1 '\" te
2 .\" To view license terms, attribution, and copyright for IP Filter, the
3 .\" default path is /usr/lib/ipf/IPFILTER.LICENCE. If the Illumos operating
4 .\" environment has been installed anywhere other than the default, modify the
5 .\" given path to access the file at the installed location.
6 .\" Portions Copyright (c) 2015, Joyent, Inc.
7 .TH IPF 4 "Mar 18, 2015"
8 .SH NAME
9 ipf, ipf.conf, ipf6.conf \- IP packet filter rule syntax
10 .SH DESCRIPTION
11 .PP
12 A rule file for \fBipf\fP may have any name or even be stdin. As
13 \fBipfstat\fP produces parsable rules as output when displaying the internal
14 kernel filter lists, it is quite plausible to use its output to feed back
15 into \fBipf\fP. Thus, to remove all filters on input packets, the following
16 could be done:
17 .nf
18
19 \fC# ipfstat \-i | ipf \-rf \-\fP
20 .fi
21 .SH GRAMMAR
22 .PP
23 The format used by \fBipf\fP for construction of filtering rules can be
24 described using the following grammar in BNF:
25 \fC
26 .nf
27 filter-rule = [ insert ] action in-out [ options ] [ tos ] [ ttl ]
28 [ proto ] ip [ group ].
29
30 insert = "@" decnumber .
31 action = block | "pass" | log | "count" | skip | auth | call .
32 in-out = "in" | "out" .
33 options = [ log ] [ tag ] [ "quick" ] [ "on" interface-name [ dup ]
34 [ froute ] [ replyto ] ] .
35 tos = "tos" decnumber | "tos" hexnumber .
36 ttl = "ttl" decnumber .
37 proto = "proto" protocol .
38 ip = srcdst [ flags ] [ with withopt ] [ icmp ] [ keep ] .
39 group = [ "head" decnumber ] [ "group" decnumber ] .
40
41 block = "block" [ return-icmp[return-code] | "return-rst" ] .
42 log = "log" [ "body" ] [ "first" ] [ "or-block" ] [ "level" loglevel ] .
43 tag = "tag" tagid .
44 skip = "skip" decnumber .
45 auth = "auth" | "preauth" .
46 call = "call" [ "now" ] function-name .
47 dup = "dup-to" interface-name [ ":" ipaddr ] .
48 froute = "fastroute" | "to" interface-name [ ":" ipaddr ] .
49 replyto = "reply-to" interface-name [ ":" ipaddr ] .
50 protocol = "tcp/udp" | "udp" | "tcp" | "icmp" | decnumber .
51 srcdst = "all" | fromto .
52 fromto = "from" [ "!" ] object "to" [ "!" ] object .
53
54 return-icmp = "return-icmp" | "return-icmp-as-dest" .
55 return-code = "(" icmp-code ")" .
56 object = addr [ port-comp | port-range ] .
57 addr = "any" | nummask | host-name [ "mask" ipaddr | "mask" hexnumber ] .
58 addr = "any" | "<thishost>" | nummask |
59 host-name [ "mask" ipaddr | "mask" hexnumber ] .
60 port-comp = "port" compare port-num .
61 port-range = "port" port-num range port-num .
62 flags = "flags" flag { flag } [ "/" flag { flag } ] .
63 with = "with" | "and" .
64 icmp = "icmp-type" icmp-type [ "code" decnumber ] .
65 return-code = "(" icmp-code ")" .
66 keep = "keep" "state" [ "(" state-options ")" ] | "keep" "frags" .
67 loglevel = facility"."priority | priority .
68
69 nummask = host-name [ "/" decnumber ] .
70 host-name = ipaddr | hostname | "any" .
71 ipaddr = host-num "." host-num "." host-num "." host-num .
72 host-num = digit [ digit [ digit ] ] .
73 port-num = service-name | decnumber .
74 state-options = state-opts [ "," state-options ] .
75
76 state-opts = "age" decnumber [ "/" decnumber ] | "strict" |
77 "no-icmp-err" | "limit" decnumber | "newisn" | "sync" .
78 withopt = [ "not" | "no" ] opttype [ withopt ] .
79 opttype = "ipopts" | "short" | "frag" | "opt" optname .
