent by or received for a 1113.Ar user . 1114A 1115.Ar user 1116may be matched by name or identification number. 1117.El 1118.Sh SETS OF RULES 1119Each rule belongs to one of 32 different 1120.Em sets 1121, numbered 0 to 31. 1122Set 31 is reserved for the default rule. 1123.Pp 1124By default, rules are put in set 0, unless you use the 1125.Cm set N 1126attribute when entering a new rule. 1127Sets can be individually and atomically enabled or disabled, 1128so this mechanism permits an easy way to store multiple configurations 1129of the firewall and quickly (and atomically) switch between them. 1130The command to enable/disable sets is 1131.Bd -ragged -offset indent 1132.Nm 1133.Cm set Oo Cm disable Ar number ... Oc Op Cm enable Ar number ... 1134.Ed 1135.Pp 1136where multiple 1137.Cm enable 1138or 1139.Cm disable 1140sections can be specified. 1141Command execution is atomic on all the sets specified in the command. 1142By default, all sets are enabled. 1143.Pp 1144When you disable a set, its rules behave as if they do not exist 1145in the firewall configuration, with only one exception: 1146.Bd -ragged -offset indent 1147dynamic rules created from a rule before it had been disabled 1148will still be active until they expire. In order to delete 1149dynamic rules you have to explicitly delete the parent rule 1150which generated them. 1151.Ed 1152.Pp 1153The set number of rules can be changed with the command 1154.Bd -ragged -offset indent 1155.Nm 1156.Cm set move 1157.Brq Cm rule Ar rule-number | old-set 1158.Cm to Ar new-set 1159.Ed 1160.Pp 1161Also, you can atomically swap two rulesets with the command 1162.Bd -ragged -offset indent 1163.Nm 1164.Cm set swap Ar first-set second-set 1165.Ed 1166.Pp 1167See the 1168.Sx EXAMPLES 1169Section on some possible uses of sets of rules. 1170.Sh STATEFUL FIREWALL 1171Stateful operation is a way for the firewall to dynamically 1172create rules for specific flows when packets that 1173match a given pattern are detected. Support for stateful 1174operation comes through the 1175.Cm check-state , keep-state 1176and 1177.Cm limit 1178options of 1179.Nm rules. 1180.Pp 1181Dynamic rules are created when a packet matches a 1182.Cm keep-state 1183or 1184.Cm limit 1185rule, causing the creation of a 1186.Em dynamic 1187rule which will match all and only packets with 1188a given 1189.Em protocol 1190between a 1191.Em src-ip/src-port dst-ip/dst-port 1192pair of addresses ( 1193.Em src 1194and 1195.Em dst 1196are used here only to denote the initial match addresses, but they 1197are completely equivalent afterwards). 1198Dynamic rules will be checked at the first 1199.Cm check-state, keep-state 1200or 1201.Cm limit 1202occurrence, and the action performed upon a match will be the same 1203as in the parent rule. 1204.Pp 1205Note that no additional attributes other than protocol and IP addresses 1206and ports are checked on dynamic rules. 1207.Pp 1208The typical use of dynamic rules is to keep a closed firewall configuration, 1209but let the first TCP SYN packet from the inside network install a 1210dynamic rule for the flow so that packets belonging to that session 1211will be allowed through the firewall: 1212.Pp 1213.Dl "ipfw add check-state" 1214.Dl "ipfw add allow tcp from my-subnet to any setup keep-state" 1215.Dl "ipfw add deny tcp from any to any" 1216.Pp 1217A similar approach can be used for UDP, where an UDP packet coming 1218from the inside will install a dynamic rule to let the response through 1219the firewall: 1220.Pp 1221.Dl "ipfw add check-state" 1222.Dl "ipfw add allow udp from my-subnet to any keep-state" 1223.Dl "ipfw add deny udp from any to any" 1224.Pp 1225Dynamic rules expire after some time, which depends on the status 1226of the flow and the setting of some 1227.Cm sysctl 1228variables. 1229See Section 1230.Sx SYSCTL VARIABLES 1231for more details. 1232For TCP sessions, dynamic rules can be instructed to periodically 1233send keepalive packets to refresh the state of the rule when it is 1234about to expire. 1235.Pp 1236See Section 1237.Sx EXAMPLES 1238for more examples on how to use dynamic rules. 