xref: /freebsd/sbin/ipfw/ipfw.8 (revision 61af1d13936ec56808f62d13dd8698f73b440dc1)
1.\"
2.\" $FreeBSD$
3.\"
4.Dd November 10, 2011
5.Dt IPFW 8
6.Os
7.Sh NAME
8.Nm ipfw
9.Nd User interface for firewall, traffic shaper, packet scheduler,
10in-kernel NAT.
11.Sh SYNOPSIS
12.Ss FIREWALL CONFIGURATION
13.Nm
14.Op Fl cq
15.Cm add
16.Ar rule
17.Nm
18.Op Fl acdefnNStT
19.Op Cm set Ar N
20.Brq Cm list | show
21.Op Ar rule | first-last ...
22.Nm
23.Op Fl f | q
24.Op Cm set Ar N
25.Cm flush
26.Nm
27.Op Fl q
28.Op Cm set Ar N
29.Brq Cm delete | zero | resetlog
30.Op Ar number ...
31.Pp
32.Nm
33.Cm set Oo Cm disable Ar number ... Oc Op Cm enable Ar number ...
34.Nm
35.Cm set move
36.Op Cm rule
37.Ar number Cm to Ar number
38.Nm
39.Cm set swap Ar number number
40.Nm
41.Cm set show
42.Ss SYSCTL SHORTCUTS
43.Pp
44.Nm
45.Cm enable
46.Brq Cm firewall | altq | one_pass | debug | verbose | dyn_keepalive
47.Nm
48.Cm disable
49.Brq Cm firewall | altq | one_pass | debug | verbose | dyn_keepalive
50.Pp
51.Ss LOOKUP TABLES
52.Nm
53.Cm table Ar number Cm add Ar addr Ns Oo / Ns Ar masklen Oc Op Ar value
54.Nm
55.Cm table Ar number Cm delete Ar addr Ns Op / Ns Ar masklen
56.Nm
57.Cm table
58.Brq Ar number | all
59.Cm flush
60.Nm
61.Cm table
62.Brq Ar number | all
63.Cm list
64.Pp
65.Ss DUMMYNET CONFIGURATION (TRAFFIC SHAPER AND PACKET SCHEDULER)
66.Nm
67.Brq Cm pipe | queue | sched
68.Ar number
69.Cm config
70.Ar config-options
71.Nm
72.Op Fl s Op Ar field
73.Brq Cm pipe | queue | sched
74.Brq Cm delete | list | show
75.Op Ar number ...
76.Pp
77.Ss IN-KERNEL NAT
78.Nm
79.Op Fl q
80.Cm nat
81.Ar number
82.Cm config
83.Ar config-options
84.Pp
85.Nm
86.Op Fl cfnNqS
87.Oo
88.Fl p Ar preproc
89.Oo
90.Ar preproc-flags
91.Oc
92.Oc
93.Ar pathname
94.Sh DESCRIPTION
95The
96.Nm
97utility is the user interface for controlling the
98.Xr ipfw 4
99firewall, the
100.Xr dummynet 4
101traffic shaper/packet scheduler, and the
102in-kernel NAT services.
103.Pp
104A firewall configuration, or
105.Em ruleset ,
106is made of a list of
107.Em rules
108numbered from 1 to 65535.
109Packets are passed to the firewall
110from a number of different places in the protocol stack
111(depending on the source and destination of the packet,
112it is possible for the firewall to be
113invoked multiple times on the same packet).
114The packet passed to the firewall is compared
115against each of the rules in the
116.Em ruleset ,
117in rule-number order
118(multiple rules with the same number are permitted, in which case
119they are processed in order of insertion).
120When a match is found, the action corresponding to the
121matching rule is performed.
122.Pp
123Depending on the action and certain system settings, packets
124can be reinjected into the firewall at some rule after the
125matching one for further processing.
126.Pp
127A ruleset always includes a
128.Em default
129rule (numbered 65535) which cannot be modified or deleted,
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,
144the firewall will have a
145.Em stateful
146behaviour, i.e., upon a match it will create
147.Em dynamic rules ,
148i.e. rules that match packets with the same 5-tuple
149(protocol, source and destination addresses and ports)
150as the packet which caused their creation.
151Dynamic rules, which have a limited lifetime, are checked
152at the first occurrence of a
153.Cm check-state ,
154.Cm keep-state
155or
156.Cm limit
157rule, and are typically used to open the firewall on-demand to
158legitimate traffic only.
159See the
160.Sx STATEFUL FIREWALL
161and
162.Sx EXAMPLES
163Sections below for more information on the stateful behaviour of
164.Nm .
165.Pp
166All rules (including dynamic ones) have a few associated counters:
167a packet count, a byte count, a log count and a timestamp
168indicating the time of the last match.
169Counters can be displayed or reset with
170.Nm
171commands.
172.Pp
173Each rule belongs to one of 32 different
174.Em sets
175, and there are
176.Nm
177commands to atomically manipulate sets, such as enable,
178disable, swap sets, move all rules in a set to another
179one, delete all rules in a set.
180These can be useful to
181install temporary configurations, or to test them.
182See Section
183.Sx SETS OF RULES
184for more information on
185.Em sets .
186.Pp
187Rules can be added with the
188.Cm add
189command; deleted individually or in groups with the
190.Cm delete
191command, and globally (except those in set 31) with the
192.Cm flush
193command; displayed, optionally with the content of the
194counters, using the
195.Cm show
196and
197.Cm list
198commands.
199Finally, counters can be reset with the
200.Cm zero
201and
202.Cm resetlog
203commands.
204.Pp
205.Ss COMMAND OPTIONS
206The following general options are available when invoking
207.Nm :
208.Bl -tag -width indent
209.It Fl a
210Show counter values when listing rules.
211The
212.Cm show
213command implies this option.
214.It Fl b
215Only show the action and the comment, not the body of a rule.
216Implies
217.Fl c .
218.It Fl c
219When entering or showing rules, print them in compact form,
220i.e., omitting the "ip from any to any" string
221when this does not carry any additional information.
222.It Fl d
223When listing, show dynamic rules in addition to static ones.
224.It Fl e
225When listing and
226.Fl d
227is specified, also show expired dynamic rules.
228.It Fl f
229Do not ask for confirmation for commands that can cause problems
230if misused,
231.No i.e. Cm flush .
232If there is no tty associated with the process, this is implied.
233.It Fl i
234When listing a table (see the
235.Sx LOOKUP TABLES
236section below for more information on lookup tables), format values
237as IP addresses. By default, values are shown as integers.
238.It Fl n
239Only check syntax of the command strings, without actually passing
240them to the kernel.
241.It Fl N
242Try to resolve addresses and service names in output.
243.It Fl q
244Be quiet when executing the
245.Cm add ,
246.Cm nat ,
247.Cm zero ,
248.Cm resetlog
249or
250.Cm flush
251commands;
252(implies
253.Fl f ) .
254This is useful when updating rulesets by executing multiple
255.Nm
256commands in a script
257(e.g.,
258.Ql sh\ /etc/rc.firewall ) ,
259or by processing a file with many
260.Nm
261rules across a remote login session.
262It also stops a table add or delete
263from failing if the entry already exists or is not present.
264.Pp
265The reason why this option may be important is that
266for some of these actions,
267.Nm
268may print a message; if the action results in blocking the
269traffic to the remote client,
270the remote login session will be closed
271and the rest of the ruleset will not be processed.
272Access to the console would then be required to recover.
273.It Fl S
274When listing rules, show the
275.Em set
276each rule belongs to.
277If this flag is not specified, disabled rules will not be
278listed.
279.It Fl s Op Ar field
280When listing pipes, sort according to one of the four
281counters (total or current packets or bytes).
282.It Fl t
283When listing, show last match timestamp converted with ctime().
284.It Fl T
285When listing, show last match timestamp as seconds from the epoch.
286This form can be more convenient for postprocessing by scripts.
287.El
288.Pp
289.Ss LIST OF RULES AND PREPROCESSING
290To ease configuration, rules can be put into a file which is
291processed using
292.Nm
293as shown in the last synopsis line.
294An absolute
295.Ar pathname
296must be used.
297The file will be read line by line and applied as arguments to the
298.Nm
299utility.
300.Pp
301Optionally, a preprocessor can be specified using
302.Fl p Ar preproc
303where
304.Ar pathname
305is to be piped through.
306Useful preprocessors include
307.Xr cpp 1
308and
309.Xr m4 1 .
310If
311.Ar preproc
312does not start with a slash
313.Pq Ql /
314as its first character, the usual
315.Ev PATH
316name search is performed.
317Care should be taken with this in environments where not all
318file systems are mounted (yet) by the time
319.Nm
320is being run (e.g.\& when they are mounted over NFS).
321Once
322.Fl p
323has been specified, any additional arguments are passed on to the preprocessor
324for interpretation.
325This allows for flexible configuration files (like conditionalizing
326them on the local hostname) and the use of macros to centralize
327frequently required arguments like IP addresses.
328.Pp
329.Ss TRAFFIC SHAPER CONFIGURATION
330The
331.Nm
332.Cm pipe , queue
333and
334.Cm sched
335commands are used to configure the traffic shaper and packet scheduler.
336See the
337.Sx TRAFFIC SHAPER (DUMMYNET) CONFIGURATION
338Section below for details.
339.Pp
340If the world and the kernel get out of sync the
341.Nm
342ABI may break, preventing you from being able to add any rules.
343This can
344adversely effect the booting process.
345You can use
346.Nm
347.Cm disable
348.Cm firewall
349to temporarily disable the firewall to regain access to the network,
350allowing you to fix the problem.
351.Sh PACKET FLOW
352A packet is checked against the active ruleset in multiple places
353in the protocol stack, under control of several sysctl variables.
354These places and variables are shown below, and it is important to
355have this picture in mind in order to design a correct ruleset.
356.Bd -literal -offset indent
357       ^    to upper layers    V
358       |                       |
359       +----------->-----------+
360       ^                       V
361 [ip(6)_input]           [ip(6)_output]     net.inet(6).ip(6).fw.enable=1
362       |                       |
363       ^                       V
364 [ether_demux]        [ether_output_frame]  net.link.ether.ipfw=1
365       |                       |
366       +-->--[bdg_forward]-->--+            net.link.bridge.ipfw=1
367       ^                       V
368       |      to devices       |
369.Ed
370.Pp
371The number of
372times the same packet goes through the firewall can
373vary between 0 and 4 depending on packet source and
374destination, and system configuration.
375.Pp
376Note that as packets flow through the stack, headers can be
377stripped or added to it, and so they may or may not be available
378for inspection.
379E.g., incoming packets will include the MAC header when
380.Nm
381is invoked from
382.Cm ether_demux() ,
383but the same packets will have the MAC header stripped off when
384.Nm
385is invoked from
386.Cm ip_input()
387or
388.Cm ip6_input() .
389.Pp
390Also note that each packet is always checked against the complete ruleset,
391irrespective of the place where the check occurs, or the source of the packet.
392If a rule contains some match patterns or actions which are not valid
393for the place of invocation (e.g.\& trying to match a MAC header within
394.Cm ip_input
395or
396.Cm ip6_input ),
397the match pattern will not match, but a
398.Cm not
399operator in front of such patterns
400.Em will
401cause the pattern to
402.Em always
403match on those packets.
404It is thus the responsibility of
405the programmer, if necessary, to write a suitable ruleset to
406differentiate among the possible places.
407.Cm skipto
408rules can be useful here, as an example:
409.Bd -literal -offset indent
410# packets from ether_demux or bdg_forward
411ipfw add 10 skipto 1000 all from any to any layer2 in
412# packets from ip_input
413ipfw add 10 skipto 2000 all from any to any not layer2 in
414# packets from ip_output
415ipfw add 10 skipto 3000 all from any to any not layer2 out
416# packets from ether_output_frame
417ipfw add 10 skipto 4000 all from any to any layer2 out
418.Ed
419.Pp
420(yes, at the moment there is no way to differentiate between
421ether_demux and bdg_forward).
422.Sh SYNTAX
423In general, each keyword or argument must be provided as
424a separate command line argument, with no leading or trailing
425spaces.
426Keywords are case-sensitive, whereas arguments may
427or may not be case-sensitive depending on their nature
428(e.g.\& uid's are, hostnames are not).
429.Pp
430Some arguments (e.g. port or address lists) are comma-separated
431lists of values.
432In this case, spaces after commas ',' are allowed to make
433the line more readable.
434You can also put the entire
435command (including flags) into a single argument.
436E.g., the following forms are equivalent:
437.Bd -literal -offset indent
438ipfw -q add deny src-ip 10.0.0.0/24,127.0.0.1/8
439ipfw -q add deny src-ip 10.0.0.0/24, 127.0.0.1/8
440ipfw "-q add deny src-ip 10.0.0.0/24, 127.0.0.1/8"
441.Ed
442.Sh RULE FORMAT
443The format of firewall rules is the following:
444.Bd -ragged -offset indent
445.Bk -words
446.Op Ar rule_number
447.Op Cm set Ar set_number
448.Op Cm prob Ar match_probability
449.Ar action
450.Op Cm log Op Cm logamount Ar number
451.Op Cm altq Ar queue
452.Oo
453.Bro Cm tag | untag
454.Brc Ar number
455.Oc
456.Ar body
457.Ek
458.Ed
459.Pp
460where the body of the rule specifies which information is used
461for filtering packets, among the following:
462.Pp
463.Bl -tag -width "Source and dest. addresses and ports" -offset XXX -compact
464.It Layer-2 header fields
465When available
466.It IPv4 and IPv6 Protocol
467TCP, UDP, ICMP, etc.
468.It Source and dest. addresses and ports
469.It Direction
470See Section
471.Sx PACKET FLOW
472.It Transmit and receive interface
473By name or address
474.It Misc. IP header fields
475Version, type of service, datagram length, identification,
476fragment flag (non-zero IP offset),
477Time To Live
478.It IP options
479.It IPv6 Extension headers
480Fragmentation, Hop-by-Hop options,
481Routing Headers, Source routing rthdr0, Mobile IPv6 rthdr2, IPSec options.
482.It IPv6 Flow-ID
483.It Misc. TCP header fields
484TCP flags (SYN, FIN, ACK, RST, etc.),
485sequence number, acknowledgment number,
486window
487.It TCP options
488.It ICMP types
489for ICMP packets
490.It ICMP6 types
491for ICMP6 packets
492.It User/group ID
493When the packet can be associated with a local socket.
494.It Divert status
495Whether a packet came from a divert socket (e.g.,
496.Xr natd 8 ) .
497.It Fib annotation state
498Whether a packet has been tagged for using a specific FIB (routing table)
499in future forwarding decisions.