80 optname = ipopts [ "," optname ] .
81 ipopts = optlist | "sec-class" [ secname ] .
82 secname = seclvl [ "," secname ] .
83 seclvl = "unclass" | "confid" | "reserv-1" | "reserv-2" | "reserv-3" |
84 "reserv-4" | "secret" | "topsecret" .
85 icmp-type = "unreach" | "echo" | "echorep" | "squench" | "redir" |
86 "timex" | "paramprob" | "timest" | "timestrep" | "inforeq" |
87 "inforep" | "maskreq" | "maskrep" | decnumber .
88 icmp-code = decumber | "net-unr" | "host-unr" | "proto-unr" | "port-unr" |
89 "needfrag" | "srcfail" | "net-unk" | "host-unk" | "isolate" |
90 "net-prohib" | "host-prohib" | "net-tos" | "host-tos" |
91 "filter-prohib" | "host-preced" | "cutoff-preced" .
92 optlist = "nop" | "rr" | "zsu" | "mtup" | "mtur" | "encode" | "ts" |
93 "tr" | "sec" | "lsrr" | "e-sec" | "cipso" | "satid" | "ssrr" |
94 "addext" | "visa" | "imitd" | "eip" | "finn" .
95 facility = "kern" | "user" | "mail" | "daemon" | "auth" | "syslog" |
96 "lpr" | "news" | "uucp" | "cron" | "ftp" | "authpriv" |
97 "audit" | "logalert" | "local0" | "local1" | "local2" |
98 "local3" | "local4" | "local5" | "local6" | "local7" .
99 priority = "emerg" | "alert" | "crit" | "err" | "warn" | "notice" |
100 "info" | "debug" .
101
102 hexnumber = "0" "x" hexstring .
103 hexstring = hexdigit [ hexstring ] .
104 decnumber = digit [ decnumber ] .
105
106 compare = "=" | "!=" | "<" | ">" | "<=" | ">=" | "eq" | "ne" | "lt" |
107 "gt" | "le" | "ge" .
108 range = "<>" | "><" .
109 hexdigit = digit | "a" | "b" | "c" | "d" | "e" | "f" .
110 digit = "0" | "1" | "2" | "3" | "4" | "5" | "6" | "7" | "8" | "9" .
111 flag = "F" | "S" | "R" | "P" | "A" | "U" .
112 .fi
113 .PP
114 This syntax is somewhat simplified for readability, some combinations
115 that match this grammar are disallowed by the software because they do
116 not make sense (such as tcp \fBflags\fP for non-TCP packets).
117 .SH FILTER RULES
118 .PP
119 The "briefest" valid rules are (currently) no-ops and are of the form:
120 .nf
121 block in all
122 pass in all
123 log out all
124 count in all
125 .fi
126 .PP
127 Filter rules are checked in order, with the last matching rule
128 determining the fate of the packet (but see the \fBquick\fP option,
129 below).
130 .PP
131 Filters are installed by default at the end of the kernel's filter
132 lists, prepending the rule with \fB@n\fP will cause it to be inserted
133 as the n'th entry in the current list. This is especially useful when
134 modifying and testing active filter rulesets. See \fBipf\fP(1M) for more
135 information.
136 .SH ACTIONS
137 .PP
138 The action indicates what to do with the packet if it matches the rest
139 of the filter rule. Each rule MUST have an action. The following
140 actions are recognised:
141 .TP
142 .B block
143 indicates that the packet should be flagged to be dropped. In response
144 to blocking a packet, the filter may be instructed to send a reply
145 packet, either an ICMP packet (\fBreturn-icmp\fP), an ICMP packet
146 masquerading as being from the original packet's destination
147 (\fBreturn-icmp-as-dest\fP), or a TCP "reset" (\fBreturn-rst\fP). An
148 ICMP packet may be generated in response to any IP packet, and its
149 type may optionally be specified, but a TCP reset may only be used
150 with a rule which is being applied to TCP packets. When using
151 \fBreturn-icmp\fP or \fBreturn-icmp-as-dest\fP, it is possible to specify
152 the actual unreachable `type'. That is, whether it is a network
153 unreachable, port unreachable or even administratively
154 prohibited. This is done by enclosing the ICMP code associated with
155 it in parenthesis directly following \fBreturn-icmp\fP or
156 \fBreturn-icmp-as-dest\fP as follows:
157 .nf
158 block return-icmp(11) ...