1239.Sh TRAFFIC SHAPER (DUMMYNET) CONFIGURATION 1240.Nm 1241is also the user interface for the 1242.Xr dummynet 4 1243traffic shaper. 1244.Pp 1245.Nm dummynet 1246operates by first using the firewall to classify packets and divide them into 1247.Em flows , 1248using any match pattern that can be used in 1249.Nm 1250rules. 1251Depending on local policies, a flow can contain packets for a single 1252TCP connection, or from/to a given host, or entire subnet, or a 1253protocol type, etc. 1254.Pp 1255Packets belonging to the same flow are then passed to either of two 1256different objects, which implement the traffic regulation: 1257.Bl -hang -offset XXXX 1258.It Em pipe 1259A pipe emulates a link with given bandwidth, propagation delay, 1260queue size and packet loss rate. 1261Packets are queued in front of the pipe as they come out from the classifier, 1262and then transferred to the pipe according to the pipe's parameters. 1263.Pp 1264.It Em queue 1265A queue 1266is an abstraction used to implement the WF2Q+ 1267(Worst-case Fair Weighted Fair Queueing) policy, which is 1268an efficient variant of the WFQ policy. 1269.br 1270The queue associates a 1271.Em weight 1272and a reference pipe to each flow, and then all backlogged (i.e., 1273with packets queued) flows linked to the same pipe share the pipe's 1274bandwidth proportionally to their weights. 1275Note that weights are not priorities; a flow with a lower weight 1276is still guaranteed to get its fraction of the bandwidth even if a 1277flow with a higher weight is permanently backlogged. 1278.Pp 1279.El 1280In practice, 1281.Em pipes 1282can be used to set hard limits to the bandwidth that a flow can use, whereas 1283.Em queues 1284can be used to determine how different flow share the available bandwidth. 1285.Pp 1286The 1287.Em pipe 1288and 1289.Em queue 1290configuration commands are the following: 1291.Bd -ragged -offset indent 1292.Cm pipe Ar number Cm config Ar pipe-configuration 1293.Pp 1294.Cm queue Ar number Cm config Ar queue-configuration 1295.Ed 1296.Pp 1297The following parameters can be configured for a pipe: 1298.Pp 1299.Bl -tag -width indent -compact 1300.It Cm bw Ar bandwidth | device 1301Bandwidth, measured in 1302.Sm off 1303.Op Cm K | M 1304.Brq Cm bit/s | Byte/s . 1305.Sm on 1306.Pp 1307A value of 0 (default) means unlimited bandwidth. 1308The unit must immediately follow the number, as in 1309.Pp 1310.Dl "ipfw pipe 1 config bw 300Kbit/s" 1311.Pp 1312If a device name is specified instead of a numeric value, as in 1313.Pp 1314.Dl "ipfw pipe 1 config bw tun0" 1315.Pp 1316then the transmit clock is supplied by the specified device. 1317At the moment only the 1318.Xr tun 4 1319device supports this 1320functionality, for use in conjunction with 1321.Xr ppp 8 . 1322.Pp 1323.It Cm delay Ar ms-delay 1324Propagation delay, measured in milliseconds. 1325The value is rounded to the next multiple of the clock tick 1326(typically 10ms, but it is a good practice to run kernels 1327with 1328.Dq "options HZ=1000" 1329to reduce 1330the granularity to 1ms or less). 1331Default value is 0, meaning no delay. 1332.El 1333.Pp 1334The following parameters can be configured for a queue: 1335.Pp 1336.Bl -tag -width indent -compact 1337.It Cm pipe Ar pipe_nr 1338Connects a queue to the specified pipe. 1339Multiple queues (with the same or different weights) can be connected to 1340the same pipe, which specifies the aggregate rate for the set of queues. 1341.Pp 1342.It Cm weight Ar weight 1343Specifies the weight to be used for flows matching this queue. 1344The weight must be in the range 1..100, and defaults to 1. 1345.El 1346.Pp 1347Finally, the following parameters can be configured for both 1348pipes and queues: 1349.Pp 1350.Bl -tag -width XXXX -compact 1351.Pp 1352.It Cm buckets Ar hash-table-size 1353Specifies the size of the hash table used for storing the 1354various queues. 1355Default value is 64 controlled by the 1356.Xr sysctl 8 1357variable 1358.Em net.inet.ip.dummynet.hash_size , 1359allowed range is 16 to 65536. 