500.El
501.Pp
502Note that some of the above information, e.g.\& source MAC or IP addresses and
503TCP/UDP ports, can be easily spoofed, so filtering on those fields
504alone might not guarantee the desired results.
505.Bl -tag -width indent
506.It Ar rule_number
507Each rule is associated with a
508.Ar rule_number
509in the range 1..65535, with the latter reserved for the
510.Em default
511rule.
512Rules are checked sequentially by rule number.
513Multiple rules can have the same number, in which case they are
514checked (and listed) according to the order in which they have
515been added.
516If a rule is entered without specifying a number, the kernel will
517assign one in such a way that the rule becomes the last one
518before the
519.Em default
520rule.
521Automatic rule numbers are assigned by incrementing the last
522non-default rule number by the value of the sysctl variable
523.Ar net.inet.ip.fw.autoinc_step
524which defaults to 100.
525If this is not possible (e.g.\& because we would go beyond the
526maximum allowed rule number), the number of the last
527non-default value is used instead.
528.It Cm set Ar set_number
529Each rule is associated with a
530.Ar set_number
531in the range 0..31.
532Sets can be individually disabled and enabled, so this parameter
533is of fundamental importance for atomic ruleset manipulation.
534It can be also used to simplify deletion of groups of rules.
535If a rule is entered without specifying a set number,
536set 0 will be used.
537.br
538Set 31 is special in that it cannot be disabled,
539and rules in set 31 are not deleted by the
540.Nm ipfw flush
541command (but you can delete them with the
542.Nm ipfw delete set 31
543command).
544Set 31 is also used for the
545.Em default
546rule.
547.It Cm prob Ar match_probability
548A match is only declared with the specified probability
549(floating point number between 0 and 1).
550This can be useful for a number of applications such as
551random packet drop or
552(in conjunction with
553.Nm dummynet )
554to simulate the effect of multiple paths leading to out-of-order
555packet delivery.
556.Pp
557Note: this condition is checked before any other condition, including
558ones such as keep-state or check-state which might have side effects.
559.It Cm log Op Cm logamount Ar number
560Packets matching a rule with the
561.Cm log
562keyword will be made available for logging in two ways:
563if the sysctl variable
564.Va net.inet.ip.fw.verbose
565is set to 0 (default), one can use
566.Xr bpf 4
567attached to the
568.Li ipfw0
569pseudo interface. There is no overhead if no
570.Xr bpf 4
571is attached to the pseudo interface.
572.Pp
573If
574.Va net.inet.ip.fw.verbose
575is set to 1, packets will be logged to
576.Xr syslogd 8
577with a
578.Dv LOG_SECURITY
579facility up to a maximum of
580.Cm logamount
581packets.
582If no
583.Cm logamount
584is specified, the limit is taken from the sysctl variable
585.Va net.inet.ip.fw.verbose_limit .
586In both cases, a value of 0 means unlimited logging.
587.Pp
588Once the limit is reached, logging can be re-enabled by
589clearing the logging counter or the packet counter for that entry, see the
590.Cm resetlog
591command.
592.Pp
593Note: logging is done after all other packet matching conditions
594have been successfully verified, and before performing the final
595action (accept, deny, etc.) on the packet.
596.It Cm tag Ar number
597When a packet matches a rule with the
598.Cm tag
599keyword, the numeric tag for the given
600.Ar number
601in the range 1..65534 will be attached to the packet.
602The tag acts as an internal marker (it is not sent out over
603the wire) that can be used to identify these packets later on.
604This can be used, for example, to provide trust between interfaces
605and to start doing policy-based filtering.
606A packet can have multiple tags at the same time.
607Tags are "sticky", meaning once a tag is applied to a packet by a
608matching rule it exists until explicit removal.
609Tags are kept with the packet everywhere within the kernel, but are
610lost when packet leaves the kernel, for example, on transmitting
611packet out to the network or sending packet to a
612.Xr divert 4
613socket.
614.Pp
615To check for previously applied tags, use the
616.Cm tagged
617rule option.
618To delete previously applied tag, use the
619.Cm untag
620keyword.
621.Pp
622Note: since tags are kept with the packet everywhere in kernelspace,
623they can be set and unset anywhere in the kernel network subsystem
624(using the
625.Xr mbuf_tags 9
626facility), not only by means of the
627.Xr ipfw 4
628.Cm tag
629and
630.Cm untag
631keywords.
632For example, there can be a specialized
633.Xr netgraph 4
634node doing traffic analyzing and tagging for later inspecting
635in firewall.
636.It Cm untag Ar number
637When a packet matches a rule with the
638.Cm untag
639keyword, the tag with the number
640.Ar number
641is searched among the tags attached to this packet and,
642if found, removed from it.
643Other tags bound to packet, if present, are left untouched.
644.It Cm altq Ar queue
645When a packet matches a rule with the
646.Cm altq
647keyword, the ALTQ identifier for the given
648.Ar queue
649(see
650.Xr altq 4 )
651will be attached.
652Note that this ALTQ tag is only meaningful for packets going "out" of IPFW,
653and not being rejected or going to divert sockets.
654Note that if there is insufficient memory at the time the packet is
655processed, it will not be tagged, so it is wise to make your ALTQ
656"default" queue policy account for this.
657If multiple
658.Cm altq
659rules match a single packet, only the first one adds the ALTQ classification
660tag.
661In doing so, traffic may be shaped by using
662.Cm count Cm altq Ar queue
663rules for classification early in the ruleset, then later applying
664the filtering decision.
665For example,
666.Cm check-state
667and
668.Cm keep-state
669rules may come later and provide the actual filtering decisions in
670addition to the fallback ALTQ tag.
671.Pp
672You must run
673.Xr pfctl 8
674to set up the queues before IPFW will be able to look them up by name,
675and if the ALTQ disciplines are rearranged, the rules in containing the
676queue identifiers in the kernel will likely have gone stale and need
677to be reloaded.
678Stale queue identifiers will probably result in misclassification.
679.Pp
680All system ALTQ processing can be turned on or off via
681.Nm
682.Cm enable Ar altq
683and
684.Nm
685.Cm disable Ar altq .
686The usage of
687.Va net.inet.ip.fw.one_pass
688is irrelevant to ALTQ traffic shaping, as the actual rule action is followed
689always after adding an ALTQ tag.
690.El
691.Ss RULE ACTIONS
692A rule can be associated with one of the following actions, which
693will be executed when the packet matches the body of the rule.
694.Bl -tag -width indent
695.It Cm allow | accept | pass | permit
696Allow packets that match rule.
697The search terminates.
698.It Cm check-state
699Checks the packet against the dynamic ruleset.
700If a match is found, execute the action associated with
701the rule which generated this dynamic rule, otherwise
702move to the next rule.
703.br
704.Cm Check-state
705rules do not have a body.
706If no
707.Cm check-state
708rule is found, the dynamic ruleset is checked at the first
709.Cm keep-state
710or
711.Cm limit
712rule.
713.It Cm count
714Update counters for all packets that match rule.
715The search continues with the next rule.
716.It Cm deny | drop
717Discard packets that match this rule.
718The search terminates.
719.It Cm divert Ar port
720Divert packets that match this rule to the
721.Xr divert 4
722socket bound to port
723.Ar port .
724The search terminates.
725.It Cm fwd | forward Ar ipaddr | tablearg Ns Op , Ns Ar port
726Change the next-hop on matching packets to
727.Ar ipaddr ,
728which can be an IP address or a host name.
729For IPv4, the next hop can also be supplied by the last table
730looked up for the packet by using the
731.Cm tablearg
732keyword instead of an explicit address.
733The search terminates if this rule matches.
734.Pp
735If
736.Ar ipaddr
737is a local address, then matching packets will be forwarded to
738.Ar port
739(or the port number in the packet if one is not specified in the rule)
740on the local machine.
741.br
742If
743.Ar ipaddr
744is not a local address, then the port number
745(if specified) is ignored, and the packet will be
746forwarded to the remote address, using the route as found in
747the local routing table for that IP.
748.br
749A
750.Ar fwd
751rule will not match layer-2 packets (those received
752on ether_input, ether_output, or bridged).
753.br
754The
755.Cm fwd
756action does not change the contents of the packet at all.
757In particular, the destination address remains unmodified, so
758packets forwarded to another system will usually be rejected by that system
759unless there is a matching rule on that system to capture them.
760For packets forwarded locally,
761the local address of the socket will be
762set to the original destination address of the packet.
763This makes the
764.Xr netstat 1
765entry look rather weird but is intended for
766use with transparent proxy servers.
767.Pp
768To enable
769.Cm fwd
770a custom kernel needs to be compiled with the option
771.Cd "options IPFIREWALL_FORWARD" .
772.It Cm nat Ar nat_nr | tablearg
773Pass packet to a
774nat instance
775(for network address translation, address redirect, etc.):
776see the
777.Sx NETWORK ADDRESS TRANSLATION (NAT)
778Section for further information.
779.It Cm pipe Ar pipe_nr
780Pass packet to a
781.Nm dummynet
782.Dq pipe
783(for bandwidth limitation, delay, etc.).
784See the
785.Sx TRAFFIC SHAPER (DUMMYNET) CONFIGURATION
786Section for further information.
787The search terminates; however, on exit from the pipe and if
788the
789.Xr sysctl 8
790variable
791.Va net.inet.ip.fw.one_pass
792is not set, the packet is passed again to the firewall code
793starting from the next rule.
794.It Cm queue Ar queue_nr
795Pass packet to a
796.Nm dummynet
797.Dq queue
798(for bandwidth limitation using WF2Q+).
799.It Cm reject
800(Deprecated).
801Synonym for
802.Cm unreach host .
803.It Cm reset
804Discard packets that match this rule, and if the
805packet is a TCP packet, try to send a TCP reset (RST) notice.
806The search terminates.
807.It Cm reset6
808Discard packets that match this rule, and if the
809packet is a TCP packet, try to send a TCP reset (RST) notice.
810The search terminates.
811.It Cm skipto Ar number | tablearg
812Skip all subsequent rules numbered less than
813.Ar number .
814The search continues with the first rule numbered
815.Ar number
816or higher.
817It is possible to use the
818.Cm tablearg
819keyword with a skipto for a
820.Em computed
821skipto, but care should be used, as no destination caching
822is possible in this case so the rules are always walked to find it,
823starting from the
824.Cm skipto .
825.It Cm call Ar number | tablearg
826The current rule number is saved in the internal stack and
827ruleset processing continues with the first rule numbered
828.Ar number
829or higher.
830If later a rule with the
831.Cm return
832action is encountered, the processing returns to the first rule
833with number of this
834.Cm call
835rule plus one or higher
836(the same behaviour as with packets returning from
837.Xr divert 4
838socket after a
839.Cm divert
840action).
841This could be used to make somewhat like an assembly language
842.Dq subroutine
843calls to rules with common checks for different interfaces, etc.
844.Pp
845Rule with any number could be called, not just forward jumps as with
846.Cm skipto .
847So, to prevent endless loops in case of mistakes, both
848.Cm call
849and
850.Cm return
851actions don't do any jumps and simply go to the next rule if memory
852can't be allocated or stack overflowed/undeflowed.
853.Pp
854Internally stack for rule numbers is implemented using
855.Xr mbuf_tags 9
856facility and currently has size of 16 entries.
857As mbuf tags are lost when packet leaves the kernel,
858.Cm divert
859should not be used in subroutines to avoid endless loops
860and other undesired effects.
861.It Cm return
862Takes rule number saved to internal stack by the last
863.Cm call
864action and returns ruleset processing to the first rule
865with number greater than number of corresponding
866.Cm call
867rule. See description of the
868.Cm call
869action for more details.
870.Pp
871Note that
872.Cm return
873rules usually end a
874.Dq subroutine
875and thus are unconditional, but
876.Nm
877command-line utility currently requires every action except
878.Cm check-state
879to have body.
880While it is sometimes useful to return only on some packets,
881usually you want to print just
882.Dq return
883for readability.
884A workaround for this is to use new syntax and
885.Fl c
886switch:
887.Pp
888.Bd -literal -offset indent
889# Add a rule without actual body
890ipfw add 2999 return via any
891
892# List rules without "from any to any" part
893ipfw -c list
894.Ed
895.Pp
896This cosmetic annoyance may be fixed in future releases.
897.It Cm tee Ar port
898Send a copy of packets matching this rule to the
899.Xr divert 4
900socket bound to port
901.Ar port .
902The search continues with the next rule.
903.It Cm unreach Ar code
904Discard packets that match this rule, and try to send an ICMP
905unreachable notice with code
906.Ar code ,
907where
908.Ar code
909is a number from 0 to 255, or one of these aliases:
910.Cm net , host , protocol , port ,
911.Cm needfrag , srcfail , net-unknown , host-unknown ,
912.Cm isolated , net-prohib , host-prohib , tosnet ,
913.Cm toshost , filter-prohib , host-precedence
914or
915.Cm precedence-cutoff .
916The search terminates.
917.It Cm unreach6 Ar code
918Discard packets that match this rule, and try to send an ICMPv6
919unreachable notice with code
920.Ar code ,
921where
922.Ar code
923is a number from 0, 1, 3 or 4, or one of these aliases:
924.Cm no-route, admin-prohib, address
925or
926.Cm port .
927The search terminates.
928.It Cm netgraph Ar cookie
929Divert packet into netgraph with given
930.Ar cookie .
931The search terminates.
932If packet is later returned from netgraph it is either
933accepted or continues with the next rule, depending on
934.Va net.inet.ip.fw.one_pass
935sysctl variable.
936.It Cm ngtee Ar cookie
937A copy of packet is diverted into netgraph, original
938packet continues with the next rule.
939See
940.Xr ng_ipfw 4
941for more information on
942.Cm netgraph
943and
944.Cm ngtee
945actions.
946.It Cm setfib Ar fibnum | tablearg
947The packet is tagged so as to use the FIB (routing table)
948.Ar fibnum
949in any subsequent forwarding decisions.
950Initially this is limited to the values 0 through 15, see
951.Xr setfib 1 .
952Processing continues at the next rule.
953It is possible to use the
954.Cm tablearg
955keyword with a setfib. If tablearg value is not within compiled FIB range packet fib is set to 0.
956.It Cm reass
957Queue and reassemble ip fragments.
958If the packet is not fragmented, counters are updated and processing continues with the next rule.
959If the packet is the last logical fragment, the packet is reassembled and, if
960.Va net.inet.ip.fw.one_pass
961is set to 0, processing continues with the next rule, else packet is allowed to pass and search terminates.
962If the packet is a fragment in the middle, it is consumed and processing stops immediately.