159 .fi
160 .PP
161 Would return a Type-Of-Service (TOS) ICMP unreachable error.
162 .TP
163 .B pass
164 will flag the packet to be let through the filter.
165 .TP
166 .B log
167 causes the packet to be logged (as described in the LOGGING section
168 below) and has no effect on whether the packet will be allowed through
169 the filter.
170 .TP
171 .B count
172 causes the packet to be included in the accounting statistics kept by
173 the filter, and has no effect on whether the packet will be allowed through
174 the filter. These statistics are viewable with ipfstat(1M).
175 .TP
176 .B call
177 this action is used to invoke the named function in the kernel, which
178 must conform to a specific calling interface. Customised actions and
179 semantics can thus be implemented to supplement those available. This
180 feature is for use by knowledgeable hackers, and is not currently
181 documented.
182 .TP
183 .B "skip <n>"
184 causes the filter to skip over the next \fIn\fP filter rules. If a rule is
185 inserted or deleted inside the region being skipped over, then the value of
186 \fIn\fP is adjusted appropriately.
187 .TP
188 .B auth
189 this allows authentication to be performed by a user-space program running
190 and waiting for packet information to validate. The packet is held for a
191 period of time in an internal buffer whilst it waits for the program to return
192 to the kernel the \fIreal\fP flags for whether it should be allowed through
193 or not. Such a program might look at the source address and request some sort
194 of authentication from the user (such as a password) before allowing the
195 packet through or telling the kernel to drop it if from an unrecognised source.
196 .TP
197 .B preauth
198 tells the filter that for packets of this class, it should look in the
199 pre-authenticated list for further clarification. If no further matching
200 rule is found, the packet will be dropped (the FR_PREAUTH is not the same
201 as FR_PASS). If a further matching rule is found, the result from that is
202 used in its instead. This might be used in a situation where a person
203 \fIlogs in\fP to the firewall and it sets up some temporary rules defining
204 the access for that person.
205 .PP
206 The next word must be either \fBin\fP or \fBout\fP. Each packet
207 moving through the kernel is either inbound (just been received on an
208 interface, and moving towards the kernel's protocol processing) or
209 outbound (transmitted or forwarded by the stack, and on its way to an
210 interface). There is a requirement that each filter rule explicitly
211 state which side of the I/O it is to be used on.
212 .SH OPTIONS
213 .PP
214 The list of options is brief, and all are indeed optional. Where
215 options are used, they must be present in the order shown here. These
216 are the currently supported options:
217 .TP
218 .B log
219 indicates that, should this be the last matching rule, the packet
220 header will be written to the \fBipl\fP log (as described in the
221 LOGGING section below).
222 .TP
223 .B tag tagid
224 indicates that, if this rule causes the packet to be logged or entered
225 in the state table, the tagid will be logged as part of the log entry.
226 This can be used to quickly match "similar" rules in scripts that post
227 process the log files for e.g. generation of security reports or accounting
228 purposes. The tagid is a 32 bit unsigned integer.
229 .TP
230 .B quick
231 allows "short-cut" rules in order to speed up the filter or override
232 later rules. If a packet matches a filter rule which is marked as
233 \fBquick\fP, this rule will be the last rule checked, allowing a
234 "short-circuit" path to avoid processing later rules for this
235 packet. The current status of the packet (after any effects of the
236 current rule) will determine whether it is passed or blocked.
237 .IP
238 If this option is missing, the rule is taken to be a "fall-through"
239 rule, meaning that the result of the match (block/pass) is saved and
240 that processing will continue to see if there are any more matches.
241 .TP
242 .B on
243 allows an interface name to be incorporated into the matching
244 procedure. Interface names are as printed by "netstat \-i". If this
245 option is used, the rule will only match if the packet is going
246 through that interface in the specified direction (in/out). If this
247 option is absent, the rule is taken to be applied to a packet
248 regardless of the interface it is present on (i.e. on all interfaces).
249 Filter rulesets are common to all interfaces, rather than having a
250 filter list for each interface.