1360.Pp 1361.It Cm mask Ar mask-specifier 1362Packets sent to a given pipe or queue by an 1363.Nm 1364rule can be further classified into multiple flows, each of which is then 1365sent to a different 1366.Em dynamic 1367pipe or queue. 1368A flow identifier is constructed by masking the IP addresses, 1369ports and protocol types as specified with the 1370.Cm mask 1371options in the configuration of the pipe or queue. 1372For each different flow identifier, a new pipe or queue is created 1373with the same parameters as the original object, and matching packets 1374are sent to it. 1375.Pp 1376Thus, when 1377.Em dynamic pipes 1378are used, each flow will get the same bandwidth as defined by the pipe, 1379whereas when 1380.Em dynamic queues 1381are used, each flow will share the parent's pipe bandwidth evenly 1382with other flows generated by the same queue (note that other queues 1383with different weights might be connected to the same pipe). 1384.br 1385Available mask specifiers are a combination of one or more of the following: 1386.Pp 1387.Cm dst-ip Ar mask , 1388.Cm src-ip Ar mask , 1389.Cm dst-port Ar mask , 1390.Cm src-port Ar mask , 1391.Cm proto Ar mask 1392or 1393.Cm all , 1394.Pp 1395where the latter means all bits in all fields are significant. 1396.Pp 1397.It Cm noerror 1398When a packet is dropped by a dummynet queue or pipe, the error 1399is normally reported to the caller routine in the kernel, in the 1400same way as it happens when a device queue fills up. Setting this 1401option reports the packet as successfully delivered, which can be 1402needed for some experimental setups where you want to simulate 1403loss or congestion at a remote router. 1404.Pp 1405.It Cm plr Ar packet-loss-rate 1406Packet loss rate. 1407Argument 1408.Ar packet-loss-rate 1409is a floating-point number between 0 and 1, with 0 meaning no 1410loss, 1 meaning 100% loss. 1411The loss rate is internally represented on 31 bits. 1412.Pp 1413.It Cm queue Brq Ar slots | size Ns Cm Kbytes 1414Queue size, in 1415.Ar slots 1416or 1417.Cm KBytes . 1418Default value is 50 slots, which 1419is the typical queue size for Ethernet devices. 1420Note that for slow speed links you should keep the queue 1421size short or your traffic might be affected by a significant 1422queueing delay. 1423E.g., 50 max-sized ethernet packets (1500 bytes) mean 600Kbit 1424or 20s of queue on a 30Kbit/s pipe. 1425Even worse effect can result if you get packets from an 1426interface with a much larger MTU, e.g. the loopback interface 1427with its 16KB packets. 1428.Pp 1429.It Cm red | gred Ar w_q Ns / Ns Ar min_th Ns / Ns Ar max_th Ns / Ns Ar max_p 1430Make use of the RED (Random Early Detection) queue management algorithm. 1431.Ar w_q 1432and 1433.Ar max_p 1434are floating 1435point numbers between 0 and 1 (0 not included), while 1436.Ar min_th 1437and 1438.Ar max_th 1439are integer numbers specifying thresholds for queue management 1440(thresholds are computed in bytes if the queue has been defined 1441in bytes, in slots otherwise). 1442The 1443.Xr dummynet 4 1444also supports the gentle RED variant (gred). 1445Three 1446.Xr sysctl 8 1447variables can be used to control the RED behaviour: 1448.Bl -tag -width indent 1449.It Em net.inet.ip.dummynet.red_lookup_depth 1450specifies the accuracy in computing the average queue 1451when the link is idle (defaults to 256, must be greater than zero) 1452.It Em net.inet.ip.dummynet.red_avg_pkt_size 1453specifies the expected average packet size (defaults to 512, must be 1454greater than zero) 1455.It Em net.inet.ip.dummynet.red_max_pkt_size 1456specifies the expected maximum packet size, only used when queue 1457thresholds are in bytes (defaults to 1500, must be greater than zero). 1458.El 1459.El 1460.Sh CHECKLIST 1461Here are some important points to consider when designing your 1462rules: 1463.Bl -bullet 1464.It 1465Remember that you filter both packets going 1466.Cm in 1467and 1468.Cm out . 1469Most connections need packets going in both directions. 1470.It 1471Remember to test very carefully. 1472It is a good idea to be near the console when doing this. 1473If you cannot be near the console, 1474use an auto-recovery script such as the one in 1475.Pa /usr/share/examples/ipfw/change_rules.sh . 