963.Pp
964Fragments handling can be tuned via
965.Va net.inet.ip.maxfragpackets
966and
967.Va net.inet.ip.maxfragsperpacket
968which limit, respectively, the maximum number of processable fragments (default: 800) and
969the maximum number of fragments per packet (default: 16).
970.Pp
971NOTA BENE: since fragments do not contain port numbers, they should be avoided with the
972.Nm reass
973rule.
974Alternatively, direction-based (like
975.Nm in
976/
977.Nm out
978) and source-based (like
979.Nm via
980) match patterns can be used to select fragments.
981.Pp
982Usually a simple rule like:
983.Bd -literal -offset indent
984# reassemble incoming fragments
985ipfw add reass all from any to any in
986.Ed
987.Pp
988is all you need at the beginning of your ruleset.
989.El
990.Ss RULE BODY
991The body of a rule contains zero or more patterns (such as
992specific source and destination addresses or ports,
993protocol options, incoming or outgoing interfaces, etc.)
994that the packet must match in order to be recognised.
995In general, the patterns are connected by (implicit)
996.Cm and
997operators -- i.e., all must match in order for the
998rule to match.
999Individual patterns can be prefixed by the
1000.Cm not
1001operator to reverse the result of the match, as in
1002.Pp
1003.Dl "ipfw add 100 allow ip from not 1.2.3.4 to any"
1004.Pp
1005Additionally, sets of alternative match patterns
1006.Pq Em or-blocks
1007can be constructed by putting the patterns in
1008lists enclosed between parentheses ( ) or braces { }, and
1009using the
1010.Cm or
1011operator as follows:
1012.Pp
1013.Dl "ipfw add 100 allow ip from { x or not y or z } to any"
1014.Pp
1015Only one level of parentheses is allowed.
1016Beware that most shells have special meanings for parentheses
1017or braces, so it is advisable to put a backslash \\ in front of them
1018to prevent such interpretations.
1019.Pp
1020The body of a rule must in general include a source and destination
1021address specifier.
1022The keyword
1023.Ar any
1024can be used in various places to specify that the content of
1025a required field is irrelevant.
1026.Pp
1027The rule body has the following format:
1028.Bd -ragged -offset indent
1029.Op Ar proto Cm from Ar src Cm to Ar dst
1030.Op Ar options
1031.Ed
1032.Pp
1033The first part (proto from src to dst) is for backward
1034compatibility with earlier versions of
1035.Fx .
1036In modern
1037.Fx
1038any match pattern (including MAC headers, IP protocols,
1039addresses and ports) can be specified in the
1040.Ar options
1041section.
1042.Pp
1043Rule fields have the following meaning:
1044.Bl -tag -width indent
1045.It Ar proto : protocol | Cm { Ar protocol Cm or ... }
1046.It Ar protocol : Oo Cm not Oc Ar protocol-name | protocol-number
1047An IP protocol specified by number or name
1048(for a complete list see
1049.Pa /etc/protocols ) ,
1050or one of the following keywords:
1051.Bl -tag -width indent
1052.It Cm ip4 | ipv4
1053Matches IPv4 packets.
1054.It Cm ip6 | ipv6
1055Matches IPv6 packets.
1056.It Cm ip | all
1057Matches any packet.
1058.El
1059.Pp
1060The
1061.Cm ipv6
1062in
1063.Cm proto
1064option will be treated as inner protocol.
1065And, the
1066.Cm ipv4
1067is not available in
1068.Cm proto
1069option.
1070.Pp
1071The
1072.Cm { Ar protocol Cm or ... }
1073format (an
1074.Em or-block )
1075is provided for convenience only but its use is deprecated.
1076.It Ar src No and Ar dst : Bro Cm addr | Cm { Ar addr Cm or ... } Brc Op Oo Cm not Oc Ar ports
1077An address (or a list, see below)
1078optionally followed by
1079.Ar ports
1080specifiers.
1081.Pp
1082The second format
1083.Em ( or-block
1084with multiple addresses) is provided for convenience only and
1085its use is discouraged.
1086.It Ar addr : Oo Cm not Oc Bro
1087.Cm any | me | me6 |
1088.Cm table Ns Pq Ar number Ns Op , Ns Ar value
1089.Ar | addr-list | addr-set
1090.Brc
1091.Bl -tag -width indent
1092.It Cm any
1093matches any IP address.
1094.It Cm me
1095matches any IP address configured on an interface in the system.
1096.It Cm me6
1097matches any IPv6 address configured on an interface in the system.
1098The address list is evaluated at the time the packet is
1099analysed.
1100.It Cm table Ns Pq Ar number Ns Op , Ns Ar value
1101Matches any IPv4 address for which an entry exists in the lookup table
1102.Ar number .
1103If an optional 32-bit unsigned
1104.Ar value
1105is also specified, an entry will match only if it has this value.
1106See the
1107.Sx LOOKUP TABLES
1108section below for more information on lookup tables.
1109.El
1110.It Ar addr-list : ip-addr Ns Op Ns , Ns Ar addr-list
1111.It Ar ip-addr :
1112A host or subnet address specified in one of the following ways:
1113.Bl -tag -width indent
1114.It Ar numeric-ip | hostname
1115Matches a single IPv4 address, specified as dotted-quad or a hostname.
1116Hostnames are resolved at the time the rule is added to the firewall list.
1117.It Ar addr Ns / Ns Ar masklen
1118Matches all addresses with base
1119.Ar addr
1120(specified as an IP address, a network number, or a hostname)
1121and mask width of
1122.Cm masklen
1123bits.
1124As an example, 1.2.3.4/25 or 1.2.3.0/25 will match
1125all IP numbers from 1.2.3.0 to 1.2.3.127 .
1126.It Ar addr Ns : Ns Ar mask
1127Matches all addresses with base
1128.Ar addr
1129(specified as an IP address, a network number, or a hostname)
1130and the mask of
1131.Ar mask ,
1132specified as a dotted quad.
1133As an example, 1.2.3.4:255.0.255.0 or 1.0.3.0:255.0.255.0 will match
11341.*.3.*.
1135This form is advised only for non-contiguous
1136masks.
1137It is better to resort to the
1138.Ar addr Ns / Ns Ar masklen
1139format for contiguous masks, which is more compact and less
1140error-prone.
1141.El
1142.It Ar addr-set : addr Ns Oo Ns / Ns Ar masklen Oc Ns Cm { Ns Ar list Ns Cm }
1143.It Ar list : Bro Ar num | num-num Brc Ns Op Ns , Ns Ar list
1144Matches all addresses with base address
1145.Ar addr
1146(specified as an IP address, a network number, or a hostname)
1147and whose last byte is in the list between braces { } .
1148Note that there must be no spaces between braces and
1149numbers (spaces after commas are allowed).
1150Elements of the list can be specified as single entries
1151or ranges.
1152The
1153.Ar masklen
1154field is used to limit the size of the set of addresses,
1155and can have any value between 24 and 32.
1156If not specified,
1157it will be assumed as 24.
1158.br
1159This format is particularly useful to handle sparse address sets
1160within a single rule.
1161Because the matching occurs using a
1162bitmask, it takes constant time and dramatically reduces
1163the complexity of rulesets.
1164.br
1165As an example, an address specified as 1.2.3.4/24{128,35-55,89}
1166or 1.2.3.0/24{128,35-55,89}
1167will match the following IP addresses:
1168.br
11691.2.3.128, 1.2.3.35 to 1.2.3.55, 1.2.3.89 .
1170.It Ar addr6-list : ip6-addr Ns Op Ns , Ns Ar addr6-list
1171.It Ar ip6-addr :
1172A host or subnet specified one of the following ways:
1173.Bl -tag -width indent
1174.It Ar numeric-ip | hostname
1175Matches a single IPv6 address as allowed by
1176.Xr inet_pton 3
1177or a hostname.
1178Hostnames are resolved at the time the rule is added to the firewall
1179list.
1180.It Ar addr Ns / Ns Ar masklen
1181Matches all IPv6 addresses with base
1182.Ar addr
1183(specified as allowed by
1184.Xr inet_pton
1185or a hostname)
1186and mask width of
1187.Cm masklen
1188bits.
1189.El
1190.Pp
1191No support for sets of IPv6 addresses is provided because IPv6 addresses
1192are typically random past the initial prefix.
1193.It Ar ports : Bro Ar port | port Ns \&- Ns Ar port Ns Brc Ns Op , Ns Ar ports
1194For protocols which support port numbers (such as TCP and UDP), optional
1195.Cm ports
1196may be specified as one or more ports or port ranges, separated
1197by commas but no spaces, and an optional
1198.Cm not
1199operator.
1200The
1201.Ql \&-
1202notation specifies a range of ports (including boundaries).
1203.Pp
1204Service names (from
1205.Pa /etc/services )
1206may be used instead of numeric port values.
1207The length of the port list is limited to 30 ports or ranges,
1208though one can specify larger ranges by using an
1209.Em or-block
1210in the
1211.Cm options
1212section of the rule.
1213.Pp
1214A backslash
1215.Pq Ql \e
1216can be used to escape the dash
1217.Pq Ql -
1218character in a service name (from a shell, the backslash must be
1219typed twice to avoid the shell itself interpreting it as an escape
1220character).
1221.Pp
1222.Dl "ipfw add count tcp from any ftp\e\e-data-ftp to any"
1223.Pp
1224Fragmented packets which have a non-zero offset (i.e., not the first
1225fragment) will never match a rule which has one or more port
1226specifications.
1227See the
1228.Cm frag
1229option for details on matching fragmented packets.
1230.El
1231.Ss RULE OPTIONS (MATCH PATTERNS)
1232Additional match patterns can be used within
1233rules.
1234Zero or more of these so-called
1235.Em options
1236can be present in a rule, optionally prefixed by the
1237.Cm not
1238operand, and possibly grouped into
1239.Em or-blocks .
1240.Pp
1241The following match patterns can be used (listed in alphabetical order):
1242.Bl -tag -width indent
1243.It Cm // this is a comment.
1244Inserts the specified text as a comment in the rule.
1245Everything following // is considered as a comment and stored in the rule.
1246You can have comment-only rules, which are listed as having a
1247.Cm count
1248action followed by the comment.
1249.It Cm bridged
1250Alias for
1251.Cm layer2 .
1252.It Cm diverted
1253Matches only packets generated by a divert socket.
1254.It Cm diverted-loopback
1255Matches only packets coming from a divert socket back into the IP stack
1256input for delivery.
1257.It Cm diverted-output
1258Matches only packets going from a divert socket back outward to the IP
1259stack output for delivery.
1260.It Cm dst-ip Ar ip-address
1261Matches IPv4 packets whose destination IP is one of the address(es)
1262specified as argument.
1263.It Bro Cm dst-ip6 | dst-ipv6 Brc Ar ip6-address
1264Matches IPv6 packets whose destination IP is one of the address(es)
1265specified as argument.
1266.It Cm dst-port Ar ports
1267Matches IP packets whose destination port is one of the port(s)
1268specified as argument.
1269.It Cm established
1270Matches TCP packets that have the RST or ACK bits set.
1271.It Cm ext6hdr Ar header
1272Matches IPv6 packets containing the extended header given by
1273.Ar header .
1274Supported headers are:
1275.Pp
1276Fragment,
1277.Pq Cm frag ,
1278Hop-to-hop options
1279.Pq Cm hopopt ,
1280any type of Routing Header
1281.Pq Cm route ,
1282Source routing Routing Header Type 0
1283.Pq Cm rthdr0 ,
1284Mobile IPv6 Routing Header Type 2
1285.Pq Cm rthdr2 ,
1286Destination options
1287.Pq Cm dstopt ,
1288IPSec authentication headers
1289.Pq Cm ah ,
1290and IPsec encapsulated security payload headers
1291.Pq Cm esp .
1292.It Cm fib Ar fibnum
1293Matches a packet that has been tagged to use
1294the given FIB (routing table) number.
1295.It Cm flow-id Ar labels
1296Matches IPv6 packets containing any of the flow labels given in
1297.Ar labels .
1298.Ar labels
1299is a comma separated list of numeric flow labels.
1300.It Cm frag
1301Matches packets that are fragments and not the first
1302fragment of an IP datagram.
1303Note that these packets will not have
1304the next protocol header (e.g.\& TCP, UDP) so options that look into
1305these headers cannot match.
1306.It Cm gid Ar group
1307Matches all TCP or UDP packets sent by or received for a
1308.Ar group .
1309A
1310.Ar group
1311may be specified by name or number.
1312.It Cm jail Ar prisonID
1313Matches all TCP or UDP packets sent by or received for the
1314jail whos prison ID is
1315.Ar prisonID .
1316.It Cm icmptypes Ar types
1317Matches ICMP packets whose ICMP type is in the list
1318.Ar types .
1319The list may be specified as any combination of
1320individual types (numeric) separated by commas.
1321.Em Ranges are not allowed .
1322The supported ICMP types are:
1323.Pp
1324echo reply
1325.Pq Cm 0 ,
1326destination unreachable
1327.Pq Cm 3 ,
1328source quench
1329.Pq Cm 4 ,
1330redirect
1331.Pq Cm 5 ,
1332echo request
1333.Pq Cm 8 ,
1334router advertisement
1335.Pq Cm 9 ,
1336router solicitation
1337.Pq Cm 10 ,
1338time-to-live exceeded
1339.Pq Cm 11 ,
1340IP header bad
1341.Pq Cm 12 ,
1342timestamp request
1343.Pq Cm 13 ,
1344timestamp reply
1345.Pq Cm 14 ,
1346information request
1347.Pq Cm 15 ,
1348information reply
1349.Pq Cm 16 ,
1350address mask request
1351.Pq Cm 17
1352and address mask reply
1353.Pq Cm 18 .
1354.It Cm icmp6types Ar types
1355Matches ICMP6 packets whose ICMP6 type is in the list of
1356.Ar types .
1357The list may be specified as any combination of
1358individual types (numeric) separated by commas.
1359.Em Ranges are not allowed .
1360.It Cm in | out
1361Matches incoming or outgoing packets, respectively.
1362.Cm in
1363and
1364.Cm out
1365are mutually exclusive (in fact,
1366.Cm out
1367is implemented as
1368.Cm not in Ns No ).
1369.It Cm ipid Ar id-list
1370Matches IPv4 packets whose
1371.Cm ip_id
1372field has value included in
1373.Ar id-list ,
1374which is either a single value or a list of values or ranges
1375specified in the same way as
1376.Ar ports .
1377.It Cm iplen Ar len-list
1378Matches IP packets whose total length, including header and data, is
1379in the set
1380.Ar len-list ,
1381which is either a single value or a list of values or ranges
1382specified in the same way as
1383.Ar ports .