251 .IP
252 This option is especially useful for simple IP-spoofing protection:
253 packets should only be allowed to pass inbound on the interface from
254 which the specified source address would be expected, others may be
255 logged and/or dropped.
256 .TP
257 .B dup-to
258 causes the packet to be copied, and the duplicate packet to be sent
259 outbound on the specified interface, optionally with the destination
260 IP address changed to that specified. This is useful for off-host
261 logging, using a network sniffer.
262 .TP
263 .B to
264 causes the packet to be moved to the outbound queue on the
265 specified interface. This can be used to circumvent kernel routing
266 decisions, and even to bypass the rest of the kernel processing of the
267 packet (if applied to an inbound rule). It is thus possible to
268 construct a firewall that behaves transparently, like a filtering hub
269 or switch, rather than a router. The \fBfastroute\fP keyword is a
270 synonym for this option.
271 .SH MATCHING PARAMETERS
272 .PP
273 The keywords described in this section are used to describe attributes
274 of the packet to be used when determining whether rules match or don't
275 match. The following general-purpose attributes are provided for
276 matching, and must be used in this order:
277 .TP
278 .B tos
279 packets with different Type-Of-Service values can be filtered.
280 Individual service levels or combinations can be filtered upon. The
281 value for the TOS mask can either be represented as a hex number or a
282 decimal integer value.
283 .TP
284 .B ttl
285 packets may also be selected by their Time-To-Live value. The value given in
286 the filter rule must exactly match that in the packet for a match to occur.
287 This value can only be given as a decimal integer value.
288 .TP
289 .B proto
290 allows a specific protocol to be matched against. All protocol names
291 found in \fB/etc/protocols\fP are recognised and may be used.
292 However, the protocol may also be given as a DECIMAL number, allowing
293 for rules to match your own protocols, or new ones which would
294 out-date any attempted listing.
295 .IP
296 The special protocol keyword \fBtcp/udp\fP may be used to match either
297 a TCP or a UDP packet, and has been added as a convenience to save
298 duplication of otherwise-identical rules.
299 .\" XXX grammar should reflect this (/etc/protocols)
300 .PP
301 The \fBfrom\fP and \fBto\fP keywords are used to match against IP
302 addresses (and optionally port numbers). Rules must specify BOTH
303 source and destination parameters.
304 .PP
305 IP addresses may be specified in one of two ways: as a numerical
306 address\fB/\fPmask, or as a hostname \fBmask\fP netmask. The hostname
307 may either be a valid hostname, from either the hosts file or DNS
308 (depending on your configuration and library) or of the dotted numeric
309 form. There is no special designation for networks but network names
310 are recognised. Note that having your filter rules depend on DNS
311 results can introduce an avenue of attack, and is discouraged.
312 .PP
313 There is a special case for the hostname \fBany\fP which is taken to
314 be 0.0.0.0/0 (see below for mask syntax) and matches all IP addresses.
315 Only the presence of "any" has an implied mask, in all other
316 situations, a hostname MUST be accompanied by a mask. It is possible
317 to give "any" a hostmask, but in the context of this language, it is
318 non-sensical.
319 .PP
320 The numerical format "x\fB/\fPy" indicates that a mask of y
321 consecutive 1 bits set is generated, starting with the MSB, so a y value
322 of 16 would give 0xffff0000. The symbolic "x \fBmask\fP y" indicates
323 that the mask y is in dotted IP notation or a hexadecimal number of
324 the form 0x12345678. Note that all the bits of the IP address
325 indicated by the bitmask must match the address on the packet exactly;
326 there isn't currently a way to invert the sense of the match, or to
327 match ranges of IP addresses which do not express themselves easily as
328 bitmasks (anthropomorphization; it's not just for breakfast anymore).
329 .PP
330 If a \fBport\fP match is included, for either or both of source and
331 destination, then it is only applied to
332 .\" XXX - "may only be" ? how does this apply to other protocols? will it not match, or will it be ignored?
333 TCP and UDP packets. If there is no \fBproto\fP match parameter,
334 packets from both protocols are compared. This is equivalent to "proto
335 tcp/udp". When composing \fBport\fP comparisons, either the service
336 name or an integer port number may be used. Port comparisons may be
337 done in a number of forms, with a number of comparison operators, or
338 port ranges may be specified. When the port appears as part of the
339 \fBfrom\fP object, it matches the source port number, when it appears
340 as part of the \fBto\fP object, it matches the destination port number.