1476.It 1477Don't forget the loopback interface. 1478.El 1479.Sh FINE POINTS 1480.Bl -bullet 1481.It 1482There are circumstances where fragmented datagrams are unconditionally 1483dropped. 1484TCP packets are dropped if they do not contain at least 20 bytes of 1485TCP header, UDP packets are dropped if they do not contain a full 8 1486byte UDP header, and ICMP packets are dropped if they do not contain 14874 bytes of ICMP header, enough to specify the ICMP type, code, and 1488checksum. 1489These packets are simply logged as 1490.Dq pullup failed 1491since there may not be enough good data in the packet to produce a 1492meaningful log entry. 1493.It 1494Another type of packet is unconditionally dropped, a TCP packet with a 1495fragment offset of one. 1496This is a valid packet, but it only has one use, to try 1497to circumvent firewalls. 1498When logging is enabled, these packets are 1499reported as being dropped by rule -1. 1500.It 1501If you are logged in over a network, loading the 1502.Xr kld 4 1503version of 1504.Nm 1505is probably not as straightforward as you would think. 1506I recommend the following command line: 1507.Bd -literal -offset indent 1508kldload /modules/ipfw.ko && \e 1509ipfw add 32000 allow ip from any to any 1510.Ed 1511.Pp 1512Along the same lines, doing an 1513.Bd -literal -offset indent 1514ipfw flush 1515.Ed 1516.Pp 1517in similar surroundings is also a bad idea. 1518.It 1519The 1520.Nm 1521filter list may not be modified if the system security level 1522is set to 3 or higher 1523(see 1524.Xr init 8 1525for information on system security levels). 1526.El 1527.Sh PACKET DIVERSION 1528A 1529.Xr divert 4 1530socket bound to the specified port will receive all packets 1531diverted to that port. 1532If no socket is bound to the destination port, or if the kernel 1533wasn't compiled with divert socket support, the packets are 1534dropped. 1535.Sh SYSCTL VARIABLES 1536A set of 1537.Xr sysctl 8 1538variables controls the behaviour of the firewall and 1539associated modules ( 1540.Nm dummynet, bridge 1541). 1542These are shown below together with their default value 1543(but always check with the 1544.Xr sysctl 8 1545command what value is actually in use) and meaning: 1546.Bl -tag -width indent 1547.It Em net.inet.ip.dummynet.expire : No 1 1548Lazily delete dynamic pipes/queue once they have no pending traffic. 1549You can disable this by setting the variable to 0, in which case 1550the pipes/queues will only be deleted when the threshold is reached. 1551.It Em net.inet.ip.dummynet.hash_size : No 64 1552Default size of the hash table used for dynamic pipes/queues. 1553This value is used when no 1554.Cm buckets 1555option is specified when configuring a pipe/queue. 1556.It Em net.inet.ip.dummynet.max_chain_len : No 16 1557Target value for the maximum number of pipes/queues in a hash bucket. 1558The product 1559.Cm max_chain_len*hash_size 1560is used to determine the threshold over which empty pipes/queues 1561will be expired even when 1562.Cm net.inet.ip.dummynet.expire=0 . 1563.It Em net.inet.ip.dummynet.red_lookup_depth : No 256 1564.It Em net.inet.ip.dummynet.red_avg_pkt_size : No 512 1565.It Em net.inet.ip.dummynet.red_max_pkt_size : No 1500 1566Parameters used in the computations of the drop probability 1567for the RED algorithm. 1568.It Em net.inet.ip.fw.autoinc_step : No 100 1569Delta between rule numbers when auto-generating them. 1570The value must be in the range 1..1000. 1571.It Em net.inet.ip.fw.curr_dyn_buckets : Em net.inet.ip.fw.dyn_buckets 1572The current number of buckets in the hash table for dynamic rules 1573(readonly). 1574.It Em net.inet.ip.fw.debug : No 1 1575Controls debugging messages produced by 1576.Nm . 1577.It Em net.inet.ip.fw.dyn_buckets : No 256 1578The number of buckets in the hash table for dynamic rules. 1579Must be a power of 2, up to 65536. 1580It only takes effect when all dynamic rules have expired, so you 1581are advised to use a 1582.Cm flush 1583command to make sure that the hash table is resized. 