1384.It Cm ipoptions Ar spec
1385Matches packets whose IPv4 header contains the comma separated list of
1386options specified in
1387.Ar spec .
1388The supported IP options are:
1389.Pp
1390.Cm ssrr
1391(strict source route),
1392.Cm lsrr
1393(loose source route),
1394.Cm rr
1395(record packet route) and
1396.Cm ts
1397(timestamp).
1398The absence of a particular option may be denoted
1399with a
1400.Ql \&! .
1401.It Cm ipprecedence Ar precedence
1402Matches IPv4 packets whose precedence field is equal to
1403.Ar precedence .
1404.It Cm ipsec
1405Matches packets that have IPSEC history associated with them
1406(i.e., the packet comes encapsulated in IPSEC, the kernel
1407has IPSEC support and IPSEC_FILTERTUNNEL option, and can correctly
1408decapsulate it).
1409.Pp
1410Note that specifying
1411.Cm ipsec
1412is different from specifying
1413.Cm proto Ar ipsec
1414as the latter will only look at the specific IP protocol field,
1415irrespective of IPSEC kernel support and the validity of the IPSEC data.
1416.Pp
1417Further note that this flag is silently ignored in kernels without
1418IPSEC support.
1419It does not affect rule processing when given and the
1420rules are handled as if with no
1421.Cm ipsec
1422flag.
1423.It Cm iptos Ar spec
1424Matches IPv4 packets whose
1425.Cm tos
1426field contains the comma separated list of
1427service types specified in
1428.Ar spec .
1429The supported IP types of service are:
1430.Pp
1431.Cm lowdelay
1432.Pq Dv IPTOS_LOWDELAY ,
1433.Cm throughput
1434.Pq Dv IPTOS_THROUGHPUT ,
1435.Cm reliability
1436.Pq Dv IPTOS_RELIABILITY ,
1437.Cm mincost
1438.Pq Dv IPTOS_MINCOST ,
1439.Cm congestion
1440.Pq Dv IPTOS_ECN_CE .
1441The absence of a particular type may be denoted
1442with a
1443.Ql \&! .
1444.It Cm ipttl Ar ttl-list
1445Matches IPv4 packets whose time to live is included in
1446.Ar ttl-list ,
1447which is either a single value or a list of values or ranges
1448specified in the same way as
1449.Ar ports .
1450.It Cm ipversion Ar ver
1451Matches IP packets whose IP version field is
1452.Ar ver .
1453.It Cm keep-state
1454Upon a match, the firewall will create a dynamic rule, whose
1455default behaviour is to match bidirectional traffic between
1456source and destination IP/port using the same protocol.
1457The rule has a limited lifetime (controlled by a set of
1458.Xr sysctl 8
1459variables), and the lifetime is refreshed every time a matching
1460packet is found.
1461.It Cm layer2
1462Matches only layer2 packets, i.e., those passed to
1463.Nm
1464from ether_demux() and ether_output_frame().
1465.It Cm limit Bro Cm src-addr | src-port | dst-addr | dst-port Brc Ar N
1466The firewall will only allow
1467.Ar N
1468connections with the same
1469set of parameters as specified in the rule.
1470One or more
1471of source and destination addresses and ports can be
1472specified.
1473Currently,
1474only IPv4 flows are supported.
1475.It Cm lookup Bro Cm dst-ip | dst-port | src-ip | src-port | uid | jail Brc Ar N
1476Search an entry in lookup table
1477.Ar N
1478that matches the field specified as argument.
1479If not found, the match fails.
1480Otherwise, the match succeeds and
1481.Cm tablearg
1482is set to the value extracted from the table.
1483.Pp
1484This option can be useful to quickly dispatch traffic based on
1485certain packet fields.
1486See the
1487.Sx LOOKUP TABLES
1488section below for more information on lookup tables.
1489.It Cm { MAC | mac } Ar dst-mac src-mac
1490Match packets with a given
1491.Ar dst-mac
1492and
1493.Ar src-mac
1494addresses, specified as the
1495.Cm any
1496keyword (matching any MAC address), or six groups of hex digits
1497separated by colons,
1498and optionally followed by a mask indicating the significant bits.
1499The mask may be specified using either of the following methods:
1500.Bl -enum -width indent
1501.It
1502A slash
1503.Pq /
1504followed by the number of significant bits.
1505For example, an address with 33 significant bits could be specified as:
1506.Pp
1507.Dl "MAC 10:20:30:40:50:60/33 any"
1508.Pp
1509.It
1510An ampersand
1511.Pq &
1512followed by a bitmask specified as six groups of hex digits separated
1513by colons.
1514For example, an address in which the last 16 bits are significant could
1515be specified as:
1516.Pp
1517.Dl "MAC 10:20:30:40:50:60&00:00:00:00:ff:ff any"
1518.Pp
1519Note that the ampersand character has a special meaning in many shells
1520and should generally be escaped.
1521.Pp
1522.El
1523Note that the order of MAC addresses (destination first,
1524source second) is
1525the same as on the wire, but the opposite of the one used for
1526IP addresses.
1527.It Cm mac-type Ar mac-type
1528Matches packets whose Ethernet Type field
1529corresponds to one of those specified as argument.
1530.Ar mac-type
1531is specified in the same way as
1532.Cm port numbers
1533(i.e., one or more comma-separated single values or ranges).
1534You can use symbolic names for known values such as
1535.Em vlan , ipv4, ipv6 .
1536Values can be entered as decimal or hexadecimal (if prefixed by 0x),
1537and they are always printed as hexadecimal (unless the
1538.Cm -N
1539option is used, in which case symbolic resolution will be attempted).
1540.It Cm proto Ar protocol
1541Matches packets with the corresponding IP protocol.
1542.It Cm recv | xmit | via Brq Ar ifX | Ar if Ns Cm * | Ar ipno | Ar any
1543Matches packets received, transmitted or going through,
1544respectively, the interface specified by exact name
1545.Ns No ( Ar ifX Ns No ),
1546by device name
1547.Ns No ( Ar if Ns Ar * Ns No ),
1548by IP address, or through some interface.
1549.Pp
1550The
1551.Cm via
1552keyword causes the interface to always be checked.
1553If
1554.Cm recv
1555or
1556.Cm xmit
1557is used instead of
1558.Cm via ,
1559then only the receive or transmit interface (respectively)
1560is checked.
1561By specifying both, it is possible to match packets based on
1562both receive and transmit interface, e.g.:
1563.Pp
1564.Dl "ipfw add deny ip from any to any out recv ed0 xmit ed1"
1565.Pp
1566The
1567.Cm recv
1568interface can be tested on either incoming or outgoing packets,
1569while the
1570.Cm xmit
1571interface can only be tested on outgoing packets.
1572So
1573.Cm out
1574is required (and
1575.Cm in
1576is invalid) whenever
1577.Cm xmit
1578is used.
1579.Pp
1580A packet might not have a receive or transmit interface: packets
1581originating from the local host have no receive interface,
1582while packets destined for the local host have no transmit
1583interface.
1584.It Cm setup
1585Matches TCP packets that have the SYN bit set but no ACK bit.
1586This is the short form of
1587.Dq Li tcpflags\ syn,!ack .
1588.It Cm sockarg
1589Matches packets that are associated to a local socket and
1590for which the SO_USER_COOKIE socket option has been set
1591to a non-zero value. As a side effect, the value of the
1592option is made available as
1593.Cm tablearg
1594value, which in turn can be used as
1595.Cm skipto
1596or
1597.Cm pipe
1598number.
1599.It Cm src-ip Ar ip-address
1600Matches IPv4 packets whose source IP is one of the address(es)
1601specified as an argument.
1602.It Cm src-ip6 Ar ip6-address
1603Matches IPv6 packets whose source IP is one of the address(es)
1604specified as an argument.
1605.It Cm src-port Ar ports
1606Matches IP packets whose source port is one of the port(s)
1607specified as argument.
1608.It Cm tagged Ar tag-list
1609Matches packets whose tags are included in
1610.Ar tag-list ,
1611which is either a single value or a list of values or ranges
1612specified in the same way as
1613.Ar ports .
1614Tags can be applied to the packet using
1615.Cm tag
1616rule action parameter (see it's description for details on tags).
1617.It Cm tcpack Ar ack
1618TCP packets only.
1619Match if the TCP header acknowledgment number field is set to
1620.Ar ack .
1621.It Cm tcpdatalen Ar tcpdatalen-list
1622Matches TCP packets whose length of TCP data is
1623.Ar tcpdatalen-list ,
1624which is either a single value or a list of values or ranges
1625specified in the same way as
1626.Ar ports .
1627.It Cm tcpflags Ar spec
1628TCP packets only.
1629Match if the TCP header contains the comma separated list of
1630flags specified in
1631.Ar spec .
1632The supported TCP flags are:
1633.Pp
1634.Cm fin ,
1635.Cm syn ,
1636.Cm rst ,
1637.Cm psh ,
1638.Cm ack
1639and
1640.Cm urg .
1641The absence of a particular flag may be denoted
1642with a
1643.Ql \&! .
1644A rule which contains a
1645.Cm tcpflags
1646specification can never match a fragmented packet which has
1647a non-zero offset.
1648See the
1649.Cm frag
1650option for details on matching fragmented packets.
1651.It Cm tcpseq Ar seq
1652TCP packets only.
1653Match if the TCP header sequence number field is set to
1654.Ar seq .
1655.It Cm tcpwin Ar win
1656TCP packets only.
1657Match if the TCP header window field is set to
1658.Ar win .
1659.It Cm tcpoptions Ar spec
1660TCP packets only.
1661Match if the TCP header contains the comma separated list of
1662options specified in
1663.Ar spec .
1664The supported TCP options are:
1665.Pp
1666.Cm mss
1667(maximum segment size),
1668.Cm window
1669(tcp window advertisement),
1670.Cm sack
1671(selective ack),
1672.Cm ts
1673(rfc1323 timestamp) and
1674.Cm cc
1675(rfc1644 t/tcp connection count).
1676The absence of a particular option may be denoted
1677with a
1678.Ql \&! .
1679.It Cm uid Ar user
1680Match all TCP or UDP packets sent by or received for a
1681.Ar user .
1682A
1683.Ar user
1684may be matched by name or identification number.
1685.It Cm verrevpath
1686For incoming packets,
1687a routing table lookup is done on the packet's source address.
1688If the interface on which the packet entered the system matches the
1689outgoing interface for the route,
1690the packet matches.
1691If the interfaces do not match up,
1692the packet does not match.
1693All outgoing packets or packets with no incoming interface match.
1694.Pp
1695The name and functionality of the option is intentionally similar to
1696the Cisco IOS command:
1697.Pp
1698.Dl ip verify unicast reverse-path
1699.Pp
1700This option can be used to make anti-spoofing rules to reject all
1701packets with source addresses not from this interface.
1702See also the option
1703.Cm antispoof .
1704.It Cm versrcreach
1705For incoming packets,
1706a routing table lookup is done on the packet's source address.
1707If a route to the source address exists, but not the default route
1708or a blackhole/reject route, the packet matches.
1709Otherwise, the packet does not match.
1710All outgoing packets match.
1711.Pp
1712The name and functionality of the option is intentionally similar to
1713the Cisco IOS command:
1714.Pp
1715.Dl ip verify unicast source reachable-via any
1716.Pp
1717This option can be used to make anti-spoofing rules to reject all
1718packets whose source address is unreachable.
1719.It Cm antispoof
1720For incoming packets, the packet's source address is checked if it
1721belongs to a directly connected network.
1722If the network is directly connected, then the interface the packet
1723came on in is compared to the interface the network is connected to.
1724When incoming interface and directly connected interface are not the
1725same, the packet does not match.
1726Otherwise, the packet does match.
1727All outgoing packets match.
1728.Pp
1729This option can be used to make anti-spoofing rules to reject all
1730packets that pretend to be from a directly connected network but do
1731not come in through that interface.
1732This option is similar to but more restricted than
1733.Cm verrevpath
1734because it engages only on packets with source addresses of directly
1735connected networks instead of all source addresses.
1736.El
1737.Sh LOOKUP TABLES
1738Lookup tables are useful to handle large sparse sets of
1739addresses or other search keys (e.g. ports, jail IDs).
1740In the rest of this section we will use the term ``address''
1741to mean any unsigned value of up to 32-bit.
1742There may be up to 128 different lookup tables, numbered 0 to 127.
1743.Pp
1744Each entry is represented by an
1745.Ar addr Ns Op / Ns Ar masklen
1746and will match all addresses with base
1747.Ar addr
1748(specified as an IP address, a hostname or an unsigned integer)
1749and mask width of
1750.Ar masklen
1751bits.
1752If
1753.Ar masklen
1754is not specified, it defaults to 32.
1755When looking up an IP address in a table, the most specific
1756entry will match.
1757Associated with each entry is a 32-bit unsigned
1758.Ar value ,
1759which can optionally be checked by a rule matching code.
1760When adding an entry, if
1761.Ar value
1762is not specified, it defaults to 0.
1763.Pp
1764An entry can be added to a table
1765.Pq Cm add ,
1766or removed from a table
1767.Pq Cm delete .
1768A table can be examined
1769.Pq Cm list
1770or flushed
1771.Pq Cm flush .
1772.Pp
1773Internally, each table is stored in a Radix tree, the same way as
1774the routing table (see
1775.Xr route 4 ) .
1776.Pp
1777Lookup tables currently support only ports, jail IDs and IPv4 addresses.
1778.Pp
1779The
1780.Cm tablearg
1781feature provides the ability to use a value, looked up in the table, as
1782the argument for a rule action, action parameter or rule option.
1783This can significantly reduce number of rules in some configurations.
1784If two tables are used in a rule, the result of the second (destination)
1785is used.
1786The
1787.Cm tablearg
1788argument can be used with the following actions:
1789.Cm nat, pipe , queue, divert, tee, netgraph, ngtee, fwd, skipto, setfib,
1790action parameters:
1791.Cm tag, untag,
1792rule options:
1793.Cm limit, tagged.
1794.Pp
1795When used with
1796.Cm fwd
1797it is possible to supply table entries with values
1798that are in the form of IP addresses or hostnames.
1799See the
1800.Sx EXAMPLES
1801Section for example usage of tables and the tablearg keyword.
1802.Pp
1803When used with the
1804.Cm skipto
1805action, the user should be aware that the code will walk the ruleset
1806up to a rule equal to, or past, the given number, and should therefore try keep the
1807ruleset compact between the skipto and the target rules.
1808.Sh SETS OF RULES
1809Each rule belongs to one of 32 different
1810.Em sets
1811, numbered 0 to 31.
1812Set 31 is reserved for the default rule.