341 See the examples for more information.
342 .PP
343 The \fBall\fP keyword is essentially a synonym for "from any to any"
344 with no other match parameters.
345 .PP
346 Following the source and destination matching parameters, the
347 following additional parameters may be used:
348 .TP
349 .B with
350 is used to match irregular attributes that some packets may have
351 associated with them. To match the presence of IP options in general,
352 use \fBwith ipopts\fP. To match packets that are too short to contain
353 a complete header, use \fBwith short\fP. To match fragmented packets,
354 use \fBwith frag\fP. For more specific filtering on IP options,
355 individual options can be listed.
356 .IP
357 Before any parameter used after the \fBwith\fP keyword, the word
358 \fBnot\fP or \fBno\fP may be inserted to cause the filter rule to only
359 match if the option(s) is not present.
360 .IP
361 Multiple consecutive \fBwith\fP clauses are allowed. Alternatively,
362 the keyword \fBand\fP may be used in place of \fBwith\fP, this is
363 provided purely to make the rules more readable ("with ... and ...").
364 When multiple clauses are listed, all those must match to cause a
365 match of the rule.
366 .\" XXX describe the options more specifically in a separate section
367 .TP
368 .B flags
369 is only effective for TCP filtering. Each of the letters possible
370 represents one of the possible flags that can be set in the TCP
371 header. The association is as follows:
372 .LP
373 .nf
374 F - FIN
375 S - SYN
376 R - RST
377 P - PUSH
378 A - ACK
379 U - URG
380 .fi
381 .IP
382 The various flag symbols may be used in combination, so that "SA"
383 would represent a SYN-ACK combination present in a packet. There is
384 nothing preventing the specification of combinations, such as "SFR",
385 that would not normally be generated by law-abiding TCP
386 implementations. However, to guard against weird aberrations, it is
387 necessary to state which flags you are filtering against. To allow
388 this, it is possible to set a mask indicating which TCP flags you wish
389 to compare (i.e., those you deem significant). This is done by
390 appending "/<flags>" to the set of TCP flags you wish to match
391 against, e.g.:
392 .LP
393 .nf
394 ... flags S
395 # becomes "flags S/AUPRFS" and will match
396 # packets with ONLY the SYN flag set.
397
398 ... flags SA
399 # becomes "flags SA/AUPRFS" and will match any
400 # packet with only the SYN and ACK flags set.
401
402 ... flags S/SA
403 # will match any packet with just the SYN flag set
404 # out of the SYN-ACK pair; the common "establish"
405 # keyword action. "S/SA" will NOT match a packet
406 # with BOTH SYN and ACK set, but WILL match "SFP".
407 .fi
408 .TP
409 .B icmp-type
410 is only effective when used with \fBproto icmp\fP and must NOT be used
411 in conjunction with \fBflags\fP. There are a number of types, which can be
412 referred to by an abbreviation recognised by this language, or the numbers
413 with which they are associated can be used. The most important from
414 a security point of view is the ICMP redirect.
415 .SH KEEP HISTORY
416 .PP
417 The second last parameter which can be set for a filter rule is whether or not
418 to record historical information for that packet, and what sort to keep. The
419 following information can be kept:
420 .TP
421 .B state
422 keeps information about the flow of a communication session. State can
423 be kept for TCP, UDP, and ICMP packets.
424 .TP
425 .B frags
426 keeps information on fragmented packets, to be applied to later
427 fragments.
428 .PP
429 allowing packets which match these to flow straight through, rather
430 than going through the access control list.
431 .SH GROUPS
432 The last pair of parameters control filter rule "grouping". By default, all
433 filter rules are placed in group 0 if no other group is specified. To add a
434 rule to a non-default group, the group must first be started by creating a
435 group \fIhead\fP. If a packet matches a rule which is the \fIhead\fP of a
436 group, the filter processing then switches to the group, using that rule as
437 the default for the group. If \fBquick\fP is used with a \fBhead\fP rule, rule
438 processing isn't stopped until it has returned from processing the group.