1584.It Em net.inet.ip.fw.dyn_count : No 3 1585Current number of dynamic rules 1586(read-only). 1587.It Em net.inet.ip.fw.dyn_keepalive : No 1 1588Enables generation of keepalive packets for 1589.Cm keep-state 1590rules on TCP sessions. A keepalive is generated to both 1591sides of the connection every 5 seconds for the last 20 1592seconds of the lifetime of the rule. 1593.It Em net.inet.ip.fw.dyn_max : No 8192 1594Maximum number of dynamic rules. 1595When you hit this limit, no more dynamic rules can be 1596installed until old ones expire. 1597.It Em net.inet.ip.fw.dyn_ack_lifetime : No 300 1598.It Em net.inet.ip.fw.dyn_syn_lifetime : No 20 1599.It Em net.inet.ip.fw.dyn_fin_lifetime : No 1 1600.It Em net.inet.ip.fw.dyn_rst_lifetime : No 1 1601.It Em net.inet.ip.fw.dyn_udp_lifetime : No 5 1602.It Em net.inet.ip.fw.dyn_short_lifetime : No 30 1603These variables control the lifetime, in seconds, of dynamic 1604rules. 1605Upon the initial SYN exchange the lifetime is kept short, 1606then increased after both SYN have been seen, then decreased 1607again during the final FIN exchange or when a RST is received. 1608Both 1609.Em dyn_fin_lifetime 1610and 1611.Em dyn_rst_lifetime 1612must be strictly lower than 5 seconds, the period of 1613repetition of keepalives. The firewall enforces that. 1614.It Em net.inet.ip.fw.enable : No 1 1615Enables the firewall. 1616Setting this variable to 0 lets you run your machine without 1617firewall even if compiled in. 1618.It Em net.inet.ip.fw.one_pass : No 1 1619When set, the packet exiting from the 1620.Xr dummynet 4 1621pipe is not passed though the firewall again. 1622Otherwise, after a pipe action, the packet is 1623reinjected into the firewall at the next rule. 1624.Pp 1625Note: bridged and layer 2 packets coming out of a pipe 1626are never reinjected in the firewall irrespective of the 1627value of this variable. 1628.It Em net.inet.ip.fw.verbose : No 1 1629Enables verbose messages. 1630.It Em net.inet.ip.fw.verbose_limit : No 0 1631Limits the number of messages produced by a verbose firewall. 1632.It Em net.link.ether.ipfw : No 0 1633Controls whether layer-2 packets are passed to 1634.Nm . 1635Default is no. 1636.It Em net.link.ether.bridge_ipfw : No 0 1637Controls whether bridged packets are passed to 1638.Nm . 1639Default is no. 1640.El 1641.Sh USING IPFW2 IN FreeBSD-STABLE 1642.Nm ipfw2 1643is standard in 1644.Fx 1645CURRENT, whereas 1646.Fx 1647STABLE still uses 1648.Nm ipfw1 1649unless the kernel is compiled with 1650.Cm options IPFW2 , 1651and 1652.Nm /sbin/ipfw 1653and 1654.Nm /usr/lib/libalias 1655are recompiled with 1656.Cm -DIPFW2 1657and reinstalled (the same effect can be achieved by adding 1658.Cm IPFW2=TRUE 1659to 1660.Nm /etc/make.conf 1661before a buildworld). 1662.Pp 1663.Sh IPFW2 ENHANCEMENTS 1664This Section lists the features that have been introduced in 1665.Nm ipfw2 1666which were not present in 1667.Nm ipfw1 . 1668We list them in order of the potential impact that they can 1669have in writing your rulesets. 1670You might want to consider using these features in order to 1671write your rulesets in a more efficient way. 1672.Bl -tag -width indent 1673.It Handling of non-IPv4 packets 1674.Nm ipfw1 1675will silently accept all non-IPv4 packets (which 1676.Nm ipfw1 1677will only see when 1678.Em net.link.ether.bridge_ipfw=1 Ns 1679). 1680.Nm ipfw2 1681will filter all packets (including non-IPv4 ones) according to the ruleset. 1682To achieve the same behaviour as 1683.Nm ipfw1 1684you can use the following as the very first rule in your ruleset: 1685.Pp 1686.Dl "ipfw add 1 allow layer2 not mac-type ip" 1687.Pp 1688The 1689.Cm layer2 1690option might seem redundant, but it is necessary -- packets 1691passed to the firewall from layer3 will not have a MAC header, 1692so the 1693.Cm mac-type ip 1694pattern will always fail on them, and the 1695.Cm not 1696operator will make this rule into a pass-all. 1697.It Address sets 1698.Nm ipfw1 1699does not supports address sets (those in the form 1700.Ar addr/masklen{num,num,...} 1701). 