1813.Pp
1814By default, rules are put in set 0, unless you use the
1815.Cm set N
1816attribute when entering a new rule.
1817Sets can be individually and atomically enabled or disabled,
1818so this mechanism permits an easy way to store multiple configurations
1819of the firewall and quickly (and atomically) switch between them.
1820The command to enable/disable sets is
1821.Bd -ragged -offset indent
1822.Nm
1823.Cm set Oo Cm disable Ar number ... Oc Op Cm enable Ar number ...
1824.Ed
1825.Pp
1826where multiple
1827.Cm enable
1828or
1829.Cm disable
1830sections can be specified.
1831Command execution is atomic on all the sets specified in the command.
1832By default, all sets are enabled.
1833.Pp
1834When you disable a set, its rules behave as if they do not exist
1835in the firewall configuration, with only one exception:
1836.Bd -ragged -offset indent
1837dynamic rules created from a rule before it had been disabled
1838will still be active until they expire.
1839In order to delete
1840dynamic rules you have to explicitly delete the parent rule
1841which generated them.
1842.Ed
1843.Pp
1844The set number of rules can be changed with the command
1845.Bd -ragged -offset indent
1846.Nm
1847.Cm set move
1848.Brq Cm rule Ar rule-number | old-set
1849.Cm to Ar new-set
1850.Ed
1851.Pp
1852Also, you can atomically swap two rulesets with the command
1853.Bd -ragged -offset indent
1854.Nm
1855.Cm set swap Ar first-set second-set
1856.Ed
1857.Pp
1858See the
1859.Sx EXAMPLES
1860Section on some possible uses of sets of rules.
1861.Sh STATEFUL FIREWALL
1862Stateful operation is a way for the firewall to dynamically
1863create rules for specific flows when packets that
1864match a given pattern are detected.
1865Support for stateful
1866operation comes through the
1867.Cm check-state , keep-state
1868and
1869.Cm limit
1870options of
1871.Nm rules .
1872.Pp
1873Dynamic rules are created when a packet matches a
1874.Cm keep-state
1875or
1876.Cm limit
1877rule, causing the creation of a
1878.Em dynamic
1879rule which will match all and only packets with
1880a given
1881.Em protocol
1882between a
1883.Em src-ip/src-port dst-ip/dst-port
1884pair of addresses
1885.Em ( src
1886and
1887.Em dst
1888are used here only to denote the initial match addresses, but they
1889are completely equivalent afterwards).
1890Dynamic rules will be checked at the first
1891.Cm check-state, keep-state
1892or
1893.Cm limit
1894occurrence, and the action performed upon a match will be the same
1895as in the parent rule.
1896.Pp
1897Note that no additional attributes other than protocol and IP addresses
1898and ports are checked on dynamic rules.
1899.Pp
1900The typical use of dynamic rules is to keep a closed firewall configuration,
1901but let the first TCP SYN packet from the inside network install a
1902dynamic rule for the flow so that packets belonging to that session
1903will be allowed through the firewall:
1904.Pp
1905.Dl "ipfw add check-state"
1906.Dl "ipfw add allow tcp from my-subnet to any setup keep-state"
1907.Dl "ipfw add deny tcp from any to any"
1908.Pp
1909A similar approach can be used for UDP, where an UDP packet coming
1910from the inside will install a dynamic rule to let the response through
1911the firewall:
1912.Pp
1913.Dl "ipfw add check-state"
1914.Dl "ipfw add allow udp from my-subnet to any keep-state"
1915.Dl "ipfw add deny udp from any to any"
1916.Pp
1917Dynamic rules expire after some time, which depends on the status
1918of the flow and the setting of some
1919.Cm sysctl
1920variables.
1921See Section
1922.Sx SYSCTL VARIABLES
1923for more details.
1924For TCP sessions, dynamic rules can be instructed to periodically
1925send keepalive packets to refresh the state of the rule when it is
1926about to expire.
1927.Pp
1928See Section
1929.Sx EXAMPLES
1930for more examples on how to use dynamic rules.
1931.Sh TRAFFIC SHAPER (DUMMYNET) CONFIGURATION
1932.Nm
1933is also the user interface for the
1934.Nm dummynet
1935traffic shaper, packet scheduler and network emulator, a subsystem that
1936can artificially queue, delay or drop packets
1937emulating the behaviour of certain network links
1938or queueing systems.
1939.Pp
1940.Nm dummynet
1941operates by first using the firewall to select packets
1942using any match pattern that can be used in
1943.Nm
1944rules.
1945Matching packets are then passed to either of two
1946different objects, which implement the traffic regulation:
1947.Bl -hang -offset XXXX
1948.It Em pipe
1949A
1950.Em pipe
1951emulates a
1952.Em link
1953with given bandwidth and propagation delay,
1954driven by a FIFO scheduler and a single queue with programmable
1955queue size and packet loss rate.
1956Packets are appended to the queue as they come out from
1957.Nm ipfw ,
1958and then transferred in FIFO order to the link at the desired rate.
1959.It Em queue
1960A
1961.Em queue
1962is an abstraction used to implement packet scheduling
1963using one of several packet scheduling algorithms.
1964Packets sent to a
1965.Em queue
1966are first grouped into flows according to a mask on the 5-tuple.
1967Flows are then passed to the scheduler associated to the
1968.Em queue ,
1969and each flow uses scheduling parameters (weight and others)
1970as configured in the
1971.Em queue
1972itself.
1973A scheduler in turn is connected to an emulated link,
1974and arbitrates the link's bandwidth among backlogged flows according to
1975weights and to the features of the scheduling algorithm in use.
1976.El
1977.Pp
1978In practice,
1979.Em pipes
1980can be used to set hard limits to the bandwidth that a flow can use, whereas
1981.Em queues
1982can be used to determine how different flows share the available bandwidth.
1983.Pp
1984A graphical representation of the binding of queues,
1985flows, schedulers and links is below.
1986.Bd -literal -offset indent
1987                 (flow_mask|sched_mask)  sched_mask
1988         +---------+   weight Wx  +-------------+
1989         |         |->-[flow]-->--|             |-+
1990    -->--| QUEUE x |   ...        |             | |
1991         |         |->-[flow]-->--| SCHEDuler N | |
1992         +---------+              |             | |
1993             ...                  |             +--[LINK N]-->--
1994         +---------+   weight Wy  |             | +--[LINK N]-->--
1995         |         |->-[flow]-->--|             | |
1996    -->--| QUEUE y |   ...        |             | |
1997         |         |->-[flow]-->--|             | |
1998         +---------+              +-------------+ |
1999                                    +-------------+
2000.Ed
2001It is important to understand the role of the SCHED_MASK
2002and FLOW_MASK, which are configured through the commands
2003.Dl "ipfw sched N config mask SCHED_MASK ..."
2004and
2005.Dl "ipfw queue X config mask FLOW_MASK ..." .
2006.Pp
2007The SCHED_MASK is used to assign flows to one or more
2008scheduler instances, one for each
2009value of the packet's 5-tuple after applying SCHED_MASK.
2010As an example, using ``src-ip 0xffffff00'' creates one instance
2011for each /24 destination subnet.
2012.Pp
2013The FLOW_MASK, together with the SCHED_MASK, is used to split
2014packets into flows. As an example, using
2015``src-ip 0x000000ff''
2016together with the previous SCHED_MASK makes a flow for
2017each individual source address. In turn, flows for each /24
2018subnet will be sent to the same scheduler instance.
2019.Pp
2020The above diagram holds even for the
2021.Em pipe
2022case, with the only restriction that a
2023.Em pipe
2024only supports a SCHED_MASK, and forces the use of a FIFO
2025scheduler (these are for backward compatibility reasons;
2026in fact, internally, a
2027.Nm dummynet's
2028pipe is implemented exactly as above).
2029.Pp
2030There are two modes of
2031.Nm dummynet
2032operation:
2033.Dq normal
2034and
2035.Dq fast .
2036The
2037.Dq normal
2038mode tries to emulate a real link: the
2039.Nm dummynet
2040scheduler ensures that the packet will not leave the pipe faster than it
2041would on the real link with a given bandwidth.
2042The
2043.Dq fast
2044mode allows certain packets to bypass the
2045.Nm dummynet
2046scheduler (if packet flow does not exceed pipe's bandwidth).
2047This is the reason why the
2048.Dq fast
2049mode requires less CPU cycles per packet (on average) and packet latency
2050can be significantly lower in comparison to a real link with the same
2051bandwidth.
2052The default mode is
2053.Dq normal .
2054The
2055.Dq fast
2056mode can be enabled by setting the
2057.Va net.inet.ip.dummynet.io_fast
2058.Xr sysctl 8
2059variable to a non-zero value.
2060.Pp
2061.Ss PIPE, QUEUE AND SCHEDULER CONFIGURATION
2062The
2063.Em pipe ,
2064.Em queue
2065and
2066.Em scheduler
2067configuration commands are the following:
2068.Bd -ragged -offset indent
2069.Cm pipe Ar number Cm config Ar pipe-configuration
2070.Pp
2071.Cm queue Ar number Cm config Ar queue-configuration
2072.Pp
2073.Cm sched Ar number Cm config Ar sched-configuration
2074.Ed
2075.Pp
2076The following parameters can be configured for a pipe:
2077.Pp
2078.Bl -tag -width indent -compact
2079.It Cm bw Ar bandwidth | device
2080Bandwidth, measured in
2081.Sm off
2082.Op Cm K | M
2083.Brq Cm bit/s | Byte/s .
2084.Sm on
2085.Pp
2086A value of 0 (default) means unlimited bandwidth.
2087The unit must immediately follow the number, as in
2088.Pp
2089.Dl "ipfw pipe 1 config bw 300Kbit/s"
2090.Pp
2091If a device name is specified instead of a numeric value, as in
2092.Pp
2093.Dl "ipfw pipe 1 config bw tun0"
2094.Pp
2095then the transmit clock is supplied by the specified device.
2096At the moment only the
2097.Xr tun 4
2098device supports this
2099functionality, for use in conjunction with
2100.Xr ppp 8 .
2101.Pp
2102.It Cm delay Ar ms-delay
2103Propagation delay, measured in milliseconds.
2104The value is rounded to the next multiple of the clock tick
2105(typically 10ms, but it is a good practice to run kernels
2106with
2107.Dq "options HZ=1000"
2108to reduce
2109the granularity to 1ms or less).
2110The default value is 0, meaning no delay.
2111.Pp
2112.It Cm burst Ar size
2113If the data to be sent exceeds the pipe's bandwidth limit
2114(and the pipe was previously idle), up to
2115.Ar size
2116bytes of data are allowed to bypass the
2117.Nm dummynet
2118scheduler, and will be sent as fast as the physical link allows.
2119Any additional data will be transmitted at the rate specified
2120by the
2121.Nm pipe
2122bandwidth.
2123The burst size depends on how long the pipe has been idle;
2124the effective burst size is calculated as follows:
2125MAX(
2126.Ar size
2127,
2128.Nm bw
2129* pipe_idle_time).
2130.Pp
2131.It Cm profile Ar filename
2132A file specifying the additional overhead incurred in the transmission
2133of a packet on the link.
2134.Pp
2135Some link types introduce extra delays in the transmission
2136of a packet, e.g. because of MAC level framing, contention on
2137the use of the channel, MAC level retransmissions and so on.
2138From our point of view, the channel is effectively unavailable
2139for this extra time, which is constant or variable depending
2140on the link type. Additionally, packets may be dropped after this
2141time (e.g. on a wireless link after too many retransmissions).
2142We can model the additional delay with an empirical curve
2143that represents its distribution.
2144.Bd -literal -offset indent
2145      cumulative probability
2146      1.0 ^
2147          |
2148      L   +-- loss-level          x
2149          |                 ******
2150          |                *
2151          |           *****
2152          |          *
2153          |        **
2154          |       *
2155          +-------*------------------->
2156                      delay
2157.Ed
2158The empirical curve may have both vertical and horizontal lines.
2159Vertical lines represent constant delay for a range of
2160probabilities.
2161Horizontal lines correspond to a discontinuity in the delay
2162distribution: the pipe will use the largest delay for a
2163given probability.
2164.Pp
2165The file format is the following, with whitespace acting as
2166a separator and '#' indicating the beginning a comment:
2167.Bl -tag -width indent
2168.It Cm name Ar identifier
2169optional name (listed by "ipfw pipe show")
2170to identify the delay distribution;
2171.It Cm bw Ar value
2172the bandwidth used for the pipe.
2173If not specified here, it must be present
2174explicitly as a configuration parameter for the pipe;
2175.It Cm loss-level Ar L
2176the probability above which packets are lost.
2177(0.0 <= L <= 1.0, default 1.0 i.e. no loss);
2178.It Cm samples Ar N
2179the number of samples used in the internal
2180representation of the curve (2..1024; default 100);
2181.It Cm "delay prob" | "prob delay"
2182One of these two lines is mandatory and defines
2183the format of the following lines with data points.
2184.It Ar XXX Ar YYY
21852 or more lines representing points in the curve,
2186with either delay or probability first, according
2187to the chosen format.
2188The unit for delay is milliseconds.
2189Data points do not need to be sorted.
2190Also, the number of actual lines can be different
2191from the value of the "samples" parameter:
2192.Nm
2193utility will sort and interpolate
2194the curve as needed.
2195.El
2196.Pp
2197Example of a profile file:
2198.Bd -literal -offset indent
2199name    bla_bla_bla
2200samples 100
2201loss-level    0.86
2202prob    delay
22030       200	# minimum overhead is 200ms
22040.5     200
22050.5     300
22060.8     1000
22070.9     1300
22081       1300
2209#configuration file end
2210.Ed
2211.El
2212.Pp
2213The following parameters can be configured for a queue:
2214.Pp
2215.Bl -tag -width indent -compact
2216.It Cm pipe Ar pipe_nr
2217Connects a queue to the specified pipe.
2218Multiple queues (with the same or different weights) can be connected to
2219the same pipe, which specifies the aggregate rate for the set of queues.
2220.Pp
2221.It Cm weight Ar weight
2222Specifies the weight to be used for flows matching this queue.
2223The weight must be in the range 1..100, and defaults to 1.
2224.El
2225.Pp
2226The following parameters can be configured for a scheduler:
2227.Pp
2228.Bl -tag -width indent -compact
2229.It Cm type Ar {fifo | wf2qp | rr | qfq}
2230specifies the scheduling algorithm to use.
2231.Bl -tag -width indent -compact
2232.It cm fifo
2233is just a FIFO scheduler (which means that all packets
2234are stored in the same queue as they arrive to the scheduler).