439 .PP
440 A rule may be both the head for a new group and a member of a non-default
441 group (\fBhead\fP and \fBgroup\fP may be used together in a rule).
442 .TP
443 .B "head <n>"
444 indicates that a new group (number n) should be created.
445 .TP
446 .B "group <n>"
447 indicates that the rule should be put in group (number n) rather than group 0.
448 .SH LOGGING
449 .PP
450 When a packet is logged, with either the \fBlog\fP action or option,
451 the headers of the packet are written to the \fBipl\fP packet logging
452 pseudo-device. Immediately following the \fBlog\fP keyword, the
453 following qualifiers may be used (in order):
454 .TP
455 .B body
456 indicates that the first 128 bytes of the packet contents will be
457 logged after the headers.
458 .TP
459 .B first
460 If log is being used in conjunction with a "keep" option, it is recommended
461 that this option is also applied so that only the triggering packet is logged
462 and not every packet which thereafter matches state information.
463 .TP
464 .B or-block
465 indicates that, if for some reason the filter is unable to log the
466 packet (such as the log reader being too slow) then the rule should be
467 interpreted as if the action was \fBblock\fP for this packet.
468 .TP
469 .B "level <loglevel>"
470 indicates what logging facility and priority, or just priority with
471 the default facility being used, will be used to log information about
472 this packet using ipmon's -s option.
473 .PP
474 See ipl(4) for the format of records written
475 to this device. The ipmon(1M) program can be used to read and format
476 this log.
477 .SH EXAMPLES
478 .PP
479 The \fBquick\fP option is good for rules such as:
480 \fC
481 .nf
482 block in quick from any to any with ipopts
483 .fi
484 .PP
485 which will match any packet with a non-standard header length (IP
486 options present) and abort further processing of later rules,
487 recording a match and also that the packet should be blocked.
488 .PP
489 The "fall-through" rule parsing allows for effects such as this:
490 .LP
491 .nf
492 block in from any to any port < 6000
493 pass in from any to any port >= 6000
494 block in from any to any port > 6003
495 .fi
496 .PP
497 which sets up the range 6000-6003 as being permitted and all others being
498 denied. Note that the effect of the first rule is overridden by subsequent
499 rules. Another (easier) way to do the same is:
500 .LP
501 .nf
502 block in from any to any port 6000 <> 6003
503 pass in from any to any port 5999 >< 6004
504 .fi
505 .PP
506 Note that both the "block" and "pass" are needed here to effect a
507 result as a failed match on the "block" action does not imply a pass,
508 only that the rule hasn't taken effect. To then allow ports < 1024, a
509 rule such as:
510 .LP
511 .nf
512 pass in quick from any to any port < 1024
513 .fi
514 .PP
515 would be needed before the first block. To create a new group for
516 processing all inbound packets on le0/le1/lo0, with the default being to block
517 all inbound packets, we would do something like:
518 .LP
519 .nf
520 block in all
521 block in quick on le0 all head 100
522 block in quick on le1 all head 200
523 block in quick on lo0 all head 300
524 .fi
525 .PP
526
527 and to then allow ICMP packets in on le0, only, we would do:
528 .LP
529 .nf
530 pass in proto icmp all group 100
531 .fi
532 .PP
533 Note that because only inbound packets on le0 are used processed by group 100,
534 there is no need to respecify the interface name. Likewise, we could further
535 breakup processing of TCP, etc, as follows:
536 .LP
537 .nf
538 block in proto tcp all head 110 group 100
539 pass in from any to any port = 23 group 110
540 .fi
541 .PP
542 and so on. The last line, if written without the groups would be:
543 .LP
544 .nf
545 pass in on le0 proto tcp from any to any port = telnet
546 .fi
547 .PP
548 Note, that if we wanted to say "port = telnet", "proto tcp" would
549 need to be specified as the parser interprets each rule on its own and
550 qualifies all service/port names with the protocol specified.
551 .SH FILES
552 /dev/ipauth
553 .br
554 /dev/ipl
555 .br
556 /dev/ipstate
557 .br
558 /etc/hosts
559 .br
560 /etc/services
561 .SH SEE ALSO
562 \fBipnat\fR(4), \fBipf\fR(1M), \fBipfstat\fR(1M), \fBipfilter\fR(5)