1702.Pp 1703.It Port specifications 1704.Nm ipfw1 1705only allows one port range when specifying TCP and UDP ports, and 1706is limited to 10 entries instead of the 15 allowed by 1707.Nm ipfw2 . 1708Also, in 1709.Nm ipfw1 1710you can only specify ports when the rule is requesting 1711.Cm tcp 1712or 1713.Cm udp 1714packets. With 1715.Nm ipfw2 1716you can put port specifications in rules matching all packets, 1717and the match will be attempted only on those packets carrying 1718protocols which include port identifiers. 1719.Pp 1720Finally, 1721.Nm ipfw1 1722allowed the first port entry to be specified as 1723.Ar port:mask 1724where 1725.Ar mask 1726can be an arbitrary 16-bit mask. 1727This syntax is of questionable usefulness and it is not 1728supported anymore in 1729.Nm ipfw2 . 1730.It Or-blocks 1731.Nm ipfw1 1732does not support Or-blocks. 1733.It keepalives 1734.Nm ipfw1 1735does not generate keepalives for stateful sessions. 1736As a consequence, it might cause idle sessions to drop because 1737the lifetime of the dynamic rules expires. 1738.It Sets of rules 1739.Nm ipfw1 1740does not implement sets of rules. 1741.It MAC header filtering and Layer-2 firewalling. 1742.Nm ipfw1 1743does not implement filtering on MAC header fields, nor is it 1744invoked on packets from 1745.Cm ether_demux() 1746and 1747.Cm ether_output_frame(). 1748The sysctl variable 1749.Em net.link.ether.ipfw 1750has no effect there. 1751.It Options 1752The following options are not supported in 1753.Nm ipfw1 1754.Pp 1755.Cm dst-ip, dst-port, layer2, mac, mac-type, src-ip, src-port. 1756.Pp 1757Additionally, the following options are not supported in 1758.Nm ipfw1 1759(RELENG_4) 1760rules: 1761.Pp 1762.Cm ipid, iplen, ipprecedence, iptos, ipttl, 1763.Cm ipversion, tcpack, tcpseq, tcpwin . 1764.It Dummynet options 1765The following option for 1766.Nm dummynet 1767pipes/queues is not supported: 1768.Cm noerror . 1769.El 1770.Sh EXAMPLES 1771There are far too many possible uses of 1772.Nm 1773so this Section will only give a small set of examples. 1774.Pp 1775.Ss BASIC PACKET FILTERING 1776This command adds an entry which denies all tcp packets from 1777.Em cracker.evil.org 1778to the telnet port of 1779.Em wolf.tambov.su 1780from being forwarded by the host: 1781.Pp 1782.Dl "ipfw add deny tcp from cracker.evil.org to wolf.tambov.su telnet" 1783.Pp 1784This one disallows any connection from the entire cracker's 1785network to my host: 1786.Pp 1787.Dl "ipfw add deny ip from 123.45.67.0/24 to my.host.org" 1788.Pp 1789A first and efficient way to limit access (not using dynamic rules) 1790is the use of the following rules: 1791.Pp 1792.Dl "ipfw add allow tcp from any to any established" 1793.Dl "ipfw add allow tcp from net1 portlist1 to net2 portlist2 setup" 1794.Dl "ipfw add allow tcp from net3 portlist3 to net3 portlist3 setup" 1795.Dl "..." 1796.Dl "ipfw add deny tcp from any to any" 1797.Pp 1798The first rule will be a quick match for normal TCP packets, 1799but it will not match the initial SYN packet, which will be 1800matched by the 1801.Cm setup 1802rules only for selected source/destination pairs. 1803All other SYN packets will be rejected by the final 1804.Cm deny 1805rule. 1806.Pp 1807If you administer one or more subnets, you can take advantage of the 1808.Nm ipfw2 1809syntax to specify address sets and or-blocks and write extremely 1810compact rulesets which selectively enable services to blocks 1811of clients, as below: 1812.Pp 1813.Dl "goodguys=\*q{ 10.1.2.0/24{20,35,66,18} or 10.2.3.0/28{6,3,11} }\*q" 1814.Dl "badguys=\*q10.1.2.0/24{8,38,60}\*q" 1815.Dl "" 1816.Dl "ipfw add allow ip from ${goodguys} to any" 1817.Dl "ipfw add deny ip from ${badguys} to any" 1818.Dl "... normal policies ..." 1819.Pp 1820The 1821.Nm ipfw1 1822syntax would require a separate rule for each IP in the above 1823example. 1824.Ss DYNAMIC RULES 1825In order to protect a site from flood attacks involving fake 1826TCP packets, it is safer to use dynamic rules: 1827.Pp 1828.Dl "ipfw add check-state" 1829.Dl "ipfw add deny tcp from any to any established" 1830.Dl "ipfw add allow tcp from my-net to any setup keep-state" 1831.Pp 1832This will let the firewall install dynamic rules only for 1833those connection which start with a regular SYN packet coming 1834from the inside of our network. 