2235FIFO has O(1) per-packet time complexity, with very low
2236constants (estimate 60-80ns on a 2Ghz desktop machine)
2237but gives no service guarantees.
2238.It Cm wf2qp
2239implements the WF2Q+ algorithm, which is a Weighted Fair Queueing
2240algorithm which permits flows to share bandwidth according to
2241their weights. Note that weights are not priorities; even a flow
2242with a minuscule weight will never starve.
2243WF2Q+ has O(log N) per-packet processing cost, where N is the number
2244of flows, and is the default algorithm used by previous versions
2245dummynet's queues.
2246.It Cm rr
2247implements the Deficit Round Robin algorithm, which has O(1) processing
2248costs (roughly, 100-150ns per packet)
2249and permits bandwidth allocation according to weights, but
2250with poor service guarantees.
2251.It Cm qfq
2252implements the QFQ algorithm, which is a very fast variant of
2253WF2Q+, with similar service guarantees and O(1) processing
2254costs (roughly, 200-250ns per packet).
2255.El
2256.El
2257.Pp
2258In addition to the type, all parameters allowed for a pipe can also
2259be specified for a scheduler.
2260.Pp
2261Finally, the following parameters can be configured for both
2262pipes and queues:
2263.Pp
2264.Bl -tag -width XXXX -compact
2265.It Cm buckets Ar hash-table-size
2266Specifies the size of the hash table used for storing the
2267various queues.
2268Default value is 64 controlled by the
2269.Xr sysctl 8
2270variable
2271.Va net.inet.ip.dummynet.hash_size ,
2272allowed range is 16 to 65536.
2273.Pp
2274.It Cm mask Ar mask-specifier
2275Packets sent to a given pipe or queue by an
2276.Nm
2277rule can be further classified into multiple flows, each of which is then
2278sent to a different
2279.Em dynamic
2280pipe or queue.
2281A flow identifier is constructed by masking the IP addresses,
2282ports and protocol types as specified with the
2283.Cm mask
2284options in the configuration of the pipe or queue.
2285For each different flow identifier, a new pipe or queue is created
2286with the same parameters as the original object, and matching packets
2287are sent to it.
2288.Pp
2289Thus, when
2290.Em dynamic pipes
2291are used, each flow will get the same bandwidth as defined by the pipe,
2292whereas when
2293.Em dynamic queues
2294are used, each flow will share the parent's pipe bandwidth evenly
2295with other flows generated by the same queue (note that other queues
2296with different weights might be connected to the same pipe).
2297.br
2298Available mask specifiers are a combination of one or more of the following:
2299.Pp
2300.Cm dst-ip Ar mask ,
2301.Cm dst-ip6 Ar mask ,
2302.Cm src-ip Ar mask ,
2303.Cm src-ip6 Ar mask ,
2304.Cm dst-port Ar mask ,
2305.Cm src-port Ar mask ,
2306.Cm flow-id Ar mask ,
2307.Cm proto Ar mask
2308or
2309.Cm all ,
2310.Pp
2311where the latter means all bits in all fields are significant.
2312.Pp
2313.It Cm noerror
2314When a packet is dropped by a
2315.Nm dummynet
2316queue or pipe, the error
2317is normally reported to the caller routine in the kernel, in the
2318same way as it happens when a device queue fills up.
2319Setting this
2320option reports the packet as successfully delivered, which can be
2321needed for some experimental setups where you want to simulate
2322loss or congestion at a remote router.
2323.Pp
2324.It Cm plr Ar packet-loss-rate
2325Packet loss rate.
2326Argument
2327.Ar packet-loss-rate
2328is a floating-point number between 0 and 1, with 0 meaning no
2329loss, 1 meaning 100% loss.
2330The loss rate is internally represented on 31 bits.
2331.Pp
2332.It Cm queue Brq Ar slots | size Ns Cm Kbytes
2333Queue size, in
2334.Ar slots
2335or
2336.Cm KBytes .
2337Default value is 50 slots, which
2338is the typical queue size for Ethernet devices.
2339Note that for slow speed links you should keep the queue
2340size short or your traffic might be affected by a significant
2341queueing delay.
2342E.g., 50 max-sized ethernet packets (1500 bytes) mean 600Kbit
2343or 20s of queue on a 30Kbit/s pipe.
2344Even worse effects can result if you get packets from an
2345interface with a much larger MTU, e.g.\& the loopback interface
2346with its 16KB packets.
2347The
2348.Xr sysctl 8
2349variables
2350.Em net.inet.ip.dummynet.pipe_byte_limit
2351and
2352.Em net.inet.ip.dummynet.pipe_slot_limit
2353control the maximum lengths that can be specified.
2354.Pp
2355.It Cm red | gred Ar w_q Ns / Ns Ar min_th Ns / Ns Ar max_th Ns / Ns Ar max_p
2356Make use of the RED (Random Early Detection) queue management algorithm.
2357.Ar w_q
2358and
2359.Ar max_p
2360are floating
2361point numbers between 0 and 1 (0 not included), while
2362.Ar min_th
2363and
2364.Ar max_th
2365are integer numbers specifying thresholds for queue management
2366(thresholds are computed in bytes if the queue has been defined
2367in bytes, in slots otherwise).
2368The
2369.Nm dummynet
2370also supports the gentle RED variant (gred).
2371Three
2372.Xr sysctl 8
2373variables can be used to control the RED behaviour:
2374.Bl -tag -width indent
2375.It Va net.inet.ip.dummynet.red_lookup_depth
2376specifies the accuracy in computing the average queue
2377when the link is idle (defaults to 256, must be greater than zero)
2378.It Va net.inet.ip.dummynet.red_avg_pkt_size
2379specifies the expected average packet size (defaults to 512, must be
2380greater than zero)
2381.It Va net.inet.ip.dummynet.red_max_pkt_size
2382specifies the expected maximum packet size, only used when queue
2383thresholds are in bytes (defaults to 1500, must be greater than zero).
2384.El
2385.El
2386.Pp
2387When used with IPv6 data,
2388.Nm dummynet
2389currently has several limitations.
2390Information necessary to route link-local packets to an
2391interface is not available after processing by
2392.Nm dummynet
2393so those packets are dropped in the output path.
2394Care should be taken to ensure that link-local packets are not passed to
2395.Nm dummynet .
2396.Sh CHECKLIST
2397Here are some important points to consider when designing your
2398rules:
2399.Bl -bullet
2400.It
2401Remember that you filter both packets going
2402.Cm in
2403and
2404.Cm out .
2405Most connections need packets going in both directions.
2406.It
2407Remember to test very carefully.
2408It is a good idea to be near the console when doing this.
2409If you cannot be near the console,
2410use an auto-recovery script such as the one in
2411.Pa /usr/share/examples/ipfw/change_rules.sh .
2412.It
2413Do not forget the loopback interface.
2414.El
2415.Sh FINE POINTS
2416.Bl -bullet
2417.It
2418There are circumstances where fragmented datagrams are unconditionally
2419dropped.
2420TCP packets are dropped if they do not contain at least 20 bytes of
2421TCP header, UDP packets are dropped if they do not contain a full 8
2422byte UDP header, and ICMP packets are dropped if they do not contain
24234 bytes of ICMP header, enough to specify the ICMP type, code, and
2424checksum.
2425These packets are simply logged as
2426.Dq pullup failed
2427since there may not be enough good data in the packet to produce a
2428meaningful log entry.
2429.It
2430Another type of packet is unconditionally dropped, a TCP packet with a
2431fragment offset of one.
2432This is a valid packet, but it only has one use, to try
2433to circumvent firewalls.
2434When logging is enabled, these packets are
2435reported as being dropped by rule -1.
2436.It
2437If you are logged in over a network, loading the
2438.Xr kld 4
2439version of
2440.Nm
2441is probably not as straightforward as you would think.
2442The following command line is recommended:
2443.Bd -literal -offset indent
2444kldload ipfw && \e
2445ipfw add 32000 allow ip from any to any
2446.Ed
2447.Pp
2448Along the same lines, doing an
2449.Bd -literal -offset indent
2450ipfw flush
2451.Ed
2452.Pp
2453in similar surroundings is also a bad idea.
2454.It
2455The
2456.Nm
2457filter list may not be modified if the system security level
2458is set to 3 or higher
2459(see
2460.Xr init 8
2461for information on system security levels).
2462.El
2463.Sh PACKET DIVERSION
2464A
2465.Xr divert 4
2466socket bound to the specified port will receive all packets
2467diverted to that port.
2468If no socket is bound to the destination port, or if the divert module is
2469not loaded, or if the kernel was not compiled with divert socket support,
2470the packets are dropped.
2471.Sh NETWORK ADDRESS TRANSLATION (NAT)
2472.Pp
2473.Nm
2474support in-kernel NAT using the kernel version of
2475.Xr libalias 3 .
2476.Pp
2477The nat configuration command is the following:
2478.Bd -ragged -offset indent
2479.Bk -words
2480.Cm nat
2481.Ar nat_number
2482.Cm config
2483.Ar nat-configuration
2484.Ek
2485.Ed
2486.Pp
2487The following parameters can be configured:
2488.Bl -tag -width indent
2489.It Cm ip Ar ip_address
2490Define an ip address to use for aliasing.
2491.It Cm if Ar nic
2492Use ip address of NIC for aliasing, dynamically changing
2493it if NIC's ip address changes.
2494.It Cm log
2495Enable logging on this nat instance.
2496.It Cm deny_in
2497Deny any incoming connection from outside world.
2498.It Cm same_ports
2499Try to leave the alias port numbers unchanged from
2500the actual local port numbers.
2501.It Cm unreg_only
2502Traffic on the local network not originating from an
2503unregistered address spaces will be ignored.
2504.It Cm reset
2505Reset table of the packet aliasing engine on address change.
2506.It Cm reverse
2507Reverse the way libalias handles aliasing.
2508.It Cm proxy_only
2509Obey transparent proxy rules only, packet aliasing is not performed.
2510.It Cm skip_global
2511Skip instance in case of global state lookup (see below).
2512.El
2513.Pp
2514Some specials value can be supplied instead of
2515.Va nat_number:
2516.Bl -tag -width indent
2517.It Cm global
2518Looks up translation state in all configured nat instances.
2519If an entry is found, packet is aliased according to that entry.
2520If no entry was found in any of the instances, packet is passed unchanged,
2521and no new entry will be created.
2522See section
2523.Sx MULTIPLE INSTANCES
2524in
2525.Xr natd 8
2526for more information.
2527.It Cm tablearg
2528Uses argument supplied in lookup table. See
2529.Sx LOOKUP TABLES
2530section below for more information on lookup tables.
2531.El
2532.Pp
2533To let the packet continue after being (de)aliased, set the sysctl variable
2534.Va net.inet.ip.fw.one_pass
2535to 0.
2536For more information about aliasing modes, refer to
2537.Xr libalias 3 .
2538See Section
2539.Sx EXAMPLES
2540for some examples about nat usage.
2541.Ss REDIRECT AND LSNAT SUPPORT IN IPFW
2542Redirect and LSNAT support follow closely the syntax used in
2543.Xr natd 8 .
2544See Section
2545.Sx EXAMPLES
2546for some examples on how to do redirect and lsnat.
2547.Ss SCTP NAT SUPPORT
2548SCTP nat can be configured in a similar manner to TCP through the
2549.Nm
2550command line tool.
2551The main difference is that
2552.Nm sctp nat
2553does not do port translation.
2554Since the local and global side ports will be the same,
2555there is no need to specify both.
2556Ports are redirected as follows:
2557.Bd -ragged -offset indent
2558.Bk -words
2559.Cm nat
2560.Ar nat_number
2561.Cm config if
2562.Ar nic
2563.Cm redirect_port sctp
2564.Ar ip_address [,addr_list] {[port | port-port] [,ports]}
2565.Ek
2566.Ed
2567.Pp
2568Most
2569.Nm sctp nat
2570configuration can be done in real-time through the
2571.Xr sysctl 8
2572interface.
2573All may be changed dynamically, though the hash_table size will only
2574change for new
2575.Nm nat
2576instances.
2577See
2578.Sx SYSCTL VARIABLES
2579for more info.
2580.Sh SYSCTL VARIABLES
2581A set of
2582.Xr sysctl 8
2583variables controls the behaviour of the firewall and
2584associated modules
2585.Pq Nm dummynet , bridge , sctp nat .
2586These are shown below together with their default value
2587(but always check with the
2588.Xr sysctl 8
2589command what value is actually in use) and meaning:
2590.Bl -tag -width indent
2591.It Va net.inet.ip.alias.sctp.accept_global_ootb_addip: No 0
2592Defines how the
2593.Nm nat
2594responds to receipt of global OOTB ASCONF-AddIP:
2595.Bl -tag -width indent
2596.It Cm 0
2597No response (unless a partially matching association exists -
2598ports and vtags match but global address does not)
2599.It Cm 1
2600.Nm nat
2601will accept and process all OOTB global AddIP messages.
2602.El
2603.Pp
2604Option 1 should never be selected as this forms a security risk.
2605An attacker can
2606establish multiple fake associations by sending AddIP messages.
2607.It Va net.inet.ip.alias.sctp.chunk_proc_limit: No 5
2608Defines the maximum number of chunks in an SCTP packet that will be parsed for a
2609packet that matches an existing association.
2610This value is enforced to be greater or equal than
2611.Cm net.inet.ip.alias.sctp.initialising_chunk_proc_limit .
2612A high value is
2613a DoS risk yet setting too low a value may result in important control chunks in
2614the packet not being located and parsed.
2615.It Va net.inet.ip.alias.sctp.error_on_ootb: No 1
2616Defines when the
2617.Nm nat
2618responds to any Out-of-the-Blue (OOTB) packets with ErrorM packets.
2619An OOTB packet is a packet that arrives with no existing association
2620registered in the
2621.Nm nat
2622and is not an INIT or ASCONF-AddIP packet:
2623.Bl -tag -width indent
2624.It Cm 0
2625ErrorM is never sent in response to OOTB packets.
2626.It Cm 1
2627ErrorM is only sent to OOTB packets received on the local side.
2628.It Cm 2
2629ErrorM is sent to the local side and on the global side ONLY if there is a
2630partial match (ports and vtags match but the source global IP does not).
2631This value is only useful if the
2632.Nm nat
2633is tracking global IP addresses.
2634.It Cm 3
2635ErrorM is sent in response to all OOTB packets on both the local and global side
2636(DoS risk).
2637.El
2638.Pp
2639At the moment the default is 0, since the ErrorM packet is not yet
2640supported by most SCTP stacks.
2641When it is supported, and if not tracking
2642global addresses, we recommend setting this value to 1 to allow
2643multi-homed local hosts to function with the
2644.Nm nat .