1835Dynamic rules are checked when encountering the first 1836.Cm check-state 1837or 1838.Cm keep-state 1839rule. 1840A 1841.Cm check-state 1842rule should usually be placed near the beginning of the 1843ruleset to minimize the amount of work scanning the ruleset. 1844Your mileage may vary. 1845.Pp 1846To limit the number of connections a user can open 1847you can use the following type of rules: 1848.Pp 1849.Dl "ipfw add allow tcp from my-net/24 to any setup limit src-addr 10" 1850.Dl "ipfw add allow tcp from any to me setup limit src-addr 4" 1851.Pp 1852The former (assuming it runs on a gateway) will allow each host 1853on a /24 network to open at most 10 TCP connections. 1854The latter can be placed on a server to make sure that a single 1855client does not use more than 4 simultaneous connections. 1856.Pp 1857.Em BEWARE : 1858stateful rules can be subject to denial-of-service attacks 1859by a SYN-flood which opens a huge number of dynamic rules. 1860The effects of such attacks can be partially limited by 1861acting on a set of 1862.Xr sysctl 8 1863variables which control the operation of the firewall. 1864.Pp 1865Here is a good usage of the 1866.Cm list 1867command to see accounting records and timestamp information: 1868.Pp 1869.Dl ipfw -at list 1870.Pp 1871or in short form without timestamps: 1872.Pp 1873.Dl ipfw -a list 1874.Pp 1875which is equivalent to: 1876.Pp 1877.Dl ipfw show 1878.Pp 1879Next rule diverts all incoming packets from 192.168.2.0/24 1880to divert port 5000: 1881.Pp 1882.Dl ipfw divert 5000 ip from 192.168.2.0/24 to any in 1883.Pp 1884.Ss TRAFFIC SHAPING 1885The following rules show some of the applications of 1886.Nm 1887and 1888.Xr dummynet 4 1889for simulations and the like. 1890.Pp 1891This rule drops random incoming packets with a probability 1892of 5%: 1893.Pp 1894.Dl "ipfw add prob 0.05 deny ip from any to any in" 1895.Pp 1896A similar effect can be achieved making use of dummynet pipes: 1897.Pp 1898.Dl "ipfw add pipe 10 ip from any to any" 1899.Dl "ipfw pipe 10 config plr 0.05" 1900.Pp 1901We can use pipes to artificially limit bandwidth, e.g. on a 1902machine acting as a router, if we want to limit traffic from 1903local clients on 192.168.2.0/24 we do: 1904.Pp 1905.Dl "ipfw add pipe 1 ip from 192.168.2.0/24 to any out" 1906.Dl "ipfw pipe 1 config bw 300Kbit/s queue 50KBytes" 1907.Pp 1908note that we use the 1909.Cm out 1910modifier so that the rule is not used twice. 1911Remember in fact that 1912.Nm 1913rules are checked both on incoming and outgoing packets. 1914.Pp 1915Should we want to simulate a bidirectional link with bandwidth 1916limitations, the correct way is the following: 1917.Pp 1918.Dl "ipfw add pipe 1 ip from any to any out" 1919.Dl "ipfw add pipe 2 ip from any to any in" 1920.Dl "ipfw pipe 1 config bw 64Kbit/s queue 10Kbytes" 1921.Dl "ipfw pipe 2 config bw 64Kbit/s queue 10Kbytes" 1922.Pp 1923The above can be very useful, e.g. if you want to see how 1924your fancy Web page will look for a residential user who 1925is connected only through a slow link. 1926You should not use only one pipe for both directions, unless 1927you want to simulate a half-duplex medium (e.g. AppleTalk, 1928Ethernet, IRDA). 1929It is not necessary that both pipes have the same configuration, 1930so we can also simulate asymmetric links. 1931.Pp 1932Should we want to verify network performance with the RED queue 1933management algorithm: 1934.Pp 1935.Dl "ipfw add pipe 1 ip from any to any" 1936.Dl "ipfw pipe 1 config bw 500Kbit/s queue 100 red 0.002/30/80/0.1" 1937.Pp 1938Another typical application of the traffic shaper is to 1939introduce some delay in the communication. 1940This can significantly affect applications which do a lot of Remote 1941Procedure Calls, and where the round-trip-time of the 1942connection often becomes a limiting factor much more than 1943bandwidth: 1944.Pp 1945.Dl "ipfw add pipe 1 ip from any to any out" 1946.Dl "ipfw add pipe 2 ip from any to any in" 1947.Dl "ipfw pipe 1 config delay 250ms bw 1Mbit/s" 1948.Dl "ipfw pipe 2 config delay 250ms bw 1Mbit/s" 1949.Pp 1950Per-flow queueing can be useful for a variety of purposes. 