2645To track global addresses, we recommend setting this value to 2 to
2646allow global hosts to be informed when they need to (re)send an
2647ASCONF-AddIP.
2648Value 3 should never be chosen (except for debugging) as the
2649.Nm nat
2650will respond to all OOTB global packets (a DoS risk).
2651.It Va net.inet.ip.alias.sctp.hashtable_size: No 2003
2652Size of hash tables used for
2653.Nm nat
2654lookups (100 < prime_number > 1000001).
2655This value sets the
2656.Nm hash table
2657size for any future created
2658.Nm nat
2659instance and therefore must be set prior to creating a
2660.Nm nat
2661instance.
2662The table sizes may be changed to suit specific needs.
2663If there will be few
2664concurrent associations, and memory is scarce, you may make these smaller.
2665If there will be many thousands (or millions) of concurrent associations, you
2666should make these larger.
2667A prime number is best for the table size.
2668The sysctl
2669update function will adjust your input value to the next highest prime number.
2670.It Va net.inet.ip.alias.sctp.holddown_time:  No 0
2671Hold association in table for this many seconds after receiving a
2672SHUTDOWN-COMPLETE.
2673This allows endpoints to correct shutdown gracefully if a
2674shutdown_complete is lost and retransmissions are required.
2675.It Va net.inet.ip.alias.sctp.init_timer: No 15
2676Timeout value while waiting for (INIT-ACK|AddIP-ACK).
2677This value cannot be 0.
2678.It Va net.inet.ip.alias.sctp.initialising_chunk_proc_limit: No 2
2679Defines the maximum number of chunks in an SCTP packet that will be parsed when
2680no existing association exists that matches that packet.
2681Ideally this packet
2682will only be an INIT or ASCONF-AddIP packet.
2683A higher value may become a DoS
2684risk as malformed packets can consume processing resources.
2685.It Va net.inet.ip.alias.sctp.param_proc_limit: No 25
2686Defines the maximum number of parameters within a chunk that will be parsed in a
2687packet.
2688As for other similar sysctl variables, larger values pose a DoS risk.
2689.It Va net.inet.ip.alias.sctp.log_level: No 0
2690Level of detail in the system log messages (0 \- minimal, 1 \- event,
26912 \- info, 3 \- detail, 4 \- debug, 5 \- max debug). May be a good
2692option in high loss environments.
2693.It Va net.inet.ip.alias.sctp.shutdown_time: No 15
2694Timeout value while waiting for SHUTDOWN-COMPLETE.
2695This value cannot be 0.
2696.It Va net.inet.ip.alias.sctp.track_global_addresses: No 0
2697Enables/disables global IP address tracking within the
2698.Nm nat
2699and places an
2700upper limit on the number of addresses tracked for each association:
2701.Bl -tag -width indent
2702.It Cm 0
2703Global tracking is disabled
2704.It Cm >1
2705Enables tracking, the maximum number of addresses tracked for each
2706association is limited to this value
2707.El
2708.Pp
2709This variable is fully dynamic, the new value will be adopted for all newly
2710arriving associations, existing associations are treated as they were previously.
2711Global tracking will decrease the number of collisions within the
2712.Nm nat
2713at a cost
2714of increased processing load, memory usage, complexity, and possible
2715.Nm nat
2716state
2717problems in complex networks with multiple
2718.Nm nats .
2719We recommend not tracking
2720global IP addresses, this will still result in a fully functional
2721.Nm nat .
2722.It Va net.inet.ip.alias.sctp.up_timer: No 300
2723Timeout value to keep an association up with no traffic.
2724This value cannot be 0.
2725.It Va net.inet.ip.dummynet.expire : No 1
2726Lazily delete dynamic pipes/queue once they have no pending traffic.
2727You can disable this by setting the variable to 0, in which case
2728the pipes/queues will only be deleted when the threshold is reached.
2729.It Va net.inet.ip.dummynet.hash_size : No 64
2730Default size of the hash table used for dynamic pipes/queues.
2731This value is used when no
2732.Cm buckets
2733option is specified when configuring a pipe/queue.
2734.It Va net.inet.ip.dummynet.io_fast : No 0
2735If set to a non-zero value,
2736the
2737.Dq fast
2738mode of
2739.Nm dummynet
2740operation (see above) is enabled.
2741.It Va net.inet.ip.dummynet.io_pkt
2742Number of packets passed to
2743.Nm dummynet .
2744.It Va net.inet.ip.dummynet.io_pkt_drop
2745Number of packets dropped by
2746.Nm dummynet .
2747.It Va net.inet.ip.dummynet.io_pkt_fast
2748Number of packets bypassed by the
2749.Nm dummynet
2750scheduler.
2751.It Va net.inet.ip.dummynet.max_chain_len : No 16
2752Target value for the maximum number of pipes/queues in a hash bucket.
2753The product
2754.Cm max_chain_len*hash_size
2755is used to determine the threshold over which empty pipes/queues
2756will be expired even when
2757.Cm net.inet.ip.dummynet.expire=0 .
2758.It Va net.inet.ip.dummynet.red_lookup_depth : No 256
2759.It Va net.inet.ip.dummynet.red_avg_pkt_size : No 512
2760.It Va net.inet.ip.dummynet.red_max_pkt_size : No 1500
2761Parameters used in the computations of the drop probability
2762for the RED algorithm.
2763.It Va net.inet.ip.dummynet.pipe_byte_limit : No 1048576
2764.It Va net.inet.ip.dummynet.pipe_slot_limit : No 100
2765The maximum queue size that can be specified in bytes or packets.
2766These limits prevent accidental exhaustion of resources such as mbufs.
2767If you raise these limits,
2768you should make sure the system is configured so that sufficient resources
2769are available.
2770.It Va net.inet.ip.fw.autoinc_step : No 100
2771Delta between rule numbers when auto-generating them.
2772The value must be in the range 1..1000.
2773.It Va net.inet.ip.fw.curr_dyn_buckets : Va net.inet.ip.fw.dyn_buckets
2774The current number of buckets in the hash table for dynamic rules
2775(readonly).
2776.It Va net.inet.ip.fw.debug : No 1
2777Controls debugging messages produced by
2778.Nm .
2779.It Va net.inet.ip.fw.default_rule : No 65535
2780The default rule number (read-only).
2781By the design of
2782.Nm , the default rule is the last one, so its number
2783can also serve as the highest number allowed for a rule.
2784.It Va net.inet.ip.fw.dyn_buckets : No 256
2785The number of buckets in the hash table for dynamic rules.
2786Must be a power of 2, up to 65536.
2787It only takes effect when all dynamic rules have expired, so you
2788are advised to use a
2789.Cm flush
2790command to make sure that the hash table is resized.
2791.It Va net.inet.ip.fw.dyn_count : No 3
2792Current number of dynamic rules
2793(read-only).
2794.It Va net.inet.ip.fw.dyn_keepalive : No 1
2795Enables generation of keepalive packets for
2796.Cm keep-state
2797rules on TCP sessions.
2798A keepalive is generated to both
2799sides of the connection every 5 seconds for the last 20
2800seconds of the lifetime of the rule.
2801.It Va net.inet.ip.fw.dyn_max : No 8192
2802Maximum number of dynamic rules.
2803When you hit this limit, no more dynamic rules can be
2804installed until old ones expire.
2805.It Va net.inet.ip.fw.dyn_ack_lifetime : No 300
2806.It Va net.inet.ip.fw.dyn_syn_lifetime : No 20
2807.It Va net.inet.ip.fw.dyn_fin_lifetime : No 1
2808.It Va net.inet.ip.fw.dyn_rst_lifetime : No 1
2809.It Va net.inet.ip.fw.dyn_udp_lifetime : No 5
2810.It Va net.inet.ip.fw.dyn_short_lifetime : No 30
2811These variables control the lifetime, in seconds, of dynamic
2812rules.
2813Upon the initial SYN exchange the lifetime is kept short,
2814then increased after both SYN have been seen, then decreased
2815again during the final FIN exchange or when a RST is received.
2816Both
2817.Em dyn_fin_lifetime
2818and
2819.Em dyn_rst_lifetime
2820must be strictly lower than 5 seconds, the period of
2821repetition of keepalives.
2822The firewall enforces that.
2823.It Va net.inet.ip.fw.enable : No 1
2824Enables the firewall.
2825Setting this variable to 0 lets you run your machine without
2826firewall even if compiled in.
2827.It Va net.inet6.ip6.fw.enable : No 1
2828provides the same functionality as above for the IPv6 case.
2829.It Va net.inet.ip.fw.one_pass : No 1
2830When set, the packet exiting from the
2831.Nm dummynet
2832pipe or from
2833.Xr ng_ipfw 4
2834node is not passed though the firewall again.
2835Otherwise, after an action, the packet is
2836reinjected into the firewall at the next rule.
2837.It Va net.inet.ip.fw.tables_max : No 128
2838Maximum number of tables (read-only).
2839.It Va net.inet.ip.fw.verbose : No 1
2840Enables verbose messages.
2841.It Va net.inet.ip.fw.verbose_limit : No 0
2842Limits the number of messages produced by a verbose firewall.
2843.It Va net.inet6.ip6.fw.deny_unknown_exthdrs : No 1
2844If enabled packets with unknown IPv6 Extension Headers will be denied.
2845.It Va net.link.ether.ipfw : No 0
2846Controls whether layer-2 packets are passed to
2847.Nm .
2848Default is no.
2849.It Va net.link.bridge.ipfw : No 0
2850Controls whether bridged packets are passed to
2851.Nm .
2852Default is no.
2853.El
2854.Pp
2855.Sh EXAMPLES
2856There are far too many possible uses of
2857.Nm
2858so this Section will only give a small set of examples.
2859.Pp
2860.Ss BASIC PACKET FILTERING
2861This command adds an entry which denies all tcp packets from
2862.Em cracker.evil.org
2863to the telnet port of
2864.Em wolf.tambov.su
2865from being forwarded by the host:
2866.Pp
2867.Dl "ipfw add deny tcp from cracker.evil.org to wolf.tambov.su telnet"
2868.Pp
2869This one disallows any connection from the entire cracker's
2870network to my host:
2871.Pp
2872.Dl "ipfw add deny ip from 123.45.67.0/24 to my.host.org"
2873.Pp
2874A first and efficient way to limit access (not using dynamic rules)
2875is the use of the following rules:
2876.Pp
2877.Dl "ipfw add allow tcp from any to any established"
2878.Dl "ipfw add allow tcp from net1 portlist1 to net2 portlist2 setup"
2879.Dl "ipfw add allow tcp from net3 portlist3 to net3 portlist3 setup"
2880.Dl "..."
2881.Dl "ipfw add deny tcp from any to any"
2882.Pp
2883The first rule will be a quick match for normal TCP packets,
2884but it will not match the initial SYN packet, which will be
2885matched by the
2886.Cm setup
2887rules only for selected source/destination pairs.
2888All other SYN packets will be rejected by the final
2889.Cm deny
2890rule.
2891.Pp
2892If you administer one or more subnets, you can take advantage
2893of the address sets and or-blocks and write extremely
2894compact rulesets which selectively enable services to blocks
2895of clients, as below:
2896.Pp
2897.Dl "goodguys=\*q{ 10.1.2.0/24{20,35,66,18} or 10.2.3.0/28{6,3,11} }\*q"
2898.Dl "badguys=\*q10.1.2.0/24{8,38,60}\*q"
2899.Dl ""
2900.Dl "ipfw add allow ip from ${goodguys} to any"
2901.Dl "ipfw add deny ip from ${badguys} to any"
2902.Dl "... normal policies ..."
2903.Pp
2904The
2905.Cm verrevpath
2906option could be used to do automated anti-spoofing by adding the
2907following to the top of a ruleset:
2908.Pp
2909.Dl "ipfw add deny ip from any to any not verrevpath in"
2910.Pp
2911This rule drops all incoming packets that appear to be coming to the
2912system on the wrong interface.
2913For example, a packet with a source
2914address belonging to a host on a protected internal network would be
2915dropped if it tried to enter the system from an external interface.
2916.Pp
2917The
2918.Cm antispoof
2919option could be used to do similar but more restricted anti-spoofing
2920by adding the following to the top of a ruleset:
2921.Pp
2922.Dl "ipfw add deny ip from any to any not antispoof in"
2923.Pp
2924This rule drops all incoming packets that appear to be coming from another
2925directly connected system but on the wrong interface.
2926For example, a packet with a source address of
2927.Li 192.168.0.0/24 ,
2928configured on
2929.Li fxp0 ,
2930but coming in on
2931.Li fxp1
2932would be dropped.
2933.Ss DYNAMIC RULES
2934In order to protect a site from flood attacks involving fake
2935TCP packets, it is safer to use dynamic rules:
2936.Pp
2937.Dl "ipfw add check-state"
2938.Dl "ipfw add deny tcp from any to any established"
2939.Dl "ipfw add allow tcp from my-net to any setup keep-state"
2940.Pp
2941This will let the firewall install dynamic rules only for
2942those connection which start with a regular SYN packet coming
2943from the inside of our network.
2944Dynamic rules are checked when encountering the first
2945.Cm check-state
2946or
2947.Cm keep-state
2948rule.
2949A
2950.Cm check-state
2951rule should usually be placed near the beginning of the
2952ruleset to minimize the amount of work scanning the ruleset.
2953Your mileage may vary.
2954.Pp
2955To limit the number of connections a user can open
2956you can use the following type of rules:
2957.Pp
2958.Dl "ipfw add allow tcp from my-net/24 to any setup limit src-addr 10"
2959.Dl "ipfw add allow tcp from any to me setup limit src-addr 4"
2960.Pp
2961The former (assuming it runs on a gateway) will allow each host
2962on a /24 network to open at most 10 TCP connections.
2963The latter can be placed on a server to make sure that a single
2964client does not use more than 4 simultaneous connections.
2965.Pp
2966.Em BEWARE :
2967stateful rules can be subject to denial-of-service attacks
2968by a SYN-flood which opens a huge number of dynamic rules.
2969The effects of such attacks can be partially limited by
2970acting on a set of
2971.Xr sysctl 8
2972variables which control the operation of the firewall.
2973.Pp
2974Here is a good usage of the
2975.Cm list
2976command to see accounting records and timestamp information:
2977.Pp
2978.Dl ipfw -at list
2979.Pp
2980or in short form without timestamps:
2981.Pp
2982.Dl ipfw -a list
2983.Pp
2984which is equivalent to:
2985.Pp
2986.Dl ipfw show
2987.Pp
2988Next rule diverts all incoming packets from 192.168.2.0/24
2989to divert port 5000:
2990.Pp
2991.Dl ipfw divert 5000 ip from 192.168.2.0/24 to any in
2992.Pp
2993.Ss TRAFFIC SHAPING
2994The following rules show some of the applications of
2995.Nm
2996and
2997.Nm dummynet
2998for simulations and the like.
2999.Pp
3000This rule drops random incoming packets with a probability
3001of 5%:
3002.Pp
3003.Dl "ipfw add prob 0.05 deny ip from any to any in"
3004.Pp
3005A similar effect can be achieved making use of
3006.Nm dummynet
3007pipes:
3008.Pp
3009.Dl "ipfw add pipe 10 ip from any to any"
3010.Dl "ipfw pipe 10 config plr 0.05"
3011.Pp
3012We can use pipes to artificially limit bandwidth, e.g.\& on a
3013machine acting as a router, if we want to limit traffic from
3014local clients on 192.168.2.0/24 we do:
3015.Pp
3016.Dl "ipfw add pipe 1 ip from 192.168.2.0/24 to any out"
3017.Dl "ipfw pipe 1 config bw 300Kbit/s queue 50KBytes"
3018.Pp
3019note that we use the
3020.Cm out
3021modifier so that the rule is not used twice.
3022Remember in fact that
3023.Nm
3024rules are checked both on incoming and outgoing packets.
3025.Pp
3026Should we want to simulate a bidirectional link with bandwidth
3027limitations, the correct way is the following:
3028.Pp
3029.Dl "ipfw add pipe 1 ip from any to any out"
3030.Dl "ipfw add pipe 2 ip from any to any in"
3031.Dl "ipfw pipe 1 config bw 64Kbit/s queue 10Kbytes"
3032.Dl "ipfw pipe 2 config bw 64Kbit/s queue 10Kbytes"
3033.Pp
3034The above can be very useful, e.g.\& if you want to see how
3035your fancy Web page will look for a residential user who
3036is connected only through a slow link.
3037You should not use only one pipe for both directions, unless
3038you want to simulate a half-duplex medium (e.g.\& AppleTalk,
3039Ethernet, IRDA).
3040It is not necessary that both pipes have the same configuration,
3041so we can also simulate asymmetric links.
3042.Pp
3043Should we want to verify network performance with the RED queue
3044management algorithm:
3045.Pp
3046.Dl "ipfw add pipe 1 ip from any to any"
3047.Dl "ipfw pipe 1 config bw 500Kbit/s queue 100 red 0.002/30/80/0.1"
3048.Pp
3049Another typical application of the traffic shaper is to
3050introduce some delay in the communication.
3051This can significantly affect applications which do a lot of Remote
3052Procedure Calls, and where the round-trip-time of the
3053connection often becomes a limiting factor much more than
3054bandwidth:
3055.Pp
3056.Dl "ipfw add pipe 1 ip from any to any out"
3057.Dl "ipfw add pipe 2 ip from any to any in"
3058.Dl "ipfw pipe 1 config delay 250ms bw 1Mbit/s"
3059.Dl "ipfw pipe 2 config delay 250ms bw 1Mbit/s"
3060.Pp
3061Per-flow queueing can be useful for a variety of purposes.
3062A very simple one is counting traffic:
3063.Pp
3064.Dl "ipfw add pipe 1 tcp from any to any"
3065.Dl "ipfw add pipe 1 udp from any to any"
3066.Dl "ipfw add pipe 1 ip from any to any"
3067.Dl "ipfw pipe 1 config mask all"
3068.Pp
3069The above set of rules will create queues (and collect
3070statistics) for all traffic.
3071Because the pipes have no limitations, the only effect is
3072collecting statistics.
3073Note that we need 3 rules, not just the last one, because
3074when
3075.Nm
3076tries to match IP packets it will not consider ports, so we
3077would not see connections on separate ports as different
3078ones.
3079.Pp
3080A more sophisticated example is limiting the outbound traffic
3081on a net with per-host limits, rather than per-network limits:
3082.Pp
3083.Dl "ipfw add pipe 1 ip from 192.168.2.0/24 to any out"
3084.Dl "ipfw add pipe 2 ip from any to 192.168.2.0/24 in"
3085.Dl "ipfw pipe 1 config mask src-ip 0x000000ff bw 200Kbit/s queue 20Kbytes"
3086.Dl "ipfw pipe 2 config mask dst-ip 0x000000ff bw 200Kbit/s queue 20Kbytes"
3087.Ss LOOKUP TABLES
3088In the following example, we need to create several traffic bandwidth
3089classes and we need different hosts/networks to fall into different classes.
3090We create one pipe for each class and configure them accordingly.
3091Then we create a single table and fill it with IP subnets and addresses.
3092For each subnet/host we set the argument equal to the number of the pipe
3093that it should use.
3094Then we classify traffic using a single rule:
3095.Pp
3096.Dl "ipfw pipe 1 config bw 1000Kbyte/s"
3097.Dl "ipfw pipe 4 config bw 4000Kbyte/s"
3098.Dl "..."
3099.Dl "ipfw table 1 add 192.168.2.0/24 1"
3100.Dl "ipfw table 1 add 192.168.0.0/27 4"
3101.Dl "ipfw table 1 add 192.168.0.2 1"
3102.Dl "..."
3103.Dl "ipfw add pipe tablearg ip from table(1) to any"
3104.Pp
3105Using the
3106.Cm fwd
3107action, the table entries may include hostnames and IP addresses.
3108.Pp
3109.Dl "ipfw table 1 add 192.168.2.0/24 10.23.2.1"
3110.Dl "ipfw table 1 add 192.168.0.0/27 router1.dmz"
3111.Dl "..."
3112.Dl "ipfw add 100 fwd tablearg ip from any to table(1)"
3113.Ss SETS OF RULES
3114To add a set of rules atomically, e.g.\& set 18:
3115.Pp
3116.Dl "ipfw set disable 18"
3117.Dl "ipfw add NN set 18 ...         # repeat as needed"
3118.Dl "ipfw set enable 18"
3119.Pp
3120To delete a set of rules atomically the command is simply:
3121.Pp
3122.Dl "ipfw delete set 18"
3123.Pp
3124To test a ruleset and disable it and regain control if something goes wrong:
3125.Pp
3126.Dl "ipfw set disable 18"
3127.Dl "ipfw add NN set 18 ...         # repeat as needed"
3128.Dl "ipfw set enable 18; echo done; sleep 30 && ipfw set disable 18"
3129.Pp
3130Here if everything goes well, you press control-C before the "sleep"
3131terminates, and your ruleset will be left active.
3132Otherwise, e.g.\& if
3133you cannot access your box, the ruleset will be disabled after
3134the sleep terminates thus restoring the previous situation.
3135.Pp
3136To show rules of the specific set:
3137.Pp
3138.Dl "ipfw set 18 show"
3139.Pp
3140To show rules of the disabled set:
3141.Pp
3142.Dl "ipfw -S set 18 show"
3143.Pp
3144To clear a specific rule counters of the specific set:
3145.Pp
3146.Dl "ipfw set 18 zero NN"
3147.Pp
3148To delete a specific rule of the specific set:
3149.Pp
3150.Dl "ipfw set 18 delete NN"
3151.Ss NAT, REDIRECT AND LSNAT
3152First redirect all the traffic to nat instance 123:
3153.Pp
3154.Dl "ipfw add nat 123 all from any to any"
3155.Pp
3156Then to configure nat instance 123 to alias all the outgoing traffic with ip
3157192.168.0.123, blocking all incoming connections, trying to keep
3158same ports on both sides, clearing aliasing table on address change
3159and keeping a log of traffic/link statistics:
3160.Pp
3161.Dl "ipfw nat 123 config ip 192.168.0.123 log deny_in reset same_ports"
3162.Pp
3163Or to change address of instance 123, aliasing table will be cleared (see
3164reset option):
3165.Pp
3166.Dl "ipfw nat 123 config ip 10.0.0.1"
3167.Pp
3168To see configuration of nat instance 123:
3169.Pp
3170.Dl "ipfw nat 123 show config"
3171.Pp
3172To show logs of all the instances in range 111-999:
3173.Pp
3174.Dl "ipfw nat 111-999 show"
3175.Pp
3176To see configurations of all instances:
3177.Pp
3178.Dl "ipfw nat show config"
3179.Pp
3180Or a redirect rule with mixed modes could looks like:
3181.Pp
3182.Dl "ipfw nat 123 config redirect_addr 10.0.0.1 10.0.0.66"
3183.Dl "			 redirect_port tcp 192.168.0.1:80 500"
3184.Dl "			 redirect_proto udp 192.168.1.43 192.168.1.1"
3185.Dl "			 redirect_addr 192.168.0.10,192.168.0.11"
3186.Dl "			 	    10.0.0.100	# LSNAT"
3187.Dl "			 redirect_port tcp 192.168.0.1:80,192.168.0.10:22"
3188.Dl "			 	    500		# LSNAT"
3189.Pp
3190or it could be split in:
3191.Pp
3192.Dl "ipfw nat 1 config redirect_addr 10.0.0.1 10.0.0.66"
3193.Dl "ipfw nat 2 config redirect_port tcp 192.168.0.1:80 500"
3194.Dl "ipfw nat 3 config redirect_proto udp 192.168.1.43 192.168.1.1"
3195.Dl "ipfw nat 4 config redirect_addr 192.168.0.10,192.168.0.11,192.168.0.12"
3196.Dl "				         10.0.0.100"
3197.Dl "ipfw nat 5 config redirect_port tcp"
3198.Dl "			192.168.0.1:80,192.168.0.10:22,192.168.0.20:25 500"
3199.Pp
3200.Sh SEE ALSO
3201.Xr cpp 1 ,
3202.Xr m4 1 ,
3203.Xr altq 4 ,
3204.Xr divert 4 ,
3205.Xr dummynet 4 ,
3206.Xr if_bridge 4 ,
3207.Xr ip 4 ,
3208.Xr ipfirewall 4 ,
3209.Xr ng_ipfw 4 ,
3210.Xr protocols 5 ,
3211.Xr services 5 ,
3212.Xr init 8 ,
3213.Xr kldload 8 ,
3214.Xr reboot 8 ,
3215.Xr sysctl 8 ,
3216.Xr syslogd 8
3217.Sh HISTORY
3218The
3219.Nm
3220utility first appeared in
3221.Fx 2.0 .
3222.Nm dummynet
3223was introduced in
3224.Fx 2.2.8 .
3225Stateful extensions were introduced in
3226.Fx 4.0 .
3227.Nm ipfw2
3228was introduced in Summer 2002.
3229.Sh AUTHORS
3230.An Ugen J. S. Antsilevich ,
3231.An Poul-Henning Kamp ,
3232.An Alex Nash ,
3233.An Archie Cobbs ,
3234.An Luigi Rizzo .
3235.Pp
3236.An -nosplit
3237API based upon code written by
3238.An Daniel Boulet
3239for BSDI.
3240.Pp
3241Dummynet has been introduced by Luigi Rizzo in 1997-1998.
3242.Pp
3243Some early work (1999-2000) on the
3244.Nm dummynet
3245traffic shaper supported by Akamba Corp.
3246.Pp
3247The ipfw core (ipfw2) has been completely redesigned and
3248reimplemented by Luigi Rizzo in summer 2002. Further
3249actions and
3250options have been added by various developer over the years.
3251.Pp
3252.An -nosplit
3253In-kernel NAT support written by
3254.An Paolo Pisati Aq piso@FreeBSD.org
3255as part of a Summer of Code 2005 project.
3256.Pp
3257SCTP
3258.Nm nat
3259support has been developed by
3260.An The Centre for Advanced Internet Architectures (CAIA) Aq http://www.caia.swin.edu.au .
3261The primary developers and maintainers are David Hayes and Jason But.
3262For further information visit:
3263.Aq http://www.caia.swin.edu.au/urp/SONATA
3264.Pp
3265Delay profiles have been developed by Alessandro Cerri and
3266Luigi Rizzo, supported by the
3267European Commission within Projects Onelab and Onelab2.
3268.Sh BUGS
3269The syntax has grown over the years and sometimes it might be confusing.
3270Unfortunately, backward compatibility prevents cleaning up mistakes
3271made in the definition of the syntax.
3272.Pp
3273.Em !!! WARNING !!!
3274.Pp
3275Misconfiguring the firewall can put your computer in an unusable state,
3276possibly shutting down network services and requiring console access to
3277regain control of it.
3278.Pp
3279Incoming packet fragments diverted by
3280.Cm divert
3281are reassembled before delivery to the socket.
3282The action used on those packet is the one from the
3283rule which matches the first fragment of the packet.
3284.Pp
3285Packets diverted to userland, and then reinserted by a userland process
3286may lose various packet attributes.
3287The packet source interface name
3288will be preserved if it is shorter than 8 bytes and the userland process
3289saves and reuses the sockaddr_in
3290(as does
3291.Xr natd 8 ) ;
3292otherwise, it may be lost.
3293If a packet is reinserted in this manner, later rules may be incorrectly
3294applied, making the order of
3295.Cm divert
3296rules in the rule sequence very important.
3297.Pp
3298Dummynet drops all packets with IPv6 link-local addresses.
3299.Pp
3300Rules using
3301.Cm uid
3302or
3303.Cm gid
3304may not behave as expected.
3305In particular, incoming SYN packets may
3306have no uid or gid associated with them since they do not yet belong
3307to a TCP connection, and the uid/gid associated with a packet may not
3308be as expected if the associated process calls
3309.Xr setuid 2
3310or similar system calls.
3311.Pp
3312Rule syntax is subject to the command line environment and some patterns
3313may need to be escaped with the backslash character
3314or quoted appropriately.
3315.Pp
3316Due to the architecture of
3317.Xr libalias 3 ,
3318ipfw nat is not compatible with the TCP segmentation offloading (TSO).
3319Thus, to reliably nat your network traffic, please disable TSO
3320on your NICs using
3321.Xr ifconfig 8 .
3322.Pp
3323ICMP error messages are not implicitly matched by dynamic rules
3324for the respective conversations.
3325To avoid failures of network error detection and path MTU discovery,
3326ICMP error messages may need to be allowed explicitly through static
3327rules.
3328.Pp
3329Rules using
3330.Cm call
3331and
3332.Cm return
3333actions may lead to confusing behaviour if ruleset has mistakes,
3334and/or interaction with other subsystems (netgraph, dummynet, etc.) is used.
3335One possible case for this is packet leaving
3336.Nm
3337in subroutine on the input pass, while later on output encountering unpaired
3338.Cm return
3339first.
3340As the call stack is kept intact after input pass, packet will suddenly
3341return to the rule number used on input pass, not on output one.
3342Order of processing should be checked carefully to avoid such mistakes.
3343