1951A very simple one is counting traffic: 1952.Pp 1953.Dl "ipfw add pipe 1 tcp from any to any" 1954.Dl "ipfw add pipe 1 udp from any to any" 1955.Dl "ipfw add pipe 1 ip from any to any" 1956.Dl "ipfw pipe 1 config mask all" 1957.Pp 1958The above set of rules will create queues (and collect 1959statistics) for all traffic. 1960Because the pipes have no limitations, the only effect is 1961collecting statistics. 1962Note that we need 3 rules, not just the last one, because 1963when 1964.Nm 1965tries to match IP packets it will not consider ports, so we 1966would not see connections on separate ports as different 1967ones. 1968.Pp 1969A more sophisticated example is limiting the outbound traffic 1970on a net with per-host limits, rather than per-network limits: 1971.Pp 1972.Dl "ipfw add pipe 1 ip from 192.168.2.0/24 to any out" 1973.Dl "ipfw add pipe 2 ip from any to 192.168.2.0/24 in" 1974.Dl "ipfw pipe 1 config mask src-ip 0x000000ff bw 200Kbit/s queue 20Kbytes" 1975.Dl "ipfw pipe 2 config mask dst-ip 0x000000ff bw 200Kbit/s queue 20Kbytes" 1976.Ss SETS OF RULES 1977To add a set of rules atomically, e.g. set 18: 1978.Pp 1979.Dl "ipfw disable set 18" 1980.Dl "ipfw add NN set 18 ... # repeat as needed" 1981.Dl "ipfw enable set 18" 1982.Pp 1983To delete a set of rules atomically the command is simply: 1984.Pp 1985.Dl "ipfw delete set 18" 1986.Pp 1987To test a ruleset and disable it and regain control if something goes wrong: 1988.Pp 1989.Dl "ipfw disable set 18" 1990.Dl "ipfw add NN set 18 ... # repeat as needed" 1991.Dl "ipfw enable set 18 ; echo done; sleep 30 && ipfw disable set 18" 1992.Pp 1993Here if everything goes well, you press control-C before the "sleep" 1994terminates, and your ruleset will be left active. Otherwise, e.g. if 1995you cannot access your box, the ruleset will be disabled after 1996the sleep terminates thus restoring the previous situation. 1997.Sh SEE ALSO 1998.Xr cpp 1 , 1999.Xr m4 1 , 2000.Xr bridge 4 , 2001.Xr divert 4 , 2002.Xr dummynet 4 , 2003.Xr ip 4 , 2004.Xr ipfirewall 4 , 2005.Xr protocols 5 , 2006.Xr services 5 , 2007.Xr init 8 , 2008.Xr kldload 8 , 2009.Xr reboot 8 , 2010.Xr sysctl 8 , 2011.Xr syslogd 8 2012.Sh BUGS 2013The syntax has grown over the years and sometimes it might be confusing. 2014Unfortunately, backward compatibility prevents cleaning up mistakes 2015made in the definition of the syntax. 2016.Pp 2017.Em !!! WARNING !!! 2018.Pp 2019Misconfiguring the firewall can put your computer in an unusable state, 2020possibly shutting down network services and requiring console access to 2021regain control of it. 2022.Pp 2023Incoming packet fragments diverted by 2024.Cm divert 2025or 2026.Cm tee 2027are reassembled before delivery to the socket. 2028The action used on those packet is the one from the 2029rule which matches the first fragment of the packet. 2030.Pp 2031Packets that match a 2032.Cm tee 2033rule should not be immediately accepted, but should continue 2034going through the rule list. 2035This may be fixed in a later version. 2036.Pp 2037Packets diverted to userland, and then reinserted by a userland process 2038(such as 2039.Xr natd 8 ) 2040will lose various packet attributes, including their source interface. 2041If a packet is reinserted in this manner, later rules may be incorrectly 2042applied, making the order of 2043.Cm divert 2044rules in the rule sequence very important. 2045.Sh AUTHORS 2046.An Ugen J. S. Antsilevich , 2047.An Poul-Henning Kamp , 2048.An Alex Nash , 2049.An Archie Cobbs , 2050.An Luigi Rizzo . 2051.Pp 2052.An -nosplit 2053API based upon code written by 2054.An Daniel Boulet 2055for BSDI. 2056.Pp 2057Work on 2058.Xr dummynet 4 2059traffic shaper supported by Akamba Corp. 2060.Sh HISTORY 2061The 2062.Nm 2063utility first appeared in 2064.Fx 2.0 . 2065.Xr dummynet 4 2066was introduced in 2067.Fx 2.2.8 . 2068Stateful extensions were introduced in 2069.Fx 4.0 . 2070.Nm ipfw2 2071was introduced in Summer 2002. 2072